Hidden Androgen Receptor Isoform Shows Promise Against Kennedy’s Disease

Scientists studying Kennedy’s disease, a rare and currently untreatable neuromuscular disorder, have uncovered a surprising ally hidden within the very protein that causes the illness. A new study published in the Journal of Advanced Research reports that a shorter, naturally occurring isoform of the androgen receptor, known as AR-A, can dramatically reduce the toxicity of the mutant protein responsible for the disease, offering a potential therapeutic route that sidesteps the severe hormonal side effects of existing strategies.

Spinal and bulbar muscular atrophy, or SBMA, is an X-linked hereditary condition that affects only males, typically manifesting between the third and fifth decades of life. It arises from an expansion of CAG trinucleotide repeats in the first exon of the androgen receptor gene. While healthy individuals carry between 9 and 36 repeats, patients carry more than 37, with up to 72 repeats reported. This expansion is translated into an abnormally long polyglutamine tract in the receptor protein, placing SBMA among the CAG/polyglutamine family of neurodegenerative diseases. The global prevalence is estimated at roughly 2.58 per 100,000 males, but the true figure is likely far higher, since population studies suggest a four-fold prevalence of carriers of pathological CAG repeats, many of whom never develop symptoms.

The disease mechanism is intimately tied to male hormones. In the absence of testosterone, the androgen receptor is held in the cytoplasm by chaperone proteins such as HSP90, HSP40 and HSP70, which keep the ligand-binding pocket in a high-affinity conformation. When testosterone or its metabolite dihydrotestosterone binds, the receptor dissociates from its chaperones, dimerizes, exposes its nuclear localization signal and moves into the nucleus to regulate transcription. In the mutant receptor, the elongated polyglutamine tract disrupts these processes, promoting misfolding and aggregation in both neuronal and muscle tissues. This toxic gain of function depletes the cell’s protein quality control machinery and alters transcription, ultimately driving the loss of lower motor neurons in the brainstem and spinal cord and the progressive wasting of skeletal muscle.

Crucially, the androgen receptor gene does not produce just one protein. Alternative use of start codons generates multiple isoforms. The canonical full-length receptor, AR-B, is translated from the first AUG codon and contains the polyglutamine tract within its N-terminal domain. A second AUG codon, located downstream of the CAG repeat, drives production of AR-A, an 87-kilodalton isoform that lacks the first 188 amino acids of the N-terminal domain, including the entire polyglutamine tract. First described in 1994 in human genital skin fibroblasts, AR-A has remained poorly characterized, largely because available antibodies cannot easily distinguish the two isoforms. In androgen-responsive tissues, AR-B predominates at a ratio of roughly 10:1, with AR-A generally accounting for less than 20 percent of total receptor.

The research team, led by investigators at the University of Milan together with collaborators in Italy, France and the United States, set out to map where AR-A is expressed and to determine whether it behaves like its toxic mutant counterpart. Using western blot analysis of tissues from wild-type mice at one, three and six months of age, they found that AR-B is abundant in the testis, prostate, skeletal muscle, heart, brainstem and spinal cord, whereas AR-A was detectable only in the spinal cord and brainstem. In these neural tissues, the AR-A to AR-B ratio was notably elevated. Extending the analysis to a knock-in mouse model carrying 113 glutamine repeats in the androgen receptor, the researchers observed a statistically significant increase in the AR-A to AR-B ratio in the spinal cord and brainstem of the diseased animals, hinting that the shorter isoform may be part of an endogenous response to the pathology.

The team then compared the biochemical behavior of AR-A, AR-B and the mutant polyglutamine-expanded receptor in NSC-34 motor neuron-like cells. Under basal conditions, all three isoforms showed diffuse cytoplasmic localization, indicating that the missing 188 amino acids do not disrupt the receptor’s native conformation or its interaction with chaperones. Upon treatment with dihydrotestosterone, AR-B translocated completely into the nucleus, while the mutant receptor formed large aggregates and only partially entered the nucleus. Strikingly, AR-A translocated fully into the nucleus and formed no aggregates whatsoever, confirming that the absence of the polyglutamine tract is both necessary and sufficient for normal behavior. In transcriptional assays using a luciferase reporter driven by androgen response elements, AR-A activated gene expression at levels comparable to the mutant receptor, roughly 60 to 70 percent of full-length AR-B activity, likely reflecting the partial loss of the N-terminal AF-1 activation domain.

