jpad journal

AND option

OR option



M. Senda1, K. Ishii2, K. Ito3, T. Ikeuchi4, H. Matsuda5, T. Iwatsubo6, A. Iwata7, R. Ihara7, K. Suzuki8, K. Kasuga4, Y. Ikari1,9, Y. Niimi6, H. Arai10, A. Tamaoka11, Y. Arahata3, Y. Itoh12, H. Tachibana13, Y. Ichimiya14, S. Washizuka15, T. Odawara16, K. Ishii17, K. Ono18, T. Yokota19, A. Nakanishi20, E. Matsubara21, H. Mori12, H. Shimada12


1. Kobe City Medical Center General Hospital, Japan; 2. Tokyo Metropolitan Institute of Gerontology, Japan; 3. National Center for Geriatrics and Gerontology, Japan; 4. Niigata University, Japan; 5. National Center of Neurology and Psychiatry, Japan; (currently, Southern Tohoku Drug Development and Cyclotron Research Center, Japan); 6. The University of Tokyo, Japan; 7. The University of Tokyo, Japan; (currently, Tokyo Metropolitan Geriatric Hospital, Japan); 8. The University of Tokyo, Japan; (currently, National Defense Medical College, Japan); 9. Osaka University, Japan; 10. Tohoku University, Japan; 11. University of Tsukuba, Japan; 12. Osaka City University, Japan; 13. Kobe University, Japan; 14. Juntendo Tokyo Koto Geriatric Medical Center, Japan; 15. Shinshu University, Japan; 16. Yokohama City University, Japan; 17. Kindai University, Japan; 18. Showa University, Japan; 19. Tokyo Medical and Dental University, Japan; 20. Osaka City Kosaiin Hospital, Japan; 21. Oita University, Japan

Corresponding Author: Michio Senda, Division of Molecular Imaging Research Kobe City Medical Center General Hospital (KCGH), 2-1-1 Minatojima-Minamimachi, Chuo-ku, Kobe 650-0047 Japan, E-mail:, Phone: 81-78-304-5212, Fax: 81-78-304-5201.



BACKGROUND: PET (positron emission tomography) and CSF (cerebrospinal fluid) provide the “ATN” (Amyloid, Tau, Neurodegeneration) classification and play an essential role in early and differential diagnosis of Alzheimer’s disease (AD).
OBJECTIVE: Biomarkers were evaluated in a Japanese multicenter study on cognitively unimpaired subjects (CU) and early (E) and late (L) mild cognitive impairment (MCI) patients.
MEASUREMENTS: A total of 38 (26 CU, 7 EMCI, 5 LMCI) subjects with the age of 65-84 were enrolled. Amyloid-PET and FDG-PET as well as structural MRI were acquired on all of them, with an additional tau-PET with 18F-flortaucipir on 15 and CSF measurement of Aβ1-42, P-tau, and T-tau on 18 subjects. Positivity of amyloid and tau was determined based on the positive result of either PET or CSF.
RESULTS: The amyloid positivity was 13/38, with discordance between PET and CSF in 6/18. Cortical tau deposition quantified with PET was significantly correlated with CSF P-tau, in spite of discordance in the binary positivity between visual PET interpretation and CSF P-tau in 5/8 (PET-/CSF+). Tau was positive in 7/9 amyloid positive and 8/16 amyloid negative subjects who underwent tau measurement, respectively. Overall, a large number of subjects presented quantitative measures and/or visual read that are close to the borderline of binary positivity, which caused, at least partly, the discordance between PET and CSF in amyloid and/or tau. Nine subjects presented either tau or FDG-PET positive while amyloid was negative, suggesting the possibility of non-AD disorders.
CONCLUSION: Positivity rate of amyloid and tau, together with their relationship, was consistent with previous reports. Multicenter study on subjects with very mild or no cognitive impairment may need refining the positivity criteria and cutoff level as well as strict quality control of the measurements.

Key words: Alzheimer’s disease, PET, CSF biomarker, amyloid, tau.




Early and differential diagnosis of Alzheimer’s disease (AD) has been drawing more and more attention these days as the target population of the therapeutic trials has shifted toward the early phases of the AD continuum. Biomarkers including PET, MRI and cerebrospinal fluid(CSF)/plasma play an essential role in such early phases, where clinical manifestation and behavioral findings are limited. Jack et al (1) extracted three markers, i.e., amyloid (A), tau (T) and neurodegeneration (N), and proposed the “ATN” classification for differential diagnosis of AD continuum. PET provides imaging and quantification of amyloid and tau deposition as well as neurodegeneration evaluable with 18F-fluorodeoxyglucose(FDG)-PET. Amyloid and tau can also be evaluated with CSF sampling, and recently with plasma as well, and MRI volumetry has also been used as a marker of neurodegeneration.
In Japan, a large-scale prospective observational study called J-ADNI (Japanese Alzheimer’s Disease Neuroimaging Initiative) was completed (2), in which a total of 537 subjects were enrolled, comprising 154 cognitively unimpaired subjects (CU), 234 MCI and 149 AD patients.
Then, a new version of J-ADNI was designed by the same group, named “AMED Preclinical AD Study”, which focused on CU and MCI and acquired amyloid-PET and FDG-PET on all subjects. Part of the subjects also underwent a tau-PET scan and/or a CSF sampling. The objective of the study was to evaluate PET and MRI images and CSF biomarkers in CU and MCI subjects in Japan, compare those biomarkers between modalities to explore their reliability and usefulness in such early-phase subjects, and obtain a rough idea of the fractions of ATN-based classifications. This report summarizes the results of the study that was recently completed.




The study was a non-randomized prospective observational study, and was designed and conducted in accordance with the ethical principles as proclaimed in the Declaration of Helsinki. The study protocol was first approved by Ethical Committee of Osaka City University Graduate School of Medicine (site of leading PI) and registered as UMIN000019926, and was later re-approved by Osaka City University Hospital Certified Review Board when the Japanese Law on Clinical Research was enacted, and was registered as jRCTs051180239. The protocol was also approved by each participating site according to the Japanese regulations and ethics guidance. The tau-PET portion of the study was designed as a nominally separate add-on study when tau-PET became available, though limited, later in the course of the research project, and was approved and registered as jRCTs051190065.
The subjects were enrolled at a total of 14 clinical sites and consisted of 26 CU subjects and 12 MCI patients (7 early (E) MCI and 5 late (L) MCI as classified below) based on the neuropsychological tests.
Written informed consent was obtained from each subject and the study partner such as a family member of the subject.
The CU subjects were 65-82 years old without any memory problem and CDR-J=0. The MMSE-J score was 25-30 (higher than or equal to 24), and the delayed recall score of WMS-R logical memory (WMS-R LM II) ranged from 4 to 19 except for one subject (See footnote of Table 1).
The MCI subjects were 65-84 years old with objective persistent memory impairment reported by the study partner and CDR-J=0.5 with memory box score being 0.5 or higher. The MMSE-J score was 27-30 (higher than or equal to 24). The WMS-R LM II was used to classify the MCI subjects into EMCI (WMS-R LM II = 3-6, 5-9, 9-11) and LMCI (≤2, ≤4, ≤8) depending on the educational years (0-7, 8-15, ≥16 years, respectively), and was 6-18 for EMCI and 0-8 for LMCI.

CSF measurements and genotyping

CSF was collected from 18 subjects by lumbar puncture and stored in polypropylene tubes at -80℃ until biochemical analysis. CSF concentration of Aβ1-42 was analyzed using V-PLEX Aβ Peptide Panel 1 kit with MESO QuickPlex SQ120 (MesoScale Discovery, Rockville, MD). CSF phosphorylated tau (P-tau) and total tau (T-tau) were measured using commercially available ELISA kits, INNOTEST hTAU and PHOSPHO-TAU (181P) (Fujirebio Europe, Belgium), respectively, according to the manufacturer’s instructions. Stability of the results was monitored in the Alzheimer’s Association QC program. Cutoff values (Aβ42<378.7 pg/mL, P-tau>29.1 pg/mL, and T-tau>88.8 pg/mL) that best discriminated PiB-PET positive AD patients from PiB-PET negative CU subjects were determined using independent J-ADNI cohort (2). Because the CSF assays used in this study were different from those used in J-ADNI study, calibration between two assays were performed.
APOE genotyping (rs429358 and rs7412) was performed by Taq-Man based assay using blood samples.

MRI imaging

The brain MRI was acquired for each subject using a 3-Tesla or 1.5-Tesla scanner. The structural 3D-T1 images (MP-RAGE or IR-SPGR) were analyzed with FreeSurfer (Ver. 6.0) to measure the regional cerebral gray-matter volumes. Because the absolute volumetry depends on the version of the software and other conditions, the regional atrophy of the subject was derived as z-score using mean and SD of the baseline scan for the 26 CU subjects of this study. The volume of 8 regions in the temporal lobe (right and left entorhinal cortex, parahippocampal gyrus, hippocampus, and amygdala) were summed up and the z score was derived as a measure of the temporal lobe atrophy for each subject.

PET image acquisition

All subjects underwent an amyloid-PET and an FDG-PET. Each PET imaging site, together with the PET camera, was qualified, in which the reconstruction parameters were determined for each PET camera so that all the PET cameras satisfied the image quality criteria with the Hoffman 3D brain phantom and the uniform cylindrical phantom (3).
For amyloid PET, either 11C-PiB (PiB), 18F-florbetapir (FBP) or 18F-flutemetamol (FMM) was used for 23, 13, and 2 subjects, respectively. The injection activity was 555MBq, 370MBq, 185MBq, the uptake time (start of emission scan post injection) was 50min, 50min, 90min, and the scan duration was 20min, 20min, 30min, for PiB, FBP, and FMM, respectively.
For the FDG scan, after at least 4 hours of fasting, the subject was administered with 185 MBq of 18F-FDG in a quiet, dimly lit room while resting in a reclining chair or bed, and the subject remained in the condition until several minutes before the start of the scanning session. The PET emission data was acquired for 30 minutes starting at 30 minutes post injection.
Tau-PET was performed with 18F-flortaucipir (FTP) on 15 subjects. Because tau-PET was not ready until late in the course of the research project, the time span from amyloid to tau-PET ranged from 1.0 to 2.0 (mean 1.56) years. The subject was administered with 240.5 MBq of FTP and a 30 min emission scan started 75 min post injection.
No adverse effects were observed at the PET scans of this study.

PET image analysis

The amyloid PET images were binary interpreted visually in a blind manner by the readers who were qualified for this study, and the adjudicator (K.I.) confirmed them. The PiB images were interpreted visually using the criteria adopted in J-ADNI (4), and the FMM and FBP images were interpreted with each vendor’s criteria.
As a quantitative analysis of the amyloid PET, mean cortical standardized uptake value ratio (mcSUVR) of PiB images was computed using the cerebellar cortex as a reference based on the method of J-ADNI, and the cutoff value of 1.5 was used to determine the quantitative positivity (4). The FMM images were analyzed with CORTEX ID (GE Healthcare) to derive mcSUVR using the pons as a reference, for which the cutoff value of 0.58 was used for the quantitative positivity (5). The FBP images were analyzed with MIMneuro (MIM Software) to derive mcSUVR using the whole cerebellum as a reference, for which the cutoff value of 1.10 was used for the quantitative positivity (6).
The FDG images together with the semiquantitative 3D-SSP results were visually interpreted by three independent readers followed by a consensus read in the same way as J-ADNI (7), and the images were classified into N1 (normal), N2 (reflecting atrophy), N3, P1 (AD pattern), P2 (FTD pattern), P3, and P1+ (DLB pattern) (8). No one presented N3 or P3 in this study. The DLB pattern criteria was interpreted in a broader sense to include cases with occipital hypometabolism extending to neighboring areas even if typical temporoparietal hypometabolism was not observed. The FDG images were also quantified with AD t-sum (9) using the module PALZ in the PMOD software package (Ver. 3.2; PMOD Technologies, Zurich, Switzerland), which were then converted into PET score [10] that reflects the severity of temporoparietal hypometabolism (AD pattern).
The FTP images were interpreted and classified into AD negative, AD+ and AD++, according to the vendor’s criteria that regards cortical uptake except anterior temporal as AD-related ( The FTP-PET was also analyzed with MUBADA-PERSI method to derive SUVR over the area affected by AD process (posterior temporal, occipital, parietal and part of frontal cortex) with white matter as a reference (11, 12).

Follow up

Whenever possible, each subject was followed up every year with a general clinical interview with neuropsychological tests, an MRI scan, and an FDG-PET scan. As a result, 1-year follow-up data were acquired on 32 subjects, and 2-year follow-up on 5 subjects.

Statistical methods

Because the number of subjects was small, descriptive results were presented in general. Proportion of positivity was compared between groups using chi-square tests, in which EMCI and LMCI were combined to increase the number of observations. Statistical tests were also performed on the Pearson correlation coefficient between two variables.



Findings of each subject

Table 1 describes findings of each subject as classified according to the ATN concept. In this study, amyloid (A) was interpreted as positive (A+) when either PET or CSF Aβ was positive. Tau (T) was interpreted as positive (T+) when either PET or CSF P-tau was positive; negative (T-) when either of them was obtained and neither of them were positive; and was “na” (not available) (Tna) when neither of them were obtained. Neurodegeneration (N) was interpreted as positive (N+) when the consensus visual read of FDG-PET showed a progressive pattern (P1, P2, or P1+), and negative (N-) when it was a non-progressive pattern (N1 or N2).

Table 1. Findings for each subject and ATN classification


The amyloid positivity rate was 13/38 overall (6/26 CU, 4/7 EMCI, 3/5 LMCI, p>0.05 between CU and MCI), while it was 8/38 based on the PET alone (3/26 CU, 2/7 EMCI, 3/5 LMCI, p>0.05 between CU and MCI).
Tau was positive for 7, negative for 2 and not available for 4 out of the 13 A+ subjects, being 2, 2, 2 and 5, 0, 2 out of the 6 A+ CU and 7 A+ MCI subjects, respectively.
FDG-PET showed a progressive pattern in 6/13 A+ subjects (3/6 CU, 1/4 EMCI, 2/3 LMCI) as compared to 3/25 A- subjects (0/20 CU, 1/3 EMCI, 2/2 LMCI). Significant difference was observed in the FDG-PET positivity (N+) proportion between A+ and A- (p<0.05) as well as between CU (3/26) and MCI (6/12) (p<0.05).
Of interest, tau was positive for as many as 8 (negative for 8, not available for 9) out of the 25 amyloid negative subjects, indicating tau deposition without AD pathological process. It should be noted that all the 8 A-T+ subjects was tau positive due to CSF test, in spite of negative tau PET for two of them.
Association of APOE genotypes with amyloid PET (p>0.6) or CSF Aβ (p>0.5), or with any other biomarkers, was not observed for the presence of E4, probably due to the small number of subjects.

Representative cases

Figures 1 (#24, LMCI) and 2 (#22, CU) depict a case with prodromal AD (A+T+N+) and preclinical AD (A+T+N-), respectively. PET and CSF were discordant for “A” and/or “T” in both cases, which may be related to visually equivocal images and near-cutoff level quantified values. In the case of Figure 2, CSF P-tau was positive while tau PET was negative, consistent with the report of earlier and more sensitive positivity of CSF P-tau than tau-PET in the AD continuum (13).
Four cases (1 CU, 1 EMCI, 2 LMCIs) showed a mild/partial DLB pattern in FDG-PET marked with “P1+” in Table 1, featuring hypometabolism in the occipital cortex extending into surrounding areas but not showing a typical AD pattern of temporo-parietal hypometabolism. Amyloid was positive for 3/4 and tau was positive for 4/4. Figure 3 (#26, CU) depicts one of them.

Figure 1. PiB, FTP and FDG-PET of a female LMCI patient in her 70s (#24) interpreted as prodromal AD

Amyloid PET with PiB was visually negative, as the left parietal mild accumulation did not reach the cortical surface (arrow). However, the subject was classified as “A+” because quantitative analysis revealed SUVR (1.57) above cutoff. The CSF Aβ was negative (399.8 pg/mL). The FTP-PET showed abnormal tau accumulation in the left posterior temporal lobe (arrow), typical of AD process. Note off-target uptake of FTP in choroid plexus (arrowheads), substantia nigra, and striatum. The FDG-PET was read as temporo-parietal hypometabolism indicating AD pattern in the baseline that progressed in two years (arrows). PET score and MRI z-score also increased in two years: from 0.76 to 1.08 and from 2.7 to 3.2, respectively.

Figure 2. PiB, FTP and FDG-PET of a female CU subject of her 70s (#22) interpreted as preclinical AD

PiB-PET revealed positive amyloid accumulation in the left temporal and parietal areas (arrows). Tau PET with FTP acquired 1 year later was negative, because mild activity along the cortical rim was interpreted as off-target uptake by the meninges (short arrows) and that the left anterior temporal uptake was considered non-pathological within the AD continuum (long arrow). CSF P-tau was positive. FDG-PET showed a normal pattern.

Figure 3. FBP, FTP and FDG-PET of a female CU subject of her 70s (#26)

FBP-PET presented negative amyloid, and tau was negative in FTP-PET, although CSF showed positive Aβ (317.9pg/mL) and P-tau (38.2pg/mL). FDG-PET revealed a DLB pattern, presenting occipital hypometabolism (long arrows) extending to the right temporal and parietal cortex (short arrows), which progressed 1 year later. Note cingulate island sign denoting preserved metabolism in the posterior cingulate cortex (arrowheads).


Association between PET and CSF

For the 18 subjects, in which CSF data were obtained, amyloid positivity by CSF agreed with that by PET in 12 cases while 6 showed a discordance (Table 1). The rate of discordance was consistent with previous reports and may be caused by various factors (13).
Quantified tau uptake (SUVR) measured with FTP-PET using MUBADA-PERSI method was significantly correlated with CSF P-tau (r=0.92, p<0.001, n=8) (Figure 4). Although the cutoff for SUVR with MUBADA-PERSI SUVR is not established yet, the visual read of the FTP-PET was positive only for two of them. In CSF P-tau, however, 7 out of the 8 subjects showed P-tau above the cutoff level, indicating a discordance in the tau positivity between PET and CSF. This is consistent with recent investigations that reported earlier or more sensitive positivity of CSF P-tau than tau-PET in the AD continuum (i.e. in amyloid positive subjects), because secretion of soluble p-tau to CSF is increased by Aβ pathology before tau begins to accumulate in the brain (14).

Figure 4. Scatter plot of tau uptake (SUVR) quantified with FTP-PET and CSF P-tau

Red marks indicate PET-positive cases by visual read. Arrow indicates cutoff value for CSF P-tau.



PET/CSF discordance for amyloid and tau

This study suffers limitations such as the small number of subjects, poor follow-up records, and lack of tau-PET and CSF measurement for a large fraction of the subjects. However, some findings are notable.
The rate of amyloid positivity based on the combination of PET and CSF (6/26=23% for CU, 7/12=58% for MCI) was consistent with previous reports including J-ADNI. Discordance of positivity between amyloid-PET and CSF Aβ was observed in 6 subjects (5 PET-/CSF+, 1 PET+/CSF-), suggesting higher sensitivity of CSF, which was also consistent with the ADNI data on CU and MCI (13).
The rate of tau positivity was 4/6 for amyloid PET-/CSF Aβ+ or amyloid PET+/CSF Aβ-, and 2/2 for amyloid PET+/CSF Aβ+ in this study (Table 1). This was agreeable with the above ADNI data, in which the former two groups presented significantly lower CSF p-tau and PET-measured tau deposition than the latter and suggested earlier manifestations of AD process (13).
It is known that CSF p-tau is quantitatively associated with PET-measured tau deposition, especially in the AD continuum, and that CSF p-tau rises in the earlier stage than the pathological uptake of FTP-PET accrues (14). The result of the present study indicated a similar association in spite of the small number of subjects (Figure 4), in which FTP-PET was quantified with MUBADA, and that CSF p-tau was more sensitive than visual read of FTP-PET. In the early phase of AD continuum, CSF p-tau and FTP-PET could be considered to reflect different pathological changes, as the former may indicate excess secretion of p-tau to CSF and the latter represents tissue tau deposition. The present study adopted the criteria of visual FTP-PET interpretation to determine the positivity, by which the anterior temporal FTP uptake was considered insignificant. Because MUBADA VOI covers wide cortical areas, it may not be a sensitive measure of early tau deposition in the AD continuum that begins in the temporal cortex. Although a recent study suggests the earliest tau deposition at the rhinal cortex located in the anteromedial temporal lobe (15), it was hard to quantify the pathological FTP uptake in that region due to off-target binding to the choroid plexus in the present study (Figure 2). In that sense, use of FTP was another limitation of this study. Tau PET drugs with little off-target binding such as 18F-MK-6240 or 18F-PI-2620 may be more suitable for evaluation of the earliest stage of AD.

Neurodegeneration marker

In this study only visual assessment of FDG-PET was used to evaluate the “N” (neurodegeneration) marker, and quantitative PET score was not used so that one “N-” subject presented a high PET score (#4). Although CSF T-tau and MRI-volumetry are also regarded as an N-marker, their association with FDG-PET remains to be investigated as they represent different pathophysiology.
Hypometabolism depicted by FDG-PET reflects reduced neuronal activity in general, regardless of pathophysiology. No subjects showed A+T-N+ in this study, which agrees with the concept of tau deposition leading to neurodegeneration in AD continuum, although such manifestation, if occurred, might have suggested combined AD and non-AD processes. Outside the AD continuum (A-), however, FDG-PET neurodegeneration was positive in 3/5 MCI (2/2 LMCI) subjects and was not observed in CU subjects (0/20), which is consistent with the above notion and agrees with previous reports (16).

Binary criteria

A large number of subjects presented quantitative measures and/or visual read that are close to the borderline of binary positivity in the present study, which caused, at least partly, discordance between PET and CSF in amyloid and/or tau, such as the cases in Figures 1 and 2. This is understandable because most of the subjects in the present study were in the early phase of AD continuum or of a non-AD disease if any, and that the current criteria and cutoff level have been derived from differential diagnosis of AD patients from CU subjects. To deal with early-phase subjects having no or very mild cognitive impairment, the criteria and cutoff level of the biomarkers may need refining, and the data acquisition may need strict quality control.

Non-AD disorders

The present study revealed 8 A-T+ subjects. Because two of them (#37, #38) were LMCI patients and showed AD or DLB pattern in FDG-PET, they are considered SNAP and to have cognitive impairment due to non-AD disorders (17). The other six are CU subjects and had non-progressive FDG pattern, and may suggest a very early stage of various non-AD tauopathy such as PART or normal aging process (17). There is also a possibility of false-positive CSF P-tau as the value was close to the cutoff, ranging 31.8-38.2 pg/mL for 5 of the 6 subjects. Since biomarkers in non-AD tauopathies are not well understood, further investigations are needed.
Another EMCI subject (#36) presented “A-T-N+” with FTD pattern in FDG-PET, and was suspected of early stage of non-amyloid non-tau FTLD.
The present study also revealed 4 subjects presenting DLB pattern in FDG-PET, with occipital hypometabolism extending to surrounding areas (Figure 3). Three of them were amyloid positive and all were tau positive. It is not clear whether they were preclinical or prodromal stage of atypical AD, or DLB with or without amyloid deposition.
These findings suggest that a significant fraction of the subjects in this study might be related to non-AD disorders such as DLB, SNAP, PART, argyrophilic grain disease (AGD), and TDP-43 proteinopathy (like LATE) (17). Even if they are amyloid positive, there is a possibility of incidental amyloid deposition. Therefore, possibility of non-AD disorders should always be considered when clinical trials targeting preclinical or prodromal AD are designed and subjects are selected based on the biomarkers.



In conclusion, this study confirmed the known changes of PET and CSF biomarkers in preclinical and prodromal AD, and at the same time, suggested difficulties of determining the criteria and cutoff level of those biomarkers to evaluate such subjects as well as the possibility of unsolicited inclusion of early-phase non-AD disorders.


Acknowledgements: We are grateful for the materials and technical supports for the PET imaging by Fujifilm Toyama Chemical, Avid Radiopharmaceuticals/Eli Lilly Japan, and GE Healthcare. PET centers that imaged the subjects but did not belong to the clinical site that enrolled the subjects are also acknowledged, including Tohoku University Cyclotron and Radioisotope Center (CYRIC), Tsukuba Advanced Imaging Center (AIC), Tokyo Metropolitan Institute of Gerontology (TMIG), Aizawa Hospital, MI Clinic, and Kobe City Medical Center General Hospital (KCGH). We thank all the people who participated in this study in the clinical and imaging sites as well as in the Core sites.

Conflict of interest: The following conflicts of interest are disclosed by the authors. Senda reports provision of devices, cassettes, and precursors from Avid/Eli Lilly Japan and GE, funding as PI of clinical trials sponsored by Eli Lilly, Eisai, Biogen, Cerveau and Merck, as well as leadership role in the Japanese Society of Nuclear Medicine as board member, congress chair and committee chair. Ikeuchi reports grants from AMED (JP19dk0207020, JP20dk0202028, JP20dm0207073). Matsuda reports a grant from AMED (19dk0207020h0005), intramural grants from National Center of Neurology and Psychiatry, and an entrusted research fund from Nihon Medi-Physics Co. Ltd. Iwatsubo reports a grant from an anonymous Foundation. Iwata reports grants from AMED (19dk0207020h0005, 16dk0207028h0001). Ikari is a full time employee of CMIC Inc. as well as graduate student of Osaka University. Washizuka reports research funding from AMED and pharmaceutical companies including Otsuka, Eisai, Pfizer, Daiichi-Sankyo, Tsumura, Mochida, Astellas, Shionogi, Takeda, Sumitomo-Dainippon, as well as honoraria from such pharmaceutical companies. Kazunari Ishii reports honoraria from Nihon Medi-Physics. Yokota reports licensing and collaboration research with Takeda Pharmaceutical Company. Nakanishi reports research funding from Eisai and Elli Lilly Japan as well as leadership role as a director in the Japan Society for Dementia Research. Shimada reports grants from AMED (19dk0207020h0005, 20dk0207028h0005). The other authors have nothing to disclose.

Funding: This study was financially supported by grants from Japan Agency for Medical Research and Development (AMED) 19dk0207020h0005, 20dk0202028h0005 and 20dm0207073h003, as well as by an anonymous Foundation.



1. Jack CR, Jr., Bennett DA, Blennow K, et al. A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 2016;87:539-547.
2. Iwatsubo T, Iwata A, Suzuki K, et al. Japanese and North American Alzheimer’s Disease Neuroimaging Initiative studies: Harmonization for international trials. Alzheimers Dement 2018;14:1077-1087.
3. Ikari Y, Akamatsu G, Nishio T, et al. Phantom criteria for qualification of brain FDG and amyloid PET across different cameras. EJNMMI physics 2016;3:23.
4. Yamane T, Ishii K, Sakata M, et al. Inter-rater variability of visual interpretation and comparison with quantitative evaluation of (11)C-PiB PET amyloid images of the Japanese Alzheimer’s Disease Neuroimaging Initiative (J-ADNI) multicenter study. Eur J Nucl Med Mol Imaging 2017;44:850-857.
5. Thurfjell L, Lilja J, Lundqvist R, et al. Automated quantification of 18F-flutemetamol PET activity for categorizing scans as negative or positive for brain amyloid: concordance with visual image reads. J Nucl Med 2014;55:1623-1628.
6. Clark CM, Pontecorvo MJ, Beach TG, et al. Cerebral PET with florbetapir compared with neuropathology at autopsy for detection of neuritic amyloid-β plaques: a prospective cohort study. The Lancet Neurology 2012;11:669-678.
7. Yamane T, Ikari Y, Nishio T, et al. Visual-statistical interpretation of (18)F-FDG-PET images for characteristic Alzheimer patterns in a multicenter study: inter-rater concordance and relationship to automated quantitative evaluation. AJNR Am J Neuroradiol 2014;35:244-249.
8. Silverman DHS, Small GW, Chang CY, et al. Positron Emission Tomography in Evaluation of DementiaRegional Brain Metabolism and Long-term Outcome. JAMA 2001;286:2120-2127.
9. Herholz K, Salmon E, Perani D, et al. Discrimination between Alzheimer dementia and controls by automated analysis of multicenter FDG PET. NeuroImage 2002;17:302-316.
10. Herholz K, Westwood S, Haense C, Dunn G. Evaluation of a calibrated (18)F-FDG PET score as a biomarker for progression in Alzheimer disease and mild cognitive impairment. J Nucl Med 2011;52:1218-1226.
11. Southekal S, Devous MD, Sr., Kennedy I, et al. Flortaucipir F 18 Quantitation Using Parametric Estimation of Reference Signal Intensity. J Nucl Med 2018;59:944-951.
12. Devous MD, Sr., Joshi AD, Navitsky M, et al. Test-Retest Reproducibility for the Tau PET Imaging Agent Flortaucipir F 18. J Nucl Med 2018;59:937-943.
13. Reimand J, Collij L, Scheltens P, Bouwman F, Ossenkoppele R, Alzheimer’s Disease Neuroimaging I. Association of amyloid-beta CSF/PET discordance and tau load 5 years later. Neurology 2020;95:e2648-e2657.
14. Mattsson-Carlgren N, Andersson E, Janelidze S, et al. Aβ deposition is associated with increases in soluble and phosphorylated tau that precede a positive Tau PET in Alzheimer’s disease. Science advances 2020;6:eaaz2387.
15. Sanchez JS, Becker JA, Jacobs HIL, et al. The cortical origin and initial spread of medial temporal tauopathy in Alzheimer’s disease assessed with positron emission tomography. Science translational medicine 2021;13:eabc0655.
16. Coutinho AM, Busatto GF, de Gobbi Porto FH, et al. Brain PET amyloid and neurodegeneration biomarkers in the context of the 2018 NIA-AA research framework: an individual approach exploring clinical-biomarker mismatches and sociodemographic parameters. Eur J Nucl Med Mol Imaging 2020;47:2666-2680.
17. Jack CR, Jr., Knopman DS, Chételat G, et al. Suspected non-Alzheimer disease pathophysiology–concept and controversy. Nature reviews Neurology 2016;12:117-124.



S. Gauthier1, P.S. Aisen2, J. Cummings3, M.J. Detke5, F.M. Longo6, R. Raman2, M. Sabbagh4, L. Schneider7, R. Tanzi8, P. Tariot9, M. Weiner10, J. Touchon11, B.Vellas12 and the EU/US CTAD Task Force*


* EU/US/CTAD TASK FORCE: Susan Abushakra (Framingham); John Alam (Boston); Sandrine Andrieu (Toulouse); Anu Bansal (Simsbury); Monika Baudler (Basel); Joanne Bell (Wilmington); Mickaël Beraud (Zaventem); Tobias Bittner (Basel); Samantha Budd Haeberlein (Cambridge); Szofia Bullain (Basel); Marc Cantillon (Gilbert); Maria Carrillo (Chicago); Carmen Castrillo-Viguera (Cambridge); Ivan Cheung (Woodcliff Lake); Julia Coelho (San Francisco); Daniel Di Giusto (Basel); Rachelle Doody (South San Francisco); John Dwyer (Washington); Michael Egan (North Wales); Colin Ewen (Slough); Charles Fisher (San Francisco); Michael Gold (North Chicago); Harald Hampel (Woodcliff Lake) ; Ping He (Cambridge) ; Suzanne Hendrix (Salt Lake City) ; David Henley (Titusville) ; Michael Irizarry (Woodcliff Lake); Atsushi Iwata (Tokyo); Takeshi Iwatsubo (Tokyo); Michael Keeley (South San Francisco); Geoffrey Kerchner (South San Francisco); Gene Kinney (San Francisco); Hartmuth Kolb (Titusville); Marie Kosco-Vilbois (Lausanne); Lynn Kramer (Westport); Ricky Kurzman (Woodcliff Lake); Lars Lannfelt (Uppsala); John Lawson (Malvern); Jinhe Li (Gilbert); Mark Mintun (Philadelphia); Vaidrius Navikas (Valby); Gerald Novak (Titusville); Gunilla Osswald (Stockholm); Susanne Ostrowitzki (South San Francisco); Anton Porsteinsson (Rochester); Ivana Rubino (Cambridge); Stephen Salloway (Providence); Rachel Schindler (New York); Hiroshi Sekiya (Malvern); Dennis Selkoe (Boston); Eric Siemers (Zionsville); John Sims (Indianapolis); Lisa Sipe (San Marcos); Olivier Sol (Lausanne); Reisa Sperling (Boston); Andrew Stephens (Berlin); Johannes Streffer (Braine-l’Alleud); Joyce Suhy (Newark); Chad Swanson (Woodcliff Lake); Gilles Tamagnan (New Haven); Edmond Teng (South San Francisco); Martin Tolar (Framingham); Martin Traber (Basel); Andrea Vergallo (Woodcliff Lake); Christian Von Hehn (Cambridge); George Vradenburg (Washington); Judy Walker (Singapore) ; Glen Wunderlich (Ridgefield); Roy Yaari (Indianapolis); Haichen Yang (North Wales); Wagner Zago (San Francisco); Thomas Zoda (Austin)

1. McGill Center for Studies in Aging, Verdun, QC, Canada; 2. Alzheimer’s Therapeutic Research Institute (ATRI), Keck School of Medicine, University of Southern California, San Diego, CA, USA; 3. Department of Brain Health, School of Integrated Health Sciences, University of Nevada Las Vegas (UNLV), USA; 4. Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas, NV, USA; 5. Cortexyme, South San Francisco, CA, USA; 6. Stanford University School of Medicine, Stanford CA USA; 7. University of Southern California Keck School of Medicine, Los Angeles, CA USA; 8. Harvard University, Boston, MA USA; 9. Banner Alzheimer’s Institute, Phoenix AZ USA; 10. University of California, San Francisco, CA USA; 11. Montpellier University, INSERM 1061, Montpellier, France; 12. Gerontopole, INSERM U1027, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France

Corresponding Author: Serge Gauthier, McGill Center for Studies in Aging, Verdun QC, Canada,

J Prev Alz Dis 2020;3(7):152-157
Published online April 6, 2020,



While amyloid-targeting therapies continue to predominate in the Alzheimer’s disease (AD) drug development pipeline, there is increasing recognition that to effectively treat the disease it may be necessary to target other mechanisms and pathways as well. In December 2019, The EU/US CTAD Task Force discussed these alternative approaches to disease modification in AD, focusing on tau-targeting therapies, neurotrophin receptor modulation, anti-microbial strategies, and the innate immune response; as well as vascular approaches, aging, and non-pharmacological approaches such as lifestyle intervention strategies, photobiomodulation and neurostimulation. The Task Force proposed a general strategy to accelerate the development of alternative treatment approaches, which would include increased partnerships and collaborations, improved trial designs, and further exploration of combination therapy strategies.

