In frontotemporal dementia, corrupted RNA caused by TDP-43 protein loss has been found in the synapse for the first time; earlier studies detected it almost exclusively in the nucleus

In frontotemporal dementia, corrupted RNA caused by TDP-43 protein loss has been found in the synapse for the first time; earlier studies detected it almost exclusively in the nucleus

Scientists at Barrow Neurological Institute in Phoenix compared synapses in the brains of ten deceased patients with frontotemporal dementia caused by a C9ORF72 gene mutation and ten healthy donors, along with neurons derived from stem cells of other patients. In the postmortem synapses, they found corrupted STMN2 RNA in eight of the ten patients and in none of the healthy donors; in the cell model, the same defect did not appear at all.

Frontotemporal dementia destroys the frontal and temporal lobes of the brain, altering behavior, speech, and the ability to plan actions. The disease typically begins between the ages of 45 and 64 and kills within 7 to 13 years; no drug that halts it exists. The most common genetic cause is an expansion of a short repeat in the C9ORF72 gene; the same mutation also causes amyotrophic lateral sclerosis (ALS), which is why the two conditions are considered part of one spectrum. In mice with ALS, where TDP-43 also fails, senolytics (drugs that kill senescent cells) have already nearly restored normal excitability to the motor cortex.

In neurons carrying this mutation, the TDP-43 protein leaves the nucleus. Under normal conditions, TDP-43 helps the cell assemble RNA correctly by excising unnecessary segments. One such segment is called a cryptic exon: without TDP-43, the cell stops excising it, and it remains in the mature RNA, corrupting the resulting protein. STMN2, KALRN, and UNC13A were already known as markers of this defect, having been found in the nuclei of patient neurons, including in earlier work from this same laboratory. Separately, it became clear that corrupted RNA does not have to remain in the nucleus, but whether it reaches the neuron's most distant functional compartment, the synapse, had not been tested.

To answer this, the authors isolated synaptosomes (fragments of nerve terminals) from the frontal cortex and ran them through protein mass spectrometry and RNA sequencing; they did the same with neurons derived from skin cells of other patients. The three genes behaved differently. Corrupted STMN2 RNA accumulated in the synapse at statistically significant levels, appearing in eight of ten patients versus none of the healthy donors. Corrupted KALRN RNA also appeared in the synapse occasionally, but the results varied too much among patients to distinguish them from healthy donors. Corrupted UNC13A RNA was found in whole brain tissue from all patients but was absent from the synapse in every case. The authors attribute this to the molecule's usual trafficking route: neurons normally transport RNA, including the healthy copy of STMN2, into distant processes for local synthesis of up to half of their proteins, and the corrupted copy likely travels the same path.

The cell model provided independent confirmation: patient neurons showed reduced density of proteins that stabilize the synaptic contact from within, and the electrical activity of the cell network declined over weeks, indicating that contacts were weakening and transmitting signals less effectively. This was unexpected: in earlier ALS models carrying the same mutation, neurons typically became more excitable. The authors attribute the difference to the fact that ALS and this form of dementia affect different neurons, motor neurons in one case and frontal cortex neurons in the other. Corrupted RNA and TDP-43 mislocalization from the nucleus did not appear in this model: it confirmed the loss of synaptic contacts but not the molecular cause observed in patient brain tissue.

The synapses were taken from the brains of deceased individuals, meaning they were contacts that survived until the patient's death; what happened to the rest can no longer be determined, so the findings may reflect the state of surviving synapses only. Current therapies targeting TDP-43 pathologies are attempting to return the protein to the nucleus; after this study, the synapse, where a portion of the corrupted RNA reaches, has been added to the list of therapeutic targets.