Simufilam in Alzheimer’s Disease: Assessment of Efficacy of a Controversial Drug in Human Neuronal Cell Culture

TL;DR

This study examines why simufilam, a drug for Alzheimer's, failed in clinical trials by testing its effects on human neuronal cells.

Contribution

The study provides new insights into simufilam's molecular mechanisms and lack of efficacy in Alzheimer's treatment.

Findings

Simufilam increased β-secretase protein at 50 µM in differentiated SH-SY5Y cells.

Simufilam reduced brain-derived neurotrophic factor protein levels in differentiated SH-SY5Y cells.

Simufilam had no effect on amyloid processing genes or mitochondrial function.

Abstract

Background/Objectives: Alzheimer’s disease (AD) is a progressive multifactorial neurodegenerative disorder. Current AD therapies offer minimal benefits and do not prevent or repair neuronal damage. More effective therapeutic approaches are needed to restore normal bioenergetics and metabolic function to AD neurons. Simufilam is a small-molecule oral drug that targets filamin A, a scaffolding protein in brain cells. Phase III clinical trials of simufilam failed to show any significant cognitive or functional improvements in AD patients. The purpose of this study is to identify and explain the molecular mechanisms that may have contributed to this drug’s lack of clinical success. Methods: Our study investigates the effects of simufilam on amyloid processing, neuronal health, and mitochondrial functioning in the SH-SY5Y human neuronal cell model. SH-SY5Y cells were differentiated into neurons using 10 µM retinoic acid. Undifferentiated and differentiated SH-SY5Y were exposed to simufilam (5 µM, 50 µM; 24 hr). Results: Simufilam did not affect the expression of genes involved in amyloid processing. Amyloid precursor protein (APP), β-secretase, and α-secretase mRNA levels in simufilam-treated SH-SY5Y cells were all unchanged compared to untreated cells. However, amyloidogenic β-secretase protein was significantly increased (fold change 1.17) at 50 µM of simufilam only in differentiated SH-SY5Y cells without affecting APP or α-secretase protein expression. Simufilam at the 50 µM concentration reduced brain-derived neurotrophic factor protein levels (fold change 0.7) only in differentiated SH-SY5Y. Further, simufilam did not improve mitochondrial genes or structure. Conclusions: Our results align with clinical outcomes and indicate that insufficient activity across multiple tests of ability to impact processes related to neuronal health can serve as a preliminary indicator of limited clinical utility.

Figures

Full text

1. Introduction

Alzheimer’s disease (AD) is one of the most common neurodegenerative disorders. Pathological hallmarks observed in the AD brain include deposition of extracellular plaques composed of aggregated amyloid-β (Aβ) protein as well as intracellular accumulation of neurofibrillary tangles consisting of hyperphosphorylated tau protein, accompanied by neuronal degeneration and loss of memory [1,2]. The primary symptoms of AD are irreversible memory loss, cognitive decline, and changes in mood and behavior, ultimately leading to incapacity and death [3]. It is estimated that more than 7 million people in America have this disease and most of them are over the age of 65 years [4]. In 2024, AD constituted the sixth leading cause of mortality in the United States, exceeding diabetes, which was the seventh leading cause. The FDA-approved AD drugs cholinesterase inhibitors and memantine provide only symptomatic relief, but several newer Aβ-targeting immunotherapies may be disease-modifying and able to delay cognitive decline [5,6,7]. The failure of numerous clinical trials of immunotherapies such as those targeting tau have led scientists to move in new directions in order to develop effective treatment interventions [8,9].