Key words: Alzheimer’s disease, dementia, tau, tauopathy, neurotrophins, neuroinflammation, lifestyle intervention, photobiomodulation, neurostimulation, geroscience.



Following a discussion on lessons learned from clinical trials of amyloid-based therapies for Alzheimer’s disease (AD) (1), on December 4, 2019, the EU/US CTAD Task Force turned their attention to alternative approaches for disease modification. These strategies do not negate the validity of the amyloid hypothesis; indeed, recently discovered genetic evidence continues to support the centrality of amyloid in the neurodegenerative processes that lead to AD (2–4). However, genetic and other studies point to additional mechanisms and pathways both upstream and downstream of amyloidogenesis, which may provide druggable therapeutic targets with potential for disease modification.
Neuropathological and imaging studies confirm the complexity and heterogeneity of AD (5) Mixed pathologies are evident in most individuals with a clinical diagnosis of AD (6), and in early clinical studies of amyloid-targeting drugs, a significant proportion of trial participants were shown to have no detectable amyloid. Nonetheless, among putative disease-modifying AD drugs in clinical trials, 40% target amyloid either with small molecules or immunotherapies. Another 18% target tau. Other mechanisms targeted for disease modification include neuroprotection, anti-inflammatory effects, growth factor promotion, and/or metabolic effects (7). Additional trials are underway assessing non-pharmacological approaches to treat AD, including lifestyle interventions and neurostimulation.


Anti-tau therapies

The microtubule-associated protein tau (MAPT, commonly referred to as tau) is the main constituent of the neurofibrillary tangles that are one of the two primary pathological hallmarks of AD. Its normal function is to stabilize microtubules and thus regulate intracellular trafficking, but in AD and other tauopathies, the protein undergoes post-translational modifications that lead to the development of a variety of oligomeric species, tangles, and neuropil threads that may be deposited as aggregates in specific brain regions, disrupting normal cytoskeletal function and protein degradation pathways (8). In the human brain, six isoforms of tau are present, which are classified as either 3R or 4R tau based on the number of repeat domains. Approximately equal levels of 3R and 4R tau are expressed in the normal brain; however, 3R:4R tau imbalances are seen in brains of individuals with tauopathies. In AD, isoform imbalances vary across brain regions and disease progression.
Unlike levels of amyloid beta protein (Aβ), which correlate poorly with cognition, tau levels are associated with both neurodegeneration and cognitive deficits (9). Tau pathology has been shown to follow a characteristic progression pathway in the brain, starting in areas responsible for learning and memory before spreading to cortical areas involved in other cognitive functions (10).
The complex progression of tau pathological events provides multiple potential opportunities for intervention. Anti-tau drugs in development target tau expression, aggregation, degradation, protein modifications (e.g. phosphatase modifiers, kinase inhibitors), microtubule stabilization, and extracellular tau inter-neuronal spread (8). As of February 2019, clinical trials were underway for 17 tau-targeting drugs – seven small molecules and 10 biologics (7). Only one drug, LMTX (TRx0237) – a reduced form of methylene blue, and a tau protein aggregation inhibitor — is currently being tested in a Phase 3 trial in early AD at 8 – 16 mg/day doses versus placebo (NCT03446001). This trial follows two Phase 3 trials in mild and mild to moderate AD (NCT01689246, NCT01689233) and a trial in behavioral variant FTD (NCT01626378) with higher doses, which showed negative results in the primary analysis of clinical efficacy. Biogen has a Phase 2 study underway of the anti-tau agent BIIB092 (gosuranemab) in participants with MCI due to AD or mild AD (NCT03352557). Phase 2 studies in biologically defined populations are also being conducted. For example, Roche/Genentech is conducting two Phase 2 studies of the anti-tau monoclonal antibody semorinemab in participants with prodromal or probable AD confirmed by amyloid positron emission tomography (PET) or cerebrospinal fluid (CSF) testing (NCT03828747). Clinical trials of anti-tau therapeutics have been conducted in other tauopathies, although two recent Phase 2 studies of anti-tau monoclonal antibody therapies (Abbvie’s AABV-8E12 and Biogen’s gosuranemab) in participants with progressive supranuclear palsy (PSP) were recently terminated for lack of efficacy (NCT2985879 and NCT03068468, respectively). Non-clinical studies of innovative anti-tau therapies are underway, such as a study that uses engineered tau-degrading intrabodies to target intracellular tau (11).
It is also theoretically possible that early anti-amyloid intervention may attenuate or even preclude downstream effects on tau. That is, non-tau-based treatments could have implications for tau and tangles.
Several challenges face developers of tau-based therapeutics. For tau reduction approaches, it is not known how much reduction is needed, how quickly and safely it can be accomplished, when different interventions might be effective during the course of the disease, and how long drug levels must be maintained to get an effect. Tau biology is complicated with numerous fragments and post-translational modifications associated with tauopathies, yet it remains unclear which tau species are toxic. Moreover, the targets, mechanisms and cellular locations through which such tau species promote degeneration remain to be identified. These issues make the design of clinical trials especially complicated and highlight the need for better tau biomarkers. Recent progress made in the development of tau ligands for PET may improve the efficiency of clinical trials, since tau-PET enables early diagnosis and tracking of disease progression, identifying individuals at risk for faster cognitive decline, and rapidly assessing pharmacodynamic effects of treatments (12). Plasma levels of total tau (t-tau) and neurofilament light (NfL) have been developed as biomarkers of neurodegeneration (13). Still needed are biomarkers that distinguish 3R from 4R tau and that quantify the many different tau species.


Neurotrophic strategies

The neurodegeneration that occurs in AD results from a complicated molecular and biochemical signaling network, likely triggered by Aβ and eventually leading to synaptic dysfunction, loss of dendritic spines, and neurite degeneration (14). Growth factors called neurotrophins regulate neuronal survival, development, and function by binding to cell surface receptors. The signaling networks regulated by these receptors have extensive overlap with those associated with neurodegeneration and modulation of neurotrophin receptors has thus been proposed as a potential therapeutic strategy (15). The Longo lab and others have zeroed in on the p75 neurotrophin receptor (p75NTR) as a therapeutic target for AD. Their working hypothesis, supported by human genomic and proteomic data, along with animal studies is that the p75NTR modulates the complex AD degenerative signaling network and that downregulating its signaling renders oligomeric Aβ unable to promote degeneration (16, 17).
Longo and colleagues have developed small molecule ligands that bind to p75NTR, activate survival-promoting signaling, and prevent Aβ-induced neurodegeneration and synaptic impairment (18). One molecule in particular, LM11A-31, has been shown to block Aβ-induced tau phosphorylation, misfolding, oligomerization and mislocalization; reverse late-stage spine degeneration; reverse synaptic impairment; prevent microglial dysfunction; and in wildtype mice suppress age-related basal forebrain cholinergic neuron degeneration (18–20). There is evidence that dendritic spine preservation is associated with cognitive resilience (21).
A Phase 2a pilot study sponsored by PharmatrophiX Inc. and funded in part by the National institute on Aging (NIA) and the Alzheimer Drug Discovery Foundation is underway, testing oral LM11A-31 in participants with mild-to-moderate AD and amyloid positivity assessed by CSF Aβ screening (NCT03069014). With an expected completion in the third quarter of 2020, the trial will assess safety and tolerability as well as cognitive, clinical, biomarker, and imaging exploratory endpoints. LM11A-31 may be effective in other disorders such as Huntington’s disease (22), diabetes-induced macular oedema (23), and traumatic brain injury (24).


Anti-microbial and anti-inflammatory strategies

Neuropathological studies of the AD brain show not only amyloid plaques and tau-based tangles but neuroinflammation as well. Indeed, according to the innate immune hypothesis, plaques, tangles, and neuroinflammation orchestrate an innate immune response that has evolved to protect the brain against microbial infection, with Aβ itself acting as an antimicrobial peptide (AMP) in the brain (25, 26). This hypothesis suggests that subclinical microbial infections in the brain rapidly ‘seed’ Aβ to trap microbes, and that this process drives Aβ neurotoxicity and opsonization (i.e, an ‘eat me’ signal for microglia to remove axons and synapses) (25). Tangles form in response to microbe invasion to block neurotropic microbe spread. AD risk genes are implicated in the innate immune protection hypothesis, which posits that AD-associated genetic risk variants were evolutionarily conserved to keep Aβ deposition, tangle formation, and gliosis/neuroinflammation on a ‘hair trigger’ as a means of protecting a subset of the human species in the advent of a major epidemic of brain infection.
The molecular pathways involved in these processes provide multiple potential therapeutic targets, including the use of anti-viral drugs, antibiotics, blockade of toxic microbial products, and immunization for prevention of subclinical infections; secretase inhibitors and immunotherapies to prevent Aβ seeding; kinase or phosphatase inhibitors to prevent the development of pathological forms of tau, and anti-inflammatories to suppress neuroinflammation. Gut microbiota may also play a role in AD pathogenesis by disrupting neuroinflammation and metabolic homeostasis, thus representing another potential intervention target (27).
One example of a bacterial hypothesis and associated strategy is based on the discovery of the bacterium Porphyromonas gingivalis (Pg), most commonly associated with periodontitis, in the brains of AD patients. Toxic virulence factors from the bacterium, proteases called gingipains, have been identified in AD brains, and gingipain levels correlated with tau and ubiquitin pathology. Oral infection of mice with Pg resulted in brain colonization, increased Aβ1-42, and loss of hippocampal neurons, effects that were blocked by COR388, a small-molecule irreversible lysine- gingipain inhibitor. COR388 significantly lowered markers of inflammation in plasma as well as AD-associated APOE fragments in CSF in a small Phase 1b study in mild-moderate AD patients (28), and a large Phase 2/3 study is underway with an interim readout expected in Q4 2020 and topline data in Q4 2021 (NCT03823404).
A retrospective cohort study showed that Herpes simplex virus (HSV)-infected subjects had a nearly 3-fold increased risk of AD but that treatment with anti-viral drugs such as acyclovir brought risk to non-infected levels (29). There is an ongoing phase 2 trial of valacyclovir for patients with mild AD and positive titers for HSV1 and HSV2 (NCT03282916). Trials in AD using doxycycline and minocycline did not show efficacy (30).
Anti-inflammatory strategies are also being pursued. A Phase 2 study underway in participants with late mild cognitive impairment (MCI) or early AD aims to protect neurons against oxidative stress using two small molecule drugs — tauroursodexycholic acid (TUDCA) and sodium phenylbutyrate — repurposed by Amylyx Pharmaceutical as AMX0035 (NCT03533257). Yet another Phase 3 study sponsored by AZTherapies, Inc. aims to reduce neuroinflammation by converting microglia from a proinflammatory to phagocytic state to promote clearance of Aβ by using a combination of two marketed drugs, cromolyn and ibuprofen, known as ALZT-OP1 (NCT02547818) (31).


Lifestyle intervention strategies and other non-pharmacological approaches

Multiple epidemiological studies in Europe, the United States, and Canada investigating an observed decline in the prevalence of dementia in recent years have suggested that dementia may be preventable by targeting lifestyle risk factors such as diabetes, hypertension, obesity, physical inactivity, smoking, depression, low education, and social isolation (32). Clinical studies are now beginning to support this assertion. The Systolic Blood Pressure Intervention Trial –Memory and Cognition in Decreased Hypertension (SPRINT MIND) study suggested that intensive blood pressure control may reduce the risk of probable dementia and mild cognitive impairment (MCI), although the results were not statistically significant, in part because the SPRINT trial was terminated early based on the significant benefits of blood pressure control on cardiovascular outcomes. The study may have been underpowered for cognitive endpoints (33). Further study is warranted given that a 10-year study in France showed that hypertension was associated with poorer cognition in middle-aged individuals (34).
Multi-domain strategies have focused on lifestyle factors. For example, the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) trial demonstrated improved or stabilized cognitive function in participants that adhered to an intervention combining diet, physical exercise, cognitive training, and vascular risk monitoring (35). The Multidomain Alzheimer Prevention Trial (MAPT) tested an intervention combining cognitive and physical intervention along with omega-3 polyunsaturated fatty acid supplementation in frail, non-demented, community dwelling adults (36, 37). While MAPT failed to demonstrate significant slowing of cognitive decline, subgroup analyses suggested that individuals with low plasma levels of docosahexaenoic acid (DHA, an omega-3 fatty acid) have more cognitive decline, which appeared to be normalized with omega-3 supplementation(38). The benefits of omega-3 supplementation appeared to be greater in amyloid-positive individuals and in those with increased cardiovascular risk scores (39, 40). Based on the results from FINGER, MAPT, and other multidomain intervention studies, many additional studies are planned, including worldwide FINGERS studies (WW-FINGERS), a network of studies throughout the world that are adapting the multidomain strategies of the FINGER trial to different populations (41).
In addition to physical and cognitive activity, other non-pharmacological strategies are being investigated for their potential to slow cognitive decline and prevent dementia. For example, photobiomodulation (PBM) has been shown to be neuroprotective. In animal models PBM improved memory and normalized markers of AD, oxidative stress and neuroinflammation (42). A pilot study is now underway in participants with probable AD (NCT03405662).
Non-invasive neurostimulation with techniques such as repetitive transcranial magnetic stimulation (rTMS) has been proposed as a treatment for AD (43). Other technological approaches including assistive technologies, smart technologies, and telemedicine may improve the treatment and care of people with AD.



Given that aging is the major risk factor for AD, therapeutic strategies aimed at the diseases of aging (e.g., frailty) may slow cognitive decline and the development of dementia (44) Considerable research is underway to investigate the relationship between biological aging and neurodegenerative disease. These efforts have coalesced in the emerging field of geroscience (44), which explores whether the physiological hallmarks of aging such as mitochondrial dysfunction, loss of proteostasis, increased cellular senescence, and stem cell exhaustion may contribute to the development of AD pathology and neurodegeneration (45). Identification of biomarkers of aging and elucidation of how the molecular pathways of aging and AD intersect could advance the identification of novel therapeutic targets and next-generation therapies, such as the use of mesenchymal stem cells (46). The links between aging and AD are being explored as one element of the INSPIRE Research Initiative (Barreto JFA in press).


Conclusions/moving forward

While the AD drug development pipeline continues to be dominated by Aβ-targeting therapies, there is increasing recognition that addressing the complexity of AD may require multiple agents and may need to start in early disease stage before pathology becomes irreversible. A “deep biology” view, such as that proposed by advocates of p75NTR modulation, posits that key ‘hub’ targets may enable modulation of multiple mechanisms (e.g. resilience to both Aβ and tau) and that key components of pathology could be reversible (e.g. spines, synaptic function). A single treatment could thus promote synaptic function and slow progression and prevent upstream tau aggregation and oligomer formation.
Given the importance of tau in the development of AD, and reflecting the recently proposed Research Framework (47), CTAD Task Force members advocated assessment of both Aβ and tau levels in all clinical trials. The A-T+N+ AD phenotype is common and should be targeted for anti-tau trials. A suggestion was made to name this phenotype Dementia Associated and Neurofibrillary tangle Neuroimaging Abnormality (DANNA). Tau imaging may provide a biological outcome, at least in Phase 2 studies, although the Task Force recognized that amyloid and/or tau PET imaging adds substantial subject and trial burden and cost. Other suggestions that could accelerate the development of anti-tau therapies include using basket designs that include participants with other tauopathies such as frontotemporal degeneration (FTD), progressive supranuclear palsy (PSP), and corticobasal degeneration (CBD). While such trials would include participants with heterogeneous presentations, an outcome assessment such as Goal Attainment Scaling (GAS) could enable capture of clinically meaningful outcomes from diverse participants. This tool enables patients, caregivers, and clinicians, to set goals for treatment using a standardized guided interview, followed by an assessment of whether those goals have been attained (48, 49).
The Task Force suggested that combination therapy may be required to tackle such a complex disease as AD (50). They also advocated employing other innovative clinical trial methodologies to accelerate development of alternative approaches.
The Task Force proposed a general strategy to accelerate the development of alternative treatment approaches, which would include:
• Increased partnerships in the pre-competitive space with increased sharing of granular level data, shared biomarkers, statistical approaches, information on site performance
• Innovative trial design
• More collaborative approaches to recruitment and retention of participants for clinical trials with a focus on participation of representative populations.


Acknowledgements: The authors thank Lisa J. Bain for assistance in the preparation of this manuscript.

Conflicts of interest: The Task Force was partially funded by registration fees from industrial participants. These corporations placed no restrictions on this work. Dr. Gauthier is a member of scientific advisory boards for Biogen, Boehringer-Ingelheim, and TauRx; and a member of the DSMB for ADCS, ATRI, and Banner Health; Dr. Aisen reports grants from Janssen, grants from NIA, grants from FNIH, grants from Alzheimer’s Association, grants from Eisai, personal fees from Merck, personal fees from Biogen, personal fees from Roche, personal fees from Lundbeck, personal fees from Proclara, personal fees from Immunobrain Checkpoint, outside the submitted work; Dr. Cummings is a consultant for Acadia, Actinogen, AgeneBio, Alkahest, Alzheon, Annovis, Avanir, Axsome, Biogen, Cassava, Cerecin, Cerevel, Cognoptix, Cortexyme, EIP Pharma, Eisai, Foresight, Gemvax, Green Valley, Grifols, Karuna, Nutricia, Orion, Otsuka, Probiodrug, ReMYND, Resverlogix, Roche, Samumed, Samus Therapeutics, Third Rock, Signant Health, Sunovion, Suven, United Neuroscience pharmaceutical and assessment companies, and the Alzheimer Drug Discovery Foundation; and owns stock in ADAMAS, BioAsis, MedAvante, QR Pharma, and United Neuroscience. Dr. Detke reports personal fees, non-financial support and other from Cortexyme, during the conduct of the study; personal fees and other from Embera, personal fees and other from Evecxia, personal fees from NIH, outside the submitted work; Dr Kramer is an employee of Eisai Company, Ltd; Dr Longo has equity in and consults for PharmatrophiX, a company focused on the development of small molecule modulators for neurotrophin receptors. He is also a co-inventor on related patent applications. Dr. Raman reports grants from NIH, grants from Eli Lilly, grants from Eisai, outside the submitted work; Dr Sabbagh reports personal fees from Allergan, personal fees from Biogen, personnal fees from Grifols, personal fees from vTV Therapeutics, personal fees from Sanofi, personal fees from Neurotrope, personal fees from Cortexyme, other from Neurotrope, other from uMethod, other from Brain Health Inc, other from Versanum Inc, other from Optimal Cognitive Health Company, outside the submitted work; Dr. Schneider reports grants and personal fees from Eli Lilly, personal fees from Avraham, Ltd, personal fees from Boehringer Ingelheim, grants and personal fees from Merck, personal fees from Neurim, Ltd, personal fees from Neuronix, Ltd, personal fees from Cognition, personal fees from Eisai, personal fees from Takeda, personal fees from vTv, grants and personal fees from Roche/Genentech, grants from Biogen, grants from Novartis, personal fees from Abbott, grants from Biohaven, grants from Washington Univ/ NIA DIAN-TU, personal fees from Samus, outside the submitted work; Dr. Tanzi is a consultant and shareholder in AZTherapies, Amylyx, Promis, Neurogenetic Pharmaceuticals, Cerevance, and DRADS Capital; Dr. Tariot reports personal fees from Acadia , personal fees from AC Immune, personal fees from Axsome, personal fees from BioXcel, personal fees from Boehringer-Ingelheim, personal fees from Brain Test Inc., personal fees from Eisai, personal fees from eNOVA, personal fees from Gerontological Society of America, personal fees from Otuska & Astex, personal fees from Syneos, grants and personal fees from Abbvie, grants and personal fees from Avanir, grants and personal fees from Biogen, grants and personal fees from Cortexyme, grants and personal fees from Genentech, grants and personal fees from Lilly, grants and personal fees from Merck & Co, grants and personal fees from Roche, grants from Novartis, grants from Arizona Department of Health Services, grants from National Institute on Aging, other from Adamas, outside the submitted work; In addition, Dr. Tariot has a patent U.S. Patent # 11/632,747, “Biomarkers of Neurodegenerative disease.” issued; Dr. Weiner is the PI of The Alzheimer’s Disease Neuroimaging Initiative (ADNI) and the Brain Health Registry. I am a Professor at University of California San Francisco; Dr. Touchon has received personnal fees from Regenlife and is JPAD associated Editor and part of the CTAD organizing committee; Dr. Vellas reports grants from Lilly, Merck, Roche, Lundbeck, Biogen, grants from Alzheimer’s Association, European Commission, personal fees from Lilly, Merck, Roche, Biogen, outside the submitted work

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Aisen, PS, Cummings, J, Doody, R, Kramer, L, Salloway, S, et al. The Future of Anti-Amyloid Trials. J Prev Alz Dis.
2. Kunkle BW, Grenier-Boley B, Sims R, Bis JC, Damotte V, Naj AC, et al. Genetic meta-analysis of diagnosed Alzheimer’s disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing. Nat Genet. 2019;51(3):414–30.
3. Suh J, Choi SH, Romano DM, Gannon MA, Lesinski AN, Kim DY, et al. ADAM10 missense mutations potentiate β-amyloid accumulation by impairing prodomain chaperone function. Neuron. 2013 Oct 16;80(2):385–401.
4. Jansen IE, Savage JE, Watanabe K, Bryois J, Williams DM, Steinberg S, et al. Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimer’s disease risk. Nat Genet. 2019;51(3):404–13.
5. Long JM, Holtzman DM. Alzheimer Disease: An Update on Pathobiology and Treatment Strategies. Cell. 2019 Oct 3;179(2):312–39.
6. Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology. 2007 Dec 11;69(24):2197–204.
7. Cummings J, Lee G, Ritter A, Sabbagh M, Zhong K. Alzheimer’s disease drug development pipeline: 2019. Alzheimers Dement (N Y). 2019;5:272–93.
8. Congdon EE, Sigurdsson EM. Tau-targeting therapies for Alzheimer disease. Nat Rev Neurol. 2018 Jul;14(7):399–415.
9. Digma LA, Madsen JR, Reas ET, Dale AM, Brewer JB, Banks SJ, et al. Tau and atrophy: domain-specific relationships with cognition. Alzheimer’s Research & Therapy. 2019 Jul 27;11(1):65.
10. Schwarz AJ, Yu P, Miller BB, Shcherbinin S, Dickson J, Navitsky M, et al. Regional profiles of the candidate tau PET ligand 18 F-AV-1451 recapitulate key features of Braak histopathological stages. Brain. 2016 May 1;139(5):1539–50.
11. Gallardo G, Wong CH, Ricardez SM, Mann CN, Lin KH, Leyns CEG, et al. Targeting tauopathy with engineered tau-degrading intrabodies. Mol Neurodegener. 2019 22;14(1):38.
12. Hanseeuw BJ, Betensky RA, Jacobs HIL, Schultz AP, Sepulcre J, Becker JA, et al. Association of Amyloid and Tau With Cognition in Preclinical Alzheimer Disease: A Longitudinal Study. JAMA Neurol. 2019 Jun 3;
13. Zetterberg H, Burnham SC. Blood-based molecular biomarkers for Alzheimer’s disease. Molecular Brain. 2019 Mar 28;12(1):26.
14. Eisenberg D, Jucker M. The amyloid state of proteins in human diseases. Cell. 2012 Mar 16;148(6):1188–203.
15. Longo FM, Massa SM. Small-molecule modulation of neurotrophin receptors: a strategy for the treatment of neurological disease. Nat Rev Drug Discov. 2013 Jul;12(7):507–25.
16. Knowles JK, Rajadas J, Nguyen T-VV, Yang T, LeMieux MC, Vander Griend L, et al. The p75 neurotrophin receptor promotes amyloid-beta(1-42)-induced neuritic dystrophy in vitro and in vivo. J Neurosci. 2009 Aug 26;29(34):10627–37.
17. Counts SE, He B, Prout JG, Michalski B, Farotti L, Fahnestock M, et al. Cerebrospinal Fluid proNGF: A Putative Biomarker for Early Alzheimer’s Disease. Curr Alzheimer Res. 2016;13(7):800–8.
18. Yang T, Knowles JK, Lu Q, Zhang H, Arancio O, Moore LA, et al. Small molecule, non-peptide p75 ligands inhibit Abeta-induced neurodegeneration and synaptic impairment. PLoS ONE. 2008;3(11):e3604.
19. Xie Y, Meeker RB, Massa SM, Longo FM. Modulation of the p75 neurotrophin receptor suppresses age-related basal forebrain cholinergic neuron degeneration. Sci Rep. 2019 Mar 27;9(1):5273.
20. Nguyen T-VV, Shen L, Vander Griend L, Quach LN, Belichenko NP, Saw N, et al. Small molecule p75NTR ligands reduce pathological phosphorylation and misfolding of tau, inflammatory changes, cholinergic degeneration, and cognitive deficits in AβPP(L/S) transgenic mice. J Alzheimers Dis. 2014;42(2):459–83.
21. Boros BD, Greathouse KM, Gentry EG, Curtis KA, Birchall EL, Gearing M, et al. Dendritic spines provide cognitive resilience against Alzheimer’s disease. Ann Neurol. 2017 Oct;82(4):602–14.
22. Simmons DA, Belichenko NP, Ford EC, Semaan S, Monbureau M, Aiyaswamy S, et al. A small molecule p75NTR ligand normalizes signalling and reduces Huntington’s disease phenotypes in R6/2 and BACHD mice. Hum Mol Genet. 2016 15;25(22):4920–38.
23. Elshaer SL, Alwhaibi A, Mohamed R, Lemtalsi T, Coucha M, Longo FM, et al. Modulation of the p75 neurotrophin receptor using LM11A-31 prevents diabetes-induced retinal vascular permeability in mice via inhibition of inflammation and the RhoA kinase pathway. Diabetologia. 2019;62(8):1488–500.
24. Shi J, Longo FM, Massa SM. A small molecule p75(NTR) ligand protects neurogenesis after traumatic brain injury. Stem Cells. 2013 Nov;31(11):2561–74.
25. Soscia SJ, Kirby JE, Washicosky KJ, Tucker SM, Ingelsson M, Hyman B, et al. The Alzheimer’s disease-associated amyloid beta-protein is an antimicrobial peptide. PLoS ONE. 2010 Mar 3;5(3):e9505.
26. Moir RD, Lathe R, Tanzi RE. The antimicrobial protection hypothesis of Alzheimer’s disease. Alzheimers Dement. 2018;14(12):1602–14.
27. Seo D-O, Holtzman DM. Gut microbiota: from the forgotten organ to a potential key player in the pathology of Alzheimer disease. J Gerontol A Biol Sci Med Sci. 2019 Nov 18;
28. Dominy SS, Lynch C, Ermini F, Benedyk M, Marczyk A, Konradi A, et al. Porphyromonas gingivalis in Alzheimer’s disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Sci Adv. 2019;5(1):eaau3333.
29. Tzeng N-S, Chung C-H, Lin F-H, Chiang C-P, Yeh C-B, Huang S-Y, et al. Anti-herpetic Medications and Reduced Risk of Dementia in Patients with Herpes Simplex Virus Infections-a Nationwide, Population-Based Cohort Study in Taiwan. Neurotherapeutics. 2018;15(2):417–29.
30. Howard R, Zubko O, Bradley R, Harper E, Pank L, O’Brien J, et al. Minocycline at 2 Different Dosages vs Placebo for Patients With Mild Alzheimer Disease: A Randomized Clinical Trial. JAMA Neurol. 2019 Nov 18;
31. Zhang C, Griciuc A, Hudry E, Wan Y, Quinti L, Ward J, et al. Cromolyn Reduces Levels of the Alzheimer’s Disease-Associated Amyloid β-Protein by Promoting Microglial Phagocytosis. Sci Rep. 2018 18;8(1):1144.
32. Livingston G, Sommerlad A, Orgeta V, Costafreda SG, Huntley J, Ames D, et al. Dementia prevention, intervention, and care. The Lancet. 2017 Dec 16;390(10113):2673–734.
33. Williamson JD, Pajewski NM, Auchus AP, Bryan RN, Chelune G, Cheung AK, et al. Effect of Intensive vs Standard Blood Pressure Control on Probable Dementia: A Randomized Clinical Trial. JAMA. 2019 Feb 12;321(6):553–61.
34. Rouch L, Cestac P, Hanon O, Ruidavets J-B, Ehlinger V, Gentil C, et al. Blood pressure and cognitive performances in middle-aged adults: the Aging, Health and Work longitudinal study. J Hypertens. 2019 Jun;37(6):1244–53.
35. Ngandu T, Lehtisalo J, Solomon A, Levälahti E, Ahtiluoto S, Antikainen R, et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. Lancet. 2015 Jun 6;385(9984):2255–63.
36. Andrieu S, Guyonnet S, Coley N, Cantet C, Bonnefoy M, Bordes S, et al. Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): a randomised, placebo-controlled trial. Lancet Neurol. 2017 May;16(5):377–89.
37. Vellas B, Carrie I, Gillette-Guyonnet S, Touchon J, Dantoine T, Dartigues JF, et al. MAPT Study: A multidomain approach for preventing Alzheimer’s disease: Design and baseline date. J Prev Alzheimers Dis. 2014 Jun;1(1):13–22.
38. Hooper C, De Souto Barreto P, Coley N, Cantet C, Cesari M, Andrieu S, et al. Cognitive Changes with Omega-3 Polyunsaturated Fatty Acids in Non-Demented Older Adults with Low Omega-3 Index. J Nutr Health Aging. 2017;21(9):988–93.
39. Delrieu J, Payoux P, Carrié I, Cantet C, Weiner M, Vellas B, et al. Multidomain intervention and/or omega-3 in nondemented elderly subjects according to amyloid status. Alzheimers Dement. 2019 Sep 23;
40. Chhetri JK, de Souto Barreto P, Cantet C, Pothier K, Cesari M, Andrieu S, et al. Effects of a 3-Year Multi-Domain Intervention with or without Omega-3 Supplementation on Cognitive Functions in Older Subjects with Increased CAIDE Dementia Scores. J Alzheimers Dis. 2018;64(1):71–8.
41. Rosenberg A, Mangialasche F, Ngandu T, Solomon A, Kivipelto M. Multidomain interventions to prevent cognitive impairment, alzheimer’s disease, and dementia: from finger to world-wide fingers. Journal of Prevention of Alzheimer’s Disease [Internet]. 2019 Mar 1 [cited 2020 Jan 30]; Available from:
42. Blivet G, Meunier J, Roman FJ, Touchon J. Neuroprotective effect of a new photobiomodulation technique against Aβ25-35 peptide-induced toxicity in mice: Novel hypothesis for therapeutic approach of Alzheimer’s disease suggested. Alzheimers Dement (N Y). 2018;4:54–63.
43. Lin Y, Jiang W-J, Shan P-Y, Lu M, Wang T, Li R-H, et al. The role of repetitive transcranial magnetic stimulation (rTMS) in the treatment of cognitive impairment in patients with Alzheimer’s disease: A systematic review and meta-analysis. J Neurol Sci. 2019 Mar 15;398:184–91.
44. Sierra F. Geroscience and the role of aging in the etiology and management of alzheimer’s disease. Journal of Prevention of Alzheimer’s Disease [Internet]. 2019 Mar 1 [cited 2020 Jan 30]; Available from:
45. Guerville F, De Souto Barreto P, Ader I, Andrieu S, Casteilla L, Dray C, et al. Revisiting the Hallmarks of Aging to Identify Markers of Biological Age. J Prev Alzheimers Dis. 2020;7(1):56–64.
46. Oliva AA, McClain-Moss L, Pena A, Drouillard A, Hare JM. Allogeneic mesenchymal stem cell therapy: A regenerative medicine approach to geroscience. Aging Med (Milton). 2019 Sep;2(3):142–6.
47. Jack CR, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement. 2018;14(4):535–62.
48. Rockwood K, Graham JE, Fay S, ACADIE Investigators. Goal setting and attainment in Alzheimer’s disease patients treated with donepezil. J Neurol Neurosurg Psychiatry. 2002 Nov;73(5):500–7.
49. Rockwood K, Fay S, Song X, MacKnight C, Gorman M. Attainment of treatment goals by people with Alzheimer’s disease receiving galantamine: a randomized controlled trial. CMAJ. 2006 Apr 11;174(8):1099–105.
50. Gauthier S, Alam J, Fillit H, Iwatsubo T, Liu-Seifert H, Sabbagh M, et al. Combination Therapy for Alzheimer’s Disease: Perspectives of the EU/US CTAD Task Force. J Prev Alzheimers Dis. 2019;6(3):164–8.

What do you want to do ?

New mail

What do you want to do ?

New mail

What do you want to do ?

