Depression and bipolar disorder of late onset may represent more than just mental health conditions. Growing evidence suggests these late-life mood disorders (LLMDs) could be not merely risk factors, but rather early warning signs of neurodegenerative diseases like dementia, even when they appear years before memory loss or other cognitive symptoms become apparent. Unfortunately, scientists have struggled to understand the connection between LLMDs and developing dementia at the biological level. While previous research suggested connections between specific disorders like late-life depression and Alzheimer’s disease, the specific neurological mechanisms involved remain mostly unclear. This knowledge gap is particularly pronounced for late-life bipolar disorder, which has rarely been investigated in relation to dementia.
On top of this, limitations in brain imaging technology have prevented researchers from detecting all the different types of abnormal proteins that might underlie these conditions. Against this backdrop, a research team led by Dr. Shin Kurose and Dr. Keisuke Takahata from the National Institutes for Quantum Science and Technology, Japan, conducted a comprehensive investigation into the brain changes associated with LLMDs. Their study explores the presence of abnormal tau protein (a hallmark of several neurodegenerative diseases) in the brains of people with late-life depression and bipolar disorder. The researchers used advanced brain imaging techniques to examine 52 participants with LLMDs and 47 healthy controls. They employed a PET scan using two different tracers, which can detect various forms of tau protein and amyloid beta accumulation.
To validate their findings, they also analyzed brain tissue samples from 208 autopsy cases, examining the relationship between late-life mood symptoms and the subsequent development of neurodegenerative diseases. The results were striking: approximately 50% of participants with LLMDs showed tau accumulation in their brains, compared to only about 15% of healthy controls. Similarly, nearly 29% of participants with LLMDs had detectable amyloid deposits versus just 2% of controls. The autopsy findings further supported these results, showing a significantly higher prevalence of diverse tau protein-related pathologies in individuals who had experienced late-life mania or depression. Another noteworthy discovery was that many participants showed tau accumulation in the frontal regions of the brain, which is crucial for emotional and cognitive function.
The study also revealed that these abnormal proteins could be detected years before traditional cognitive symptoms of dementia appeared. As revealed by the autopsy cases, mood symptoms preceded cognitive or motor symptoms by at least 7 years earlier. The insights uncovered in this study have important implications for clinical practice, as some cases of late-life depression and bipolar disorder could likely benefit from an evaluation for underlying neurodegenerative diseases. Timely identification of these conditions would allow for earlier intervention with disease-modifying treatments. Moreover, the researchers also highlight the value of the tracer molecules used in their PET scans as effective biomarkers for detecting these diverse tau-related pathologies in living patients.
With any luck, these efforts will help solidify our understanding of how neurodegenerative diseases first manifest, leading to earlier diagnosis and potentially better outcomes. Indeed, Alzheimer’s disease is the most common cause of dementia. The disease has proven difficult to develop new treatments for, and available treatment options are limited. Considering its sanitary burden worldwide, more therapies are needed to improve patients’ quality of life and reduce the burden on the health care system and family caregivers. Scientists at Sanford Burnham Prebys and elsewhere have recently reported real-world links in medical records associating common HIV drugs with a reduced incidence of Alzheimer’s disease though previously no plausible explaniation was given.
The studies showed patients were at less risk of developing Alzheimer’s disease if they were taking drugs to block a famous enzyme called reverse transcriptase (REV), which copies RNA into DNA, opposite to the classical process. REV is best known from being an essential enzyme allowing HIV and other retroviruses to replicate in host cells, and FDA-approved REV inhibitor drugs prevent HIV reproduction. To better understand the links between Alzheimer’s disease risk and people taking prescribed REV inhibitor drugs, scientists looked for evidence of actual REV activity in the aging human brain and in brains affected by Alzheimer’s disease, identifying REV enzymatic activity, and novel RNAs that encode brain REVs especially in neurons of the aging human brain.
Team’s prior landmark publication in Nature in 2018 described how REV-mediated somatic gene recombination of the amyloid-beta precursor protein (APP) gene can occur in neurons of the human brain including those from the most common non-familial or sporadic form of Alzheimer’s disease. Rare familial mutations in the APP gene cause a form of Alzheimer’s disease that can be inherited in families, whereas sporadic disease lacks this inheritance but can be affected by non-inherited “somatic” mutations produced by RET. So scientists asked themselves a basic question: is there actually any RET activity in the aging human brain? And, if there is, where does it come from and which brain cells are affected? Scientists examined post-mortem brain tissue from donors died from Alzheimer’s disease and compared it to control samples without obvious disease.
REV activity was found within every brain sample, with a trend towards reductions in REV activity in the brains from terminal Alzheimer’s disease. This is consistent with the neuronal degeneration that is a hallmark of Alzheimer’s disease.To investigate the origins of this REV activity further, the scientists assessed multiple possible sources and identified long interspersed nuclear element-1 (LINE1), an ancient genetic sequence so common in mammalian genomes that it makes up nearly one-fifth of all human DNA. It is normally inactive, but scientists have found rare forms that are active, using their own REVs to copy and paste themselves elsewhere in the genome. The prevailing deem among scientists was that LINE1 can only function if expressed from an intact, bicistronic mRNA copy. Instead, there were found thousands of truncated versions of LINE1 expressed in the human brain, including hundreds of sequences not annotated in the human genome.
In addition to uncovering abbreviated versions of LINE1, the scientists found that most of these variations contained only one of the two protein-coding regions that appear on a full-length transcript. The scientists addressed their second major question regarding the types of cells with REV activity by comparing samples of neuron-rich gray matter with white matter that contains mostly glial cells. RET activity was significantly higher in gray matter; this is consistent with RET activity being predominantly found in neurons and has potentially widespread implications as our post-mitotic neurons accumulate DNA changes over lifetime. Given the proven safety of FDA-approved RET inhibitors, scientists suggest that physicians and scientists should pursue prospective clinical trials studying these drugs’ effects on persons with early Alzheimer’s disease as a near-term approach.
- Edited by Dr. Gianfrancesco Cormaci, PhD, specialist in Clinical Biochemistry.
Scientific references
Shi SM et al. Nature. 2025; 639(8056):985-994.
Dancy C et al. Int J Mol Sci. 2024; 25(15):8404.
Nicodemus J et al. J Neurosci. 2025 May:e2298242025.
Kurose S et al. Alzheimers Dement. 2025; 21(6):e70195.
Suzuki H et al. J Alzheimers Dis. 2024; 102(4):1271-1285.