Among the attempts to devise new AD treatments, simufilam is a small-molecule drug candidate that moved from inception to clinical testing but ultimately failed to show efficacy [10]. The theoretical rationale for treating AD with simufilam was its ability to bind to and restore the native conformation of filamin A (FLNA), which is structurally abnormal in the AD brain [11]. FLNA is a scaffolding protein and regulator of the actin cytoskeleton [12]. Altered FLNA links with Aβ and the α7 nicotinic acetylcholine receptor (α7nAChR), thus resulting in toxic signaling within neurons, which triggers tau phosphorylation and synaptic dysfunction, neurodegeneration and intracellular amyloid plaques [13,14,15]. The binding of simufilam to FLNA dissociates its binding to Aβ and α7nAChR, thus disrupting toxic pathways and preventing toll-like receptor (TLR)4-mediated neuroinflammation [13,16]. Several preclinical studies have demonstrated potentially beneficial effects of simufilam in AD models, including reversal of abnormal FLNA receptor interactions associated with pathology [16]. In addition, data suggesting potentially positive effects of the drug have been reported in experimental systems and analytical reviews [12,17]. Oral administration of simufilam for two months improved synaptic plasticity and reduced receptor dysfunction in transgenic AD murine models [15].

In November 2024, the biopharma company Cassava Sciences announced that in the REFOCUS-ALZ Phase 3 clinical trial (NCT05026177), simufilam did not demonstrate a statistically significant improvement in cognitive function compared with placebo after 76 weeks [18]. Simufilam did not meet its prespecified primary, secondary, or exploratory endpoints according to the company although it was generally well-tolerated with no serious adverse effects reported [19]. In addition to a lack of symptom reduction, there were no measurable changes in biomarkers associated with AD. The company has also faced scrutiny following allegations concerning data integrity, which led to charges from the Securities and Exchange Commission [20].

Conceptually, the application of simufilam to promote restoration of the native conformation of FLNA and attenuation of pathological receptor interactions associated with Aβ toxicity is scientifically plausible [21]. However, the modest results of anti-amyloid antibody therapy in humans underscore the need to consider a multi-faceted approach to AD treatment encompassing tau pathology, neuroinflammation, synaptic disruption, vascular insufficiency, and cytoskeletal dysfunction [22]. An Aβ mitigation strategy involving FLNA would likely have a place in this regimen.

The goal of the present study is to gain insight into the factors that contribute to the lack of efficacy of simufilam by exploring its effect on the expression of genes pertaining to AD and neuronal health using a human neuronal cell culture model. Mechanistic clarity can inform future studies and deepen the understanding of critical pathways necessary for improving the prognosis in AD.

2. Results

2.1. Simufilam Treatment and Genes Involved Aβ Formation

The effect of simufilam on amyloidogenic pathway genes was assessed by measuring the expression of amyloid precursor protein (APP), β-secretase (BACE)1 and α-secretase (ADAM10) in both undifferentiated and differentiated SHSY-5Y cells at both the mRNA and protein levels. Compared to control, simufilam treatment at 5 µm and 50 µM concentrations did not alter the mRNA abundance of APP, BACE1 or ADAM10 in either the undifferentiated or differentiated SHSY-5Y cells (Figure 1A,B). Further, immunoblotting was employed to assess the expression of amyloidogenic APP and BACE1 and non-amyloidogenic ADAM10 at the protein level in undifferentiated and differentiated SHSY-5Y cells in the presence and absence of simufilam. Western blots showed that simufilam did not affect APP or ADAM10 protein levels but did significantly elevate BACE1 protein at the 50 µM concentration in differentiated cells (Figure 1C,D).

2.2. Simufilam Treatment and Genes Involved in Neuronal Health

To investigate the role of simufilam in neuronal health, we checked the expression of synaptophysin and brain-derived neurotrophic factor (BDNF) in undifferentiated and differentiated SHSY-5Y cells with (5 µm, 50 µM) and without (control) simufilam treatment. Real-time PCR analysis showed that mRNA levels of synaptophysin and BDNF did not change significantly with simufilam exposure in either the undifferentiated or differentiated SHSY-5Y cells (Figure 2A,B). We also quantified the protein levels of synaptophysin and BDNF in both undifferentiated and differentiated cells. Western blot analysis showed that simufilam did not change the protein levels of synaptophysin at either of the concentrations, but BDNF protein levels were significantly reduced at the 50 µM of simufilam concentration in differentiated SHSY-5Y cells (Figure 2C,D).