New mail



H.-Y.Wang1,2, Z. Pei1, K.-C. Lee1, E. Lopez-Brignoni3, B. Nikolov3, C.A. Crowley4, M.R. Marsman4, R. Barbier4, N. Friedmann4, L.H. Burns4

1. Department of Molecular, Cellular and Biomedical Sciences, City University of New York School of Medicine, New York, NY, USA; 2. Department of Biology and Neuroscience, Graduate School of the City University of New York, New York, NY, USA; 3. IMIC, Inc., Palmetto Bay, FL, USA; 4. Cassava Sciences, Inc., Austin, TX, USA

Corresponding Author: Lindsay H. Burns, PhD, Cassava Sciences, Inc., 7801 N. Capital of Texas Hwy, Ste. 260, Austin, TX 78746, Phone: 512-501-2484, Email:    

J Prev Alz Dis 2020;4(7):256-264
Published online February 7, 2020,


BACKGROUND: The most common dementia worldwide, Alzheimer’s disease is often diagnosed via biomarkers in cerebrospinal fluid, including reduced levels of Aβ1–42, and increases in total tau and phosphorylated tau-181.  Here we describe results of a Phase 2a study of a promising new drug candidate that significantly reversed all measured biomarkers of Alzheimer’s disease, neurodegeneration and neuroinflammation. PTI-125 is an oral small molecule drug candidate that binds and reverses an altered conformation of the scaffolding protein filamin A found in Alzheimer’s disease brain. Altered filamin A links to the α7-nicotinic acetylcholine receptor to allow Aβ42’s toxic signaling through this receptor to hyperphosphorylate tau. Altered filamin A also links to toll-like receptor 4 to enable Aβ-induced persistent activation of this receptor and inflammatory cytokine release. Restoring the native shape of filamin A prevents or reverses filamin A’s linkages to the α7-nicotinic acetylcholine receptor and toll-like receptor 4, thereby blocking Aβ42’s activation of these receptors. The result is reduced tau hyperphosphorylation and neuroinflammation, with multiple functional improvements demonstrated in transgenic mice and postmortem Alzheimer’s disease brain.
OBJECTIVES: Safety, pharmacokinetics, and cerebrospinal fluid and plasma biomarkers were assessed following treatment with PTI-125 for 28 days. Target engagement and mechanism of action were assessed in patient lymphocytes by measuring 1) the reversal of filamin A’s altered conformation, 2) linkages of filamin A with α7-nicotinic acetylcholine receptor or toll-like receptor 4, and 3) levels of Aβ42 bound to α7-nicotinic acetylcholine receptor or CD14, the co-receptor for toll-like receptor 4.
DESIGN: This was a first-in-patient, open-label Phase 2a safety, pharmacokinetics and biomarker study.
SETTING: Five clinical trial sites in the U.S. under an Investigational New Drug application.
PARTICIPANTS: This study included 13 mild-to-moderate Alzheimer’s disease patients, age 50-85, Mini Mental State Exam ≥16 and ≤24 with a cerebrospinal fluid total tau/Aβ42 ratio ≥0.30.
INTERVENTION: PTI-125 oral tablets (100 mg) were administered twice daily for 28 consecutive days.
MEASUREMENTS: Safety was assessed by electrocardiograms, clinical laboratory analyses and adverse event monitoring. Plasma levels of PTI-125 were measured in blood samples taken over 12 h after the first and last doses; cerebrospinal fluid levels were measured after the last dose. Commercial enzyme linked immunosorbent assays assessed levels of biomarkers of Alzheimer’s disease in cerebrospinal fluid and plasma before and after treatment with PTI-125. The study measured biomarkers of pathology (pT181 tau, total tau and Aβ42), neurodegeneration (neurofilament light chain and neurogranin) and neuroinflammation (YKL-40, interleukin-6, interleukin-1β and tumor necrosis factor α). Plasma levels of phosphorylated and nitrated tau were assessed by immunoprecipitation of tau followed by immunoblotting of three different phospho-epitopes elevated in AD (pT181-tau, pS202-tau and pT231-tau) and nY29-tau. Changes in conformation of filamin A in lymphocytes were measured by isoelectric focusing point. Filamin A linkages to α7-nicotinic acetylcholine receptor and toll-like receptor 4 were assessed by immunoblot detection of α7-nicotinic acetylcholine receptor and toll-like receptor 4 in anti-filamin A immunoprecipitates from lymphocytes. Aβ42 complexed with α7-nicotinic acetylcholine receptor or CD14 in lymphocytes was also measured by co-immunoprecipitation. The trial did not measure cognition.
RESULTS: Consistent with the drug’s mechanism of action and preclinical data, PTI-125 reduced cerebrospinal fluid biomarkers of Alzheimer’s disease pathology, neurodegeneration and neuroinflammation from baseline to Day 28. All patients showed a biomarker response to PTI-125.  Total tau, neurogranin, and neurofilament light chain decreased by 20%, 32% and 22%, respectively. Phospho-tau (pT181) decreased 34%, evidence that PTI-125 suppresses tau hyperphosphorylation induced by Aβ42’s signaling through α7-nicotinic acetylcholine receptor. Cerebrospinal fluid biomarkers of neuroinflammation (YKL-40 and inflammatory cytokines) decreased by 5-14%. Biomarker effects were similar in plasma. Aβ42 increased slightly – a desirable result because low Aβ42 indicates Alzheimer’s disease. This increase is consistent with PTI-125’s 1,000-fold reduction of Aβ42’s femtomolar binding affinity to α7-nicotinic acetylcholine receptor. Biomarker reductions were at least p ≤ 0.001 by paired t test. Target engagement was shown in lymphocytes by a shift in filamin A’s conformation from aberrant to native: 93% was aberrant on Day 1 vs. 40% on Day 28. As a result, filamin A linkages with α7-nicotinic acetylcholine receptor and toll-like receptor 4, and Aβ42 complexes with α7-nicotinic acetylcholine receptor and CD14, were all significantly reduced by PTI-125. PTI-125 was safe and well-tolerated in all patients. Plasma half-life was 4.5 h and approximately 30% drug accumulation was observed on Day 28 vs. Day 1.
CONCLUSIONS:  PTI-125 significantly reduced biomarkers of Alzheimer’s disease pathology, neurodegeneration, and neuroinflammation in both cerebrospinal fluid and plasma. All patients responded to treatment. The magnitude and consistency of reductions in established, objective biomarkers imply that PTI-125 treatment counteracted disease processes and reduced the rate of neurodegeneration. Based on encouraging biomarker data and safety profile, approximately 60 patients with mild-to-moderate AD are currently being enrolled in a Phase 2b randomized, placebo-controlled confirmatory study to assess the safety, tolerability and efficacy of PTI-125.

Key words: tau, neurofilament light chain, neurogranin, YKL40, neuroinflammation.



Alzheimer’s disease (AD) is the most common dementia, afflicting fifty million people worldwide (1). PTI-125 is a new small molecule drug candidate for AD with a novel mechanism of action: it binds and restores to normal an altered conformation of the scaffolding protein filamin A (FLNA) (2-4). PTI-125’s target, an altered conformation of FLNA, is a known proteopathy in AD brain (2, 3). FLNA is a large, 280-KDa intracellular scaffolding protein best known for cross-linking actin to regulate cell structure and motility and is highly expressed in brain. FLNA dimerizes by a domain in its membrane-bound C terminal, protruding into the cytoplasm as an inverse V shape to interact with at least 90 different proteins (5), indicating potential involvement in numerous intracellular processes. In the cell membrane, FLNA constitutively associates with certain receptors such as the insulin receptor and the mu opioid receptor, and FLNA regulates insulin receptor signaling and mu opioid receptor – G-protein coupling (6). FLNA does not normally link to α7-nicotinic acetylcholine receptor (α7nAChR) or toll-like receptor 4 (TLR4); however, in AD brain, the altered conformation of FLNA enables FLNA’s association with both (3). Importantly, the aberrant FLNA linkages of altered FLNA to α7nAChR and TLR4 promote toxic signaling of soluble Aβ42 via each receptor, contributing to AD pathology (3, 4).
The extremely tight binding (high femtomolar affinity) of Aβ42 to α7nAChR, first demonstrated two decades ago (7), is reinforced by altered FLNA’s linkage to this receptor (3). Aβ42 signals via α7nAChR to hyperphosphorylate tau protein by activating kinases ERK and JNK1 (8). Tau hyperphosphorylation is a hallmark of AD pathology and disrupts tau’s function of stabilizing microtubules, promoting degeneration. Hyperphosphorylated tau also initiates neurofibrillary lesions and tau protofibril formation, leading to eventual fibrillar tau-rich tangles. This toxic signaling pathway of Aβ42 via α7nAChR has been confirmed by multiple laboratories under conditions that maintain Aβ42 as soluble monomers or small oligomers (9-11).
Aβ42 activates TLR4 by binding the TLR4 co-receptor CD14 (12). Aβ42 binding to CD14 promotes a sustained activation of TLR4 and persistent release of inflammatory cytokines such as IL-6, IL-1β and TNFα (13). Like Aβ42 signaling through α7nAChR, this Aβ42-induced TLR4 activation requires the linkage of altered FLNA to TLR4 (4). The chronic activation of TLR4 by amyloid in AD leads to neuroinflammation and exacerbates neurodegeneration.
By preferentially binding the altered conformation of FLNA and restoring its native shape, PTI-125 releases FLNA from α7nAChR and TLR4, reducing Aβ42-driven tau hyperphosphorylation and neuroinflammation, thereby attenuating neurodegeneration (2). In triple transgenic AD mice, PTI-125 restored α7nAChR, NMDAR and insulin receptor signaling, improved synaptic plasticity, reduced amyloid deposits and neurofibrillary lesions, robustly attenuated inflammatory cytokine levels, and improved cognition (3). In vitro PTI-125 incubation of postmortem human AD brain (or age-matched control brain treated with exogenous Aβ42) also reduced FLNA linked to α7nAChR or TLR4, decreased Aβ42 – α7nAChR complex levels, decreased Aβ42-induced tau hyperphosphorylation, and again improved synaptic plasticity and receptor function. IC50s for these effects were nanomolar, and significant effects were seen at concentrations as low as 1 picomolar (3, 4).
This Phase 2a clinical trial follows favorable safety and tolerability data from a Phase 1 study of 50, 100, or 200 mg of PTI-125 in healthy volunteers. In this first-in-patient, Phase 2a clinical trial, thirteen mild-to-moderate AD patients received 100 mg oral PTI-125 b.i.d. for 28 days. Patients were age 50-85, MMSE ≥ 16 and ≤ 24, with a CSF T-tau/Aβ42 ratio ≥ 0.30. The 100 mg dose was selected because it is equivalent by body surface area conversion (the accepted method of determining equivalent doses between species) to effective daily doses of PTI-125 in AD mouse models (3, 4). The T-tau/Aβ42 ratio was selected as a low cutoff to confirm AD diagnosis based on biomarker determinations from samples obtained from the Swedish bioFINDER and ADNI (Alzheimer’s Disease Neuroimaging Initiative) cohorts (14), recognizing that we used commercial enzyme-linked immunosorbent assays (ELISAs) and that study used ElectroChemiLuminescence Immunoassays (ECLIAs).  Although these assays are not directly comparable, values from each are reported in pg/mL. Safety was monitored by electrocardiograms, clinical labs, adverse event monitoring and physical examinations. Change from baseline was measured for CSF and plasma biomarkers of AD pathology (T-tau, P-tau and Aβ42), neurodegeneration (neurofilament light (NfL) chain and neurogranin) and neuroinflammation (YKL-40 and inflammatory cytokines IL-6, IL-1β and TNFα). Pre-dose and Day 28 samples were tested in triplicate in a single ELISA plate according to manufacturers’ instructions.
A recent meta-analysis showed that in CSF of AD patients vs. age-matched controls, T-tau and P-tau are increased respectively by 2.5- and 1.9-fold and Aβ42 is reduced by half (15). NfL, expressed predominantly in large caliber axons and indicating axonal damage, is increased by 2.3-fold in CSF of AD patients relative to controls (15). NfL may track disease progression and is emerging as a plasma AD biomarker (16). Neurogranin, a post-synaptic protein in dendritic spines, is elevated in AD and represents synaptic/dendritic destruction (17-19). The secreted glycoprotein YKL-40, an inflammation mediator, is thought to accompany microglial activation and extracellular tissue remodeling and is also elevated in AD (18, 20, 21).




Of 19 patients screened, 13 enrolled. All 13 were mild-to-moderate AD patients, age 50-85, MMSE ≥ 16 and ≤ 24, with a CSF T-tau/Aβ42 ratio ≥ 0.30. Although the protocol stated 12 patients would be enrolled, 13 were enrolled due to multiple patients in screening near the end of the study.

Lymphocyte and plasma preparation

To prepare lymphocytes, 8 ml venous blood collected in EDTA-containing tubes was layered onto 8 ml Histopaque-1077 at 25°C and centrifuged (400 g, 30 min, 25°C) to yield plasma (top layer) and lymphocytes (opaque interface). Plasma fractions were aliquoted into Eppendorf centrifuge tubes and stored at -80°C until assay. The obtained lymphocytes were washed twice by mixing with 10 ml phosphate-buffered saline (PBS) followed by centrifugation at 250 g for 10 min. The final lymphocyte pellet was resuspended in 600 µl cell freezing medium (DMEM, 5% DMSO, 10% fetal bovine serum), aliquoted and held at -80°C until assay.

Treatment of CSF and plasma for biomarkers

CSF and plasma were thawed on ice and immediately treated with 20X protease inhibitor cocktail (Complete mini EDTA-free protease inhibitors, Roche, 04693159001) and protein phosphatase inhibitor cocktail (Phosphostop phosphatase inhibitors, Roche, 04906837001).

Determination of CSF and plasma biomarkers

Levels of biomarkers in protease and protein phosphatase inhibitor treated CSF and plasma were measured by enzyme-linked immunosorbent assays (ELISA) in triplicate against standards, according to manufacturer’s instruction.  The ELISA kits (Table 1) from Invitrogen or Lifespan were solid phase sandwich ELISAs (except neurogranin, a solid phase indirect ELISA) that detect endogenous levels of biomarkers with a specific detecting antibody followed by an anti-species IgG, horseradish peroxidase (HRP)-linked antibody to recognize the bound detection antibody. HRP substrate tetramethylbenzidine was added to develop color. Absorbance for the developed color is proportional to the quantity of protein. Absorbances were analyzed against standards by linear regression using GraphPad Prism.

Table 1. ELISA kits used to measure CSF and plasma biomarkers

Table 1. ELISA kits used to measure CSF and plasma biomarkers


Measurement of FLNA – α7nAChR/TLR4 linkages and Aβ42 –α7nAChR/CD14 complexes in lymphocytes and ex vivo Aβ42 treatment

Levels of FLNA – α7nAChR/TLR4 and Aβ42 – α7nAChR/CD14 interactions were assessed in patient lymphocytes.  Lymphocytes (200 μg) from patients at the indicated dosing days were incubated at 37°C with oxygenated protease inhibitors containing Kreb’s Ringer (K-R) or 0.1 μM Aβ42 for 30 min (250 μl total incubation volume). The assay mixtures were aerated with 95%O2/5%CO2 for 1 min every 10 min. Reactions were terminated by adding ice-cold Ca2+-free K-R containing protease and protein phosphatase inhibitors (Roche) and centrifuging. The obtained lymphocytes were homogenized in 250 μl ice-cold immunoprecipitation buffer (25 mM HEPES, pH 7.5, 200 mM NaCl, 1 mM EDTA, 0.2% 2-mercaptoethanol with protease and protein phosphatase inhibitors) by sonication for 10 s on ice, and solubilized by nonionic detergents (0.5% NP-40/0.2% Na cholate/0.5% digitonin) for 60 min (4°C) with end-to-end rotation. The obtained lysate was centrifuged at 20,000 g for 30 min (4°C) and the resultant supernatant (0.25 ml) was diluted 4x with 0.75 ml immunoprecipitation buffer. Aβ42 – α7nAChR complexes were immunoprecipitated with immobilized anti-FLNA (SC-58764 + SC-17749, Santa Cruz) or anti-Aβ42 (AB5078P, Millipore Sigma) + anti-actin (SC-8432+ SC-376421) antibodies onto protein A/G-conjugated agarose beads (#20421, Thermo). Resultant immunocomplexes were pelleted by centrifugation (4°C), washed 3x with ice-cold PBS, pH 7.2, containing 0.1% NP-40, and centrifuged. Immunocomplexes were solubilized by boiling 5 min in 100 μl SDS-PAGE sample preparation buffer (62.5 mM Tris-HCl, pH 6.8; 10% glycerol, 2% SDS; 5% 2mercaptoethanol, 0.1% bromophenol blue) and centrifuged to remove antibody-protein A/G agarose beads. Levels of α7nAChRs (SC-58607) and β-actin (SC-47778) were determined by immunoblotting, with FLNA and β-actin levels serving as the indicators of immunoprecipitation efficiency and gel loading (3, 4, 22). Blots were then stripped and re-probed with specific antibodies against TLR4 (SC-302972) and CD14 (SC-1182) to assess levels of FLNA – TLR4 and Aβ42 – CD14 associations, respectively. Blots of anti-Aβ42 immunoprecipitates size-fractionated on 20% SDS-PAGE were used to survey the molecular mass of Aβ42 using anti-Aβ42 (AB5078P, Millipore Sigma).

Measurement of phosphorylated and nitrated tau in plasma

Phosphorylated and nitrated tau in plasma were assessed using an established method (3, 8, 23). Total tau proteins were immunoprecipitated by 1 µg immobilized anti-tau (SC-65865 and SC-166060), which does not discriminate between phosphorylation states. Levels of phosphorylated tau (pS202tau, pT231tau and pT181tau), nitrated tau (nY29tau) and total tau precipitated (loading control) were assessed by immunoblotting with antibodies specific to each phosphoepitope (pS202tau: AT8 [MN1020], pT231tau: AT180 [MN1040] and pT181tau: AT270 [MN1050], all from Thermo Invitrogen), anti-nY29tau (SC-66177), and anti-tau (SC-32274).

Isoelectric point assessment

Lymphocyte FLNA was isolated using a procedure established for brain tissues (3) with slight modification. Lymphocytes (200 µg) were sonicated for 10 s on ice in 200 µl modified hypotonic solution (50 mM Tris HCl, pH 8.0, 11.8 mM NaCl, 0.48 mM KCl, 0.13 mM CaCl2, 0.13 mM Mg2SO4, 2.5 mM NaHCO3) with a cocktail of protease and protein phosphatase inhibitors. The treated lymphocytes were solubilized using 0.5% digitonin/0.2% sodium cholate/0.5% NP-40 (4°C) with end-over-end rotation for 1 h. Following centrifugation to remove insoluble debris, the lysate was treated with 1% sodium dodecyl sulfate (SDS) for 1 min to dissociate FLNA-associated proteins, diluted 10x with immunoprecipitation buffer, and immunopurified with immobilized anti-FLNA (SC-58764 + SC-17749). Resultant FLNA was eluted using 200 µl antigen-elution buffer (#21004, Thermo), neutralized immediately with 100 mM Tris HCl (pH 9.0), diluted to 500 µl with 50 mM Tris HCl, pH 7.5, and passed through a 100-kD cut-off filter to remove low-molecular weight FLNA fragments. Once purified, FLNA was suspended in 100 µl isoelectric focusing sample buffer. Samples (50 µl) were loaded onto pH 3-10 isoelectric focusing gels and proteins fractionated (100 V for 1 h, 200 V for 1 h, and 500 V for 30 min). Separated proteins were then electrophoretically transferred to nitrocellulose membranes. FLNA was identified by immunoblotting with anti-FLNA (SC-57864) antibodies.

Pharmacokinetic methods

Blood samples (4 mL) were drawn into a Vacutainer® tube containing K2EDTA, placed on ice, and centrifuged 1000 xg for 15 min. Plasma was split into two aliquots, transferred to polypropylene tubes and stored at -20°C or below until analysis.
Plasma PK parameters for PTI-125 were calculated using non-compartmental methods in WinNonlin. The peak drug concentration (Cmax), the time to peak drug concentration (Tmax), Tlast and Clast, the time to the last quantifiable drug concentration, were obtained directly from the data without interpolation. The following parameters were calculated: the elimination rate constant (λz), the terminal elimination half-life (T1/2), the AUC from time zero to the time of the last quantifiable concentration (AUClast), the AUC from time zero extrapolated to infinity (AUCinf), and the percentage of AUCinf based on extrapolation (AUCextrap(%)), Cl/F, the apparent oral clearance, and Vz/F, apparent volume of distribution. Accumulation was estimated by the ratios of the AUC on the last day of dosing to the corresponding AUC the first day of dosing.
Below limit of quantitation (BLQ) concentrations were treated as zero from time-zero up to the time at which the first quantifiable concentration was observed; embedded and/or terminal BLQ concentrations were treated as “missing.” Full precision concentration data and actual sample times were used for all pharmacokinetic and statistical analyses.

Statistics and blinding

All CSF and plasma ELISA biomarker data were analyzed by two-sided paired t test by an independent statistician. Plasma tau phosphorylation and lymphocyte data including FLNA conformation were analyzed by one-way ANOVA with post-hoc two-sided t test (unpaired) on all 13 patients, with one missing (baseline) value. All biomarker assessments were performed blind to treatment day; samples were coded prior to testing.




This first-in-patient Phase 2a trial enrolled 13 patients: 9 females, 4 males; 3 black, 10 white; 6 Hispanic, 7 non-Hispanic. Both CSF and plasma/lymphocyte data were n=12 (or n=13 with one missing value) because one patient declined the second CSF draw, and the baseline plasma/lymphocyte sample was missing for another patient. Additionally, one patient stopped dosing on Day 21, was tested positive for cocaine on that day, did not resume dosing, but returned for the CSF draw and whole blood sample on Day 28. This patient’s CSF and plasma biomarker data are included.

Safety and pharmacokinetics

In this Phase 2a trial, PTI-125 was safe and well tolerated with no drug-related adverse events. Half-life was 4.5 h, and approximately 30% accumulation was observed on Day 28 compared to Day 1 exposure. PK parameters are shown in Table 2.  The CSF to plasma ratio of the PTI-125 analyte used plasma levels from the nearest PK time point to the time of CSF draw on Day 28, or the average of two if in the middle. Ratios ranged from 0.09 to 1.2, with CSF draw times ranging from 1.15 to 7.75 h post-dose, with higher ratios trending to later time points.

Table 2. Mean PK parameters of PTI-125 100 mg b.i.d. in AD patients (± SD)

Table 2. Mean PK parameters of PTI-125 100 mg b.i.d. in AD patients (± SD)


CSF and plasma biomarkers

Consistent with PTI-125’s mechanism of action and preclinical data, PTI-125 treatment reduced CSF biomarkers of neurodegeneration and AD pathology from baseline to Day 28 (Fig. 1). T-tau decreased 20%, neurogranin decreased 32%, and NfL decreased 22%. P-tau (pT181) decreased 34%, evidence that PTI-125 suppresses tau hyperphosphorylation induced by Aβ42’s signaling through α7nAChR. CSF biomarkers of neuroinflammation were also reduced: YKL-40, IL-6, IL-1β and TNFα decreased by 9%, 14%, 11% and 5%, respectively. Plasma NfL, neurogranin, T-tau and YKL-40 levels were similarly reduced (Fig. 1). All reductions were of slightly lower magnitude in plasma except for neurogranin, which was reduced 40.7% in plasma. In contrast to the consistent and highly significant reductions of all other biomarkers, Aβ42 increased slightly in both CSF and plasma – a desirable result because low Aβ42 in CSF and plasma indicates AD. This increase, significant only in plasma due to variability in CSF, is consistent with PTI-125’s mechanism: Aβ42 bound to α7nAChR is released due to a profound reduction in Aβ42’s affinity for α7nAChR when PTI-125 binds altered FLNA and restores its native shape (2-4).

Figure 1. PTI-125 treatment reduces CSF and plasma biomarkers

Figure 1. PTI-125 treatment reduces CSF and plasma biomarkers

All CSF biomarkers elevated in AD were significantly reduced in CSF after PTI-125 treatment. T-tau, NfL, neurogranin (Nrgrn) and YKL-40 were also significantly reduced in plasma. The slight increase in Aβ42 was significant in plasma but not in CSF due to variability. Inflammatory cytokines and P-tau in plasma were not assessed. *p < 0.0001, +p < 0.001, # p < 0.01 in paired t test comparing Day 28 to pre-dose baseline. N=12. Error bars are SEM.


Spaghetti plots of individual CSF biomarker values (pg/mL) show that each patient responded to PTI-125 treatment on virtually all biomarkers (Fig. 2). Interestingly, the two patients with high Aβ42 levels showed a decrease in this biomarker post-treatment. Familial AD mutations were not assessed but may have contributed. Although cytokines can be difficult to measure, the lower limits of quantitation (2x background) were 3.9 pg/mL for IL-6 (R2 value 0.8864) and 7.8 pg/mL for both IL-1β and TNFα (R2 values 0.9374 and 0.8767, respectively).

Figure 2. Individual patient changes in CSF biomarkers

Figure 2. Individual patient changes in CSF biomarkers

Each spaghetti plot shows reductions from baseline for each patient in one of nine biomarkers of Alzheimer’s disease, neurodegeneration or neuroinflammation. Levels of Aβ42, usually low in AD patients, increased after treatment with PTI-125, except for two patients with high baseline Aβ42.  Data are plotted as pg/mL.

Because levels of phosphorylated tau defined by individual phospho-epitopes are low in CSF and even lower in plasma (as reflected in total tau levels), plasma levels of phosphorylated tau were assessed by immunoprecipitation of tau with anti-tau followed by immunoblotting of three different phospho-epitopes known to be elevated in AD. Tau phosphorylation at these sites (pT181-tau, pS202-tau and pT231-tau) was significantly reduced in plasma by 12.5%, 14.0% and 16.3%, respectively, following PTI-125 treatment (Fig. 3), corroborating pT181-tau results in CSF. Because tau is immunoprecipitated with a tau-specific antibody, tau levels serve as the control for phospho-tau and nitrated tau levels. The ratios for each P-tau epitope and nY to total tau were adjusted by the average reduction in total tau reduction (0.955 for Day 14 and 0.887 for Day 28; loading could not be adjusted in the experiment due to blinding). The higher than expected molecular weight for tau may be due to additional phosphorylation in blood; the anti-tau antibody used to detect tau is commonly used. The reduction in phosphorylated tau, together with 20.4% lower nitrated tau (nY29-tau; Fig. 3), suggests that PTI-125 can reduce tau hyperphosphorylation and oxidative stress to stabilize mitochondria and attenuate neurofibrillary lesions and neurodegeneration.

Figure 3. PTI-125 treatment reduces phosphorylated and nitrated tau in plasma

Figure 3. PTI-125 treatment reduces phosphorylated and nitrated tau in plasma

a, b, Reductions in tau phosphorylation and tau nitration found in plasma following PTI-125 treatment.  Reductions were evident by 14 days and stronger by 28 days of dosing, demonstrated by anti-tau antibody immunoprecipitation and immunoblotting with antibodies specific for each phosphorylation or nitration site. Blots (a) were evaluated by densitometric quantitation (b). *p < 0.001 vs. dosing day 0, +p < 0.01 vs. dosing day 14. N=13 with one missing value. Error bars are SEM


Mechanism and target engagement in patient lymphocytes

Patient lymphocytes, which express FLNA (24), α7nAChR and TLR4, were used to demonstrate target engagement and mechanism of action of PTI-125. Confirming the altered and acidic FLNA in AD brain (3), FLNA in patients’ lymphocytes had an isoelectric focusing point (pI) of 5.3 prior to treatment. PTI-125 treatment reverted FLNA’s pI from almost exclusively 5.3 on Day 0 to mostly 5.9 on Day 28 (Fig. 4a,b), indicating the reversal of pathological to physiological, native conformation (3).
The benefit of this shift in FLNA conformation is shown by reduced FLNA linkages to α7nAChR and TLR4 in patient lymphocytes (by 45.4% for each, Fig. 4c,d) assessed by immunoblot detection of α7nAChR and TLR4 in anti-FLNA immunoprecipitates, as previously described for assessments in AD mouse models and AD postmortem human brain (3, 4).  Finally, PTI-125 treatment benefit is also corroborated by reduced binding of Aβ42 to both α7nAChR and CD14, the co-receptor for TLR4, by 54.6% and 40.1%, respectively, demonstrated by immunoprecipitation with a specific anti-Aβ42 antibody and subsequent immunoblot detection of α7nAChR or CD14 in the immunoprecipitate  (Fig. 5a, b). Immunoblot detection of Aβ42 itself in the immunoprecipitate showed a <10 KDa species that increases following treatment, indicating small oligomers or monomers, as well as a band of 57 KDa, representing Aβ42 monomers tightly bound to α7nAChR or CD14 (7, 25), which decreased with treatment (Fig. 5c, d). The reduction in Aβ42 bound to α7nAChR is consistent with PTI-125’s 1000-10,000-fold reduction in binding affinity of Aβ42 to α7nAChR (4).

Figure 4. PTI-125 restores FLNA’s native shape and reduces FLNA linkages to α7nAChR and TLR4

Figure 4. PTI-125 restores FLNA’s native shape and reduces FLNA linkages to α7nAChR and TLR4

a, b Restoration of FLNA’s native shape. Isoelectric focusing gel (a) and its quantitation (b) show that 93% of FLNA isolated from lymphocytes prior to treatment is in the altered conformation (pI 5.3), with just 7% in the native shape (pI 5.9).  PTI-125 treatment for 28 days shifts this distribution to 40% in the altered and 60% in the native conformation. *p < 0.0001 comparing 5.9 to 5.3, #p < 0.0001 vs. dosing day 0, +p < 0.0001 vs. dosing day 14. c,d, Reductions in FLNA linkages to TLR4 and α7nAChR found in lymphocytes. Reductions are illustrated by TLR4 or α7nAChR levels detected using immunoblotting with specific antibodies in solubilized anti-FLNA antibody immunoprecipitates of lymphocytes. Additionally, exogenous Aβ42 added in vitro to lymphocytes reversed these reductions in FLNA associations, returning levels to pre-treatment baseline. Blots (c) were assessed by densitometric quantitation (d). *p < 0.001 vs. dosing day 0, +p < 0.01 and ++p < 0.05 vs. dosing day 14. N=13 with one missing value. Error bars are SEM.

The fact that reduced FLNA linkages and Aβ42 binding to α7nAChR/CD14 can be reversed by adding exogenous Aβ42 (0.1 µM) illustrates the dynamic nature of this Aβ42-mediated pathogenesis in AD. Because CSF biomarkers were notably reduced, these findings in lymphocytes of PTI-125-treated patients can be inferred also to occur in brain. In support, reductions in FLNA linkages to α7nAChR and TLR4 in lymphocytes (unpublished data) of transgenic AD mice treated with PTI-125 tracked similar reductions we reported in brains of these mice (3).

Figure 5. PTI-125 treatment reduces levels of Aβ42 bound to α7nAChR and CD14

Figure 5. PTI-125 treatment reduces levels of Aβ42 bound to α7nAChR and CD14

a, b, Reductions in levels of complexes of Aβ42 with α7nAChR and CD14 found in lymphocytes. Reductions are illustrated by α7nAChR and CD14 levels detected using immunoblotting with specific antibodies in anti-Aβ42 antibody immunoprecipitates of lymphocytes. c, d, The 57-KDa band representing Aβ42 tightly bound to CD14 or α7nAChR is progressively reduced over dosing days. Aβ42 in lymphocytes (endogenous or exogenously added) is predominantly monomeric, illustrated by size < 10 KDa (c). As with FLNA linkages, exogenous Aβ42 added in vitro to lymphocytes reversed these reductions in Aβ42 – α7nAChR/CD14 complexes, returning levels to pre-treatment baseline. Actin simultaneously immunoprecipitated by anti-actin antibodies was used as a loading control for both blots. Blots (a, c) were assessed by densitometric quantitation (b, d). *p < 0.001 vs. dosing day 0, +p < 0.01 and ++p < 0.05 vs. dosing day 14. N=13 with one missing value. Error bars are SEM


Discussion/Future development plans

In a first-in-patient clinical trial of PTI-125, CSF and plasma biomarkers of AD pathology, neurodegeneration and neuroinflammation were markedly improved following 28-day oral treatment with PTI-125. All patients showed a biomarker response to PTI-125.  The drug was well tolerated, with no observable drug-related adverse events. PTI-125’s effects in patients are consistent with its mechanism of action and published preclinical data. Target engagement and mechanism of action of PTI-125 were demonstrated in patient lymphocytes by reduced associations of FLNA with α7nAChR and TLR4, reduced binding of Aβ42 to α7nAChR or CD14 and a shift back to FLNA’s native shape, visible by isoelectric focusing point. The magnitude and consistency of reductions in several established, objective biomarkers following treatment with PTI-125 at 100 mg twice daily for 28 days imply a slower rate of neurodegeneration or a suppression of disease processes in AD.
Cognition and function were not assessed in this small study. However, elevated CSF biomarkers of P-tau and total tau/Aβ42 ratio have previously been correlated with worse performance on a wide range of memory and sustained attention assessments (26) and define the disease state if not also progression. We therefore hypothesize that decreasing these markers will favorably impact patient cognition and function or their rates of decline. Additionally, other research has shown that 11-13% decreases in CSF neurogranin and P-tau and a slower increase in CSF NfL compared to placebo over 18 months is associated with a slower rate of cognitive decline in prodromal AD patients (27).
Based on these encouraging safety and biomarker data, patients are currently being enrolled in a Phase 2b study to assess the safety, tolerability and effects of PTI-125 in patients with mild-to-moderate AD. This blinded, randomized, placebo-controlled, clinical trial will enroll approximately 60 patients with mild-to-moderate Alzheimer’s disease.  In the Phase 2b study, patients are administered PTI-125 100 mg, 50 mg or matching placebo, twice daily, for 28 continuous days. The primary endpoint is improvement in biomarkers of Alzheimer’s disease from baseline to Day 28. Although Phase 2b is too small (N=60) to generate statistically meaningful data in cognition, a cognition scale (beyond MMSE) is being utilized to guide statistical considerations for future, large-scale clinical investigations with PTI-125. Unambiguous improvements in cognition and function is a key efficacy criterion for FDA approval of a new drug in AD (28), a hurdle which, to date, no drug candidate for AD has met with clear and compelling clinical data. Ultimately, to demonstrate disease modification in AD, future investigations must correlate improvements in biomarkers by PTI-125 with beneficial effects on cognition and function.


Funding:  This trial was funded by grant award AG060878 from the National Institute on Aging at NIH.

Acknowledgements: We sincerely thank NIA for their support of our work in Alzheimer’s disease. We thank the clinical investigators and patients who have participated in the clinical program for PTI-125. We thank consultants Chuck Davis for statistics on ELISA biomarkers and Jeff Stark for PK analyses.

Conflict of interest: LHB, CAC, RB and NF are employees of Cassava Sciences, Inc. HYW and MRM are science advisors to Cassava Sciences, Inc. ELB and BN are employees of IMIC, Inc., an independent clinical site that participated in this study.

Role of the sponsor: Cassava Sciences, Inc. provided all drug supply and material support for this clinical research, designed the study in consultation with its advisors and monitored the conduct of the study and data collection. Biomarker assays were conducted blind to treatment day by HWY and his lab at CUNY Medical school. LHB assisted in the interpretation of the data and wrote the manuscript together with RB and HWY.

Ethical standards: All participants and their caregivers provided written informed consent. The protocol, informed consent forms and clinical sites were all approved by Advarra IRB.