Filter retardation assays revealed that while the mutant receptor accumulated abundant high-molecular-weight oligomeric species upon androgen treatment, AR-A formed very few such species under any condition. The researchers also dissected the degradative pathways acting on AR-A: the inactive monomeric form was cleared primarily by the proteasome, whereas the activated dimeric form was preferentially degraded by autophagy, mirroring the pattern previously established for the full-length receptor. Phosphorylation analysis added another layer of complexity. AR-A was phosphorylated at serine 25, the counterpart of serine 215 in AR-B, more than 15-fold in response to androgen, compared with a 3.3-fold increase at the corresponding site in AR-B. This heightened phosphorylation, possibly due to greater accessibility of the exposed N-terminus to the kinase Akt, appears to modulate AR-A function, since the coactivator PRMT6 failed to enhance AR-A transcriptional activity but strongly boosted a phosphorylation-defective S25A mutant.

The most striking findings emerged when the two isoforms were co-expressed. Because AR-A can heterodimerize with the mutant receptor, the researchers tested different ratios of the two proteins. Co-expression of AR-A reduced the formation of insoluble mutant receptor aggregates across micro-aggregates, macro-oligomeric species and detergent-insoluble mature forms, as quantified by filter retardation assays and a flow cytometric technique called FloIT. Fluorescence microscopy showed that cells co-expressing both isoforms had a 90 percent reduction in the average number of aggregates per cell, dropping from about 179 to fewer than 24, and a 35 percent reduction in the proportion of aggregate-bearing cells. Photobleaching experiments revealed that the remaining aggregates were far more dynamic: fluorescence recovery rose from 34 percent to 49 percent, with a 15 percent increase in the mobile fraction. This suggests that AR-A keeps mutant receptor assemblies in a liquid-like condensate state, preventing their maturation into rigid insoluble inclusions and facilitating clearance by the cell’s degradative systems. Importantly, AR-A exerted no dominant-negative effect on transcription, and heterodimers retained roughly the same activity as mutant homodimers.

Finally, the team validated these results in vivo using a fruit fly model of SBMA in which expression of the mutant receptor in the eye causes progressive degeneration. Flies expressing AR-A alone showed no eye degeneration, confirming the isoform’s inherent harmlessness. When AR-A was co-expressed with the mutant receptor, neurodegeneration was reduced by 50 percent, and filter retardation assays of fly eye tissue confirmed a significant reduction in high-molecular-weight aggregates. The authors propose that strategies to redirect translation from the first AUG codon to the second, or to deliver AR-A directly through viral vectors, antisense oligonucleotides or small molecules, could boost the isoform’s protective presence in affected cells without suppressing androgen signaling. Such an approach would avoid the endocrine, cardiovascular, metabolic and behavioral side effects that plague current androgen-deprivation strategies, including leuprorelin, bicalutamide and antisense-based receptor knockdown. The researchers caution that their findings rest on cellular and fly models, and that validation in human induced pluripotent stem cell-derived motor neurons and muscle cells, followed by mammalian in vivo studies, will be essential before AR-A induction can advance toward the clinic. They also note that modulating full-length receptor transcription through AR-A may hold relevance beyond SBMA, potentially informing therapeutic design in androgen-responsive cancers such as prostate cancer.

Subject of Research: The protective role of the alternative androgen receptor isoform AR-A in spinal and bulbar muscular atrophy

Article Title: The alternative androgen receptor isoform A mitigates toxicity of polyglutamine-elongated mutant androgen receptor in spinal and bulbar muscular atrophy

Article References: Chierichetti, M., Andreotti, R., Tedesco, B., Ferrari, V., Cornaggia, L., Pramaggiore, P., Cozzi, M., Mohamed, A., Magdalena, R., Piccolella, M., Boarolo, G., Crippa, V., Rusmini, P., Galbiati, M., Rinaldi, C., Anderson, E. N., Pandey, U. B., Pennuto, M., Cristofani, R., & Poletti, A. (2026). The alternative androgen receptor isoform A mitigates toxicity of polyglutamine-elongated mutant androgen receptor in spinal and bulbar muscular atrophy. Journal of Advanced Research, 88, 959-976. https://doi.org/10.1016/j.jare.2026.01.051

Image Credits: AI Generated

DOI: Not provided

Keywords: spinal and bulbar muscular atrophy, Kennedy's disease, androgen receptor, AR-A isoform, polyglutamine, protein aggregation, motor neurons, neurodegeneration, autophagy, proteasome, gene therapy, Drosophila model