2.3. Simufilam Treatment and Genes Involved in Mitochondrial Function

To document changes in the expression of genes relevant to mitochondrial functioning induced by simufilam treatment, we measured the expression of TFAM, NRF1 and PGC1α at the mRNA and protein levels in the presence and absence of the drug. Compared to control, treatment of simufilam to undifferentiated and differentiated SHSY-5Y cells has no effect on mRNA and protein levels of TFAM, NRF1 and PGC1α at any concentration (Figure 3).

2.4. Simufilam Treatment and MitoTracker Staining

To assess mitochondrial health and biogenesis, we stained undifferentiated and differentiated SHSY-5Y cells with MitoTracker staining after simufilam treatment. Images showed a decrease in active mitochondria at 50 µM dose in undifferentiated SHSY-5Y cells but no difference in active mitochondria in differentiated SHSY-5Y cells following simufilam treatment (Figure 4).

2.5. Electron Microscopy

To determine mitochondrial morphology, transmission electron microscopy was done after 24 h of simufilam treatment versus untreated control. Transmission electron microscopy images show that simufilam treatment did not affect mitochondrial morphology in undifferentiated and differentiated SHSY-5Y cells. We also measured the mitochondrial area in the TEM images. Simufilam treatment did not change the mitochondrial area in undifferentiated and differentiated SHSY-5Y cells (Figure 5).

3. Discussion

In the present study, we evaluated whether simufilam exhibits neuroprotective effects on undifferentiated and differentiated SHSY-5Y human neuronal cells via examination of critical gene expression and morphological parameters. Although our assessment of mitochondrial morphology is limited to changes in area, this measurement is widely used to assess mitochondrial health, dynamics, and cellular stress [23,24,25]. Exposure to simufilam did not incite any positive change in neuronal health and mitochondrial functioning in SHSY-5Y cells, irrespective of their differentiation state. Further, a lack of positive change in BDNF is a robust indicator of a lack of effectiveness since BDNF is considered a key barometer of nerve health and synaptic function [26]. To our knowledge, our study of SH-SY5Y cells treated with simufilam is the first to shed light on the lack of positive changes in neuronal health and mitochondrial functioning with this drug in human neurons. The SH-SY5Y cell type is commonly used in vitro as a model for the behavior of human brain neurons in response to drug treatments [27]. The differentiated and undifferentiated SH-SY5Y cells are extensively utilized in neurological studies related to Alzheimer’s and Parkinson’s diseases [28,29,30,31,32].

Simufilam is an oral drug candidate for mild-to-moderate AD that disrupts the upstream pathway involved in hyperphosphorylation of tau proteins and aggregation of amyloid as well as neuroinflammatory pathways [12,33]. Simufilam was postulated to work by restoring normal conformation to the scaffolding protein filamin A, which is altered in AD neuropathogenesis. The binding of simufilam restored filamin A to native conformation thus preventing destructive AD-associated protein interactions [16]. Wang et al. showed that simufilam reduced Aβ-42 binding to α7nAChR and reduced inflammatory cytokine release in human astrocytes stimulated by Aβ-42 [16]. Simufilam improved insulin sensitivity in lymphocytes of AD patients by reducing mTOR signaling [12].

Cassava Sciences invested tremendous resources into simufilam. They started a Phase 1 clinical trial of simufilam (also known as PTI-125) in 24 healthy adults in 2017. In 2019, the company ran an NIH-funded Phase 2a trial in people with mild-to-moderate AD. Wang et al. in 2020 published the results of the Phase 2a trial of simufilam showing a reduction in biomarkers, including t-tau and p-tau181, in cerebrospinal fluid (CSF) over 28 days in persons with mild-to-moderate AD [33]. Simufilam suppressed hyperphosphorylation of tau protein induced by Aβ-42’s signaling through α7-nicotinic acetylcholine receptor. The decrease in CSF total and phosphorylated tauT181 (pT181) and inflammatory biomarkers were observed with treatment [33]. The company ran an NIH-funded Phase 2b study from September 2019 to March 2020 in 64 participants with mild-to-moderate AD. The data showed that both 50 and 100 mg daily doses led to improvements in all CSF biomarkers including total tau, p-tau181, and neurogranin. Aβ-42 in CSF was reported to be increased by 10%.