1.     World Alzheimer Report 2018. 2018: Alzheimer’s Disease International, London.
2.    Burns, L.H. and H.-Y. Wang, Altered filamin A enables amyloid beta-induced tau hyperphosphorylation and neuroinflammation in Alzheimer’s disease. Neuroimmunol Neuroinflammation, 2017. 4: p. 263-71.
3.    Wang, H.-Y., et al., PTI-125 binds and reverses an altered conformation of filamin A to reduce Alzheimer’s disease pathogenesis. Neurobiol Aging, 2017. 55: p. 99-114.
4.    Wang, H.-Y., et al., Reducing amyloid-related Alzheimer’s disease pathogenesis by a small molecule targeting filamin A. J Neurosci, 2012. 32: p. 9773-84.
5.    Nakamura, F., T. Stossel, and J. Hartwig, The filamins: organizers of cell structure and function. Cell Adh Migr, 2011. 5: p. 160-9.
6.    Wang, H.-Y., M. Frankfurt, and L.H. Burns, High-affinity naloxone binding to filamin A prevents mu opioid receptor – Gs coupling underlying opioid tolerance and dependence. PLoS One, 2008. 3(2): p. e1554.
7.    Wang, H.-Y., et al., β-Amyloid1-42 binds to α7 nicotinic acetylcholine receptor with high affinity:Implication for Alzheimer’s disease pathology. J Biol Chem, 2000. 275: p. 5626-5632.
8.    Wang, H.-Y., et al., α7 nicotinic acetylcholine receptors mediate β-amyloid peptide-induced tau protein phosphorylation. J Biol Chem, 2003. 278: p. 31547-31553.
9.    Dineley, K., et al., β-Amyloid peptide activates α7 nicotinic acetylcholine receptors expressed in xenopus oocytes. J Biol Chem, 2002. 227: p. 25056–61.
10.    Nagele, R., et al., Accumulation of beta-amyloid1-42 in neurons is facilitated by the alpha7 nicotinic acetylcholine receptor in Alzheimer’s disease. Neuroscience 2002. 110: p. 199-211.
11.    Wang, H.-Y., et al., Amyloid peptide Aβ1-42 binds selectively and with picomolar affinity to α7 nicotinic acetylcholine receptors. J Neurochem, 2000. 75: p. 1155-1161.
12.    Gambuzza, M., et al., Toll-like receptors in Alzheimer’s disease: a therapeutic perspective. CNS Neurol Disord Drug Targets, 2014. 13: p. 1542-58.
13.    Reed-Geaghan, E., et al., CD14 and Toll-like receptors 2 and 4 are required for fibrillar Aβ-stimulated microglial activation. J Neurosci, 2009. 29: p. 11982–11992.
14.    Hansson, O., et al., CSF biomarkers of Alzheimer’s disease concord with amyloid-b PETand predict clinical progression: A study of fully automated immunoassays in BioFINDER and ADNI cohorts. Alzheimer’s Dement, 2018. 14: p. 1470-1481.
15.    Olsson, B., et al., CSF and blood biomarkers for the diagnosis of Alzheimer’s disease: a systematic review and meta-analysis. Lancet Neurol, 2016. 15: p. 673-684.
16.    Bacioglu, M., et al., Neurofilament light chain in blood and CSF as marker of disease progression in mouse models and in neurodegenerative diseases. Neuron, 2016. 91: p. 1-11.
17.    Hampel, H., et al., Alzheimer’s disease biomarker-guided diagnostic workflow using the added value of six combined cerebrospinal fluid candidates: Aβ1–42, total-tau, phosphorylated-tau, NFL, neurogranin, and YKL-40. Alzheimer’s & Dementia, 2018. 14: p. 492-501.
18.    Hellwig, K., et al., Neurogranin and YKL-40: Independent markers of synaptic degeneration and neuroinflammation in Alzheimer’s disease. Alzheimers Res Ther, 2015. 7: p. 74.
19.    Lista, S. and H. Hampel, Synaptic degeneration and neurogranin in the pathophysiology of Alzheimer’s disease. Expert Rev Neurother 2017. 17: p. 47-57.
20.    Llorens, F., et al., YKL-40 in the brain and cerebrospinal fluid of neurodegenerative dementias. Mol Neurodegener, 2017. 12: p. 83.
21.    Nordengen, K., et al., Glial activation and inflammation along the Alzheimer’s disease continuum. J Neuroinflammation, 2019. 16: p. 46.
22.    Wang, H., et al., Increased Aβ42-α7-like nicotinic acetylcholine receptor complex level in lymphocytes is associated with apolipoprotein E4-driven Alzheimer’s disease pathogenesis. Alzheimers Res Ther., 2017. 9: p. 54.
23.    Wang, H.-Y., et al., S 24795 limits β-amyloid – α7 nicotinic receptor interaction and reduces Alzheimer’s disease-like pathologies. Biol Psychiatry 2010; 67: 522-530. Biol Psychiatry, 2010. 67: p. 522-530.
24.    Hayashi, K. and A. Altman, Filamin A Is Required for T Cell Activation Mediated by Protein Kinase C-θ. J Immunol, 2006. 177: p. 1721-1728.
25.    Wang, H.-Y., et al., Dissociating beta-amyloid from alpha 7 nicotinic acetylcholine receptor by a novel therapeutic agent, S 24795, normalizes alpha 7 nicotinic acetylcholine and NMDA receptor function in Alzheimer’s disease brain. J Neurosci, 2009. 35: p. 10961-73.
26.    Nathan, P., et al., Association between CSF biomarkers, hippocampal volume and cognitive function in patients with amnestic mild cognitive impairment (MCI). Neurobiol Aging, 2017. 17: p. 1-10.
27.    Swanson, C. Effect of BAN2401 on underlying AD pathophysiology. in CTAD. 2018. Barcelona.
28.    US FDA, Early Alzheimer’s Disease:  Developing Drugs for Treatment, Guidance for Industry 2018.

What do you want to do ?

New mail

What do you want to do ?

New mail



S. Gauthier1, J. Alam2, H. Fillit3, T. Iwatsubo4, H. Liu-Seifert5, M. Sabbagh6, S. Salloway7, C. Sampaio8, J.R. Sims5, B. Sperling9, R. Sperling10, K.A. Welsh-Bohmer11, J. Touchon12, B. Vellas13, P. Aisen14 and the EU/US/CTAD Task Force*


* EU/US/CTAD TASK FORCE: Bjorn Aaris Gronning (Valby); Paul Aisen (San Diego); John Alam (Cambridge); Sandrine Andrieu (Toulouse), Randall Bateman (St. Louis); Monika Baudler (Basel);  Joanne Bell (Wilmington); Kaj Blennow (Mölndal); Claudine Brisard (Blue Bell); Samantha Budd-Haeberlein (USA); Szofia Bullain (Basel) ; Marc Cantillon (Princeton) ; Maria Carrillo (Chicago);  Gemma Clark (Princeton); Jeffrey Cummings (Las Vegas); Daniel Di Giusto (Basel); Rachelle Doody (Basel); Sanjay Dubé (Aliso Viejo); Michael Egan (North Wales); Howard Fillit (New York); Adam Fleisher (Philadelphia); Mark Forman (North Wales); Cecilia Gabriel-Gracia (Suresnes); Serge Gauthier (Verdun); Jeffrey Harris (South San Francisco); Suzanne Hendrix (Salt Lake City); Dave Henley (Titusville); David Hewitt (Blue Bell); Mads Hvenekilde (Basel); Takeshi Iwatsubo (Tokyo); Keith Johnson (Boston); Michael Keeley (South San Francisco); Gene Kinney (South San Francisco); Ricky Kurzman (Woodcliffe Lake); Valérie Legrand (Nanterre); Stefan Lind (Valby); Hong Liu-Seifert (Indianapolis); Simon Lovestone (Oxford); Johan Luthman (Woodcliffe); Annette Merdes (Munich); David Michelson (Cambridge); Mark Mintun (Philadelphia); José Luis Molinuevo (Barcelona); Susanne Ostrowitzki (South San Francisco); Anton Porsteinsson (Rochester);  Martin Rabe (Woodcliffe Lake); Rema Raman (San Diego); Elena Ratti (Cambridge);  Larisa Reyderman (Woodcliffe Lake); Gary Romano (Titusville); Ivana Rubino (Cambridge); Marwan Noel Sabbagh (Las Vegas);  Stephen Salloway (Providence); Cristina Sampaio (Princeton); Rachel Schindler (USA); Peter Schüler (Langen); Dennis Selkoe (Boston); Eric Siemers (New York);  John Sims (Indianapolis); Heather Snyder (Chicago); Georgina Spence (Galashiels); Bjorn Sperling (Valby); Reisa Sperling (Boston); Andrew Stephens (Berlin); Joyce Suhy (Newark); Gilles Tamagnan (New Haven); Edmond Teng (South San Francisco); Gary Tong (Valby); Jan Torleif Pedersen (Valby); Jacques Touchon (Montpellier); Bruno Vellas (Toulouse ); Vissia Viglietta (Cambridge) ; Christian Von Hehn (Cambridge); Philipp Von Rosenstiel (Cambridge) ; Michael Weiner (San Francisco); Kathleen Welsh-Bohmer (Durham);  Iris Wiesel (Basel); Haichen Yang (North Wales);  Wagner Zago (South San Francisco); Beyhan Zaim (Woodcliffe Lake); Henrik Zetterberg (Mölndal)  

1. McGill Center for Studies in Aging, Verdun QC, Canada;  2. EIP Pharma Inc., Cambridge MA, USA; 3. The Alzheimer’s Drug Discovery Foundation, New York NY, USA; 4. University of Tokyo, Japan; 5. Eli Lilly and Company, Indianapolis IN, USA; 6L Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas NV, USA; 7. The Warren Alpert Medical School of Brown University, Providence RI, USA; 8. CHDI Foundation, Princeton NJ, USA; 9. Lundbeck, Valby 2500 Denmark; 10. Brigham and Women’s Hospital, Boston MA, USA; 11. Duke University, Durham NC, USA; 12. University Hospital of Montpellier, 34025 Montpellier Cedex 5, and INSERM 1061, France; 13. Gerontopole, INSERM U1027, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France; 14. Alzheimer’s Therapeutic Research Institute (ATRI), Keck School of Medicine, University of Southern California, San Diego CA, USA

Corresponding Author: Serge Gauthier, McGill Center for Studies in Aging, Verdun QC, Canada,

J Prev Alz Dis
Published online April 18, 2019,



Combination therapy is expected to play an important role for the treatment of Alzheimer’s disease (AD). In October 2018, the European Union-North American Clinical Trials in Alzheimer’s Disease Task Force (EU/US CTAD Task Force) met to discuss scientific, regulatory, and logistical challenges to the development of combination therapy for AD and current efforts to address these challenges. Task Force members unanimously agreed that successful treatment of AD will likely require combination therapy approaches that target multiple mechanisms and pathways. They further agreed on the need for global collaboration and sharing of data and resources to accelerate development of such approaches.

Key words: Alzheimer’s disease, amyloid, tau, therapeutics, trial design.



Combination therapy has resulted in improved outcomes for many of the world’s most significant and complex diseases, including cancer, AIDS, and cardiovascular disease, and the prospect of combination therapy has also gained traction in the Alzheimer’s disease (AD) field (1-3). The reasons for pursuing combination therapy for AD go beyond the disappointing track record in developing effective treatments for this disease that is likely to affect more than 150 million people worldwide by 2050 (4-6). As with many other complex diseases, AD arises from a series of pathological changes and the involvement of many pathogenic pathways that begin well before symptoms appear (7), suggesting that effective treatment will require targeting multiple pathways, either simultaneously or sequentially. However, the complexity of AD pathophysiology also introduces substantial hurdles to the development of combinatorial approaches. To better understand current efforts to develop such approaches and the steps that need to be taken to expedite this process, the European Union-North American Clinical Trials in Alzheimer’s Disease Task Force (EU/US CTAD Task Force) discussed combination therapy for AD at its 2018 meeting. The Task Force brings together investigators from industry, academia, and regulatory agencies to build consensus and promote collaboration and information sharing on issues important for the development of effective Alzheimer’s treatments. Many Task Force members expect combination therapy to play an important role in treating AD and call for global collaboration to develop combination therapies (8, 9), but agree that the path forward has yet to be clearly defined.


Best candidates for combination therapy

Combination therapy for AD could involve interrupting a single important pathogenic pathway (such as amyloid or tau) at multiple points or targeting two or more pathways together (such as amyloid plus tau). Despite the many disappointing clinical trials of disease-modifying therapies targeting amyloid, it remains a promising target for disease modification, in particular for prevention studies. The rationale for targeting amyloid is strong (10). Most known genetic mutations related to AD are involved in amyloid production or processing. This includes mutations in the Presenilin 1 and 2 and amyloid precursor protein (APP) genes, and Down syndrome, the most common cause of early-onset AD, which is caused by a trisomy of chromosome 21 where the APP gene resides. In addition, a mutation in the APP gene known as the Icelandic mutation (A673T) has been shown to be protective against AD and cognitive decline (11).
Moreover, there is abundant evidence that Aβ oligomers and amyloid plaques are toxic (12, 13), and encouraging although preliminary evidence that removing plaques may be associated with improved cognition and clinical outcomes.
The APP molecule undergoes sequential cleavage via β- and γ-secretases to produce amyloidogenic fragments. Amyloid peptides take on monomeric, oligomeric, and fibrillar forms that may cause toxicity through a variety of mechanisms including oxidative stress, excitotoxicity, synaptic failure, and other mechanisms associated with neuronal death (14). This complex pathway from APP to toxicity thus creates multiple potential therapeutic targets (Figure 1). Antibodies directed at different amyloid fragments have been developed as potential treatments against AD with varying degrees of success at removing amyloid and halting the disease process; secretase inhibitors have also been effective at reducing amyloid load but have been associated with cognitive worsening and other adverse events (15, 16).

Figure 1. Opportunities for amyloid-based combination therapies based on therapeutics currently in clinical development

Figure 1. Opportunities for amyloid-based combination therapies based on therapeutics currently in clinical development


A workgroup of the National Institutes on Aging and the Alzheimer’s Association (NIA-AA) recently published a research framework that defines and stages the disease according to the presence of Amyloid (A), Tau (T) and neurodegeneration (N) biomarkers (17).  Yet while the AD disease-modifying drug development pipeline continues to reflect the predominance of the amyloid pathway, there has recently been an increase in the number of drug trials testing non-amyloid mechanisms (18, 19). In agreement with the NIA-AA Research Framework, the Task Force recognized the need to add biomarkers of other pathologies commonly seen in the brains of people with AD, such as vascular pathology (V), inflammation (I), and Lewy bodies (L).


Possible combination trial designs that target amyloid

Preclinical AD is marked primarily by amyloid accumulation, with cognitive performance and biomarkers of neurodegeneration, tau, and cerebral metabolism increasing markedly only in the clinical stages of disease (20). This suggests that a vigorous attack on amyloid using multiple agents simultaneously to target different steps in the amyloid pathway may slow, stop, or reverse the progression of AD.
However, an even more promising approach may be attacking the amyloid pathway sequentially at different times and disease stages. Sequential therapy offers efficiency advantages by enabling the assessment of individual adverse events and benefits more readily. One potential sequential therapy design using an induction/maintenance approach would be to start treating with an inhibitor of Aβ production, such as a beta-secretase inhibitor (BACEi), before there is any detectable amyloid; and then introducing amyloid-reducing antibodies when amyloid becomes elevated but before neuronal damage has begun. This approach could reduce the number of anti-amyloid antibody infusions required, thus saving costs and reducing exposure. However, designing a trial using this strategy could become very complicated.
An alternative would be to start with an anti-amyloid antibody first to induce an amyloid-free state for 3 months to 1 year (long enough to see cognitive benefit in early stage), and then push backwards and treat with BACEi as maintenance therapy. Although BACEi have shown significant adverse events in several trials, a lower dose (e.g. inhibiting only ≤30% of BACE) may improve the risk/benefit calculation. Other secretase modulators, antibodies that target diffusible amyloid, or amyloid active vaccine may also be used for maintenance.
A combination study including both anti-tau and anti-amyloid drugs also has been suggested although many questions remain about the efficacy of anti-tau agents, the best tau epitopes to target, the optimal stage of disease to treat, how to establish target engagement, and how to design anti-tau trials (31).  Another combination clinical trial that combines two non-amyloid approaches is also underway at Amylyx Pharmaceuticals in partnership with the Alzheimer’s Drug Discovery Foundation and the Alzheimer’s Association. This Phase 2 trial of AMX0035 combines  sodium phenylbutyrate, which is approved for the treatment of urea cycle disorders, and tauroursodeoxycholic acid (TUDCA), a bile acid that supports mitochondrial energetics (19). The combination is expected to protect neurons from inflammation and oxidative stress.


Best target populations and study designs

For clinical trials of combination therapies such as those described above, the stage of disease and study design for proof-of-concept and Phase 3 studies will be determined by a medication’s mode of action on disease pathophysiology. For example, trials designed to treat patients in early disease stages, i.e., symptomatic with a CDR 0.5 or 1, should maximize the likelihood of detecting disease progression during the trial and demonstrating a slowing of progression if the treatment is efficacious. Enabling optimal designs and optimizing treatment assignment will require that participants have adequate biological characterization with biomarkers.
The most informative trial design for a two-agent combination therapy trial would employ a 2 x 2 factorial structure where each agent is tested alone and in combination (21). A more efficient approach, however, would be a 2-arm trial of the combination vs. placebo, with deconvolution of the contribution of each agent should the initial approach be successful. In either case, selecting dose and treatment regimens for combination studies is complicated and often leads investigators to take shortcuts, which can lead to misleading results or unacceptable risks to participants. The statistical and regulatory implications of various trial designs are discussed below.
For trials in patients with AD dementia, since many individuals will already be taking acetylcholinesterase inhibitors (e.g., donepezil) and/or memantine (22), add-on designs that combine the standard treatment plus the disease-modifying agent being tested may be necessary. To test combinations in early AD patients, a different type of add-on design could provide more precision. For this type of study, participants would be randomized first to induction therapy with an agent that targets the most prominent apparent pathology (amyloid for most, tau for a few); then after a pre-determined time period (e.g., 6 months), a second treatment is added that targets the second most predominant pathology (e.g., tau, amyloid, inflammation, Lewy bodies, or vascular).
Open perpetual platform trials using a master protocol with defined inclusion and exclusion criteria may be the most efficient way to conduct combination trials. The Dominantly Inherited Alzheimer’s Network Trials Unit (DIAN-TU) has developed such a platform for testing a variety of therapeutics in people with autosomal dominant AD (23). Such platforms enable testing of multiple active treatment arms with shared control arm, and they allow for: 1) pooling of placebo groups, 2) the discontinuation of arms for futility, 3) the addition of new arms including either new drugs or new doses, 4) adaptive randomization, and 5) personalization of arms to specific subgroups (24).


Regulatory issues

Regulatory authorities encourage innovative development approaches for delivering combination therapies for AD. In 2013, the U.S. Food and Drug Administration (FDA) published guidance for co-development of two or more new investigational drugs for use in combination (25). According to this guidance, combination therapy is justified for treating serious diseases with unmet medical needs when there is a strong rationale and strong preclinical data for the combination, and when there is a compelling reason for developing the two drugs in tandem rather than independently.
Selecting agents to combine begins with assessing and characterizing whether the interaction between the components is additive, synergistic, or antagonistic. In addition, since most amyloid treatments activate the immune system, nonclinical studies are needed to assess the interaction of combinations with immune mechanisms. How the effectiveness of the combination is defined affects the study design and may depend on the stages of development of the components.  Thus, if one component is already approved, it may be sufficient to demonstrate how much greater is the effect of the combination of new drug plus the approved drug compared to the effect of the approved drug alone. If both components are novel, however, a full factorial design may be needed to understand contributions of the different agents to the treatment response. Additive or synergistic effects may be demonstrated.
Both FDA and the European Medicines Agency (EMA) require preclinical studies of the combination. In some cases, toxicology of the combination will need to be tested, although there have been some studies where regulators were sufficiently confident that a combination would be safe and allowed advancing to Phase 2.


Blazing the trail to combination therapy

In December 2017, Lilly launched the multi-site TRAILBLAZER-ALZ Phase 2 trial, which combined a BACEi with the anti-amyloid monoclonal antibody LY3002813 (NCT03367403), a humanized IgG1 antibody directed at N3pG, an amyloid epitope that is present only in amyloid plaques (26). In preclinical studies, LY3002813 was shown to remove amyloid plaque through microglial-mediated clearance (27). In the PDAPP mouse model, an antiN3pG plus a BACEi removed most pre-existing plaque and improved neuronal health in a synergistic, dose-dependent manner (28).
A phase 1 study of the monoclonal antibody demonstrated a significant reduction in brain amyloid by florbetapir positron emission tomography (PET); and a phase 1 study of the BACEi demonstrated a lowering of cerebrospinal fluid (CSF) Aβ, with no safety or tolerability concerns. These results in Phase 1, combined with the preclinical data, prompted Lilly to plan the TRAILBLAZER trial that would include three arms: 1) placebo, 2) N3pG monoclonal antibody alone, and 3) N3pG mAb plus BACEi. Rather than using a full factorial design, external data from multiple ongoing BACEi studies would be used to demonstrate the efficacy of the BACEi alone.
The study enrolled participants with early symptomatic AD who are amyloid positive with a low-to-medium tau burden, randomized to the three arms. These inclusion/exclusion criteria were selected to produce a relatively homogeneous population. A composite scale of cognition and function was selected as the primary outcome, and a robust biomarker strategy was planned to demonstrate the contribution of each component of the combination (29). Other cognitive and functional measures as well as amyloid PET, tau PET, and volumetric magnetic resonance imaging (MRI) were included as secondary outcome measures.
The combination BACEi plus N3pG mAb arm of the trial was subsequently discontinued based on data from multiple sources that raised concerns about the risk/benefit profile of BACEi (30). Nonetheless, the study design holds lessons for future trials of combination therapies, including the use of preclinical data in animal models to demonstrate synergy, the use of robust Phase 1 data to simplify Phase 2 combination designs, and the importance of early interaction with regulators to design toxicology and clinical studies.


Moving forward

Despite the discontinuation of the combination therapy arm in the TRAILBLAZER trial, Task Force members unanimously agreed that successful treatment of AD will require combination approaches that target multiple mechanisms and pathways. However, many questions remain regarding how best to move forward in the development of combination therapies.
Task Force members suggested several steps that should be taken to expedite the development of combination therapies:
•    Establish thresholds for pathologies beyond amyloid and tau, including inflammation and vascular load.
•    Pool observational studies to determine natural history of various combinations of pathologies.
•    Negotiate a DIAN-like structure with global resources from companies and academia, for example through the European Prevention of Alzheimer’s Dementia Consortium (EPAD).
•    Enlarge the dialogue about combination therapy to include disease modifying as well as symptomatic treatments and mechanisms that address the neurodegenerative process.
•    Pool resources, for instance by testing add-on compounds in participants enrolled in preclinical or Phase 2 clinical trials with a single agent, having completed the double-blind placebo-controlled phase of the study.

Task Force members also agreed that patient engagement is key to the development of combination therapies, particularly for treatments intended for the presymptomatic stages of the disease.


Acknowledgements: The authors thank Lisa J. Bain for assistance in the preparation of this manuscript.

Conflicts of interest: The Task Force was partially funded by registration fees from industrial participants. These corporations placed no restrictions on this work. Dr. Gauthier reports personal fees from TauRx, Lundbeck Institute, and Esai; and grants from Lilly and Roche, outside the submitted work. Dr. Alam reports personal fees (employment) from EIP Pharma, Inc,  outside the submitted work. Dr. Fillit discloses the following consulting relationships during the past 3 years: Axovant, vTv, Lundbeck, Otsuka, Lilly, RTI, Roche, Genentech, Merck, Samus, Pfizer. He reports no conflicts of interest related to these disclosures that are relevant to this publication. Dr. Iwatsubo has nothing to disclose. Dr. Liu-Seifert reports other from Lilly,  outside the submitted work. Dr. Sabbagh has consulted for Allergan, Biogen, Bracket, Neurotrope, Cortexyme, Roche, Grifols, Sanofi, VTV therapeutic, and Alzheon. Dr Sslloway has nothing to disclose; Dr. Sims, employee of Eli Lilly and Company and holder of stock in Eli Lilly and Company. Dr. Sperling is an employee of H. Lundbeck A/S,  outside the submitted work. Dr. Sperling reports grants from Janssen, during the conduct of the study; personal fees from AC Immune, personal fees from Biogen, personal fees from Roche, personal fees from Eisai, personal fees from Insightec, personal fees from Takeda, personal fees from Merck, personal fees from General Electric, outside the submitted work. Dr. Welsh-Bohmer has contracts with Takeda Pharmaceutical Company and with VeraSci where she is the VP for Neurodegenerative Disorders. Dr. Touchon has nothing to disclose; Dr. Vellas reports grants from Lilly, Merck, Roche, Lundbeck, Biogen, grants from Alzheimer’s Association, European Commission, personal fees from Lilly, Merck, Roche, Biogen, outside the submitted work; Dr. Aisen reports grants from Lilly, personal fees from Proclara, other from Lilly, other from Janssen, other from Eisai, grants from Janssen, grants from NIA, grants from FNIH, grants from Alzheimer’s Association, personal fees from Merck, personal fees from Roche, personal fees from Lundbeck, personal fees from Biogen, personal fees from ImmunoBrain Checkpoint,  outside the submitted work.



1.    Hendrix JA, Bateman RJ, Brashear HR, et al. Challenges, solutions, and recommendations for Alzheimer’s disease combination therapy. Alzheimers Dement 2016;12:623-630.
2.    Perry D, Sperling R, Katz R, et al. Building a roadmap for developing combination therapies for Alzheimer’s disease. Expert Rev Neurother 2015;15:327-333.
3.    Stephenson D, Perry D, Bens C, et al. Charting a path toward combination therapy for Alzheimer’s disease. Expert Rev Neurother 2015;15:107-113.
4.    Alzheimer’s Association. 2018 Alzheimer’s disease facts and figures. Alzheimers Dement 2018;14:367-429.
5.    Alzheimer’s Disease International. World Alzheimer’s Report 2018. The state of the art of dementia research: New frontiers. London2018.
6.    Cummings J, Lee G, Mortsdorf T, Ritter A, Zhong K. Alzheimer’s disease drug development pipeline: 2017. Alzheimers Dement (N Y) 2017;3:367-384.
7.    Bateman RJ, Xiong C, Benzinger TLS, et al. Clinical, cognitive, and biomarker changes in the Dominantly Inherited Alzheimer Network. The New England journal of medicine 2012;367:795-804.
8.    Schindler RJ. Study design considerations: conducting global clinical trials in early Alzheimer’s disease. J Nutr Health Aging 2010;14:312-314.
9.    Sperling R, Cummings J, Donohue M, Aisen P. Global Alzheimer’s Platform Trial Ready Cohorts for the Prevention of Alzheimer’s Dementia. J Prev Alzheimers Dis 2016;3:185-187.
10.    Selkoe DJ, Hardy J. The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol Med 2016;8:595-608.
11.    Jonsson T, Atwal JK, Steinberg S, et al. A mutation in APP protects against Alzheimer’s disease and age-related cognitive decline. Nature 2012;488:96-99.
12.    Serrano-Pozo A, Betensky RA, Frosch MP, Hyman BT. Plaque-Associated Local Toxicity Increases over the Clinical Course of Alzheimer Disease. Am J Pathol 2016;186:375-384.
13.    Benilova I, Karran E, De Strooper B. The toxic Abeta oligomer and Alzheimer’s disease: an emperor in need of clothes. Nat Neurosci 2012;15:349-357.
14.    O’Brien RJ, Wong PC. Amyloid precursor protein processing and Alzheimer’s disease. Annu Rev Neurosci 2011;34:185-204.
15.    Sims JR, Selzler KJ, Downing AM, et al. Development Review of the BACE1 Inhibitor Lanabecestat (AZD3293/LY3314814). J Prev Alzheimers Dis 2017;4:247-254.
16.    Budd Haeberlein S, O’Gorman J, Chiao P, et al. Clinical Development of Aducanumab, an Anti-Abeta Human Monoclonal Antibody Being Investigated for the Treatment of Early Alzheimer’s Disease. J Prev Alzheimers Dis 2017;4:255-263.
17.    Jack CR, Jr., Bennett DA, Blennow K, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 2018;14:535-562.
18.    Cummings J, Lee G, Ritter A, Zhong K. Alzheimer’s disease drug development pipeline: 2018. Alzheimers Dement (N Y) 2018;4:195-214.
19.    Hara Y, McKeehan N, Fillit HM. Translating the biology of aging into novel therapeutics for Alzheimer disease. Neurology 2019;92:84-93.
20.    Jack CR, Jr., Knopman DS, Weigand SD, et al. An operational approach to National Institute on Aging-Alzheimer’s Association criteria for preclinical Alzheimer disease. Ann Neurol 2012;71:765-775.
21.    Tomaszewski S, Gauthier S, Wimo A, Rosa-Neto P. Combination Therapy of Anti-Tau and Anti-Amyloid Drugs for Disease Modification in Early-stage Alzheimer’s Disease: Socio-economic Considerations Modeled on Treatments for Tuberculosis, HIV/AIDS and Breast Cancer. J Prev Alzheimers Dis 2016;3:164-172.
22.    Hendrix S, Ellison N, Stanworth S, Otcheretko V, Tariot PN. Post Hoc Evidence for an Additive Effect of Memantine and Donepezil: Consistent Findings from DOMINO-AD Study and Memantine Clinical Trial Program. J Prev Alzheimers Dis 2015;2:165-171.
23.    Bateman RJ, Benzinger TL, Berry S, et al. The DIAN-TU Next Generation Alzheimer’s prevention trial: Adaptive design and disease progression model. Alzheimers Dement 2017;13:8-19.
24.    Saville BR, Berry SM. Efficiencies of platform clinical trials: A vision of the future. Clin Trials 2016;13:358-366.
25.    Food and Drug Administration. Guidance for industry: Codevelopment of two or more unmarketed investigational drugs for use in combination.. Accessed online 4/13/14 at, 2013.
26.    Irizarry MC, Fleisher AS, Hake AM, et al. TRAILBLAZER-ALZ (NCT03367403): A Phase 2 disease-modification combination therapy trial targeting multiple mechanisms of action along the amyloid pathway. Alzheimer Dement 2018;14:P1622-P1623.
27.    DeMattos RB, Lu J, Tang Y, et al. A plaque-specific antibody clears existing beta-amyloid plaques in Alzheimer’s disease mice. Neuron 2012;76:908-920.
28.    DeMattos R, May P, Racke M, et al. Combination therapy with a plaque-specific Abeta antibody and BACE inhibitor results in dramatic plaque lowering in aged PDAPP transgenic mice. Alzheimer Dement 2014;10:P149.
29.    Wessels AM, Siemers ER, Yu P, et al. A Combined Measure of Cognition and Function for Clinical Trials: The Integrated Alzheimer’s Disease Rating Scale (iADRS). J Prev Alzheimers Dis 2015;2:227-241.
30.    Panza F, Lozupone M, Solfrizzi V, et al. BACE inhibitors in clinical development for the treatment of Alzheimer’s disease. Expert Rev Neurother 2018;18:847-857.
31.    Vellas B, Bain LJ, Touchon J, et al. Advancing Alzheimer’s Disease Treatment: Lessons from CTAD 2018 . J Prev Alzheimers Dis. 2019; in press



J. Cummings1, K. Blennow2, K. Johnson3, M. Keeley4, R.J. Bateman5, J.L. Molinuevo6, J. Touchon7, P. Aisen8, B. Vellas9 and the EU/US/CTAD Task Force*


* EU/US/CTAD TASK FORCE: EU/US/CTAD TASK FORCE: Bjorn Aaris Gronning (Valby); Paul Aisen (San Diego); John Alam (Cambridge); Sandrine Andrieu (Toulouse), Randall Bateman (St. Louis); Monika Baudler (Basel);  Joanne Bell (Wilmington); Kaj Blennow (Mölndal); Claudine Brisard (Blue Bell); Samantha Budd-Haeberlein (USA); Szofia Bullain (Basel) ; Marc Cantillon (Princeton) ; Maria Carrillo (Chicago);  Gemma Clark (Princeton); Jeffrey Cummings (Las Vegas); Daniel Di Giusto (Basel); Rachelle Doody (Basel); Sanjay Dubé (Aliso Viejo); Michael Egan (North Wales); Howard Fillit (New York); Adam Fleisher (Philadelphia); Mark Forman (North Wales); Cecilia Gabriel-Gracia (Suresnes); Serge Gauthier (Montreal); Jeffrey Harris (South San Francisco); Suzanne Hendrix (Salt Lake City); Dave Henley (Titusville); David Hewitt (Blue Bell); Mads Hvenekilde (Basel); Takeshi Iwatsubo (Tokyo); Keith Johnson (Boston); Michael Keeley (South San Francisco); Gene Kinney (South San Francisco); Ricky Kurzman (Woodcliffe Lake); Valérie Legrand (Nanterre); Stefan Lind (Valby); Hong Liu-Seifert (Indianapolis); Simon Lovestone (Oxford); Johan Luthman (Woodcliffe); Annette Merdes (Munich); David Michelson (Cambridge); Mark Mintun (Philadelphia); José Luis Molinuevo (Barcelona); Susanne Ostrowitzki (South San Francisco); Anton Porsteinsson (Rochester);  Martin Rabe (Woodcliffe Lake); Rema Raman (San Diego); Elena Ratti (Cambridge);  Larisa Reyderman (Woodcliffe Lake); Gary Romano (Titusville); Ivana Rubino (Cambridge); Marwan Noel Sabbagh (Las Vegas);  Stephen Salloway (Providence); Cristina Sampaio (Princeton); Rachel Schindler (USA); Peter Schüler (Langen); Dennis Selkoe (Boston); Eric Siemers (Indianapolis);  John Sims (Indianapolis); Heather Snyder (Chicago); Georgina Spence (Galashiels); Bjorn Sperling (Valby); Reisa Sperling (Boston); Andrew Stephens (Berlin); Joyce Suhy (Newark); Gilles Tamagnan (New Haven); Edmond Teng (South San Francisco); Gary Tong (Valby); Jan Torleif Pedersen (Valby); Jacques Touchon (Montpellier); Bruno Vellas (Toulouse ); Vissia Viglietta (Cambridge) ; Christian Von Hehn (Cambridge); Philipp Von Rosenstiel (Cambridge) ; Michael Weiner (San Francisco); Kathleen Welsh-Bohmer (Durham);  Iris Wiesel (Basel); Haichen Yang (North Wales);  Wagner Zago (South San Francisco); Beyhan Zaim (Woodcliffe Lake); Henrik Zetterberg (Mölndal)

1. University of Nevada Las Vegas, School of Allied Health Sciences and Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas, Nevada, USA; 2. Inst. of Neuroscience and Physiology, University of Gothenburg, Sahlgrenska University Hospital, Mölndal, Sweden; 3. Massachusetts General Hospital, Harvard Medical School, Boston MA, USA; 4. Genentech Research and Early Development, So. San Francisco, CA, USA; 5. Washington University School of Medicine, St. Louis, MO, USA; 6. BarcelonaBeta Brain Research Center Pasqual Maragall Foundation, Barcelona, Spain; 7. Montpellier University, and INSERM U1061, Montpellier, France; 8. Alzheimer’s Therapeutic Research Institute (ATRI), Keck School of Medicine, University of Southern California, San Diego, CA, USA; 9. Gerontopole, INSERM U1027, Alzheimer’s Disease Research and Clinical Center, Toulouse University Hospital, Toulouse, France

Corresponding Author: Jeffrey Cummings, University of Nevada Las Vegas, School of Allied Health Sciences and Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas, Nevada, USA,

J Prev Alz Dis 2019;
Published online April 18, 2019,



Efforts to develop effective disease-modifying treatments for Alzheimer’s disease (AD) have mostly targeted the amyloid β (Aβ) protein; however, there has recently been increased interest in other targets including phosphorylated tau and other forms of tau. Aggregated tau appears to spread in a characteristic pattern throughout the brain and is thought to drive neurodegeneration. Both neuropathological and imaging studies indicate that tau first appears in the entorhinal cortex and then spreads to the neocortex. Anti-tau therapies currently in Phase 1 or 2 trials include passive and active immunotherapies designed to prevent aggregation, seeding, and spreading, as well as small molecules that modulate tau metabolism and function. EU/US/CTAD Task Force members support advancing the development of anti-tau therapies, which will require novel imaging agents and biomarkers, a deeper understanding of tau biology and the dynamic interaction of tau and Aβ protein, and development of multiple targets and candidate agents addressing the tauopathy of AD. Incorporating tau biomarkers in AD clinical trials will provide additional knowledge about the potential to treat AD by targeting tau.

