Research Overview

Aberrant biomolecular condensates are implicated in multiple incurable neurological disorders, including Amyotrophic Lateral Sclerosis (ALS), Frontotemporal Dementia (FTD), and DYT1 dystonia. However, the role of condensates in driving disease etiology remains poorly understood. We identified myeloid leukemia factor 2 (MLF2) as a disease-agnostic biomarker for phase transitions, including stress granules and nuclear condensates associated with dystonia. Exploiting fluorophore-derivatized MLF2 constructs, we developed a high-content platform and computational pipeline to screen modulators of NE condensates across chemical and genetic space (Poch et al. 2025, BioRxiv). We identified RNF26 and ZNF335 as protective factors that prevent the buildup of nuclear condensates sequestering K48-linked polyubiquitinated proteins. Chemical screening identified several FDA-approved drugs that potently modulate condensates by resolving polyubiquitinated cargo and MLF2 accumulation. Application of our platform to a genome-wide CRISPR KO screen identified strong enrichment of candidate genes linked to primary microcephaly and related neurodevelopmental disorders.

Future directions include (i) screening >100k novel small molecules for their ability to rescue proteotoxic phase-transitions and proteomic target identification, (ii) exploring the functional roles of RNF26 and ZNF335 in safeguarding against the accumulation of nuclear condensates, and (iii) determine what role condensates may have in driving the molecular pathogenesis of neurodevelopmental conditions such as dystonia and microcephaly.

Co-translational folding is the process by which proteins begin to fold as they are being synthesized, reducing the risk of harmful aggregation and avoiding the need for energy-intensive unfolding steps. We explore how this process connects to nuclear import, proposing that import factors (karyopherins) and flexible nucleoporins not only safeguard the nuclear pore barrier but also foster a supportive environment, the “folding phase”, which allows proteins to fold correctly as they enter the nucleus (Mallik et al. 2025). This perspective offers new insight into how disruptions in nuclear transport and protein quality control may contribute to neurological disease.

Current work is focused on (i) using in vitro biochemical assays to probe how FG-nucleoporins contribute to protein folding and/or prevent the aggregation of protein misfolding. Additionally, we aim to (ii) establish cellular FRET-based assays that allow real-time measurements of protein folding rates within the nuclear pore complex of live cells. Our findings may pave the ground for novel biotechnological advances concerning protein stability and targeted therapeutics that prevent the disruption of protein folding.

The endoplasmic reticulum (ER) is a dynamic network of sheets and tubules that must stay intact even under mechanical strain. Our work focuses on NOMO, an ER-resident protein highly expressed in muscle, which helps preserve ER structure and withstand physical forces. We discovered a key structural interface in NOMO that enables it to maintain ER morphology and measured the tiny forces it experiences using built-in molecular tension sensors. These features proved critical for muscle formation, as mutations disrupting this interface prevented NOMO from supporting myogenesis. NOMO loss also impaired movement in nematodes, highlighting its broader importance in muscle physiology.

Our future work aims to discover NOMO’s dynamicity and conformational intermediates as well as its’ molecular functions in force buffering and sensing. Using both in vitro biochemical as well as cell models, we plan on investigating ER dynamics and its role in cell fate and signaling in response to mechanical input. By pushing these questions further, we hope to discover how the ER can sense, transmit, and adapt to physical forces, especially in cell states, where cells undergo extensive remodeling and are subjected to substantial mechanical challenges, like muscle differentiation and contraction.

Torsin ATPases are unusual AAA+ ATPases with critical medical relevance. A single amino acid deletion (E303) in TorsinA causes DYT1 dystonia, a debilitating movement disorder. Our lab pioneered the first functional in vitro system to study Torsins, leading to several unexpected discoveries about their unique regulation and activity. We demonstrated that Torsin ATPases are outliers of the AAA+ superfamily of ATPases in the human genome in that they lack significant basal ATPase activity and that their catalytic activity is strictly reliant on LAP1 and LULL1, two accessory cofactors that accelerate the hydrolysis step by several orders of magnitude (P.N.A.S. 110(17):E1545-54) by virtue of an active site complementation mechanism (P.N.A.S. 111(45):E4822-3). Importantly, we found that this activation mechanism is offset by disease-causing mutations, which is the first direct demonstration of a loss-of-function mechanism for DYT1 Dystonia.

Nuclear envelope (NE) blebbing is the phenotypic hallmark resulting from Torsin manipulation. Our group has engineered an altogether Torsin-deficient cell line through CRISPR genome editing technology that recapitulates the aberrant NE structures that are the phenotypic hallmark resulting from Torsin manipulation in a wide variety of metazoans (Laudermilch et al.). We pioneered a live cell imaging platform to show that NE blebs represent stalled intermediates in nuclear pore biogenesis (Rampello et al. 2020). Our Nature Cell Biology article (Prophet et al. 2022) reported that these aberrant NE condensates perturb protein homeostasis by sequestering a chaperone network of proteins (DNAJB6, HSPA1A, HSC70, etc.) and allowing proteins to avoid proteasomal degradation, suggesting a potentially causal link to the etiology of DYT1 dystonia.

Many human diseases are caused by mutations in nuclear envelope proteins, yet the mechanisms of protein quality control at the nuclear envelope remain poorly understood. We discovered that the Lamin B receptor (LBR) has an essential role in cholesterol synthesis, which is strongly perturbed by LBR mutations responsible for the congenital disorders Greenberg Dysplasia and Pelger-Huët anomaly (Tsai et al. 2016). Additionally, using a short-lived Lamin B receptor disease variant as a model, we conducted a genetic screen to identify factors required for NE protein turnover (Tsai et al. 2022). This work uncovered the ubiquitin-conjugating enzymes Ube2G2 and Ube2D3, as well as the membrane-resident ubiquitin ligases RNF5 and HRD1, along with the poorly characterized protein TMEM33. These results reveal how the ubiquitin-proteasome system partitions across cellular compartments to adapt to local demands, establishing a robust quality control network at the nuclear envelope.