In November 2021, the company began two Phase 3 clinical trials: ReThink-ALZ and ReFocus-ALZ. In November 2024, the company announced that simufilam did not show any significant reduction in cognitive decline in patients with mild-to-moderate AD when compared to placebo in the ReThink-ALZ Phase 3 study. Simufilam failed to meet primary, secondary, or exploratory endpoints in the Phase 3 ReThink-ALZ study. Similarly, in the REFOCUS-ALZ trial, simufilam failed to show any significant cognitive or functional improvements. In March 2025, the company decided that it would discontinue all efforts to develop simufilam for mild-to-moderate AD and phase out the program by the end of June 2025 [10].

Simufilam was of special interest because it is one of the few putative AD treatments not solely targeting either amyloid or tau that reached the stage of human trials. While accumulation of Aβ plaques within the brain is a major hallmark of AD, more and more research shows that it is increasingly clear that Aβ accumulation is not the primary causal mechanism of AD [34,35,36]. In a recent editorial, Forlenza et al. address the repeated failure of clinical trials of anti-amyloid drugs in AD treatment [37].

Alternative strategies to bring about a disease-modifying change in the course of AD could attain success by exerting neuroprotective or anti-inflammatory effects [38,39]. The impairment of mitochondrial functioning is a major factor associated with the pathogenesis of AD. This includes oxidative stress-induced respiratory chain dysfunction, loss of mitochondrial biogenesis, mitochondrial DNA mutations, and impaired mitochondrial transport and mitophagy [2,40,41,42]. These processes induce accumulation of damaged mitochondria, failure of bioenergetics, and neuroinflammation, which further leads to neurodegeneration [41]. Thus, emerging evidence suggests that targeting mitochondrial dysfunction works as an alternative strategy for the treatment of AD. Reduced levels of mitochondrial proteins such as NRF and TFAM have been reported in mouse models of AD as well as post mortem brain tissue of AD patients [43,44,45]. Overexpression or pharmacological activation of mitochondrial proteins such as NRF2 and TFAM reduces oxidative stress and increases oxidative phosphorylation activity and mitochondrial mass in AD mouse models [46,47]. In the present study we did not detect any positive effect of simufilam on mitochondrial biogenesis, morphology, or functioning in SH-SY5Y neuronal cells. In fact, mitochondrial biogenesis was reduced in undifferentiated SH-SY5Y cells upon exposure to simufilam at the 50 µM concentration. We conducted a similar study of the cancer therapeutic drug nilotinib, a tyrosine kinase inhibitor that is being considered for repurposing in AD treatment because it can reduce Aβ and tau in animal models. In a Phase 2 clinical trial, Turner et al. demonstrated that nilotinib is safe and reduces amyloid burden in mild-to-moderate AD patients [48]. Our group recently published an analysis of nilotinib that found no effect on amyloid formation, mitochondrial function, or neuronal health in our human neuronal cell model [49]. Results of the clinical trial are still pending [50].

Our study found that simufilam did not improve key features of neuronal health such as mitochondrial activity or morphology nor did it increase BDNF or synaptophysin and we hypothesize that this is the reason it is ineffective despite targeting amyloid and inflammation in prior preclinical studies [15,16]. Changes in amyloid or tau alone may be effective in mouse AD models but translation to human studies has yielded limited success and many attempts to modify AD trajectory via neuroinflammatory pathways have not produced tangible benefits [51,52]. It is not uncommon for drugs with early promise based on mechanism to exhibit outcome dissociation between molecular target engagement and clinical efficacy in AD. Other examples include the neuroprotective peptide davunetide, which reduces tau hyperphosphorylation and the microtubule stabilizer epothilone D, both of which had positive effects in preclinical AD models, but did not show benefits in human AD trials [53,54,55,56,57].