Key words: Alzheimer’s disease, tau, tauopathy, therapeutics, biomarkers.



No new drugs have been approved by the US Food and Drug Administration (FDA) for the treatment of Alzheimer’s disease (AD) since 2003 (1) despite the fact that an estimated 5.7 million Americans and 50 million people worldwide have AD today, and the prevalence is expected to grow to 152 million worldwide by 2050 (2, 3).  AD clinical trials have failed at a very high rate: between 2002 and 2012, 99.6% of AD drugs tested failed to demonstrate clinical efficacy (1). Possible reasons for the high failure rate include targeting the wrong pathology or the wrong stage of disease (4, 5). Inappropriately designed trials and other methodological or unknown factors may have also contributed to treatment failures (6).
Despite the disappointments of the past 20 years, many experts in the Alzheimer’s community see reasons for optimism, including the emergence of novel drugs addressing a broader array of mechanisms than in the past (7). A recent report on the status of the AD drug development pipeline identified 112 agents: 26 in Phase 3 studies, 75 in Phase 2 studies, and 23 in Phase 1 studies (8). Moreover, whereas most of the negative studies in recent years targeted brain amyloidosis and amyloid β (Aβ), current studies are targeting a broader repertoire of mechanisms, including tau pathology. Of the 26 agents in Phase 3, only one targets tau, while 9 of the agents in Phase 2 (5 immunotherapies and 4 anti-aggregation agents) target tau (8).


Biology of tau and anti-tau therapeutics

The microtubule-associated protein tau, commonly referred to simply as tau, is found in a hyperphosphorylated form as insoluble, filamentous tangles and neuropil threads as well as dystrophic neurites in the AD brain (9). Along with plaques made up of aggregated Aβ protein, neurofibrillary tangles (NFTs) represent one of the hallmark pathologies of AD. Like Aβ, tau is found in several forms in the brain including monomers, oligomers, and fibrillary tangles (10).  Tau pathology also plays a central role in other neurodegenerative diseases known collectively as tauopathies, including the primary tauopathies frontotemporal lobar degeneration with tau inclusions (FTLD-tau), Pick’s disease, progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), argyrophilic grain disease (AGD), and chronic traumatic encephalopathy (CTE) (11).
Six isoforms of tau exist; it binds to several other proteins; and it undergoes many post-translational modifications, all of which contribute to its multiple functions in the brain. Tau protein plays important roles in cytoskeletal stability, cell signaling, synaptic plasticity, and neurogenesis (12, 13). In the AD brain, NFTs and neuropil threads composed of aggregated hyperphosphorylated tau are thought to be the primary drivers of neurodegeneration, although the mechanisms underlying the pathogenic process and the exact relationships of tau to Aβ remain unclear. Evidence also strongly suggests that tau propagates or spreads between cells (14) and that neuroinflammation triggered by microglial activation and astrogliosis contributes to tau-associated pathogenesis. Microglia may contribute to tau spreading (15). While postmortem and tau positron emission tomography (PET) studies indicate that tau spreading is associated with disease progression (16, 17), there are many unanswered questions regarding the rate of seeding or the effects of tau spreading on neuronal biology. If the spread of tau is driving clinical and cognitive changes, this would support intervening at the earliest stages of the tau-related disease process.
Neuropathological and imaging studies using PET suggest that tau aggregates are found in the entorhinal cortex and then the neocortex.  If and how this drives neurodegeneration, what forms of tau are toxic, and the relationship of tau to amyloid in terms of toxicity remain unanswered questions. Tau pathology correlates much more closely to cognitive decline than does amyloid pathology (18, 19), and a recent study suggests that tau aggregation is linked to neurodegeneration and clinical manifestations of AD (20).
The complexity of tau biology provides many potential therapeutic targets to prevent tau production, aggregation, or spread at the level of transcription, phosphorylation, depolymerization, and transport. For example, preclinical studies indicate that antibodies against tau can prevent the trans-synaptic transmission of tau between neurons (21). A Phase 1 study of the humanized monoclonal antibody ABBV-8E12 showed acceptable safety and tolerability, which provided the basis for initiating a Phase 3 study in PSP patients to assess dose-related efficacy (22). The antibody is intended to prevent the trans-neuronal spread of the tau protein. Other monoclonal antibodies being assessed in early phase studies and targeting aspects of the tau protein include BIIB092, LY3303560, and RO7105705 (Table 1).

Table 1. Phase 1 and 2 clinical Trials targeting tau in AD populations

Table 1. Phase 1 and 2 clinical Trials targeting tau in AD populations

ADAS-cog = Alzheimer’s Disease Assessment Scale, cognition subscale, ADCS-ADL = Alzheimer’s Disease Cooperative Study – Activities of Daily Living Inventory, AEs= adverse events, CDR-SB = Clinical Dementia Rating Scale Sum of Boxes, CGIC = Clinical Global Impressions Scale, FAQ = functional activities questionnaire, fMRI = functional magnetic resonance imaging, HAM-D = Hamilton Psychiatric Rating Scale for Depression, iADRS=integrated AD rating scale, MMSE = Mini-mental state examination, NPI = neuropsychiatric inventory, NPS battery = neuropsychiatric symptoms battery, RBANS= Repeatable Battery for the Assessment of Neuropsychological Status, TEAEs = treatment-emergent adverse events. UPSA=University of California Performance Based Skills Assessment, brief version


Current status of anti-tau therapies in the AD treatment pipeline

One putative anti-tau agent, TRx0237 was studied in a Phase 3 trial and failed to show a difference between different doses.  Studies in mouse models suggested that the agent functioned as an aggregation inhibitor and reduced the number of tau positive neurons (23); no target engagement biomarker was included in trial to determine if this was achieved in humans (24). Subgroup analyses suggest that some patients may have benefited from therapy and further studies of this compound are underway (24). Table 1 summarizes the anti-tau agents that are currently being tested in Phase 1 or Phase 2 clinical trials. These include both passive and active immunotherapies with monoclonal antibodies as well as drugs that affect the molecular structure of tau to modulate its function or prevent phosphorylation.
Other anti-tau drugs are also in development for AD including epigallocatechin-3 gallate (EGCG), a polyphenolic flavanoid extracted from green tea 25, and AC Immune’s tau morphomers, small molecules designed to inhibit aggregation and seeding and disaggregate already formed tau aggregates. Preclinical studies suggest that tau morphomers reduce pathological tau, improve cognition and function, and reduce microglia activation. Importantly, they are capable of crossing the blood-brain barrier.


Outcome measures and biomarkers

Tau PET imaging

Tau and Aβ aggregates in the brain have been investigated in several cohort studies, both neuropathologically at autopsy and in living people using PET (26-28). The overall picture emerging from these studies is that among cognitively normal individuals, about one-third have high amyloid, and among those with high amyloid about half also have high tau loads. A minority of cognitively normal individuals have sub-threshold levels of amyloid and high tau. The anatomic location of tau deposition may be important. These observations raise the possibility that quantifying progression of tau pathology may provide an early indicator of disease.
Johnson and colleagues have investigated the anatomical variability of amyloid and tau deposition in more than 400 individuals. These data indicate that distribution of tau in the rhinal cortex correlates with amyloid burden and that low amyloid individuals just starting to show elevations in tau are those most likely to be on the way to neocortical tauopathy. By the time tau levels have increased in the inferior temporal cortex, individuals may show significant impairments. These data support the hypothesis that amyloid is associated with tau spread.
Longitudinal data also provide support for these measures as useful for staging in order to establish a basis on which to measure change in serial imaging that could be useful in the clinical and clinical trial settings. Four stages were proposed:
Stage 0 – No signal exceeding background, consistent with Braak 0.
Stage 1 – Rhinal cortex signal emerging in a minority of low-amyloid clinically unimpaired individuals (allocortex, MTL) consistent with Braak I/II
Stage 2 – Inferior temporal signal emerging in the presence of high levels of fibrillar amyloid in clinically unimpaired individuals (corresponding to Braak stages III/VI)
Stage 3 – Additional neocortical binding in mild cognitive impairment (MCI) and AD patients (beyond inferior temporal; corresponding to Braak stages V/VI)
Figure 1 provides an example of images with high and low tau burden.

Figure 1. Flortaucipir images with low (Braak I/II) and high (Braak III/VI) levels of tau. The individual whose image is shown on the left had low amyloid levels; the one shown on the right had high amyloid levels (images courtesy of Keith Johnson)

Figure 1. Flortaucipir images with low (Braak I/II) and high (Braak III/VI) levels of tau. The individual whose image is shown on the left had low amyloid levels; the one shown on the right had high amyloid levels (images courtesy of Keith Johnson)


CSF and blood biomarkers of tau

A systematic review and meta-analysis of cerebrospinal fluid (CSF) and blood biomarkers showed that CSF levels of total tau (T-tau), phosphorylated tau (p-tau), Aβ42, and neurofilament light (NfL), and plasma levels of T-tau were associated with AD and MCI due to AD but with quite pronounced, assay-dependent variation between studies, and no or only weak correlation with CSF T-tau levels (29-31). With regard to P-tau, a semi-sensitive assay for tau phosphorylated at threonine 181 (similar to the most employed CSF test) with electrochemiluminescence detection has been developed (32). Using this assay, plasma P-tau concentration was higher in AD dementia patients than controls. Plasma P-tau concentration was associated with both Aβ and tau PET and more AD-associated than the corresponding plasma T-tau test, which are promising results in need of replication. While conventional plasma measures of Aβ42 and Aβ40 by ELISA do not show a consistent change in clinically diagnosed AD cases as compared with cognitively unimpaired elderly (29), recent studies of blood Aβ using single molecule array (Simoa) or mass spectrometry have shown a relationship between blood levels of Aβ 40/42 ratios and the brain burden of Aβ (33-35). NfL indicates axonal damage and can also be measured in blood (36). Blood NfL shows particular promise as a biomarker of neurodegeneration in AD (37, 38) but high levels are also found in many other disorders characterized by neurodegeneration (39, 40). Given that NfL is a general neurodegeneration marker and not specifically involved in AD pathophysiology, it may give more unbiased information than tau biomarkers in clinical trials. Furthermore, the correlation between CSF and blood levels of NfL is very high (36), which is not the case for blood measures of tau (30). Synaptic proteins, including dendritic protein neurogranin and the pre-synaptic growth-associated protein 43 (GAP-43), show marked increases in CSF and are seemingly specific for AD (41, 42). Emerging CSF biomarkers including neuron-specific enolase (NSE), visinin-like protein 1 (VLP-1), heart fatty acid binding protein (HFPAP), and YKL-40 (a marker of glial activation) show moderate associations with AD (29, 43).
CSF tau comprises many different tau fragments that reflect processing of secreted tau, and some of these fragments may prove to be useful diagnostically (44) or provide information about tau kinetics in neurons (45). New assays are being developed to measure additional endogenous tau fragments that may correlate with tau pathology. For example, one of these tau fragments, tau368, results from cleavage of tau by asparagine endopeptidase (AEP) at position 368. The result of this is tau hyperphosphorylation, impaired microtubule assembly, and aggregation of truncated tau in neurofibrillary tangles (46).  Inhibiting AEP may represent a novel therapeutic strategy for neurodegenerative disease (47).  Tau368 can be measured in CSF and a first small study shows an association with longitudinal increase in tau PET tracer retention (48). Further, mass spectrometry studies show that CSF tau is specifically cleaved to a mid-domain fragment between amino acids 222-225 (45). Using an assay based on an end-specific tau x-224 monoclonal antibody, increased CSF levels were found in AD, while tau224 levels were low in other tauopathies (49). Exosomal tau has been evaluated as a biomarker but the studies have not been replicated and it is presently not possible to draw any conclusion on whether or not exosomal tau is a biomarker for AD.
The varying measures of tau report on different aspects of AD biology. In the amyloid, tau, neurodegeneration (ATN) Framework for AD diagnosis (50), tau PET and CSF p-tau are viewed as reporters of the presence and spread of tau pathology, whereas CSF t-tau, fluorodeoxyglucose PET, and MRI atrophy are seen as reporters of neurodegeneration. Recent evidence suggests that the soluble forms of tau are increased in production with greater amyloid plaque burden (45), while aggregated forms of tau appear at later stages of AD pathophysiology, closer to symptom onset. Tau markers — tau PET, p-tau, t-tau — measure different aspects of AD from this perspective.
For use in clinical trials of anti-tau agents, CSF biomarkers of amyloid and tau are needed to provide evidence of target engagement, enable enrichment of trials with appropriate participants, and show downstream effects of treatment (51). Lowering of CSF p-tau may suggest an effect on tau phosphorylation; however, more studies are needed to evaluate how CSF p-tau relates to brain pathology. Biomarker studies in recent clinical trials of the anti-amyloid antibodies bapineuzumab, gantenerumab, and BAN-2401 suggest that declines in CSF p-tau, t-tau, neurogranin and NfL indicate a downstream effect of Aβ immunotherapy on neurodegeneration, tau pathology, and synaptic degeneration (52-54).
Fully automated CSF immunoassays of AD biomarkers are now available, and in a study comparing fully automated CSF immunoassay with amyloid PET imaging, a multinational group of investigators found that the CSF tau/Aβ ratio was as accurate as amyloid PET in predicting clinical progression among patients with MCI (55) .


Challenges and unanswered questions

While the development of tau-targeted therapies is seen by many in the AD research community as one of the highest priority efforts, the complexity of tau protein processing gives rise to many challenges that have slowed development of tau-based therapies (56). Among the questions raised by the Task Force were these:
•    What is known about the normal physiological function of tau, and are there potential negative/untoward consequences of reducing tau?
•    What is the effect of tau suppression on spatiotemporal deposition of tau?
•    What degree of tau lowering should be targeted to achieve an optimal therapeutic effect?
•    What other factors may contribute to tau-based neurodegeneration (e.g., inflammation, aging, or vascular factors?)
•    What is the relationship of amyloid-beta and tau?
•    What is the relationship of soluble forms of tau and aggregated tau deposits?
•    What is the role of microglia activation in the development of tau pathology?
•    Since most tau is intracellular, will targeting it extracellularly be sufficient; or is there a window of time during which limiting extracellular tau would show a treatment benefit?
•    What happens downstream when an antibody binds to tau? Is it sequestered or disposed of through cellular mechanisms or the glymphatic system 57 and does this result in downstream preservation of neurons?
•    What are the best tau epitopes or tau fragments to target?
•    Which tau fragments correlate best with AD-type neurodegeneration in CSF or in plasma?
•    Which p-tau variants in CSF or blood correlate best with tau pathology in AD, or can differentiate AD from other tauopathies?
•    Are there differential rates of change in tau deposition across the anatomy?
•    What regions should tau PET target to demonstrate target engagement, and how should tau PET be developed for use in clinical trials to predict treatment response or measure treatment effect?
•    What will be required to make tau PET useful clinically for diagnosis, prognosis, or prediction of treatment response?
•    Do trials for anti-tau agents require similar structures as for Aβ-targeting agents even though the dynamics of the protein are different?
•    What is the best population, taking into account the ATN stage, to target?
•    Should anti-tau clinical trials focus on subpopulations and if so, which subpopulations?
•    Would the best path forward for anti-tau agents be to test them in combination trials with Aβ-targeting agents or drugs that target other pathologies such as neuroinflammation?
•    How can tau-PET be used to stage AD?
•    What are the best tau-related outcomes for AD trials?



Anti-tau therapies are beginning to populate the AD drug development pipeline, mostly in Phase 1 and Phase 2 trials. However, anti-tau treatments have not yet shown evidence of a treatment effect in patients. The Task Force concluded that the development of anti-tau treatment will be determined by multiple trials and will require contributions from industry, academia, and advocacy groups.
The Task Force also called for incorporating CSF tau measures in all anti-tau trials. At a later date, tau PET may also be a viable option. For a biomarker to accurately assess target engagement and for pharmacodynamic studies, assays need to be designed specifically for the therapeutic antibody in addition to general tau-based assays. Such assays would enable exploration of whether a change in a specific tau species indicates that the therapeutic antibody binds tau in the brain parenchyma and if bound tau is secreted into the CSF.
Most Task Force members agreed that anti-tau trials are justified because AD symptoms are likely driven by the spread of tau and its degenerative effects, as well as by amyloid. However, most members also agreed that the specific tau-based mechanisms that will likely provide a treatment effect from anti-tau therapy are unclear and that significant observational and trial related studies will help better inform which tau targets will be most effective.


Acknowledgements: The authors thank Lisa J. Bain for assistance in the preparation of this manuscript.

Conflicts of interest: The Task Force was partially funded by registration fees from industrial participants. These corporations placed no restrictions on this work. Dr. Cummings is the Chief Scientific Officer of CNS Innovations. He acknowledges funding from the National Institute of General Medical Sciences (Grant: P20GM109025) and support from Keep Memory Alive; Consultation for Pharmaceutical Companies: Dr. Cummings has provided consultation to Acadia, Accera, Actinogen, AgeneBio, Alkahest, Allergan, Alzheon, Avanir, Axsome, Binomics, BiOasis Technologies, Biogen, Bracket, Denali, Diadem, EIP Pharma, Eisai, Genentech, Green Valley, Grifols, Hisun, Idorsia, Lundbeck, MedAvante, Merck, Otsuka, Pain Therapeutics, Probiodrug, Proclara, QR, Resverlogix, Roche, Samumed, Shinkei Therapeutics, Sunovion, Suven, Takeda, and United Neuroscience pharmaceutical and assessment companies. Consultation for Foundations: Dr. Cummings has provided consultation to Global Alzheimer Platform (GAP). Stock: Dr. Cummings owns stock in ADAMAS, BioAsis, Prana, MedAvante, Neurokos, and QR Pharma. Board member: None. Speaker/lecturer: None. Other: Dr. Cummings owns the copyright of the Neuropsychiatric Inventory (NPI). Dr. Cummings is the Chief Scientific Officer of CNS Innovations.Expert witness/legal consultation: None. NIH support: COBRE grant # P20GM109025; TRC-PAD # R01AG053798; DIAGNOSE CTE # U01NS093334.Research Support: None. Spousal ownership or significant financial interest in a relevant company: CNS Innovations. Dr. Johnson has consulted for Merck, Eli Lilly, Novartis, Biogen, Takeda, Roche, Eisai, Piramal, and GE. Dr. Keeley reports that he is an employee of Genentech. Dr. Bateman reports grants from BrightFocus Foundation, Pharma Consortium (Abbvie, AstraZeneca, Biogen, Eisai, Eli Lilly and Co., Hoffman La-Roche Inc., Janssen, Pfizer, Sanofi-Aventi),  the Tau SILK/PET Consortium (Biogen/Abbvie/Lilly), Association for Frontotemporal Degeneration FTD Biomarkers Initiative, Anonymous Foundation, GHR Foundation, NIH, Alzheimer’s Association, Lilly, Rainwater Foundation Tau Consortium, and Cure Alzheimer’s Fund, grants, personal fees and non-financial support from Roche and Janssen, personal fees and non-financial support from Pfizer, Eisai, and Merck, and non-financial support from Avid Radiopharmaceuticals outside the submitted work. Washington University, Dr. Bateman, and David Holtzman have equity ownership interest in C2N Diagnostics and receive royalty income based on technology (stable isotope labeling kinetics and blood plasma assay) licensed by Washington University to C2N Diagnostics. RJB receives income from C2N Diagnostics for serving on the scientific advisory board. Washington University, with RJB as co-inventor, has submitted the US nonprovisional patent application “Methods for Measuring the Metabolism of CNS Derived Biomolecules In Vivo” and provisional patent application “Plasma Based Methods for Detecting CNS Amyloid Deposition”. Dr. Molinuevo reports personal fees from Alergan, from Oryzon, from Genentech, from Novartis, from Lundbeck, from Biogen, from Lilly, from Janssen, Green Valley, from MSD, from Eisai, from Alector and from Raman Health,  outside the submitted work. Dr. Touchon has nothing to disclose. Dr. Aisen reports grants from Lilly, personal fees from Proclara, other from Lilly, other from Janssen, other from Eisai, grants from Janssen, grants from NIA, grants from FNIH, grants from Alzheimer’s Association, personal fees from Merck, personal fees from Roche, personal fees from Lundbeck, personal fees from Biogen, personal fees from ImmunoBrain Checkpoint,  outside the submitted work. Dr. Vellas reports grants from Lilly, Merck, Roche, Lundbeck, Biogen, grants from Alzheimer’s Association, European Commission, personal fees from Lilly, Merck, Roche, Biogen,  outside the submitted work.



1.    Cummings JL, Morstorf T, Zhong K. Alzheimer’s disease drug-development pipeline: few candidates, frequent failures. Alzheimers Res Ther 2014;6:37.
2.    Alzheimer’s Association. 2018 Alzheimer’s disease facts and figures. Alzheimers Dement 2018;14:367-429.
3.    Alzheimer’s Disease International. World Alzheimer’s Report 2018. The state of the art of dementia research: New frontiers. London2018.
4.    Sperling RA, Jack CR, Jr., Aisen PS. Testing the right target and right drug at the right stage. Sci Transl Med 2011;3:111cm133.
5.    Vellas B, Sampaio C, Bateman R, et al. EU./U.S. CTAD Task Force on Alzheimer’s Trial Populations. J Prev Alzheimers Dis 2014;1:110-116.
6.    Ousset PJ, Cummings J, Delrieu J, et al. Is Alzheimer’s Disease Drug Development Broken? What Must Be Improved. J Prev Alzheimers Dis 2014;1:40-45.
7.    Aisen PS. Continuing Progress in Alzheimer’s Disease Trials: Cause for Optimism. J Prev Alzheimers Dis 2017;4:211-212.
8.    Cummings J, Lee G, Ritter A, Zhong K. Alzheimer’s disease drug development pipeline: 2018. Alzheimers Dement (N Y) 2018;4:195-214.
9.    Serrano-Pozo A, Frosch MP, Masliah E, Hyman BT. Neuropathological alterations in Alzheimer disease. Cold Spring Harb Perspect Med 2011;1:a006189.
10.    Cardenas-Aguayo Mdel C, Gomez-Virgilio L, DeRosa S, Meraz-Rios MA. The role of tau oligomers in the onset of Alzheimer’s disease neuropathology. ACS Chem Neurosci 2014;5:1178-1191.
11.    Holtzman DM, Carrillo MC, Hendrix JA, et al. Tau: From research to clinical development. Alzheimers Dement 2016;12:1033-1039.
12.    Guo T, Noble W, Hanger DP. Roles of tau protein in health and disease. Acta Neuropathol 2017;133:665-704.
13.    Morris M, Maeda S, Vossel K, Mucke L. The many faces of tau. Neuron 2011;70:410-426.
14.    Mudher A, Colin M, Dujardin S, et al. What is the evidence that tau pathology spreads through prion-like propagation? Acta Neuropathol Commun 2017;5:99.
15.    Leyns CEG, Holtzman DM. Glial contributions to neurodegeneration in tauopathies. Mol Neurodegener 2017;12:50.
16.    Braak H, Del Tredici K. Spreading of Tau Pathology in Sporadic Alzheimer’s Disease Along Cortico-cortical Top-Down Connections. Cereb Cortex 2018;28:3372-3384.
17.    Schwarz AJ, Yu P, Miller BB, et al. Regional profiles of the candidate tau PET ligand 18F-AV-1451 recapitulate key features of Braak histopathological stages. Brain 2016;139:1539-1550.
18.    van Rossum IA, Vos SJ, Burns L, et al. Injury markers predict time to dementia in subjects with MCI and amyloid pathology. Neurology 2012;79:1809-1816.
19.    Roe CM, Fagan AM, Grant EA, et al. Amyloid imaging and CSF biomarkers in predicting cognitive impairment up to 7.5 years later. Neurology 2013;80:1784-1791.
20.    Ossenkoppele R, Schonhaut DR, Scholl M, et al. Tau PET patterns mirror clinical and neuroanatomical variability in Alzheimer’s disease. Brain 2016;139:1551-1567.
21.    Lee SH, Le Pichon CE, Adolfsson O, et al. Antibody-Mediated Targeting of Tau In Vivo Does Not Require Effector Function and Microglial Engagement. Cell Rep 2016;16:1690-1700.
22.    West T, Hu Y, Verghese PB, et al. Preclinical and Clinical Development of ABBV-8E12, a Humanized Anti-Tau Antibody, for Treatment of Alzheimer’s Disease and Other Tauopathies. J Prev Alzheimers Dis 2017;4:236-241.
23.    Melis V, Magbagbeolu M, Rickard JE, et al. Effects of oxidized and reduced forms of methylthioninium in two transgenic mouse tauopathy models. Behav Pharmacol 2015;26:353-368.
24.    Gauthier S, Feldman HH, Schneider LS, et al. Efficacy and safety of tau-aggregation inhibitor therapy in patients with mild or moderate Alzheimer’s disease: a randomised, controlled, double-blind, parallel-arm, phase 3 trial. Lancet 2016;388:2873-2884.
25.    Gueroux M, Fleau C, Slozeck M, Laguerre M, Pianet I. Epigallocatechin 3-Gallate as an Inhibitor of Tau Phosphorylation and Aggregation: A Molecular and Structural Insight. J Prev Alzheimers Dis 2017;4:218-225.
26.    Hyman BT, Phelps CH, Beach TG, et al. National Institute on Aging-Alzheimer’s Association guidelines for the neuropathologic assessment of Alzheimer’s disease. Alzheimers Dement 2012;8:1-13.
27.    Jack CR, Jr., Wiste HJ, Weigand SD, et al. Age-specific and sex-specific prevalence of cerebral beta-amyloidosis, tauopathy, and neurodegeneration in cognitively unimpaired individuals aged 50-95 years: a cross-sectional study. Lancet Neurol 2017;16:435-444.
28.    Sepulcre J, Grothe MJ, Sabuncu M, et al. Hierarchical Organization of Tau and Amyloid Deposits in the Cerebral Cortex. JAMA Neurol 2017;74:813-820.
29.    Olsson B, Lautner R, Andreasson U, et al. CSF and blood biomarkers for the diagnosis of Alzheimer’s disease: a systematic review and meta-analysis. Lancet Neurol 2016;15:673-684.
30.    Mattsson N, Zetterberg H, Janelidze S, et al. Plasma tau in Alzheimer disease. Neurology 2016;87:1827-1835.
31.    Zetterberg H, Wilson D, Andreasson U, et al. Plasma tau levels in Alzheimer’s disease. Alzheimers Res Ther 2013;5:9.
32.    Mielke MM, Hagen CE, Xu J, et al. Plasma phospho-tau181 increases with Alzheimer’s disease clinical severity and is associated with tau- and amyloid-positron emission tomography. Alzheimers Dement 2018;14:989-997.
33.    Nakamura A, Kaneko N, Villemagne VL, et al. High performance plasma amyloid-beta biomarkers for Alzheimer’s disease. Nature 2018;554:249-254.
34.    Ovod V, Ramsey KN, Mawuenyega KG, et al. Amyloid beta concentrations and stable isotope labeling kinetics of human plasma specific to central nervous system amyloidosis. Alzheimers Dement 2017;13:841-849.
35.    Janelidze S, Stomrud E, Palmqvist S, et al. Plasma beta-amyloid in Alzheimer’s disease and vascular disease. Sci Rep 2016;6:26801.
36.    Gisslen M, Price RW, Andreasson U, et al. Plasma Concentration of the Neurofilament Light Protein (NFL) is a Biomarker of CNS Injury in HIV Infection: A Cross-Sectional Study. EBioMedicine 2016;3:135-140.
37.    Lewczuk P, Ermann N, Andreasson U, et al. Plasma neurofilament light as a potential biomarker of neurodegeneration in Alzheimer’s disease. Alzheimers Res Ther 2018;10:71.
38.    Zetterberg H, Skillback T, Mattsson N, et al. Association of Cerebrospinal Fluid Neurofilament Light Concentration With Alzheimer Disease Progression. JAMA Neurol 2016;73:60-67.
39.    Bacioglu M, Maia LF, Preische O, et al. Neurofilament Light Chain in Blood and CSF as Marker of Disease Progression in Mouse Models and in Neurodegenerative Diseases. Neuron 2016;91:56-66.
40.    Hansson O, Janelidze S, Hall S, et al. Blood-based NfL: A biomarker for differential diagnosis of parkinsonian disorder. Neurology 2017;88:930-937.
41.    Portelius E, Olsson B, Hoglund K, et al. Cerebrospinal fluid neurogranin concentration in neurodegeneration: relation to clinical phenotypes and neuropathology. Acta Neuropathol 2018;136:363-376.
42.    Sandelius A, Portelius E, Kallen A, et al. Elevated CSF GAP-43 is Alzheimer’s disease specific and associated with tau and amyloid pathology. Alzheimers Dement 2019;15:55-64.
43.    Molinuevo JL, Ayton S, Batrla R, et al. Current state of Alzheimer’s fluid biomarkers. Acta Neuropathol 2018;136:821-853.
44.    Meredith JE, Jr., Sankaranarayanan S, Guss V, et al. Characterization of novel CSF Tau and ptau biomarkers for Alzheimer’s disease. PLoS One 2013;8:e76523.
45.    Sato C, Barthelemy NR, Mawuenyega KG, et al. Tau Kinetics in Neurons and the Human Central Nervous System. Neuron 2018;97:1284-1298 e1287.
46.    Zhang Z, Song M, Liu X, et al. Cleavage of tau by asparagine endopeptidase mediates the neurofibrillary pathology in Alzheimer’s disease. Nat Med 2014;20:1254-1262.
47.    Zhang Z, Kang SS, Liu X, et al. Asparagine endopeptidase cleaves alpha-synuclein and mediates pathologic activities in Parkinson’s disease. Nat Struct Mol Biol 2017;24:632-642.
48.    Leuzy A, Cicognola C, Chiotis K, et al. Longitudinal tau and metabolic PET imaging in relation to novel CSF tau measures in Alzheimer’s disease. Eur J Nucl Med Mol Imaging 2019.
49.    Cicognola C, Brinkmalm G, Wahlgren J, et al. Novel tau fragments in cerebrospinal fluid: relation to tangle pathology and cognitive decline in Alzheimer’s disease. Acta Neuropathol 2019;137:279-296.
50.    Jack CR, Jr., Bennett DA, Blennow K, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 2018;14:535-562.
51.    Blennow K, Zetterberg H. Amyloid and Tau Biomarkers in CSF. J Prev Alzheimers Dis 2015;2:46-50.
52.    Blennow K, Zetterberg H, Rinne JO, et al. Effect of immunotherapy with bapineuzumab on cerebrospinal fluid biomarker levels in patients with mild to moderate Alzheimer disease. Arch Neurol 2012;69:1002-1010.
53.    Ostrowitzki S, Lasser RA, Dorflinger E, et al. A phase III randomized trial of gantenerumab in prodromal Alzheimer’s disease. Alzheimers Res Ther 2017;9:95.
54.    Salloway S, Sperling R, Fox NC, et al. Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer’s disease. N Engl J Med 2014;370:322-333.
55.    Hansson O, Seibyl J, Stomrud E, et al. CSF biomarkers of Alzheimer’s disease concord with amyloid-beta PET and predict clinical progression: A study of fully automated immunoassays in BioFINDER and ADNI cohorts. Alzheimers Dement 2018;14:1470-1481.
56.    Khanna MR, Kovalevich J, Lee VM, Trojanowski JQ, Brunden KR. Therapeutic strategies for the treatment of tauopathies: Hopes and challenges. Alzheimers Dement 2016;12:1051-1065.
57.    Rasmussen MK, Mestre H, Nedergaard M. The glymphatic pathway in neurological disorders. Lancet Neurol 2018;17:1016-1024.



