TARDBP Motor-Neuron Review Examines Which Phenotypes Can Support Drug Screens

A screening model needs a reproducible phenotype

Motor neurons derived from induced pluripotent stem cells (iPSCs) can preserve a patient's genetic background while making disease-associated cellular changes accessible to experiments. For amyotrophic lateral sclerosis (ALS) linked to TARDBP, the gene encoding TDP-43, a central question is which changes are consistent enough to serve as drug-screening endpoints. A systematic review by researchers in France and Germany compares those phenotypes alongside the methods used to measure them.

The authors searched PubMed for iPSC, TDP43, TARDBP and motor-neuron terms. From more than 60 papers published since 2012, they included 36 studies of patient-derived cells or control lines edited to carry TARDBP mutations. TDP-43 overexpression and knockdown studies were excluded to focus on disease-relevant mutation models. The review notes that its defined search could still miss relevant publications.

Cell identity and analytical choices complicate comparison

Of the included studies, 34 used small molecules to direct motor-neuron differentiation and 2 used transcription factors. Reported purity depended on marker choice, antibody methods and neuronal age. A marker such as HB9 can change with maturation, so its positive-cell percentage is not a uniform purity measure across protocols. The models predominantly represent lower motor neurons, with current differentiation approaches tending toward cervical identities.

The review could not establish consistent correlations between discrepant phenotypes and differentiation method, purity, substrate or maturation. Incomplete metadata and non-equivalent analysis methods limited that comparison. Measures described broadly as viability, for example, included membrane leakage, ATP-dependent metabolic activity, apoptotic staining and nuclear morphology. These assays interrogate different aspects or stages of injury.

Baseline survival illustrates the problem. Some studies found increased death in mutant cultures, whereas others found no difference over extended observation. Neurite shortening and swelling were also variable and could depend on when neurons were measured or how they extended within the culture. Those outcomes can be useful within a well-characterized experiment, but their inconsistency makes them uncertain default endpoints across TARDBP screening studies.

Three areas show greater convergence

Stress susceptibility was more consistent than baseline mortality. Across the reviewed literature, mutant motor neurons showed increased vulnerability under defined challenges, although the effective stressor and exposure paradigm differed. Some individual stressors produced no genotype difference, so the convergent finding concerns susceptibility across the literature rather than universal sensitivity to every challenge. The review favors defined stress and recovery windows with multiple injury measurements.

Mitochondrial abnormalities form a second recurring area. Studies reported changes in structure, transport, respiration and oxidative stress, but the precise morphological direction was not uniform: both enlarged, elongated mitochondria and smaller mitochondrial profiles were described. Reduced basal respiration was reported in several studies, while other respiratory measures could remain unchanged. The review leaves open whether the mitochondrial changes initiate pathology or follow other disease processes.

RNA metabolism was the third broad area of convergence. Transcriptomic and proteomic studies showed extensive dysregulation, but enrichment analyses disagreed about the direction of particular RNA-processing pathways. Mutant neurons also only partly recapitulated the cryptic-splicing signature associated with loss of TDP-43 function. Some established targets were unchanged at baseline, and additional stress or inhibition of RNA surveillance was required to reveal particular transcripts.

TDP-43 pathology is not a uniform quantitative endpoint

Sixteen studies examined TDP-43 distribution by immunofluorescence. More than half reported no cytoplasmic mislocalization. Other studies detected small localization shifts or aggregates, and biochemical measurements of soluble and insoluble protein were likewise influenced by different extraction and quantification procedures. The review found no evidence that studies reporting mislocalization simply used older neurons.

Electrophysiology adds a time-dependent dimension. Some cultures progressed from early hyperexcitability to reduced firing, whereas other studies found no clear electrophysiological alteration. Stress-granule formation and persistence also depended on the stressor and recovery condition. These patterns support measuring a defined process at a defined stage, with acquisition settings and segmentation criteria available for comparison.

Cellular rescue and clinical translation remain separate tests

The reviewed drug studies used diverse endpoints. An early multiphenotype screen tested 1,232 approved drugs across ALS backgrounds, then retested 95 candidates in TARDBP neurons; 9 compounds rescued phenotypes in both the initial and follow-up settings. Ropinirole was prioritized for further development. The review reports that a 20-participant sporadic-ALS trial found no benefit during its double-blind phase, while exploratory open-label observations required further investigation. It does not identify an approved ALS treatment arising from these TARDBP screens.

Other examples targeted stress responses, mitochondrial transport or protein-associated phenotypes. Their cellular effects remain tied to the selected assay. The authors recommend prioritizing stress susceptibility, mitochondrial dysfunction and RNA-related abnormalities, together with isogenic controls, multiple independent lines, shared differentiation batches and transparent automated analysis.

The review also considers glial, vascular, muscle and neuromuscular models as ways to investigate non-cell-autonomous effects. Its main screening conclusions, however, are specific to TARDBP-mutant iPSC-derived motor neurons. They provide an endpoint-selection argument for this model family, with the conflicting and negative findings retained as part of that decision.