In order for a treatment to modulate progression of AD it is critical that it prevents neuronal loss, likely by preserving respiration and bioenergetics dependent upon healthy mitochondria. Our evidence suggests that simufilam has not demonstrated effects. The importance of energy metabolism and mitochondria in AD neurodegeneration is supported by clinical and post mortem data [58]. AD brains show compromised mitochondrial respiration and early and substantial declines in the activity of complex IV of the mitochondrial electron transport chain (cytochrome c oxidase) and oxygen metabolism in affected regions [59,60,61]. Early claims of disease modification without reproducible, mechanistically coherent human cell and biomarker data can be misleading and, in this case, led to consumption of valuable medical and scientific resources. There is a consensus that multimodal treatment will be necessary if the goal is true neuronal preservation rather than modest symptom slowing and, in this setting, simufilam and FLNA might have a role [62]. The Alzheimer’s Disease Neuroprotection Research Initiative (ADNRI) supports the perspective that neuronal death prevention via multiple approaches is critical for AD treatment [63].

The limitations of the current study are related to the use of a cell culture approach performed in isolated neuronal cultures. Further, we did not examine neuronal–glial co-cultures despite the potential for glial cells to play a role in AD-modulating effects of simufilam as glial cells are involved in generating neuroinflammation and maintaining neuronal function [64]. The physiological relevance of this work would be further enhanced with studies involving Aβ42-induced stress to determine the impact of the drug against what is considered a key pathological process in AD. We also note that mitochondrial morphology was assessed by measuring area only. Our dataset (based on two-dimensional transmission electron microscopy images) captures only cross-sections of mitochondria and was not optimized for robust quantification of three-dimensional shape metrics such as the length-to-width ratio or fragmentation index.

4. Materials and Methods

4.1. Cell Culture and Simufilam Treatment

SHSY5Y cells (American Type Culture Collection), a human neuroblastoma cell line, were maintained in DMEM/F-12 media at 37 °C in a 5% CO_2_ atmosphere. Media was supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine and 1% penicillin–streptomycin. Each well was seeded with 300,000 cells, which were then differentiated into neurons using 10 µM retinoic acid with 2% FBS for 6 days. Undifferentiated and differentiated SHSY-5Y cells were treated with 0 (control), 5 and 50 µM simufilam (SELLECK Chemicals Houston, TX, USA) for 24 h. Simufilam concentrations of 5 μM and 50 μM in cell culture were determined by aligning with human pharmacokinetics while accounting for experimental constraints. Concentrations were based on the simufilam clinical protocol from Cassava Sciences, which projected a maximum plasma concentration of 1100 ng/mL (4.2 µM) [65]. We used this as a baseline and added a higher dose to compensate for the short exposure time compared to the situation in human studies where drugs are administered over weeks to months, aiming to achieve comparable pharmacodynamic effects despite reduced cumulative exposure. Following 24 h of simufilam exposure at either concentration, the cells showed a normal, spread and adherent morphology without rounding, blebbing, or detachment.

4.2. RNA Isolation and Real-Time PCR

Total RNA from simufilam treated SHSY-5Y cells was isolated using TRIzol reagent and reconstituted in DEPC water. The quality and quantity of the total RNA was analyzed using a Nanodrop Spectrophotometer to measure absorption at 260–280 nm. An amount of 1 µg of total RNA along with 25 µM MgCl2, 10XPCR buffer, 10 nM dNTP, 50 µM hexamers, RNAse A inhibitor, and M-MuLV reverse transcriptase were used to reverse transcribe into cDNA in the Eppendorf Mastercycler nexus thermocycler PCR machine (Millipore-Sigma, St Louis, MO, USA). The PCR protocol was set as: primer annealing at 25 °C for 5 min, reverse transcription at 42 °C for 60 min, enzyme inactivation at 80 °C for 5 min, hold upon completion at 4 °C. cDNA were then subsequently used for quantitative real-time PCR analysis using the FastStart SYBR Green Reagents Kit (Millipore-Sigma, St Louis, MO, USA). The samples were plated in duplicates on a MicroAmp Optical 96-well plate (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and run on the QuantStudio 3 PCR (Thermo Fisher Scientific, Inc., Waltham, MA, USA) with 40 cycles of amplification. Human primers utilized in this study are listed in Table 1 with their respective melting temperatures (Tm). Expression levels were analyzed using the double delta Ct method (2^−ΔΔCt^) and normalized to the housekeeping gene GAPDH.