P. Novak1, N. Zilka2, M. Zilkova2, B. Kovacech2, R. Skrabana2, M. Ondrus1, L. Fialova2, E. Kontsekova2, M. Otto3, M. Novak4


1. AXON Neuroscience CRM Services SE, Dvorakovo nabrezie 10, 811 02 Bratislava, Slovakia; 2. AXON Neuroscience R&D Services SE, Dvorakovo nabrezie 10, 811 02 Bratislava, Slovakia; 3. Ulm University, Department of Neurology, Oberer Eselsberg 45, 89081 Ulm, Germany; 4. AXON Neuroscience SE, Arch. Makariou & Kalogreon 4, 6016 Larnaca, Cyprus

Corresponding Author: P. Novak, Axon Neuroscience CRM Services SE, Slovakia, +421911187237,

J Prev Alz Dis 2019;1(6):63-69
Published online December 14, 2018,



Neurofibrillary tau protein pathology is closely associated with the progression and phenotype of cognitive decline in Alzheimer’s disease and other tauopathies, and a high-priority target for disease-modifying therapies. Herein, we provide an overview of the development of AADvac1, an active immunotherapy against tau pathology, and tau epitopes that are potential targets for immunotherapy. The vaccine leads to the production of antibodies that target conformational epitopes in the microtubule-binding region of tau, with the aim to prevent tau aggregation and spreading of pathology, and promote tau clearance. The therapeutic potential of the vaccine was evaluated in transgenic rats and mice expressing truncated, non mutant tau protein, which faithfully replicate of human tau pathology. Treatment with AADvac1 resulted in reduction of neurofibrillary pathology and insoluble tau in their brains, and amelioration of their deleterious phenotype. The vaccine was highly immunogenic in humans, inducing production of IgG antibodies against the tau peptide in 29/30 treated elderly patients with mild-to-moderate Alzheimer’s. These antibodies were able to recognise insoluble tau proteins in Alzheimer patients’ brains. Treatment with AADvac1 proved to be remarkably safe, with injection site reactions being the only adverse event tied to treatment. AADvac1 is currently being investigated in a phase 2 study in Alzheimer’s disease, and a phase 1 study in non-fluent primary progressive aphasia, a neurodegenerative disorder with a high tau pathology component.

Key words: tau, Alzheimer’s disease, tauopathy, immunotherapy, clinical trial.



Neurofibrillary pathology, composed of misfolded tau protein, is a hallmark of Alzheimer’s disease (AD) and a range of non-AD tauopathies (1-4). Long neglected by a large portion of the AD field and seen as a by-product of amyloid β pathology, neurofibrillary tau pathology has been established as the closest correlate of cortical atrophy (5) and clinical progression in AD (6-8). Pathological changes on tau protein became a tempting target for disease-modifying therapy, with the first tau-targeted immunotherapy entering clinical development in 2013 (9). The steps taken along the way were recently summarised by Iqbal et al. (10) and the current state of drug development thoroughly reviewed by Li & Götz (11).
The function of microtubule-associated protein tau is regulated by alternative splicing, localisation, oligomerisation, phosphorylation and other posttranslational modifications, accompanied by conformational change (10). Tau is a vital component of the cytoskeleton and a multifunctional molecule involved in dynamics of neuronal microtubules, neuronal polarisation and synapse formation (12). Recent research indicates that it possesses diverse other functions, such as ribosomal DNA transcription, nucleolar transcriptional regulation, RNA metabolism and brain insulin signalling, morphological and synaptic maturation of new-born hippocampal granule neurons, and trafficking of cellular components (13, 14). In disease, tau protein undergoes a transformation that leads to the loss of vital functions and a toxic gain of function including microtubule mis-assembly, cytoskeleton disruption, mitochondrial impairment, oxidative stress, DNA damage, and neuroinflammation (11, 12, 15-18).
The post-translational modifications of tau protein are legion, with phosphorylation and truncation being specifically associated with pathogenesis of Alzheimer’s disease and other tauopathies (19-21). Specifically, it was shown that phosphorylation and truncation lead to enhancement of abnormal tau-tau interaction and tau oligomerisation in vitro and in vivo, generating precursors of oligomeric pathological tau (22-24). These modifications bestow novel conformations and epitope structure upon pathological moieties of tau (25, 26), which in turn make them immunologically distinguishable from healthy tau proteins, and thus an ideal target for immunotherapy. Tau epitopes in disease can arise due to aberrant phosphorylation (20), truncation (27), spontaneous conformational switch and/or template-assisted misfolding (28, 29). A crucial aspect is the distribution of these neo-epitopes among pathological tau species. As tau aggregation in AD takes place via the microtubule-binding repeat domain (30), this domain will be present in all pathological tau oligomers, and the neo-epitopes therein are likely to be conserved in the course of the disease. Since tau in the oligomers is frequently truncated in AD (21), neo-epitopes at the molecule’s termini are cleaved off in a subset of pathological tau forms. Such species become untargetable by therapies aimed at the termini (29). It is known that tau hyperphosphorylation at individual epitopes waxes and wanes over the course of the disease (31), making targeting of certain individual phospho-sites less effective.
Tau aggregates have the intriguing ability to spread from neurons affected by tau pathology to their healthy neighbours and perpetuate neurofibrillary degeneration therein, behaving in essence in a prion like fashion (32). These particles named «tauons», were first proposed in 1994 (33) and shown experimentally in 2009 (34). Locally, the spreading of tau aggregates can occur via diffusion based on proximity, while the spreading pattern to distant brain regions is based on neuronal connectivity (35, 36). The common immunological denominators of these propagating tau species are logical drug targets, and preventing the spread of tauons will halt neurofibrillary pathology and the progression of neurodegeneration in the patients’ brains.


Animal model design

An essential step in the development of disease-modifying therapies is the development of a suitable model that recapitulates biochemical features of human tau pathology faithfully. Only such a model has a high predictive value, making it more likely that an investigational medicinal product will perform as well in humans as it did in animals. The development of models for tau pathology has often taken a shortcut via using mutant tau (22, 37). However, no tau mutations have been identified in Alzheimer’s disease so far (4), and it is known that mutant tau filaments are conformationally different from healthy tau (28). In order to prepare a genuine model of Alzheimer tau pathology, we have instead opted for transgenes conferring expression of truncated 3R and 4R tau derived from sporadic Alzheimer’s disease (38). When expressed in brains of rats and mice, the truncated tau proteins induce extensive neurofibrillary degeneration that fulfils the criteria for human NFT pathology (thioflavin-S reactive, Congo-red birefringent, argyrophilic). Sarkosyl insoluble tau in these animals is composed of both endogenous and transgenic tau, and featuring both low- and high-molecular-weight filament and oligomeric tau species and multiple truncated forms. The pathology is accompanied by neuroinflammation, oxidative stress, synaptic abnormalities, and progressive neuronal dysfunction resulting in neurobehavioural impairment and death of the animals, with lifespan depending on transgene expression level (39-46).

Figure 1. Rat models expressing truncated tau 151-391/4R faithfully recapitulate human neurofibrillary tau pathology

Figure 1. Rat models expressing truncated tau 151-391/4R faithfully recapitulate human neurofibrillary tau pathology

Neurofibrillary tangles produced by transgenic animals were A) argyrophilic, B) Congo-red positive, C) thioflavin S reactive, and D) AT8-reactive, as one would expect of human NFTs. The animals displayed neuroinflammation, with increase in microglial numbers, altered morphology, and shift towards a phagocytic morphology (arrows) (E, F).


Preclinical development of AADvac1

In our efforts to identify a functionally important common denominator of pathological tau protein, we have initially generated a panel of anti-tau antibodies, including antibody DC8E8, which displayed a range of highly desirable traits. The DC8E8 antibody recognises three or four individual epitopes in the microtubule-binding region of 3R- and 4R-tau protein, respectively. The recognition is phosphorylation-independent. Importantly, the accessibility of the epitopes is highly increased in truncated tau, which translates into a pronounced preference of DC8E8 for the precursors of pathological tau over physiological tau. The antibody was able to recognise both early (pre tangle), intermediate (intracellular tangle) and late (ghost tangle) manifestations of tau pathology, with no off-target binding in a cross reactivity study in a range of normal human tissues. In Western blot, it recognised the entire ladder of pathological tau protein moieties extracted from AD brains (47). Functionally, DC8E8 was found to inhibit tau aggregation in vitro, probably by action of markedly flexible CDRH3 and CDRL1 loops (47). We studied in details the DC8E8 binding site topology by X-ray crystallography. DC8E8 possesses a 10 Å-deep binding pocket, extending over 18 × 14 Å of surface. The shape of this pocket indicates that the DC8E8 epitope on tau 299HVPGGG304 adopts a fold protruding into this space to bind in the DC8E8 combining site, creating a 180° turn on the tau chain (47).
It is worthy of note that several potential antibody candidates that, while highly specific for pathological, AD-derived forms of tau, have promoted tau aggregation instead of inhibiting it (47), highlighting the need to thoroughly evaluate the properties of any promising tau immunotherapy agents.
Subsequently, DC8E8 was used as a template to design an immunogen that would stimulate the production of antibodies with DC8E8-like properties. The peptide 294KDNIKHVPGGGS305, comprising one of the DC8E8 epitopes, was found to fulfil these requirements. Interestingly, an X-ray crystallography study (47) of the DC8E8 binding site suggests that the peptide forms a sharply protruding turn even in the unbound state in solution. Thus, AADvac1 contains the peptide hapten with a turn motif at the immunologically dominant residues, which promotes production of antibodies with DC8E8-like binding properties.
This peptide hapten was coupled to keyhole limpet haemocyanin (KLH) carrier and formulated with aluminium hydroxide adjuvant to yield the highly immunogenic vaccine AADvac1. The carrier serves to provide necessary T-cell epitopes without eliciting a T-cell response against tau, thus circumventing the problems seen with early anti-amyloid-β immunotherapy (i.e. the meningoencephalitis observed with AN1792) (48). Active anti-amyloid immunotherapies have similarly used hapten-carrier conjugates to address the same problem (49).
Similarly to what was observed with DC8E8 treatment in mice (22, 47), administration of AADvac1 to transgenic rats expressing truncated tau proved to be highly efficacious. The vaccine induced high IgG1-dominated antibody titres; no tau-directed T-cell response was observed. The amount of neurofibrillary tangles in the rodents’ brains was reduced (see Figure 2); sarcosyl-insoluble tau in the animals’ brains was reduced by ~70%, and pathological phospho-tau species by up to 95% (see Figure 3). This highlights the fact that while AADvac1 targets a conformational epitope, it also affects phospho-tau species. The proposed mechanism behind this is effect the antibody-mediated elimination of abnormal tau seeds that promote the spreading of tau pathology. Thus, the substrate for kinases – full length, truncated and oligomerised tau molecules – are removed, which results in reduced amount of hyperphosphorylated tau. Along with reduction of neurofibrillary pathology, the neurobehavioural phenotype of the animals was similarly improved (9).

Figure 2. Active vaccination reduced the number of transgenic rats developing extensive neurofibrillary pathology

Figure 2. Active vaccination reduced the number of transgenic rats developing extensive neurofibrillary pathology

Immunostaining with AT8, pT212 and pS214 shows low numbers of neurofibrillary tangles in the brainstem of treated transgenic rats (B), (E) and (H) compared with untreated transgenic rats (A), (D) and (G). Immunisation lowered the number of transgenic rats with extensive neurofibrillary degeneration by 55% (C) and (F) or by 77% (I). Modified from [9] originally published by BioMed Central, licensed under CC BY 4.0.

Figure 3. Immunisation with AADvac1 vaccine reduced tau oligomers and tau hyperphosphorylation

Figure 3. Immunisation with AADvac1 vaccine reduced tau oligomers and tau hyperphosphorylation

Western blot analysis with pan-tau monoclonal antibody DC25 showed reduction in oligomeric tau in the brain of transgenic rats treated with tau peptide vaccine (A). The monomeric endogenous rat tau proteins run between 43 and 68 kDa marker bands, whereas monomeric transgenic tau comprises multiple phospho-species between 29 and 43 kDa marker bands. In the vaccine-treated animals, there are only remnants of non phosphorylated transgene running just above the 29 kDa marker band. Western blot analysis revealed significant reduction of hyperphosphorylated tau species phosphorylated at Thr217 (monoclonal antibody (mAb) DC217) (B), pThr231 (mAb DC209) (C), pSer202/pThr205 (mAb AT8) (D) and pThr181 (mAb DC179) (E). (F) The graph represents the quantification and statistical evaluation of the difference between animals treated with vaccine and those treated with adjuvant only; *P < 0.05, **P < 0.01. Modified from (9) originally published by BioMed Central, licensed under CC BY 4.0.


Finally, GLP toxicology studies of AADvac1 were conducted in mice, rats, dogs, and rabbits, with the vaccine being well-tolerated at all administered dose levels in all tested species throughout the course of all studies. Single-dose toxicity studies were carried out in Wistar rats, with doses of up to 160 μg Axon Peptide 108 (coupled to KLH). Three individual chronic toxicity studies were conducted in rabbits, with doses of up to 200 μg Axon Peptide 108 (coupled to KLH) per dose, and up to 12 individual doses being administered over the course of 34 weeks. CNS safety pharmacology studies were carried out in mice (Irwin screen test); cardiorespiratory safety pharmacology studies were carried out in Beagle dogs.


Clinical development

Phase 1 (AD)

The first-in man clinical study of AADvac1 was initiated in 2013 and concluded in 2015 (50). As AADvac1 was the first tau-targeting immunotherapy investigated in humans, the enrolment of the initial 8 patients was performed in a stepwise manner. These patients were observed for at least 3 months before recruiting the remaining 22 patients. The trials’ primary purpose, safety assessment, provided encouraging results: the only adverse events clearly tied to treatment were (mostly mild) reversible injection site reactions.
The study’s duration was 24 weeks. Patients were administered 6 individual doses of AADvac1 (40 μg Axon Peptide 108 (coupled to KLH) with aluminium hydroxide (containing 0·5 mg Al³+) in a phosphate buffer volume of 0·3 mL) in 4 week intervals; patients allocated to placebo have received 3 doses of placebo, followed by a switch to AADvac1 treatment.
The study’s 18 month follow-up trial, conducted to assess the persistence of the antibody response, response to booster doses, and long-term effects of AADvac1 on safety, MRI volumetry, and cognition was concluded in 2016 (data are not published yet). Two booster doses of AADvac1 at 6-month intervals were administered over the course of the follow-up; including the initial vaccination regimen, patients have received a total of 8 AADvac1 doses over the course of 96 weeks.
One hurdle that any active immunotherapy for AD has to overcome is the generally reduced immune fitness of the elderly population (51). In this regard, the results were highly encouraging, since the peptide-KLH conjugate in AADvac1 was able to induce the desired IgG antibody response against the tau peptide in 29 out of 30 patients enrolled in the study. IgG titres continued to increase over the initial 6 dose regimen.
The antibodies generated by the patients’ immune systems were able to target also recombinant misdisordered truncated tau protein aa151-391/4R, and tau pathology in brain samples from AD patients, as shown by Western blotting).

Figure 4. IgG antibody response against the tau peptide component of AADvac1 continues to rise over the initial 6 dose regimen

Figure 4. IgG antibody response against the tau peptide component of AADvac1 continues to rise over the initial 6 dose regimen

The assay’s lower limit of quantification at a titre of 100 stands in for «no response». Upper limit of quantification at the titre of 204800. Non-responder indicated by red rectangle. Error bars denote geometric mean and 95% CI.


Following the initial 6-dose treatment regimen, the IgG titres in responders ranged from 1:4925 to >1:204800, as a function of the patients’ degree of immune competence (Figure 4). Patients with higher titres generally had high CD4+ lymphocyte counts, and low neutrophil counts. The sole non-responder displayed poor results in haematological assessments (low CD4+ T-helper cell counts, low absolute and relative lymphocyte counts, high absolute and relative neutrophil counts). This variability in the strength of the antibody response needs to be taken into account in any immunogenicity studies of active vaccines in seniors and should be reflected in adequate sample sizes.
Cognition, measured by the ADAS-Cog11, Letter fluency, and Category fluency tests, was stable over the initial 6 months of treatment. Due to the limited sample size of the study and the short duration of observation, this is no hard proof of efficacy, but is compatible with what would be observed with an efficacious compound.

Phase 2 (AD)

Based on the high immunogenicity, good safety profile, and cognitive trends observed in the phase 1 clinical trial, we have initiated the phase 2 study «ADAMANT» in mild Alzheimer’s disease (EudraCT 2015 000630-30, NCT 02579252). The trial is a double-blinded, randomised, placebo controlled, parallel group safety and efficacy study, enrolling patients with biomarker evidence of hippocampal atrophy and/or pathological tau protein and amyloid profiles in the CSF, in an MMSE range of 20-26.

Figure 5. : Sera of selected patients treated with AADvac1 detect high- and low-molecular-weight pathological tau species in all assessed brain extracts

Figure 5. : Sera of selected patients treated with AADvac1 detect high- and low-molecular-weight pathological tau species in all assessed brain extracts

(A) Sera from different AADvac1-treated patients detect the same AD brain extract. (B) Intensity of the staining is proportional to the antibody titre generated by the patients against pathological tau.


The patients will receive 11 vaccinations over the period of two years. In comparison to the phase 1 study and its follow up, the trial features an adapted treatment regimen, with booster doses administered at 3 month intervals (instead of 6 month intervals).
With 208 enrolled patients, the trial is powered to detect a 47% slowing of patient decline, as measured by the CDR-SB, as statistically significant. Beside the CDR-SB, we have implemented also a cognitive test battery that is tailored to the mild dementia population, avoiding issues like ceiling and floor effects that were seen in attempts to repurpose cognitive assessment tools initially intended mainly for moderate dementia for milder AD populations (52-54). The battery is composed of tests with proven utility in mild AD, measuring processing speed, verbal immediate and delayed recall and recognition, as well as visual memory via computerised Cogstate tests, verbal fluency and executive function via the letter fluency and category fluency tests, and finally executive function and working memory via the digit-symbol coding test. This composite is expected to be especially sensitive to cognitive decline at the mild dementia stage.
Extensive MRI measures (volumetry, diffusion tensor imaging, and fMRI) in all subjects, as well as CSF biomarker assessments and FDG-PET evaluation of a subset of study subjects were implemented to detect interaction between AADvac1 treatment and the AD pathophysiological process.
The study is expected to conclude in summer 2019.

Phase 1 (nfvPPA)

Non-fluent primary progressive aphasia (nfvPPA) is characterised by impairment in grammar and motor speech (apraxia of speech and dysarthria), along with predominant atrophy of the left posterior frontal lobe and insula (55). This phenotypic manifestation of the underlying neuronal degeneration is caused, in a majority of nfvPPA cases, by tau pathology, morphologically similar to AD, CBD or PSP (56-58). Therefore, patients with this well defined tauopathy are well suited for treatment with the AADvac1 vaccine.
The ongoing study “AIDA” A 24-month randomised parallel group single-blinded multi-centre phase 1 pilot study of AADvac1 in patients with non fluent primary progressive aphasia (EudraCT 2017-000643-41, NCT 03174886) is primarily a safety and immunogenicity trial, comparing two dosage strengths of AADvac1 in patients with mild-to-moderate nfvPPA. Efficacy is evaluated in an exploratory fashion, though some markers, e.g., neurofilament light chain protein (59, 60), appear to be very sensitive indicators of disease severity in frontotemporal dementias, and are thus potentially theragnostic. The cognitive and functional assessment were tailored to the population under study, seeking to capture impairment in language, behaviour, cognition, and possibly motor function. We have implemented a language-focused custom cognitive test battery (61). Additionally, the Addenbrooke’s Cognitive Examination test (ACE-III) was chosen as an established scale that assesses multiple cognitive domains, and is well-targeted to the patient population. The FTLD version of the Clinical Dementia Rating (Sum of Boxes) is used as a functional assessment; in comparison to the version used in AD, it incorporates also the domains «Language» and  «Behaviour, comportment and personality» (62); everyday functioning assessed by the Amsterdam instrumental activities of living scale (63). Behavioural pathology is assessed in greater depth by the frontal systems behaviour scale (FrSBe). Finally, to assess the entire spectrum of symptoms, the UPDRS Part III is used to evaluate Parkinson-like motor symptoms that can occur with the progression of the disease.
The thorough coverage of potential manifestations of nfvPPA clinical symptoms will allow the selection of most suitable endpoints for later studies.



AADvac1 targets a functionally and structurally outstanding epitope in the microtubule-binding region of tau protein, present once in each of the microtubule-binding repeats. The peptide hapten component of AADvac1 forms a conformational neo-epitope, identical to that present on truncated tau precursors of pathological oligomerised tau. This neo-epitope constitutes a common denominator of all analysed pathological tau species and is present in all assessed AD and non-AD tauopathy brains.
Using truncated tau as found in sporadic AD, we have generated transgenic rats and mice that faithfully replicate human tau pathology. Targeting the abovementioned epitope via passive or active immunisation resulted in reduction of neurofibrillary pathology, and improvement of the neurobehavioural phenotype of both transgenic mice and rats.
Treatment with AADvac1 induced IgG antibodies in 29 of 30 treated AD patients. These antibodies were able to target tau pathology in AD brain tissue slides. The encouraging safety profile observed in the phase 1 study marks AADvac1 as suitable for long term treatment, or even preventive application in the future.
The ADAMANT phase 2 double-blind safety and efficacy study is expected to conclude in mid-2019.
Meanwhile, the development of AADvac1 in non-AD tauopathies has begun with the phase 1 «AIDA» study in nfvPPA patients; results after 1 year of treatment are expected in 2019.
Should AADvac1 prove efficacious, its nature as an active vaccine will naturally lend itself both to the treatment of manifest AD, but especially also to the prevention of dementia, i.e. prevention of the accumulation of pathological tau proteins at the asymptomatic or early symptomatic stages of AD, as described in the FDA’s guidance for early Alzheimer’s disease (draft issued 2018), and in cases with family history of tau pathology (MAPT gene mutation carriers), as well as in carriers of mutations in the amyloid pathway (64).


Conflict of interest: Authors affiliated with AXON Neuroscience or one of its subsidiaries are employees of these companies. The employer of Markus Otto, Univeristy of Ulm, is receiving payments from AXON Neuroscience for the conduct of clinical studies on a per-patient per-visit basis.

Ethical standards: All experiments on animals were carried out according to institutional animal care guidelines conforming to international standards and were approved by the State Veterinary and Food Committee of the Slovak Republic. All human clinical trials are conducted according to the Declaration of Helsinki, and the ICH guidance on good clinical practice. All studies were approved by the responsible ethics committees and competent regulatory authorities. All patients have provided informed consent, and agreed to the publication of trial data.



1.    Grundke-Iqbal, I., et al., Microtubule-associated protein tau. A component of Alzheimer paired helical filaments. J Biol Chem, 1986. 261(13): p. 6084-9.
2.    Wischik, C.M., et al., Isolation of a fragment of tau derived from the core of the paired helical filament of Alzheimer disease. Proc Natl Acad Sci U S A, 1988. 85(12): p. 4506-10.
3.    Poorkaj, P., et al., Tau is a candidate gene for chromosome 17 frontotemporal dementia. Ann Neurol, 1998. 43(6): p. 815-25.
4.    Spillantini, M.G. and M. Goedert, Tau pathology and neurodegeneration. Lancet Neurol, 2013. 12(6): p. 609-22.
5.    Whitwell, J.L., et al., MRI correlates of neurofibrillary tangle pathology at autopsy: a voxel-based morphometry study. Neurology, 2008. 71(10): p. 743-9.
6.    Nelson, P.T., et al., Correlation of Alzheimer disease neuropathologic changes with cognitive status: a review of the literature. J Neuropathol Exp Neurol, 2012. 71(5): p. 362-81.
7.    Murray, M.E., et al., Clinicopathologic and 11C-Pittsburgh compound B implications of Thal amyloid phase across the Alzheimer’s disease spectrum. Brain, 2015. 138(Pt 5): p. 1370-81.
8.    Braak, H. and E. Braak, Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol, 1991. 82(4): p. 239-59.
9.    Kontsekova, E., et al., First-in-man tau vaccine targeting structural determinants essential for pathological tau-tau interaction reduces tau oligomerisation and neurofibrillary degeneration in an Alzheimer’s disease model. Alzheimers Res Ther, 2014. 6(4): p. 44.
10.    Iqbal, K., F. Liu, and C.X. Gong, Tau and neurodegenerative disease: the story so far. Nat Rev Neurol, 2016. 12(1): p. 15-27.
11.    Li, C. and J. Gotz, Tau-based therapies in neurodegeneration: opportunities and challenges. Nat Rev Drug Discov, 2017. 16(12): p. 863-883.
12.    Guo, T., W. Noble, and D.P. Hanger, Roles of tau protein in health and disease. Acta Neuropathol, 2017. 133(5): p. 665-704.
13.    Sotiropoulos, I., et al., Atypical, non-standard functions of the microtubule associated Tau protein. Acta Neuropathol Commun, 2017. 5(1): p. 91.
14.    Violet, M., et al., A major role for Tau in neuronal DNA and RNA protection in vivo under physiological and hyperthermic conditions. Front Cell Neurosci, 2014. 8: p. 84.
15.    Paholikova, K., et al., N-terminal Truncation of Microtubule Associated Protein Tau Dysregulates its Cellular Localization. J Alzheimers Dis, 2014.
16.    Zilka, N., et al., Who fans the flames of Alzheimer’s disease brains? Misfolded tau on the crossroad of neurodegenerative and inflammatory pathways. J Neuroinflammation, 2012. 9: p. 47.
17.    Alonso, A.C., et al., Role of abnormally phosphorylated tau in the breakdown of microtubules in Alzheimer disease. Proc Natl Acad Sci U S A, 1994. 91(12): p. 5562-6.
18.    Alonso, A.C., I. Grundke-Iqbal, and K. Iqbal, Alzheimer’s disease hyperphosphorylated tau sequesters normal tau into tangles of filaments and disassembles microtubules. Nat Med, 1996. 2(7): p. 783-7.
19.    Kovacech, B. and M. Novak, Tau truncation is a productive posttranslational modification of neurofibrillary degeneration in Alzheimer’s disease. Curr Alzheimer Res, 2010. 7(8): p. 708-16.
20.    Grundke-Iqbal, I., et al., Abnormal phosphorylation of the microtubule-associated protein tau (tau) in Alzheimer cytoskeletal pathology. Proc Natl Acad Sci U S A, 1986. 83(13): p. 4913-7.
21.    Zilka, N., et al., The self-perpetuating tau truncation circle. Biochemical Society transactions, 2012. 40(4): p. 681-6.
22.    Skrabana, R., et al., Neuronal Expression of Truncated Tau Efficiently Promotes Neurodegeneration in Animal Models: Pitfalls of Toxic Oligomer Analysis. J Alzheimers Dis, 2017. 58(4): p. 1017-1025.
23.    Al-Hilaly, Y.K., et al., Alzheimer’s Disease-like Paired Helical Filament Assembly from Truncated Tau Protein Is Independent of Disulfide Crosslinking. Journal of molecular biology, 2017. 429(23): p. 3650-3665.
24.    Alonso, A., et al., Hyperphosphorylation induces self-assembly of tau into tangles of paired helical filaments/straight filaments. Proc Natl Acad Sci U S A, 2001. 98(12): p. 6923-8.
25.    Vechterova, L., et al., DC11: a novel monoclonal antibody revealing Alzheimer’s disease-specific tau epitope. Neuroreport, 2003. 14(1): p. 87-91.
26.    Kovacech, B., R. Skrabana, and M. Novak, Transition of tau protein from disordered to misordered in Alzheimer’s disease. Neurodegener Dis, 2010. 7(1-3): p. 24-7.
27.    Novak, M., J. Kabat, and C.M. Wischik, Molecular characterization of the minimal protease resistant tau unit of the Alzheimer’s disease paired helical filament. EMBO J, 1993. 12(1): p. 365-70.
28.    Frost, B., et al., Conformational diversity of wild-type Tau fibrils specified by templated conformation change. J Biol Chem, 2009. 284(6): p. 3546-51.
29.    Skrabana, R., J. Sevcik, and M. Novak, Intrinsically disordered proteins in the neurodegenerative processes: formation of tau protein paired helical filaments and their analysis. Cellular and molecular neurobiology, 2006. 26(7-8): p. 1085-97.
30.    Fitzpatrick, A.W.P., et al., Cryo-EM structures of tau filaments from Alzheimer’s disease. Nature, 2017. 547(7662): p. 185-190.
31.    Augustinack, J.C., et al., Specific tau phosphorylation sites correlate with severity of neuronal cytopathology in Alzheimer’s disease. Acta Neuropathol, 2002. 103(1): p. 26-35.
32.    Goedert, M., D.S. Eisenberg, and R.A. Crowther, Propagation of Tau Aggregates and Neurodegeneration. Annual review of neuroscience, 2017. 40: p. 189-210.
33.    Novak, M., Truncated tau protein as a new marker for Alzheimer’s disease. Acta Virol, 1994. 38(3): p. 173-89.
34.    Clavaguera, F., et al., Transmission and spreading of tauopathy in transgenic mouse brain. Nat Cell Biol, 2009. 11(7): p. 909-13.
35.    Ahmed, Z., et al., A novel in vivo model of tau propagation with rapid and progressive neurofibrillary tangle pathology: the pattern of spread is determined by connectivity, not proximity. Acta Neuropathol, 2014. 127(5): p. 667-83.
36.    Hu, W., et al., Hyperphosphorylation determines both the spread and the morphology of tau pathology. Alzheimers Dement, 2016. 12(10): p. 1066-1077.
37.    Zilka, N., M. Korenova, and M. Novak, Misfolded tau protein and disease modifying pathways in transgenic rodent models of human tauopathies. Acta Neuropathol, 2009. 118(1): p. 71-86.
38.    Masters, C.L., et al., Alzheimer’s disease. Nat Rev Dis Primers, 2015. 1: p. 15056.
39.    Zilka, N., et al., Truncated tau from sporadic Alzheimer’s disease suffices to drive neurofibrillary degeneration in vivo. FEBS Lett, 2006. 580(15): p. 3582-8.
40.    Koson, P., et al., Truncated tau expression levels determine life span of a rat model of tauopathy without causing neuronal loss or correlating with terminal neurofibrillary tangle load. Eur J Neurosci, 2008. 28(2): p. 239-46.
41.    Zimova, I., et al., Human Truncated Tau Induces Mature Neurofibrillary Pathology in a Mouse Model of Human Tauopathy. J Alzheimers Dis, 2016. 54(2): p. 831-43.
42.    Novak, P., et al., Stress-Induced Alterations of Immune Profile in Animals Suffering by Tau Protein-Driven Neurodegeneration. Cell Mol Neurobiol, 2018. 38(1): p. 243-259.
43.    Kovac, A., et al., Misfolded truncated protein tau induces innate immune response via MAPK pathway. J Immunol, 2011. 187(5): p. 2732-9.
44.    Cente, M., et al., Expression of a truncated human tau protein induces aqueous-phase free radicals in a rat model of tauopathy: implications for targeted antioxidative therapy. J Alzheimers Dis, 2009. 17(4): p. 913-20.
45.    Filipcik, P., et al., Cortical and hippocampal neurons from truncated tau transgenic rat express multiple markers of neurodegeneration. Cell Mol Neurobiol, 2009. 29(6-7): p. 895-900.
46.    Cente, M., et al., Expression of a truncated tau protein induces oxidative stress in a rodent model of tauopathy. Eur J Neurosci, 2006. 24(4): p. 1085-90.
47.    Kontsekova, E., et al., Identification of structural determinants on tau protein essential for its pathological function: novel therapeutic target for tau immunotherapy in Alzheimer’s disease. Alzheimers Res Ther, 2014. 6(4): p. 45.
48.    Gilman, S., et al., Clinical effects of Abeta immunization (AN1792) in patients with AD in an interrupted trial. Neurology, 2005. 64(9): p. 1553-62.
49.    Vandenberghe, R., et al., Active Abeta immunotherapy CAD106 in Alzheimer’s disease: A phase 2b study. Alzheimers Dement (N Y), 2017. 3(1): p. 10-22.
50.    Novak, P., et al., Safety and immunogenicity of the tau vaccine AADvac1 in patients with Alzheimer’s disease: a randomised, double-blind, placebo-controlled, phase 1 trial. Lancet Neurol, 2017. 16(2): p. 123-134.
51.    Lang, P.O., et al., Immune senescence and vaccination in the elderly. Curr Top Med Chem, 2013. 13(20): p. 2541-50.
52.    Hobart, J., et al., Putting the Alzheimer’s cognitive test to the test I: traditional psychometric methods. Alzheimers Dement, 2013. 9(1 Suppl): p. S4-9.
53.    Hobart, J., et al., Putting the Alzheimer’s cognitive test to the test II: Rasch Measurement Theory. Alzheimers Dement, 2013. 9(1 Suppl): p. S10-20.
54.    Raghavan, N., et al., The ADAS-Cog revisited: novel composite scales based on ADAS-Cog to improve efficiency in MCI and early AD trials. Alzheimers Dement, 2013. 9(1 Suppl): p. S21-31.
55.    Gorno-Tempini, M.L., et al., Classification of primary progressive aphasia and its variants. Neurology, 2011. 76(11): p. 1006-14.
56.    Josephs, K.A., et al., Neuropathological background of phenotypical variability in frontotemporal dementia. Acta Neuropathol, 2011. 122(2): p. 137-53.
57.    Santos-Santos, M.A., et al., Features of Patients With Nonfluent/Agrammatic Primary Progressive Aphasia With Underlying Progressive Supranuclear Palsy Pathology or Corticobasal Degeneration. JAMA Neurol, 2016. 73(6): p. 733-42.
58.    Grossman, M., Primary progressive aphasia: clinicopathological correlations. Nat Rev Neurol, 2010. 6(2): p. 88-97.
59.    Rohrer, J.D., et al., Serum neurofilament light chain protein is a measure of disease intensity in frontotemporal dementia. Neurology, 2016. 87(13): p. 1329-36.
60.    Steinacker, P., et al., Neurofilament as a blood marker for diagnosis and monitoring of primary progressive aphasias. Neurology, 2017. 88(10): p. 961-969.
61.    Semler, E., et al., A language-based sum score for the course and therapeutic intervention in primary progressive aphasia. Alzheimers Res Ther, 2018. 10(1): p. 41.
62.    Knopman, D.S., et al., Development of methodology for conducting clinical trials in frontotemporal lobar degeneration. Brain, 2008. 131(Pt 11): p. 2957-68.
63.    Sikkes, S.A., et al., Validation of the Amsterdam IADL Questionnaire(c), a new tool to measure instrumental activities of daily living in dementia. Neuroepidemiology, 2013. 41(1): p. 35-41.
64.    Rafii, M.S., et al., PET Imaging of Tau Pathology and Relationship to Amyloid, Longitudinal MRI, and Cognitive Change in Down Syndrome: Results from the Down Syndrome Biomarker Initiative (DSBI). J Alzheimers Dis, 2017. 60(2): p. 439-450.



T. West1, Y. Hu1, P.B. Verghese1, R.J. Bateman2, J.B. Braunstein1, I. Fogelman1, K. Budur3, H. Florian3, N. Mendonca4, D.M. Holtzman2


1. C2N Diagnostics LLC, Saint Louis, MO, USA; 2. Washington University, St. Louis, MO, USA; 3. AbbVie Inc., North Chicago, IL, USA; 4. AbbVie Deutschland GmbH & Co. KG, Ludwigshafen, Germany

Corresponding Author: Tim West, PhD, C2N Diagnostics, 20 S Sarah St, Saint Louis, MO 63108, Email:

J Prev Alz Dis 2017;4(4):236-241
Published online September 27, 2017,



Tau neurofibrillary tangles are found in the brains of patients suffering from Alzheimer’s disease and other tauopathies. The progressive spreading of tau pathology from one brain region to the next is believed to be caused by extracellular transsynaptic transmission of misfolded tau between neurons. Preclinical studies have shown that antibodies against tau can prevent this transfer of misfolded tau between cells. Thus, antibodies against tau have the potential to stop or slow the progression of tau pathology observed in human tauopathies. To test this hypothesis, a humanized anti-tau antibody (ABBV-8E12) was developed and a phase 1 clinical trial of this antibody has been completed. The double-blind, placebo-controlled phase 1 study tested single doses of ABBV-8E12 ranging from 2.5 to 50 mg/kg in 30 patients with progressive supranuclear palsy (PSP). ABBV-8E12 was found to have an acceptable safety profile with no clinically concerning trends in the number or severity of adverse events between the placebo and dosed groups. Pharmacokinetic modelling showed that the antibody has a plasma half-life and cerebrospinal fluid:plasma ratio consistent with other humanized antibodies, and there were no signs of immunogenicity against ABBV-8E12. Based on the acceptable safety and tolerability profile of single doses of ABBV-8E12, AbbVie is currently enrolling patients into two phase 2 clinical trials to assess efficacy and safety of multiple doses of ABBV-8E12 in patients with early Alzheimer’s disease or PSP.

Key words: Tau, immunotherapy, Alzheimer’s disease, tauopathy, therapeutic.



Tauopathies refer to a set of neurodegenerative disorders characterized by the pathological aggregation of microtubule-associated protein tau (MAPT) in neurons and glial cells in the human brain. The most common tauopathy is Alzheimer’s disease (AD), estimated to affect more than 5 million Americans and 46 million people worldwide. Less common tauopathies include progressive supranuclear palsy (PSP), cortical basal degeneration (CBD), Pick’s disease and frontotemporal dementia with parkinsonism linked to chromosome 17, with the total number of tauopathies being greater than 20 (1). In some of these tauopathies (such as PSP, CBD, and Pick’s), aggregation of tau in the brain is the predominant form of brain pathology, while for other diseases (such as AD and Niemann Pick type C) other brain pathologies are involved in the disease alongside tau.
The human tau gene contains 16 exons and alternative spicing of exons 2, 3 and 10 results in six different tau protein isoform (2). Based on the number of inserts near the amino-terminal and the carboxy-terminal repeat domains, the isoforms are referred to as 0N3R, 1N3R, 2N3R, 0N4R, 1N4R and 2N4R. Under normal conditions, tau is predominantly localized within neurons and more specifically axons, although non-neuronal cells can have trace amounts (3). Tau was originally identified as a microtubule-associated protein that functions to promote assembly of microtubule protein subunit tubulin into microtubules and stabilize their structure (4). More recently, novel functions of tau have been discovered, such as iron transport, neurogenesis, synaptic plasticity and neuronal DNA protection (5).
Data from biochemical and animal studies of the tau protein suggest a working model that can serve as a basic framework for studying the pathophysiological functions of tau as well as for developing tau therapeutics. Tauopathy pathogenesis includes molecular events such as hyperphosphorylation and aggregation of tau. Hyperphosphorylation of tau can reduce its binding to microtubules, and thereby cause microtubule disassembly and axonal transport impairment and eventually synaptic dysfunction (6, 7). Hyperphosphorylation, as well as some other modifications of tau (such as truncation and O-glcNAcylation) also promote tau aggregation (5).
Aggregated tau can assume the form of either soluble oligomers, insoluble paired helical filaments (PHFs), or insoluble straight filaments. PHFs manifest as neurofibrillary tangles (NFT) in the brain and are one of the major histopathological hallmarks of tauopathies. While NFTs were originally assumed to be toxic and the main cause of neurodegeneration, growing evidence suggests that NFT are neither necessary nor sufficient to cause neurodegeneration (8–10).
Recent evidence suggests that of the various aggregated tau species, tau oligomers can be toxic to neurons (11–15) and that release of tau oligomers from neurons can lead to transmission of tau pathology between cells. The tau oligomers are taken up by synaptically connected neurons, causing the normal tau in the recipient neuron to aggregate – thus resulting in spreading of tau pathology from one neuron to the next through neuronal connectivity. Calvaguera et al. found that intracerebral injection of brain extract from mice with filamentous tau pathology induces the formation and spreading of aggregates made of hyperphosphorylated tau in mice expressing human wild-type tau (16). Subsequently, more studies have reported similar findings in vivo, confirming this “prion-like” property of pathological tau species isolated from human tissue or transgenic mice (17–21). This mode of transcellular propagation suggests that extracellular transfer of tau between cells may be a susceptible target for antibody-mediated therapies. In support of this hypothesis, tau immunotherapy has emerged as a promising therapeutic strategy for tauopathies and this approach has been shown to reduce tau pathology and improve behavioral deficits in animal models (22–27).


Preclinical data

The laboratory of Dr. Holtzman at Washington University School of Medicine generated a library of anti-human tau antibodies to test the hypothesis that propagation of aggregated tau between cells can be prevented by using anti-tau antibodies and to further assess the mechanisms of action of such antibodies (28). The library originated from mice immunized with full-length human tau protein (2N4R) and antibodies recognizing various epitopes of human tau were identified. Using an in vitro cell based assay, the anti-tau antibodies were found to specifically and dose dependently block uptake of misfolded tau from brain lysates into neuronal cells (19, 28).
To assess the in vivo activity of the antibodies, three of the anti-tau antibodies and one control antibody were administered intracerebroventricularly (ICV) into P301S tau transgenic mice. The P301S transgenic mouse carries a mutated human tau gene that causes early onset frontotemporal dementia in humans. These mice develop brain tau pathology as evidenced by presence of NFTs and phosphorylated tau as well as behavioral deficits consistent with human tau pathology (29). Antibodies were infused continuously starting just after the time when tau pathology starts to develop in these mice (6 months of age). At nine months, associative learning was assessed and the brains of the animals were assessed for a variety of histological and biochemical measures. Of the three antibodies tested in vivo, one antibody (HJ8.5) demonstrated a consistent effect on all of the outcome measures as compared to the control antibody. This antibody showed a reduction in phosphorylated tau by both biochemical and histopathological measures, and an improvement in associative learning (28).

In addition to evaluating central delivery of the antibody, the biological effects of administering HJ8.5 peripherally were assessed in a separate study. Here, six-month old P301S mice received weekly intraperitoneal (IP) doses of PBS, or 10 or 50 mg/kg of HJ8.5 for three months (30). After three months of treatment, there was a significant reduction of tau pathology in the hippocampus in both the 10 mg/kg and 50 mg/kg dose group compared with control (Figure 1A and B). This reduction in tau pathology was matched by a reduction in the formic acid insoluble tau measured in the cortex of P310S mice in the 50 mg/kg dose group (Figure 1C). Peripheral HJ8.5 administration also improved sensorimotor function in these mice as measured by inverted screen and ledge tests. Importantly, peripheral antibody administration significantly attenuated brain volume loss observed in P301S mice over the three-month timeframe. This finding is significant since brain atrophy in human tauopathies associates strongly with tau accumulation in the brain regions affected, and thus it appears that the reduction in tau pathology by the antibody treatment results in protection against brain atrophy.

Figure 1. Anti-tau antibody decreased phospho-tau staining in the hippocampal CA1 cell layer

Figure 1. Anti-tau antibody decreased phospho-tau staining in the hippocampal CA1 cell layer


(A) Representative coronal sections of biotinylated AT8 antibody staining of phosphorylated tau in the hippocampal CA1 cellular region of 9-month-old P301S mice treated for three months with vehicle and HJ8.5 at 50 mg/kg. The lower images are higher power views of the CA1 region in the uppers panels. Red arrows indicate the area magnified in the lower image. Black arrows indicate the hippocampal CA1 cell layer. (B) Quantification of biotinylated AT8 antibody staining of abnormally phosphorylated tau revealed a significant decrease in AT8 staining in mice treated with HJ8.5 at 50 mg/kg in the hippocampal CA1 cellular layer compared to vehicle-treated mice (P = 0.035). Values represent mean ± SEM. (C) Levels of formic acid soluble tau determined by ELISA. HJ8.5 treatment at 50 mg/kg significantly decreased insoluble human tau (P < 0.0001) compared to vehicle-treated mice. Values represent mean ± SEM. ****P < 0.0001. Figure is modified from (30) with permission.


At the end of the peripheral dosing study, the concentration of tau in the plasma of P310S mice was measured using a tau ELISA. The concentration of tau in plasma was significantly increased in P310S animals that had been injected with HJ8.5, and the increase was higher in the animals that received the higher dose of HJ8.5 (30). Further studies indicated that the antibody dependent increase of plasma tau resulted primarily from CNS derived tau and that plasma tau concentrations appeared to reflect soluble, extracellular tau in the brain (31).
Recent studies utilized adenoassociated virus (AAV) to express single chain fragment variable domains (scFv) derived from HJ8.5 directly in the brain of P301S transgenic mice (32). This treatment also decreased tau pathology, demonstrating that the Fc domain of the antibody is not required for a therapeutic effect.
In summary, these preclinical studies show that anti-tau antibodies have the potential to engage tau present in the brain’s extracellular space as well as in plasma. Administration of the anti-tau antibody HJ8.5 in these animal models via central (ICV) and peripheral (IP) routes markedly reduced tau pathology, neuronal loss, and brain atrophy, resulting in cognitive and sensorimotor preservation compared to control treated mice.


Clinical data

Based on the promising results from the transgenic mouse model studies, a variety of humanized anti-tau antibodies were generated to test the hypothesis that anti-tau antibodies can provide therapeutic benefits in human tauopathies. Following requisite cell line development and toxicology testing, clinical trials with the lead compound (C2N-8E12, now known as ABBV-8E12) were initiated during the summer of 2015. A phase 1 clinical trial (NCT02494024) was designed to test the safety and tolerability of a single dose of ABBV-8E12 in patients with PSP. Key inclusion and exclusion criteria for this study are listed in Table 1. Subjects were randomized in blocks of four in a double-blind manner to receive a single intravenous dose of ABBV-8E12 or placebo in a three to one ratio (drug:placebo). Using a continual reassessment method that pre-specified algorithms for dose escalation, each block of subjects was assigned to one of the five dose cohorts (2.5, 7.5, 15, 25, and 50 mg/kg) tested in the study. Dose escalation was implemented only after available safety data from lower doses had been reviewed by the data safety monitoring committee (DSMC). Safety was monitored for 84 days post-dosing and included AEs, laboratory analyses, MRI assessments, ECG evaluations, physical/neurological examinations, vital signs, mental health assessments and a brain MRI at two weeks after dosing.

Table 1. Key inclusion and exclusion criteria for the phase 1 clinical trial (NCT02494024)

Table 1. Key inclusion and exclusion criteria for the phase 1 clinical trial (NCT02494024)


A total of 38 subjects were screened for the phase 1 trial, with 30 subjects enrolling. Of the enrolled subjects, 7 were assigned to the placebo arm and 23 were assigned to one of the 5 dose arms (Table 2).The safety profile at all doses supported dose escalation to the maximum dose (50 mg/kg), which was administered to 10 subjects. Twenty-seven subjects completed the 84-day follow-up and one subject withdrew from the study due to an adverse event (AE). AEs occurred in 21 of the 30 (70%) study participants. Table 3 provides an overview of the AEs observed in the study. The majority of the AEs were rated by the blinded investigators as mild or moderate in severity. Only two AEs were rated as severe – one case of headache and one case of agitation. Treatment-relatedness of the AEs was also rated by the investigators, with the majority of the AEs being rated as unrelated to treatment.

Table 2. Patient demographics and disease characteristics at screening

Table 2. Patient demographics and disease characteristics at screening


Table 3. Adverse event summary

Table 3. Adverse event summary

Events are listed by MedDRA preferred term. Number of patients experiencing an event is shown in the table with the percent incidence in parentheses.


There were no clinically concerning trends observed in the number or severity of AEs between the placebo and ABBV-8E12 dose groups. Nearly half (44%) of the observed AEs resolved within the first two days of onset and 77% of AEs resolved within the first two weeks. At the day 14 MRI, no clinically significant radiographic abnormal findings were observed. Three serious adverse events (SAEs) were reported during the study. One subject, in the 15 mg/kg dose group, with a history of experiencing several falls at baseline, had subdural hematoma. One subject, in the 25 mg/kg dose group, with a history of anxiety and agitation around stressful events including medical procedures, reported an increase in anxiety, agitation and perseverative behaviors. One subject, in the 50 mg/kg dose group, with a history of hypertension had hypertensive cerebrovascular disease. These events did not indicate an emerging safety issue. Thus, ABBV-8E12, when administered in single doses of up to 50 mg/kg, appears to have an acceptable safety and tolerability profile.
While assessment of single-dose safety and tolerability were the primary objectives of this phase 1 study, both plasma and CSF samples were acquired for assessment of pharmacokinetics (PK) of ABBV-8E12 in plasma and penetration of drug into the brain. Figure 2 shows the average PK profiles for ABBV-8E12 in plasma for each dose cohort. Across the dose range of 2.5 mg/kg to 50 mg/kg IV, the ABBV-8E12 area under the curve (AUC) increased in a dose-proportional manner. The harmonic mean plasma half-life ranged from approximately 27 days to 37 days. This plasma half-life is consistent with what has been reported for other monoclonal antibodies.

Figure 2. Pharmacokinetic profiles of ABBV-8E12 in plasma. Plasma samples were collected for up to 84 days post dosing and drug concentrations measured using a validated PK assay

Figure 2. Pharmacokinetic profiles of ABBV-8E12 in plasma. Plasma samples were collected for up to 84 days post dosing and drug concentrations measured using a validated PK assay

Data shown as mean with error bars representing the standard deviation.


Cerebrospinal fluid (CSF) was sampled at screening and 14 days after drug administration. Measurement of ABBV-8E12 in CSF on day 14 revealed that CSF ABBV-8E12 concentrations increased with dose. Comparison of CSF concentrations to plasma concentrations of ABBV-8E12 on the same day showed that the CSF/plasma ratio ranged from 0.181% to 0.385%, consistent with what has been observed in other studies of monoclonal antibodies. Plasma samples collected out to day 84 were also assayed for the presence of anti-ABBV-8E12 antibodies using a validated anti-drug antibody assay. No anti-drug antibodies were detected in the post-dose plasma samples analyzed.


Future development plan/discussion

Based on the acceptable safety and tolerability profile of single doses up to 50 mg/kg in PSP patients, AbbVie is currently enrolling patients into two phase 2 clinical trials that assess the efficacy and safety of multiple doses of ABBV-8E12 in patients with early AD or PSP. In these studies, patients will be dosed for 96 weeks (early AD) or 52 weeks (PSP). The first cohort of 48 and 30 subjects each, in AD and PSP studies respectively, will undergo intensive safety monitoring.
The phase 2 AD study is a randomized, double-blind, placebo-controlled trial designed to evaluate the efficacy and safety of ABBV-8E12 in patients with early AD. For the purposes of this study, early AD is defined as a score of 22 or higher on the mini-mental state examination, a global score of 0.5 on clinical dementia rating, a score of 85 or lower on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS)-delayed memory index, and a positive amyloid positron emission tomography (PET) scan.
The phase 2 AD study consists of a screening period of up to 8 weeks, a 96 week double-blind treatment period and a follow-up period of approximately 20 weeks following the last study drug administration. Approximately 400 subjects will be enrolled to meet the study objectives. Eligible subjects are between 55 to 85 years of age and have early AD in the absence of concurrent diseases that could confound safety or efficacy evaluations. Study participants are allowed to use concomitant medications to treat symptoms related to AD, if they are on a stable dose for at least 12 weeks prior to randomization. Upon completion of screening and baseline procedures, eligible subjects are randomized to one of the 3 ABBV-8E12 dose arms (low, medium and high dose) or placebo. Doses are administered every 4 weeks via IV infusion.
The primary efficacy measure for the AD study is the Clinical Dementia Rating scale – Sum of Boxes (CDR-SB). Secondary efficacy and exploratory outcomes include a variety of clinical measures, biologic markers, and neuroimaging measures. Safety will be monitored by adverse event reports, physical examination, laboratory tests, and imaging.
The phase 2 PSP study is a multiple dose, multicenter, multinational, randomized, double-blind, placebo-controlled trial designed to evaluate the efficacy and safety of ABBV-8E12 in patients with PSP. Eligible subjects are randomized to one of the two ABBV-8E12 dose arms (low and high dose) or placebo and ABBV-8E12 is administered every 4 weeks via IV infusion for a total of a 52-week treatment period. The study starts with a screening period of up to 8 weeks and ends with a follow-up period of approximately 20 weeks following the last study drug administration. Approximately 180 subjects will be enrolled to meet the study objectives. Eligible subjects are 40 years of age or older and meet PSP clinical criteria in the absence of concurrent diseases that could confound safety or efficacy evaluations. Study participants are required to have PSP symptoms for less than 5 years and to be able to walk 5 steps with minimal assistance.
The primary efficacy measure of this PSP phase 2 study is the PSP Rating Scale (PSPRS) (33). Secondary efficacy measures include a variety of biologic markers and clinical and neuroimaging outcome measures. Safety is assessed by adverse event reports, physical examination, laboratory tests and imaging.


Funding: This work was funded by C2N Diagnostics and AbbVie Inc., and supported by a grant from the Alzheimer’s Association (PCTR-15-330406) made possible by Part the Cloud™.

Acknowledgements: The team at C2N Diagnostics and AbbVie Inc. would like to sincerely thank the clinical investigators and the patients in the phase 1 clinical study for all their hard work and dedication to this clinical trial.

Conflict of interest: TW, IHH, PBV, and IF are full time employees and/or advisors of C2N Diagnostics, receiving stock and/or stock options. RJB is a co-founder of C2N Diagnostics. RJB and Washington University in St. Louis have equity ownership interest in C2N Diagnostics and may receive royalty income based on technology licensed by Washington University to C2N Diagnostics. RJB receives lab research funding from the Tau SILK Consortium (AbbVie, Biogen, and Eli Lilly and Co.), Eli Lilly and Co, Hoffman La-Roche, Janssen, Avid Radiopharmaceuticals, and the DIAN Pharma Consortium (Abbvie, Amgen, AstraZeneca, Biogen, Eisai, Eli Lilly and Co, Hoffman La-Roche, Janssen, Pfizer, and Sanofi). RJB has received honoraria from Janssen and Pfizer as a speaker and from Merck and Pfizer as an Advisory Board member. In addition, RJB receives income from C2N Diagnostics for serving on the Scientific Advisory Board. JBB is a co-founder and employee of C2N Diagnostics, receiving stock and/or stock options. KB, HF, and NM are employees of AbbVie, receiving stock and/or stock options. DMH co-founded and is on the scientific advisory board of C2N Diagnostics. DMH is an inventor on a submitted patent “Antibodies to Tau”, PCT/US2013/049333, that is licensed by Washington University to C2N Diagnostics. This patent was subsequently licensed to AbbVie. DMH receives research grants from the C2N Diagnostics, AbbVie, Eli Lilly, and Denali. DMH consults for Genentech, AbbVie, Eli Lilly, Proclara Biosciences, Glaxosmithkline, and Denali.



1. Williams DR. Tauopathies: classification and clinical update on neurodegenerative diseases associated with microtubule-associated protein tau. Intern Med J 2006;36:652–60. doi:10.1111/j.1445-5994.2006.01153.x.
2. Andreadis A, Broderick JA, Kosik KS. Relative exon affinities and suboptimal splice site signals lead to non-equivalence of two cassette exons. Nucleic Acids Res 1995;23:3585–93.
3. Schoenfeld TA, Obar RA. Diverse distribution and function of fibrous microtubule-associated proteins in the nervous system. Int Rev Cytol 1994;151:67–137. doi:10.1016/S0074-7696(08)62631-5.
4. Drubin DG, Kirschner MW. Tau protein function in living cells. J Cell Biol 1986;103:2739–46. doi:10.1083/jcb.103.6.2739.
5. Wang Y, Mandelkow E. Tau in physiology and pathology. Nat Rev Neurosci 2015;17:22–35. doi:10.1038/nrn.2015.1.
6. Cho JH, Johnson GVW. Glycogen synthase kinase 3β phosphorylates tau at both primed and unprimed sites: Differential impact on microtubule binding. J Biol Chem 2003;278:187–93. doi:10.1074/jbc.M206236200.
7. Schneider A, Biernat J, Von Bergen M, Mandelkow E, Mandelkow EM. Phosphorylation that detaches tau protein from microtubules (Ser262, Ser214) also protects it against aggregation into Alzheimer paired helical filaments. Biochemistry 1999;38:3549–58. doi:10.1021/bi981874p.
8. Santacruz K, Lewis J, Spires T, Paulson J, Kotilinek L, Ingelsson M, et al. Tau suppression in a neurodegenerative mouse model improves memory function. Science 2005;309:476–81. doi:10.1126/science.1113694.
9. Sydow A, Van der Jeugd A, Zheng F, Ahmed T, Balschun D, Petrova O, et al. Tau-Induced Defects in Synaptic Plasticity, Learning, and Memory Are Reversible in Transgenic Mice after Switching Off the Toxic Tau Mutant. J Neurosci 2011;31:2511–25. doi:10.1523/JNEUROSCI.5245-10.2011.
10. Van Der Jeugd A, Hochgräfe K, Ahmed T, Decker JM, Sydow A, Hofmann A, et al. Cognitive defects are reversible in inducible mice expressing pro-aggregant full-length human Tau. Acta Neuropathol 2012;123:787–805. doi:10.1007/s00401-012-0987-3.
11. Lasagna-Reeves C a, Castillo-Carranza DL, Sengupta U, Sarmiento J, Troncoso J, Jackson GR, et al. Identification of oligomers at early stages of tau aggregation in Alzheimer’s disease. FASEB J 2012;26:1946–59. doi:10.1096/fj.11-199851.
12. Maeda S, Sahara N, Saito Y, Murayama M, Yoshiike Y, Kim H, et al. Granular tau oligomers as intermediates of tau filaments. Biochemistry 2007;46:3856–61. doi:10.1021/bi061359o.
13. Tian H, Davidowitz E, Lopez P, Emadi S, Moe J, Sierks M. Trimeric tau is toxic to human neuronal cells at low nanomolar concentrations. Int J Cell Biol 2013. doi:10.1155/2013/260787.
14. Flach K, Hilbrich I, Schiffmann A, Gärtner U, Krüger M, Leonhardt M, et al. Tau oligomers impair artificial membrane integrity and cellular viability. J Biol Chem 2012;287:43223–33. doi:10.1074/jbc.M112.396176.
15. Tepper K, Biernat J, Kumar S, Wegmann S, Timm T, Hübschmann S, et al. Oligomer formation of tau protein hyperphosphorylated in cells. J Biol Chem 2014;289:34389–407. doi:10.1074/jbc.M114.611368.
16. Clavaguera F, Bolmont T, Crowther RA, Abramowski D, Frank S, Probst A, et al. Transmission and spreading of tauopathy in transgenic mouse brain. Nat Cell Biol 2009;11:909–13. doi:10.1038/ncb1901.
17. Clavaguera F, Akatsu H, Fraser G, Crowther RA, Frank S, Hench J, et al. Brain homogenates from human tauopathies induce tau inclusions in mouse brain. Proc Natl Acad Sci U S A 2013;110:9535–40. doi:10.1073/pnas.1301175110.
18. Sanders DW, Kaufman SK, DeVos SL, Sharma AM, Mirbaha H, Li A, et al. Distinct Tau Prion Strains Propagate in Cells and Mice and Define Different Tauopathies. Neuron 2014:1–18. doi:10.1016/j.neuron.2014.04.047.
19. Holmes BB, Furman JL, Mahan TE, Yamasaki TR, Mirbaha H, Eades WC, et al. Proteopathic tau seeding predicts tauopathy in vivo. Proc Natl Acad Sci U S A 2014;111:E4376-85. doi:10.1073/pnas.1411649111.
20. Guo JL, Narasimhan S, Changolkar L, He Z, Stieber A, Zhang B, et al. Unique pathological tau conformers from Alzheimer’s brains transmit tau pathology in nontransgenic mice. J Exp Med 2016;213:2635 LP-2654. doi:10.1084/jem.20160833.
21. Liu L, Drouet V, Wu JW, Witter MP, Small SA, Clelland C, et al. Trans-synaptic spread of tau pathology in vivo. PLoS One 2012;7. doi:10.1371/journal.pone.0031302.
22. Sigurdsson EM. Immunotherapy targeting pathological tau protein in Alzheimer’s disease and related tauopathies. J Alzheimer’s Dis 2008;15:157–168.
23. Sigurdsson EM. Tau-focused immunotherapy for Alzheimer’s disease and related tauopathies. Curr Alzheimer Res 2009;6:446–50.
24. Asuni A a, Boutajangout A, Quartermain D, Sigurdsson EM. Immunotherapy targeting pathological tau conformers in a tangle mouse model reduces brain pathology with associated functional improvements. J Neurosci 2007;27:9115–29. doi:10.1523/JNEUROSCI.2361-07.2007.
25. Boimel M, Grigoriadis N, Lourbopoulos A, Haber E, Abramsky O, Rosenmann H. Efficacy and safety of immunization with phosphorylated tau against neurofibrillary tangles in mice. Exp Neurol 2010;224:472–85. doi:10.1016/j.expneurol.2010.05.010.
26. Troquier L, Caillierez R, Burnouf S, Fernandez-Gomez FJ, Grosjean M-E, Zommer N, et al. Targeting phospho-Ser422 by active Tau Immunotherapy in the THYTau22 mouse model: a suitable therapeutic approach. Curr Alzheimer Res 2012;9:397–405. doi:10.2174/156720512800492503.
27. Bi M, Ittner A, Ke YD, Götz J, Ittner LM. Tau-targeted immunization impedes progression of neurofibrillary histopathology in aged P301L tau transgenic mice. PLoS One 2011;6. doi:10.1371/journal.pone.0026860.
28. Yanamandra K, Kfoury N, Jiang H, Mahan TE, Ma S, Maloney SE, et al. Anti-Tau Antibodies that Block Tau Aggregate Seeding In Vitro Markedly Decrease Pathology and Improve Cognition In Vivo. Neuron 2013;80:402–14. doi:10.1016/j.neuron.2013.07.046.
29. Yoshiyama Y, Higuchi M, Zhang B, Huang S-M, Iwata N, Saido TC, et al. Synapse loss and microglial activation precede tangles in a P301S tauopathy mouse model. Neuron 2007;53:337–51. doi:10.1016/j.neuron.2007.01.010.
30. Yanamandra K, Jiang H, Mahan TE, Maloney SE, Wozniak DF, Diamond MI, et al. Anti-tau antibody reduces insoluble tau and decreases brain atrophy. Ann Clin Transl Neurol 2015;2:278–88. doi:10.1002/acn3.176.
31. Yanamandra K, Patel TK, Jiang H, Schindler S, Ulrich JD, Boxer AL, et al. Anti-tau antibody administration increases plasma tau in transgenic mice and patients with tauopathy. Sci Transl Med 2017;9. doi:10.1126/scitranslmed.aal2029.
32. Ising C, Gallardo G, Leyns CEG, Wong CH, Stewart F, Koscal LJ, et al. AAV-mediated expression of anti-tau scFvs decreases tau accumulation in a mouse model of tauopathy. J Exp Med 2017;214:1227–38. doi:10.1084/jem.20162125.
33. Golbe LI, Ohman-Strickland P a. A clinical rating scale for progressive supranuclear palsy. Brain 2007;130:1552–65. doi:10.1093/brain/awm032.



M. Guéroux1, C. Fleau1, M. Slozeck1, M. Laguerre2, I. Pianet1,3


1. Institut des Sciences Moléculaires, CNRS, Université de Bordeaux, 351 cours de la Libération 33405 Talence, France; 2. Chimie et Biologie des Membranes et Nanoobjets, CNRS, Université de Bordeaux, 1 allée Geoffroy St Hilaire, Pessac, France; 3. Institut de Recherche sur les Archéomatériaux, CNRS, Université de Bordeaux-Montaigne, Esplanade des Antilles, Pessac, France.

Corresponding Author: Isabelle Pianet, Institut de Recherche sur les Archéomatériaux, CNRS, Université Bordeaux-Montaigne, Esplanade des Antilles 33600 Pessac,; tel: (+33)557126755; fax:  (+33)557124550

J Prev Alz Dis 2017;4(4):218-225
Published online September 27, 2017,



Polyphenols such as Epigallocatechin-3 gallate (EGCG) are currently bearer of hope to prevent or at least to slow down the deleterious effect of Tauopathies such as Alzheimer disease. One of the main effects of these neurodegenerative pathologies is the hyperphosphorylation and consequent aggregation of the Tau protein that leads to the irremediable neuronal cells death. In the present paper, we show how EGCG can play a crucial role to prevent Tau aggregation: (i) in binding Tau in its phosphorylation region with an affinity of the same order of magnitude than kinases (0.5 mM), hindering their access to the protein and (ii) in modifying the 3D-structure of Tau whose preferential conformation changes in the presence of EGCG. For this purpose, two peptides were synthesized, one of 20 residues long issued from the first Proline-rich region of Tau (171Ile-190Lys), the second of 50 residues long (171Ile-220Thr) corresponding to more than 50% of the Tau Proline rich domaine. The total attribution of all the 1H, 13C and 15N resonances of the two peptides has been achieved thanks to a “divide and conquer” strategy leading to their 3D structure preference and their affinity towards EGCG.

Key words:  Tau, EGCG, interactions, NMR, molecular modeling.

Abbreviations: AD: Alzheimer’s disease; B3: (+) catechin 4α-8 (+) catechin; Catechin: (+)-Cyanidol-3, (2R,3S)-2-(3,4-Dihydroxyphenyl)-3,4-dihydro-1(2H)-benzopyran-3,5,7-triol; DOSY : Diffusion Ordered Spectroscopy; EGCG : (-) epigallocatechin gallate (−)-cis-2-(3,4,5-Trihydroxyphenyl)-3,4-dihydro-1(2H)-benzopyran-3,5,7-triol 3-gallate, (−)-cis-3,3′,4′,5,5′,7-Hexahydroxy-flavane-3-gallate; NOESY: Nuclear Overhauser Effect Spectroscopy; STD: Saturation Transfer Difference; TOCSY: Total Correlation Spectroscopy.



Polyphenols, which are secondary metabolites widespread in plants, have interested physicians and pharmacists since the beginning of the millennium. Epigallocatechin 3-gallate (EGCG, Scheme 1A), found in high quantites in green tea (around 20% of the leaves dry weight) is one of the most popular in terms of health promotion (1) and is currently under clinical trials for treatment of Alzheimer’s disease (phases 2 and 3 (2)). These clinical trials follow the numerous publications dealing with tauopathies and polyphenols, and the numerous explanations leading to the understanding of their beneficial effects. Key parameters are the antioxidant potential of these compounds (3, 4), their ability to chelate metals (5, 6), reduce β-amyloid fibrillation (7) and Tau aberrant aggregation (8-10). Inhibition of Tau hyperphosphorylation and aggregation is one of the most promising new potential therapeutics (11, 12) and most of the previous work highlights polyphenols in general, and EGCG in particular, as good candidates to play this role.

Scheme 1. A. Chemical structures of polyphenols used for titrations. EGCG (left), and B3 (right). B. Schematic representation of Tau (Top) and amino acid of the longest isoform of Tau 2N4R (bottom). E1: acidic region, PRR: Proline-Rich Region, R: Repeat Region. The PR1 peptide is underlined in green and the PR1PR2 peptide in purple

Scheme 1. A. Chemical structures of polyphenols used for titrations. EGCG (left), and B3 (right). B. Schematic representation of Tau (Top) and amino acid of the longest isoform of Tau 2N4R (bottom). E1: acidic region, PRR: Proline-Rich Region, R: Repeat Region. The PR1 peptide is underlined in green and the PR1PR2 peptide in purple

These observations are the starting point of our work, which brings a molecular and atomistic interpretation of the role that EGCG might play to fight tauopathies such as Alzheimer’s Disease (AD). For this purpose, we started working on two synthesized peptides issued from the Proline-rich Domaine of the Tau protein (2N4R isoform, 441 residues, Scheme 1B). We target this domain because most of the phosphosphorylations at the origin of the aberrant aggregation of Tau takes place in the Proline-rich region of Tau: there are 85 putative phosphorylation sites, 45 of which are phosphorylated vs 9 for diseased and healthy neurons, respectively (13). Most of the phos-phorylation sites are located on Serine and/or Threonine residues, particularly on Ser-Pro and Thr-Pro motives (13), while 20 of the 45 phosphorylation sites listed for diseased neurons are located in the Proline Rich Regions.
The two model peptides used in this study correspond to the 171Ile-190Lys (PR1, underlined in green, Scheme 1B) and 171Ile-220Thr (PR1PR2, underlined in purple, Scheme 1B). This choice was guided by the level of proline residues and Ser/Thr-Pro motifs. Moreover, previous studies were performed on the 201Gly-220Thr fragment (14, 15) (PR2) whose previous total NMR resonance assignment permits attribution and defines the conformational preference of the longest peptide PR1PR2 using the “divide and conquer” strategy. The choice of EGCG and the catechin dimer B3 comes from these previous studies in which we demonstrated its higher affinity towards PR2 compared to other procyanidins (15). The binding of EGCG and B3 on PR1 and PR1PR2 has been characterized by using different NMR techniques (chemical shift variations via 1D 1H and 2D NMR, variation of the Diffusion coefficient of the peptides using DOSY, EGCG epitope-mapping using STD), and molecular modeling.