4.3. Western Blotting

After 24 h of simufilam treatment, SHSY-5Y cell lysates were collected using the RIPA lysis buffer system supplemented with protease and phosphatase inhibitors (Santa Cruz Biotechnology, Dallas, TX, USA). Protein was extracted from cell lysates and concentration was measured using the Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA). Protein was resolved using 8–12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and then transferred onto polyvinylidene difluoride membranes. The membranes were blocked in 5% non-fat dry milk (Cell Signaling Technology, Danvers, MA, USA) for 1 h and then subsequently incubated in primary antibody overnight at 4 °C, followed by appropriate secondary antibody incubation for 2 h at room temperature. Target proteins were detected using ECL Western blotting detection reagents on the Bio-Rad ChemiDoc Touch Imaging System (Hercules, CA, USA). GAPDH was used as an internal loading control to validate the loading of each lane. Protein levels were quantified using ImageJ software (version 1.53 National Institutes of Health, Bethesda, MD, USA). The band intensity of all corresponding bands of protein was measured by using ImageJ software. For densitometry normalization, the ratio of band intensity of protein of interest to internal control was calculated followed by determining the fold change relative to control values. The final normalized values were used to generate graphic representations.

4.4. Transmission Electron Microscopy

Simufilam-treated SHSY-5Y cell culture samples were fixed using freshly prepared 2.5% glutaraldehyde in distilled water and buffered with 0.1 M sodium cacodylate (pH 7.5). After fixation, samples were washed in 0.1 M sodium cacodylate buffer and stained with 1% osmium tetroxide buffered in sodium cacodylate. Cells were then removed from the culture dish, centrifuged, and stained with a saturated solution of uranyl acetate in 40% ethanol. The sample was then dehydrated in a graded series of ethanol, infiltrated in propylene oxide with Epon epoxy resin (Embed812, Electron Microscopy Sciences, Hatfield, PA, USA), and embedded. The sample embedded blocks were then sectioned with a Reichert Ultracut microtome at 70 nm. Sections were picked up on fine, perforated 300 mesh copper grids for support and stability. They were then dried and post-stained for enhanced contrast with heavy metal stains consisting of 1% aqueous uranyl acetate followed by 0.5% aqueous lead citrate. Stained grids were examined and photographed on a transmission electron microscope (model Zeiss EM 900) retrofitted with an SIA L3C digital camera (Scientific Instruments and Applications, Duluth, GA, USA).

4.5. MitoTracker Staining

Following 24 h of incubation in simufilam or vehicle, live cells were washed with Hanks’ Balanced Salt Solution (HBSS) and stained with 250 µM of MitoTracker red dye (Thermo Fisher Scientific, M7512) for 20 min. After staining, three washes with HBSS were performed on the cells. Images were acquired by fluorescence microscopy at a magnification of 10×.

4.6. Statistical Analysis

Data are reported as mean ± standard deviation (SD). Statistical significance was determined through one-way ANOVA followed by Bonferroni’s multiple comparison test. Data analysis was conducted using GraphPad Prism (GraphPad Software, Version 10.6.1, La Jolla, CA, USA). Fold change was calculated between control and simufilam-treated groups. Residuals of the fit models were visually inspected to confirm compliance with parametric assumptions. p values < 0.05 were considered significant.

5. Conclusions

Based on our evaluation of simufilam, we hypothesize that its disappointing clinical performance reflects its lack of potency in changing gene expression related to key pathways of mitochondrial robustness, amyloid generation and neuroprotection in neuronal cells of human origin. The only mitochondrial effect of note was a negative one in which the higher concentration of simufilam diminished mitochondrial biogenesis in undifferentiated SHSY-5Y. In the future, it might be helpful to test and screen new drug candidates for AD clinical trials in a human cell culture model in order to accumulate valuable data prior to committing huge resources to large-scale clinical studies. Developing new and effective strategies for AD treatment will require the proper design of experiments, appropriate selection of eligible patients, optimized drug dosage and better compounds.