Experimental section

Peptides and procyanidins synthesis and purification

Unlabelled PR1 and PR1PR2 were synthesized by Genecust (Luxembourg) and purified by HPLC (Water 2487 dual wavelength absorbance detector) using a Nova Pak C18 column. Purity was controlled by ESI mass spectrometry. EGCG was provided by Sigma Aldrich (purity >95%) and  B3 ((+)-catechin 4α (+)-catechin) was synthesized and purified using a strategy previously described (16).

Sample preparation

For the structure determination, a 4 mM solution of peptides PR1/ PR1PR2 was prepared in a H2O/D2O 80/20, v/v solution or with neat D2O (to record 1H-13C HSQC experiment and for proline assignments). Samples were buffered with 10 mM of KH2PO4 to obtain a pH between 6 and 6.2. For titration experiments procyanidins are progressively added to a 0.5 or 1 mM of peptide in a KH2PO4 buffer, pH 6.2, dissolved in H2O/D2O: 90/10, v/v to scan a final concentration ranging from 1 to 12 mM. In order to keep the peptide concentration constant, tannins were added as a lyophilized powder. The pH (or pD) value of the solution was controlled at the end of the titration and remains constant.

Acquisition of NMR Spectra

For peptides 3D structure, regular 1D and 2D spectra were recorded at 293 K on a Bruker Avance III 600 (PR1) spectrometer equipped with a 5mm gradient inverse broadband probe with 2H lock (CESAMO facility) and a Bruker Avance III 800 (PR1PR2, IECB facility) equipped with a 5 mm cryoprobe TCI (1H/13C/15N/2H) with Z-gradient. In all experiments, water suppression was done using a watergate sequence. The following parameters were used: for TOCSY NMR experiments, spectral width, 7692 Hz in both dimensions; time domain data points, 2048; t1 increments, 256; relaxation delay, 2 s; 16 scans per t1 increment; spinlock, 100 ms; NOESY spectra were recorded with a 7164 Hz spectral width, a 4096 time domain data points, and 512 t1 increments, a mixing time of 300 ms, a 2 s relaxation delay and 64 scans per t1 increment ; The 1H-13C HSQC experiment was acquired using a spectral width of 6010 Hz in the 1H dimension and 20000 Hz in the 13C dimension, a 4096 time domain data points, 256 t1 increments, a 2 s relaxation delay, and 16 scans per t1 increment. 1H- 15N HSQC was acquired using a spectral width of 10000 Hz in the 1H dimension and 3250 Hz in the 15N dimension, a 2048 time domain data points, 256 t1 increments, a 2 s relaxation delay, and 16 scans per t1 increment. All data processing was performed using the Topspin software version 3.1 (Bruker).
For titration experiments, 1D proton, TOCSY and DOSY spectra were recorded for each EGCG or B3 (for PR1 only), using the following parameters: for TOCSY, spectral width, 7200 Hz in both dimensions; time domain data points, 1024; t1 increments, 128; relaxation delay, 2 s; 16 scans per t1 increment; spinlock, 100ms; for diffusion measurement we used a stimulated echo with bipolar gradient pulses sequence, with the following parameters: spectral width, 7000 Hz, scan number, 138; recycling delay, 2s; intergradient delay Δ, 150ms; gradient pulse duration δ, 1.5 ms (corresponding to a δ value of 2ms). The pulse gradients (G) were incremented from 2 to 95% of the maximum gradient strength in a linear ramp in 16 steps. The diffusion coefficient, D can be obtained by fitting a specific resonance area, I, obtained at different gradient powers (G) using the Bruker Topspin 2.1 software.
The diffusion constant of water (2.3×10-9 m2/s at 298K) was used to calibrate the instrument.
STD NMR experiments were recorded for PR1PR2 (0.5 mM) with EGCG (4 mM) with 256 scans. Selective saturations of protein on resonance at -0.5 ppm and off resonance at 35 ppm were performed using a 1D sequence for saturation transfer difference, with shaped pulse train for saturation alternating between on and off resonance, and with spoil sequence to destroy unwanted magnetization. STD NMR spectra were acquired at 298 K using a series of 40 equally spaced 50-ms gaussian-shaped pulses for selective saturation, with a 1s delay between pulses, which corresponds to a total saturation time of 2 s. A Watergate sequence was used to suppress the residual HDO signal and a spin-lock filter, with a 5 kHz strength and 10 ms duration, was applied to suppress protein back-ground. To ensure the specificity of STDs eliciting resonances, another STD NMR experiment was recorded by adding to the sample a capillary with 1 mM glucose solution.

NMR Data Analysis

For titration experiments, chemical shifts variations of some protons or Diffusion coefficient of peptides were analyzed using equation 2 (17):
Aobs = 1/2Amax(1 + Kd/n*P0) + Ti)/n*P0)) – {(1+ Kd/n*P0) + Ti)/n*P0))2 -4Ti)/n*P0)}1/2)
Where Aobs is Δδi, the chemical shift variation (ppm) or ΔDobs, the diffusion coefficient variation, Amax is Δδmax (or ΔDmax), the chemical shift (diffusion D) difference between the chemical shift (or D) of the peptide alone and saturated with tannins, Kd is the dissociation constant expressed in mM, Ti, the total concentration of polyphenol able to fix the peptide in mM, P0), the total concentration of peptide in mM, and n the number of polyphenols binding sites. Kd, n, and Δδmax or ΔDmax were calculated using a least-squares-fitting routine within the software program Microsoft EXCEL.

Molecular Modeling and Dynamics

Molecular modeling calculations were performed according to a previously published procedure on a Linux workstation running MacroModel version 6.5 (Schrodinger Inc.) (18). Conformational minima were found using the modified AMBER* (1991 parameters) force field as implemented and completed in the MacroModel program. Built structures were minimized to a final RMSD gradient <0.005 kJ.Ǻ-1.mol-1 via the truncated Newton conjugate gradient (TNCG) method (1000 cycles). Calculations were performed using distance constraints according to the NOE correlation intensities: strong, 2.2±0.4Å; medium, 3.5±0.9 Å; and weak, 5.0±0.5Å. After full minimization, the 20 lowest energy conformers were selected. Molecular dynamics calculations were performed using GROMACS 4.0 and theGROMOS96 force field (19). The EGCG molecule was parameterized using the Dundee PRODRG2 server website version 2.5 (http:// davapc1. bioch. /prodrg/). The peptide PR1 and 4 EGCG were set randomly in a simple point charge (SPC) cubic water box with dimensions (100:100:100) Å. PR1 was neutralized without any salt added. Molecular dynamics (MD) runs were performed at constant temperature (300 K, time constant for coupling, 0.1 ps) and pressure (1 bar) with a Berendsen coupling as previously described (20). Calculation of amphiphilic surfaces was obtained as described in Simon, et al.(18).



1H, 13C and 15N assignments, and 3D structure of Peptides

The efficient “divide and conquer” strategy was employed to complete the 1H, 13C and 15N assignments of the naturally unfolded peptide PR1PR2 (171Ile-220Thr) representative of the Proline Rich Region of Tau, consisting of using splicing information extracted from spectra of smaller fragments. For this purpose, the total assignment of the shorter peptide PR2 (201Gly-220Thr) was already performed (15), and another peptide PR1 corresponding to the 171Ile-190Lys region is presented herein. The NMR characterization of PR1 was achieved using the Wüthrich strategy (21). TOCSY informs about the resonance pattern of each residue. The NH-Hα region shows the presence of three Lysines (K130, K174 and K190), two Alanines (A173, A178), two Serines (S184, S185) two Threonines (T175, T181), one Glycine (G186) and one Glutamic acid (E187). The assignment of the eight Prolines was completed with TOCSY experiments specifically recorded in D2O in order to decrease the residual water signal which partly hides their Hα resonances. Three set of signals were distinguishable: the resonances centered at 4.63 ppm is assigned to the Hα of prolines 176, 182 and 188 because this chemical shift value is specific of a proline Hα when followed by another proline (22). The second set of signals corresponding to Hα resonating at 4.4 ppm and can be assigned to P172 and P179, while the third set of signals resonating at 4.38 ppm can be as-signed to the three prolines preceding another proline: P177, 183 and 188.
The sequential assignment was achieved by NOESY and permitted the assignment of all the 1H resonances, and consequently, the attribution of 13C and 15N resonances thanks to 1H-13C and 1H-15N HSQC, respec-tively.
From these NMR data, the expected 11 3J Hα-NH and only 17 NOEs classified with respect to their intensity and exclusively corresponding to (i : i+1) contacts were collected. These data served as constraint files to minimize the 3D structure of this proline-rich region of Tau. As expected by the sharp NH and CHα resonances of this peptide, the level of splicing and the lack of inter-residues NOEs, the result of molecular minimization con-verges to an unfolded model (figure 1).

Figure 1. The 10 best conformations of PR1 obtained from molecular dynamics under NMR constraints. The red bold trace represents the lowest energy conformation

Figure 1. The 10 best conformations of PR1 obtained from molecular dynamics under NMR constraints. The red bold trace represents the lowest energy conformation

The NMR characterization of PR1PR2 may be achieved despite the severe resonance overlapping through the superposition of the TOCSY/NOESY spectrum of PR1PR2, PR1 and PR2. Figure 2a displays the superposed NH-Hα region of PR1PR2 with PR1 and PR2. Except for the NH resonance of Lys 190 (K190) and Thr 220 (T220), which are shifted in PR1 and PR2 with respect to PR1PR2, due to their C terminal position, a high level of resonances superposition is observed between the three different peptides. Concerning the attribution of the 15 prolines, the same method was used. Figure 2b displays the TOCSY spectrum of PR1PR2, especially the region where all the prolines resonances are, with the superposition of PR1 and PR2:  the high level of resonances superposition also permits the assignment of three distinct groups of prolines depending on their vicinity. Using this strategy, the assignment of the remaining ten residues (Ser 191 to Pro 200) was easily achieved. 1H-13C and 1H-15N HSQC were also acquired in order to assign all the 15N resonances corresponding to the NH backbone, and all the 13C except the quaternary carbons.

Figure 2. Left: Superposition of TOCSY spectra: NH-Hα region of PR1PR2 (black) PR1 (red) and PR2 (green). Spectra were recorded on a 4 mM peptide solution (H2O/D2O : 90/10; v/v, pH 6.2, KH2PO4). For clarity, the S and T Hβ were pointed instead of Hα. Right: Proton assignment of PR1PR2 prolines. Superimposition of TOCSY spectra (proline region) recorded on a 4 mM peptide solution (D2O; pD 6.6, KH2PO4)

Figure 2. Left: Superposition of TOCSY spectra: NH-Hα region of PR1PR2 (black) PR1 (red) and PR2 (green). Spectra were recorded on a 4 mM peptide solution (H2O/D2O : 90/10; v/v, pH 6.2, KH2PO4). For clarity, the S and T Hβ were pointed instead of Hα. Right: Proton assignment of PR1PR2 prolines. Superimposition of TOCSY spectra (proline region) recorded on a 4 mM peptide solution (D2O; pD 6.6, KH2PO4)

From these NMR data, 23 3J Hα-NH of the 35 expected were obtained and 34 NOEs were classified with respect to their intensity and corresponding exclusively to (i – i+1) contacts. These data combined with the sharp NH and CHα resonances of this peptide, the level of splicing between the two fragments and the 50 residues peptide PR1PR2, and the lack of inter-residue NOEs that converge to an unfolded model as previously reported on the proline rich domains of Tau protein (23-27), were not used to obtain a time-consuming and expected unfolded structure.

Interactions with Procyanidins

Procyanidins- peptides interactions have been viewed on both sides through two types of titration experiments: for PR1 and PR1PR2, the peptide chemical shift changes were followed by 1D, 2D, and the peptide translational diffusion changes using DOSY NMR; the procyanidins contact points were checked using STD NMR experi-ments on the longest peptide. Moreover, these different approaches give rise to the physicochemical quantities governing the interactions, notably the dissociation constant, Kd and the number of binding sites, n.
Titration experiments were performed on PR1 peptide with 2 different procyanidins: the galloylated monomer EGCG and the catechin dimer B3. The progressive addition of procyanidins showed significant chemical shift variations in two main regions of PR1 –Ala178 to Thr 181 and Ser185 to Lys190 hallmarking the recognition place of polyphenols on the peptide. Chemical shift variations through the titration experiments can be fitted using eq. 2, which describes the binding of polyphenols at multiple sites n on the peptide (Figure 3). The dissociation constant (Kd) and the number of binding sites (n) for each procyanidin were averaged from fitting of different chemical shift variations of the most responsive residues: Kd values were estimated at 1.1±0.1 and 2.0±0.3 mM, with a number of binding sites close to 2.1±0.3 and 2.6±0.4, for EGCG and B3, respectively. The relationship between Kd values and the chemical nature of polyphenols highlights the effect of the presence of a galloyl group upon affinity. Concerning the number of binding sites (n) it seems to be close to 2 whatever the polyphenol tested.

Figure 3. Chemical shift variations (Δδ) of PR1 (NH of 181T) with increasing EGCG (blue) and B3 (green). PR1 concentration was adjusted to 1 mM, and polyphenols were added to obtain the final concentration indicated on the x-axis. And chemical shift variations (Δδ) of PR1PR2 (NH of Thr212) with increasing [EGCG]

Figure 3. Chemical shift variations (Δδ) of PR1 (NH of 181T) with increasing EGCG (blue) and B3 (green). PR1 concentration was adjusted to 1 mM, and polyphenols were added to obtain the final concentration indicated on the x-axis. And chemical shift variations (Δδ) of PR1PR2 (NH of Thr212) with increasing [EGCG]

DOSY NMR experiments were recorded on PR1 peptide at various concentrations of B3. The D value evolves from 1.6×10-10m2.s-1 to 1.27×10-10 m2.s-1, above 4 mM of B3, confirming the association process and in accordance with the association of approximately 2 polyphenol molecules for 1 PRP1 peptide and a Kd around 2 mM: using the approximation of the Graham’s law (even if the relation should be more complex (28) depending on the spatial structure that the molecule adopts in a given solvent) for which the rate of diffusion coefficients, D, are inversely proportional to the square roots of their molecular weights,  a value of 2.3 molecules of B3 were bound to PR1).
In fine, molecular dynamics calculations were performed on a system composed of one peptide PR1 and 4 EGCG molecules, randomly disposed in a box of (100Å)3 full of water. All the calculations were run at least three times with different random seeds. The scenario observed is as follow: within the first 10 ns two different events appear, the formation of EGCG aggregates and the binding of EGCG to the peptide. At the end of the calcula-tion, a complex between the 4 EGCG (whose only three are directly fixed to the peptide), and PR1 was formed. The Ser184, Glu187 and Lys190 residues appear to be the most involved in the binding process:  in 80% of cases, EGCG plays the role of a hydrogen bond acceptor while the peptide is the hydrogen bond donor. This binding mostly takes place in the hydrophilic moiety of the peptide (in blue in Fig.4) between the carbonyl function of EGCG and the OH and/or NH3 groups of Ser184 and Lys190. However, a hydrophobic interaction is also observed, while less common, between the aromatic rings of polyphenols and the hydrophobic region of PR1 (Fig. 4).

Figure 4. Molecular lipophilicity potential contours of PR1 (dark blue ribbon) and 3 EGCG. Left : Hydrophilic part of the peptide (in blue). Right: hydrophobic part of the peptide (in red) in which one polyphenol is positioned (dark blue). The interface between the lipo- and hydrophilic region of the peptide is represented by the cream isosurface

Figure 4. Molecular lipophilicity potential contours of PR1 (dark blue ribbon) and 3 EGCG. Left : Hydrophilic part of the peptide (in blue). Right: hydrophobic part of the peptide (in red) in which one polyphenol is positioned (dark blue). The interface between the lipo- and hydrophilic region of the peptide is represented by the cream isosurface

For the PR1PR2 peptide, titration experiments were performed using the same strategy: EGCG was progressively added to a solution of 0.5 mM PR1PR2 to reach a final concentration kept below its CMC (between 0 and 5 mM). The progressive addition of EGCG induces 1H chemical shift variations mainly in two domains of PR1PR2, Ala178/Lys180/Thr181 and Gly207/Thr212 (S 11), which tag the PR1PR2 peptide binding sites. These binding sites were already observed for the two shortest peptides PR1 and PR2. Fitting chemical shift variations of the most responsive residues (mainly, Ala173/ Lys174/ Thr175/ Ala178/ Lys180/ Thr181/ Gly207/ Thr212/ Leu215 and Thr217) with Eq. 2 gives rise to the estimation of the dissociation constant, Kd (0.5±0.1 mM) and the number of fixation sites n (1.5±0.5) (Figure 3, orange). If the affinity of EGCG appears to be higher for the PR1PR2 peptide than for PR1, or PR2 (0.5 mM vs 1.1 mM and 0.9 mM, respectively (15)),the number of EGCG binding sites is quite low (1.5 vs 2.1 for PR1 and 3.1 for PR2).
The DOSY NMR spectra recorded at each EGCG concentration confirms the small amount of EGCG bond to PR1PR2, since no difference was observed between the peptide alone and the peptide with EGCG (close to 1.2±0.15×10-10 m2.s-1) : the difference between the two rates of diffusion between PR1PR2 and PR1PR2 + 1 EGCG is expected to 0.1×10-10 m2.s-1  i.e. within the error of the measurement.
Saturation Transfer Difference (STD) NMR technique was finally used to follow the interactions between EGCG and the PR1PR2 peptide. This technique commonly used for several years to identify binding events (29) by looking at the resonance signals of the ligand, EGCG in our case, permits to elucidate the epitope group of the ligand involved in the interaction. This NMR technique, applicable only on the longest peptide, associated with chemical-shift changes, gives rise to a complete picture of the interaction.
Figure 5 displays the different spectra recorded with this technique: in order to confirm the specificity of the binding, some tests were performed with the peptide and the polyphenol alone (Fig. 5, Spectra (a) and (c)) and their corresponding STD spectra (b) and (d) showing no resonance). Spectrum (g) shows the STD spectrum of a solution containing both EGCG and PR1PR2 in which mainly the resonances of the protons of the two galloyl groups of EGCG (H2’, H6’, H2”, H6”) and to a lesser extend those of the catechol ring (H6, H8) are elicited. A control test has been performed on the same solution in which 1 mM of glucose was added (spectrum (e) for the 1H spectrum of reference and (f) for the corresponding STD spectrum) confirming the specificity of the interaction of EGCG with PR1PR2 observed using STD. STD methods can be used to identify the part of the ligand in contact to the peptide (the nearest protons are more saturated when the saturation transfer occurs): in the present case, the higher level of transfer occurring on the protons of the galloyl groups suggests the involvement of the ester and/or the phenolic OH in the binding with PR1PR2.

Figure 5. STD spectra of a mixture of 0.5mM PR1PR2 and 3 mM EGCG in H2O/D2O at pH 6.2, and chemical structures of EGCG. (a) 1H NMR of the peptide alone, (b) STD NMR of PR1 alone, (c) 1H NMR of EGCG alone, (d) STD NMR of EGCG alone, (e) 1H NMR of EGCG/Peptide/Glucose mixture, (f) STD NMR of EGCG/Peptide/Glucose mixture and (g) STD spectrum of EGCG/peptide mixture.

Figure 5. STD spectra of a mixture of 0.5mM PR1PR2 and 3 mM EGCG in H2O/D2O at pH 6.2, and chemical structures of EGCG. (a) 1H NMR of the peptide alone, (b) STD NMR of PR1 alone, (c) 1H NMR of EGCG alone, (d) STD NMR of EGCG alone, (e) 1H NMR of EGCG/Peptide/Glucose mixture, (f) STD NMR of EGCG/Peptide/Glucose mixture and (g) STD spectrum of EGCG/peptide mixture.



The first part of the present work deals with the total 1H, 13C and 15N NMR assignment of a consequent part of the proline-rich domaine (50 residues over 93, the richest part in proline residues) of the Tau protein that confirms its unfolded nature. This work completes assignments performed first by Smet, Lippens et al.,(23, 30) followed by those done by Narayanan et al.(31) that reach 92 to 96 % assignment depending on the Tau isoforms studied. Another strategy has been developed on the full length Tau 441 protein by Harbison et al.,(24) that allows the assignment of 93% of its backbone. However, all these strategies, while done on full length Tau isoforms, required at least a partial 13C and 15N labeling, and do not permit the assignment of most of the proline residues.
The choice of studying a zone overlapping the two proline regions of Tau (P1 and P2) embedding most of the proline residues (15 over 22), alone or in pairs (4 proline pairs over 6) was guided by two occurrences: it is (i) the place of many biological events whose deregulation entails neurodegenerative pathologies and (ii) a potential target for drugs susceptible to protect the protein against deleterious kinases involved in its hyperphosphorylation and its consequent aggregation under neurofilament tangles (32).
Polyphenols, and especially EGCG, have been reported to have many beneficial effects upon health and specifically to prevent neurodegenerative diseases belonging from tauopathies (33). Several explanations are proposed to explain the beneficial effect of EGCG to fight tauopathies such as Alzheimer. It can act directly on the amyloid fibril formation by binding the native polypeptides and preventing their conversion into toxic amyloid fibrils (34). It was also shown to inhibit Tau aggregation both in vitro (35) and in vivo (8). The potential of EGCG to benefit AD patients, that is currently under investigation (36),  requires more data to understand its action process. This is the second goal of this work, which furnishes molecular or even atomic details of the role that procyandins may play with respect to Tau. In a first study, we tested different galloylated and non galloylated procyanidins (15) on a peptide fragment belonging from the P2 domaine of Tau (PR2 : Gly201-Thr220) : galloylated polyphenols, and mainly EGCG presented a higher affinity towards the peptide, and at least 3 molecules were able to bind PR2 by mean of hydrogen bonds in two distinct regions involving threonines : Thr205-Gly207 and Thr212-Lys215. In the present study, results obtained for the PR1 peptide (Ile171-Lys180) are somehow different: the difference of affinity between non galloylated vs galloylated polyphenols is not as much marked as for PR2. However, EGCG still remains the polyphenol with the highest affinity, and one major binding site was identified at the level of Ser184 and Lys190 mostly through hydrogen bonding as shown by dynamics calculations. Using the longest PR1PR2 peptide, the scenario highlights some interesting differences for EGCG: a higher affinity is observed (Kd close to 0.5 mM) but a number of binding sites reduced to 1.5. Such a phenomenon has been already described by us (37) for saliva proline-rich proteins and explained by the unfolded nature of the peptide which, when long enough, becomes able to “wrap around” the polyphenol, and then reduces its number of binding sites. For the PR1PR2 peptide, the main binding sites were identified at the level of Lys180 and Thr212 as suggested by the maximal amplitude of their NH chemical shifts through EGCG titration. However, we can estimate that the wrapping movement of the peptide around the polyphenol is quite dynamic since no supplementary NOE effect was observed neither between protons belonging from the peptide nor between EGCG and the peptide. On the EGCG side, the interaction mainly occurs at the level of the ester function, especially, a hydrogen bond appears between the oxygen of the carbonyl function acting as an H acceptor and the OH of the Thr or Ser residues or the ε NH2 of Lys acting as H donors. This information obtained from dynamic calculations is strengthened by STD NMR experiments since the protons receiving the highest degree of saturation define the functional group involved in binding. The present study shows that the most responsive protons with respect to STD NMR are those of the galloyl group (H2”/H6”, see fig. 5) at the tight proximity of the carbonyl suspected to be involved.
Galloylated polyphenols, and especially EGCG, obviously exhibit a higher affinity towards proline rich proteins (15, 38) than their non-galloylated counterpart. The analysis of hydrogen bond formations between the proline-rich peptide PR2 and different polyphenols during the dynamic runs performed evidences unexpectedly the greatest capacity of the ester carbonyl of galloylated polyphenols to accept an hydrogen bond as regards to amide carbonyls (15) and can explain the difference observed in affinity. However, that designates the green Tea EGCG as a good candidate to carry on further studies, and notably its ability to prevent peptide phosphorylation and precipitation under kinase attack (work in progress).
Finally, the behavior of the longest PR1PR2 Tau peptide in the presence of EGCG, i.e. its tendency to embed the polyphenol is noteworthy per se, since EGCG can act as a double-edged sword against AD: by preventing kinases attack since their target is the same (Thr/Ser residues in the vicinity of Pro) and by preventing the formation of neurofibrillary tangles due to changes occurring in its 3D structure. These results confirm what Wobsts et al. suspected on a long Tau fragment (39).
In conclusion, this work gives some molecular insights to understand the beneficial effect of polyphenols, and especially EGCG, against tauopathies such as AD. In fact, polyphenols in general, and more specifically EGCG, have interested the scientific community since the beginning of the millennium in their therapeutic efficacy, mainly in dementia, but few reports aim to explain the molecular basis of their action. In light of our work, it is noteworthy that polyphenols of the procyanidin family are able to bind the proline-rich domaine of Tau. While depending on their chemical nature (15), their affinity is weak but suitably relevant to compete all the more with deleterious kinases (40) since binding takes place in the region targeted by kinases. Lastly, the conformation preference, that the longest peptide is able to adapt when complexing the polyphenol, may help to modulate its pathologic mechanism of aggregation.


Acknowledgements: The authors want to thank the Région Aquitaine for supporting equipment of CESAMO and IECB, the Université de Bordeaux and the CNRS for their financial helps and the TGIR-RMN-THC Fr3050 CNRS for conducting the 800 MHz NMR experiments. Marie Guéroux and Charlotte Fleau gratefully acknowledge the Ministère de l’Education Nationale for their PhD grants. Isabelle Pianet gratefully thanks Nathan McClenaghan for the English improvement of the manuscript. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflict of Interest: The authors declare to have no conflict of interests.



1.     Khan N, Mukhtar H. Tea polyphenols for health promotion. Life Sci. 2007;81:519-33.
2.    Sunphenon EGCg (Epigallocatechin-Gallate) in the Early Stage of Alzheimer´s Disease (SUN-AK) [Internet]. 2009.
3.    Singh NA, Mandal AKA, Khan ZA. Potential neuroprotective properties of epigallocatechin-3-gallate (EGCG) Nutrition J. 2016;15.
4.    Rosini M, Simoni E, Milelli A, Minarini A, Melchiorre C. Oxidative Stress in Alzheimer’s Disease: Are we connecting the dots? J Med Chem. 2014;57:2821-31.
5.    Belaidi AA, Bush AI. Iron neurochemistry in Alzheimer’s disease and Parkinson’s disease: targets for therapeutics. J Neurochem. 2016;139:179-97.
6.    Ward RJ, Dexter DT, Crichton RR. Chelating agents for neurodegenerative diseases. Curr Med Chem. 2012;19:2760-72.
7.    Sohma Y. Medicinal chemistry focusing on aggregation of amyloid-beta. Chem Pharm Bull. 2016;64:1-7.
8.    Wang J-Z, Santa-Maria I, Ho L, Ksiezak-Reding H, Ono K, Teplow DB, et al. Inhibition of tau aggregation by grape derived polyphenols in a mouse model of Alzheimer’s disease. J Alzheimers Dis. 2010;22:653-61.
9.    Pasinetti GM. Novel role of Red-wine-derived polyphenols in the prevention of Alzheimer’s disease dementia and brain pathology: experimental approaches and clinical implications. Planta Med. 2012;78:1614-9.
10.    Bulic B, Pickhardt M, Schmidt B, Mandelkow E-M, Waldmann H, Mandelkow E. Development of tau aggregation inhibitors for Alzheimer’s disease. Angew Chem Int Ed. 2009;48:1740-52.
11.    Simic G, Leko MB, Wray S, Harrington C, Delalle I, Jovanov-Milosevic N, et al. Tau Protein Hyperphosphorylation and Aggregation in Alzheimer’s Disease and Other Tauopathies, and Possible Neuroprotective Strategies. Biomolecules. 2016;6:1-28.
12.    Bulic B, Pickhardt M, Mandelkow E. Progress and developments in Tau aggregation inhibitors for Alzheimer Disease. J Med Chem. 2013;56:4135-55.
13.    Bertrand J, Plouffe V, Sénéchal P, Leclerc N. The pattern of human tau phosphorylation is the result of priming and feedback events in primary hippocampal neurons Neuroscience. 2010;168:322-34.
14.    Guéroux M, Laguerre M, Slozek-Pinaud M, Fouquet E, Pianet I. Polyphenols and Alzheimer’s Disease: Tau/Polyphenol interactions investigated by NMR and Molecular Modeling. Nutrition & Aging. 2012;1:201-6.
15.    Guéroux M, Pinaud-Szlosek M, Fouquet E, de Freitas V, Laguerre M, Pianet I. How wine polyphenols can fight Alzheimer disease progression: Towards a molecular explanation. Tetrahedron. 2015;71(20):3163-70.
16.    Fabre S, Guéroux M, Nunes E, Slozek-Pinaud M, Pianet I, Fouquet E. From the gold-catalysed benzylation of arenes to the regio- and stereoselective synthesis of procyanidins dimers. Tetrahedron. 2015;71:3045-51.
17.    Charlton AJ, Baxter NJ, Lilley TH, Haslam E, McDonald CJ, Williamson MP. Tannin interactions with a full-length human salivary proline-rich protein display a stronger affinity than with single proline-rich repeats. FEBS Letters. 1996;382:289-92.
18.    Simon C, Barathieu K, Laguerre M, Schmitter JM, Fouquet E, Pianet I, et al. Three-Dimensional Structure and Dynamics of Wine Tannin-Saliva PProtein Complexes. A Multitechnique Approach. Biochemistry. 2003;42:10385-95.
19.    Van Gunsteren WF, Berendsen HJC. Gromos-87 Manual. In: Biomos BV, editor. Groningen, The Netherlands1987.
20.    Cala O, Pinaud N, Simon C, Fouquet E, Laguerre M, Dufourc EJ, et al. NMR and Molecular modeling of wine tannins binding to saliva proteins: revisiting astringency from molecular and colloidal prospects. FASEB J. 2010;24(11):4281-90.
21.    Wüthrich K. The way to NMR structure of proteins. Nature Struct Biol. 2001;8:923-5.
22.    Wishart D, Bigam CG, Holm A, Hodges RS, Sykes BD. 1H, 13C and 15N random coil chemical shifts of the common amino acids. I. Investigations of nearest-neighbor effects. J Biomol NMR. 1995;5:67-81.
23.    Smet C, Leroy A, Sillen A, Wieruszeski J-M, Landrieu I, Lippens G. Accepting its random coil nature allows a partial NMR assignment of the neuronal tau protein. ChemBioChem. 2004;5:1639-46.
24.    Harbison NW, Bhattacharya S, Eliezer D. Assigning Backbone NMR resonances for full length Tau isomeforms: Efficient compromise between manual assignments and reduced dimensionality. PLoS ONE. 2012;7(4):e34679.
25.    Narayanan RL, Dürr UHN, Bibow S, Biernat J, Mandelkow E, Zwecstetter M. Automatic assignment of the intrinsically disordered protein Tau with 441-residues. J Am Chem Soc. 2010;132:11906-7.
26.    Kadavath H, Jaremko M, Jaremko L, Biernat J, Mandelkow E, Zwecstetter M. Folding of the Tau protein on microtubules. Angew Chem Int Ed. 2015;54:10347-51.
27.    Mukrasch MD, Bibow S, Korukottu J, Jeganathan S, Biernat J, Griesinger C, et al. Structural polymorphism of 441-residue Tau at single residue resolution. PLoS ONE. 2009;7(2):e1000034.
28.    Augé S, Schmitt P-O, Crutchfield A, Islam MT, Harris DJ, Durand E, et al. NMR Measure of Translational Diffusion and Fractal Dimension. Application to Molecular Mass Measurement. J Phys Chem B. 2009;113:1914-8.
29.    Meyer B, Peters T. NMR Spectroscopy Techniques for screening and identifying ligand binding to protein receptors. Angew Chem Int Ed. 2003;42:864-90.
30.    Lippens G, Sillen A, Smet C, Wieruszeski J-M, Leroy A, Buée L, et al. Studying the natively unfolded neuronal Tau protein by solution NMR spectroscopy. Protein Pept Lett. 2006;13(3):235-46.
31.    Narayanan RL, Dürr UHN, Bibow S, Biernat J, Mandelkow E, Zweckstetter M. Automatic Assignment of the Intrinsically Disordered Protein Tau with 441-Residues. J Am Chem Soc. 2010;132:11906-7.
32.    Buée L, Bussière T, Buée-Scherrer V, Delacourte A, Hof PR. Tau protein isoforms, phosphorylation and role in neurodegenerative disorders. Brain Res Rev. 2000;33:95-130.
33.    De La Torre R, Dierssen M. Therapeutic approaches in the improvement of cognitive performance in Down syndrome: past, present, and future. Progress Brain Res. 2012;197:1-14.
34.    Ehrnhoefer DE, Bieschke J, Boeddrich A, Herbst M, Masino L, Lurz R, et al. EGCG redirects amyloidogenic polypeptides into unstructured, off-pathway oligomers. Nat Struct Mol Biol. 2008;15(6):558-66.
35.    Santa-Maria I, Diaz-Ruiz C, Ksiezak-Reding H, Chen A, Ho L, Wang J, et al. GSPE interferes with tau aggregation in vivo: implication for treating tauopathy. Neurobiol Aging. 2012;33(2072-2081).
36.    Molino s, Dossena M, Buo o, D., Ferrari F, Venturi L, Ricevuti MD, et al. Polyphenols in dementia: from molecular basis to clinical trials. Life Sciences. 2016;15:69-77.
37.    Cala O, Dufourc EJ, Fouquet E, Manigand C, Laguerre M, Pianet I. The colloidal state of tannins impacts the nature of their interaction with proteins: The case of Salivary Proline-Rich Protein/procyanidins binding. . Langmuir. 2012;28(50):17410-8.
38.    Faurie B, Dufourc EJ, Laguerre M, Pianet I. Monitoring the Interactions of a Ternary Complex Using NMR Spectroscopy: The Case of Sugars, Polyphenols, and Proteins. Anal Chem. 2017;88(24):12470-8.
39.    Wobst HJ, Sharma A, Diamond MI, Wanker EE, Bieschke J. The green tea polyphenol epigallocatechin gallate prevents the aggregation of Tau protein into toxic oligomers at substoichiometric ratios. FEBS Lett. 2015;589:77-83.
40.    Ball LJ, Khüne R, Schneider-Mergener J, Oschkinat H. Recognition of proline-rich motifs by protein-protein-interaction Domains. Angew Chem Int Ed. 2005;44:2852-69.