ABCDEFGHIJK
1
TimestampEmail AddressTITLE (140 character limit): (i.e., High-content imaging approaches for rare diseases)AUTHOR(S): (first and last names only, i.e., Anna Greka, Jillian Shaw)ABSTRACT (1000 character limit): Please provide your abstract below as a concise, narrative paragraph up to ~8 sentences (note: strict 1000 character limit).
20
1fsuripay@broadinstitute.orgTMED10 is a novel cargo receptor of ATZ and therapeutic target for lowering ATZ accumulation in Alpha-1 Antitrypsin Deficiency.Fabian Suri-Payer, Wei Tseng, Juan Pablo, Anna GrekaAlpha-1 Antitrypsin Deficiency (AATD) is a rare genetic disease caused by SERPINA1 variants which manifests as both a loss-of-function and toxic gain-of-function disorder. There is currently no cure, and only supportive treatment available for AATD. The most common AATD allele, the Z allele (E342K, known as ATZ) – destabilizes AAT, making it prone to misfold and polymerize. ATZ accumulation within the endoplasmic reticulum (ER) of liver cells causes liver damage, while the corresponding absence of protective AAT in the lung leads to lung disease. It remains unclear what decides the cellular fate of ATZ within the ER, and how this contributes to ATZ being insufficiently degraded in disease. Here we identify ATZ to be a novel cargo of the TMED cargo receptor family. Crucially, TMED2 and TMED10 knockout lessens ATZ accumulation, with TMED10 knockout further being protective of ATZ accumulation under cellular stress. Together these data identify TMED10 as a potential new therapeutic target for the treatment of chronic liver disease in AATD.
21
2rarafeh@broadinstitute.orgMapping the genetic landscape of KRAS mutant cancer cells Rand Arafeh, Laura Chang, Xiang Zhang, Arshia Hassan, Lydia Sawyer, Helen Wang, James McFarland, Peter DeWeirdt, Amy Hermin, Yenarae Lee, John Doench, Chad Myers and William C. HahnThe success of precision oncology depends on our ability to translate accumulating genomic data into actionable treatment options in a personalized manner. We desperately need to identify new targets so that we can develop new drugs to deliver on the promise of precision cancer medicine. In addition, we need to develop a rational way to combine drugs that does not depend on trial and error. To systematically interrogate such genetic interactions, we designed a dual CRISPR-Cas9 perturbation library targeting the top 1000 genes upregulated in KRAS mutant cancers. We performed these combinatorial knockout screens on 100K unique gene pairs and identified 27 pairs whose co-disruption results in a loss of cellular fitness. We also applied this system to map buffering interactions, and 16 pairs of genes were identified. Overall, our findings will provide insight into potential combinational targets in KRAS mutant tumors and will highlight the synergistic effects that occurs among the novel targets, and other proliferation, metastasis, immune and metabolism modules.
22
3pierce@broadinstitute.orgPrime editing-mediated conversion of an endogenous tRNA gene into a suppressor tRNA for disease-agnostic therapeutic genome editingSarah E. Pierce*, Steven Erwood*, Keyede Oye, Meirui An, Nicholas Krasnow, Emily Zhang, Aditya Raguram, Davis Seelig, Mark J. Osborn, David R. LiuPrecise genome editing technologies such as base editing and prime editing have the potential to directly correct the vast majority of known pathogenic human gene variants. Their widespread clinical application, however, is limited by the effort required to develop each therapeutic editing agent, which typically can only treat patients with a specific mutation. Suppressor tRNAs (sup-tRNAs) offer a more general approach to rescue nonsense mutations that result in premature stop codons. Current therapeutic efforts to use sup-tRNAs suffer from modest potency or require repeated administration. Here, we present a strategy to rescue pathogenic nonsense mutations in a disease-agnostic manner by using prime editing to convert dispensable endogenous tRNAs into optimized sup-tRNAs. Through iterative screening efforts that investigated thousands of variants of all 418 high-confidence human tRNA genes, we identified endogenous tRNAs with the strongest potential to serve as sup-tRNAs, discovered variants of these tRNAs that greatly increase nonsense suppression potency, and developed prime editing agents that efficiently install these optimized sup-tRNAs in human and mouse cells. A single prime editor resulting from these efforts supports the permanent, efficient readthrough of premature termination codons in human cell models of Batten disease, Tay-Sachs disease, Niemann-Pick disease type C1, and cystic fibrosis. In vivo delivery of a single prime editor programmed to convert an endogenous mouse tRNA into an optimized suppressor tRNA rescued a premature stop codon in a mouse reporter gene, and extensively rescued disease pathology in a mouse model of Hurler syndrome. This approach integrates the flexibility and permanence of prime editing with the broad applicability of nonsense suppression, suggesting the potential feasibility of disease-agnostic therapeutic genome editing approaches that require the development of only a single composition of matter to treat a variety of unrelated genetic diseases.
23
4mriedlkh@broadinstitute.orgTMED7 is a critical determinant of misfolded mutant Mucin-1 (MUC1-fs) accumulation in the GolgiMagdalena Riedl Khursigara, Alissa Goss, Christine De Mata, Ranjithmenon Muraleedharan, Matthew Brown, Keith Keller, Maria Kost-Alimova, Juan Pablo, Anna Greka Mucin 1 kidney disease (MKD) is caused by a frameshift mutation in the MUC1 gene (MUC1-fs), resulting in MUC1-fs misfolding and subsequent toxic accumulation in the Golgi compartment. This entrapment is mediated by members of the p24 cargo receptor family (TMEDs). Upon knock-out (KO) of TMEDs in patient-derived kidney epithelial cells we observed that TMED2, 7, 9 and 10 are involved in the toxic MUC1-fs accumulation, whereas TMED1 and 5 aid in MUC1-fs degradation. TMED2, 9 and 10 are important for the stability of the TMED complex, whereas TMED7 seems to be the important TMED for the toxic accumulation of MUC1-fs in the Golgi. We discovered that TMED7 strongly interacts with Golgin-45 and Grasp-55, two integral proteins of the Golgi.This interaction was mediated by the cytoplasmic tail of TMED7. In summary, we have identified differential TMED regulation of MUC1-fs retention in the Golgi apparatus and identified TMED7 as a new drug target for patients with MKD.
24
5sditroia@broadinstitute.orgThe Rare Genomes Project: Improving access to genomic sequencing and identifying causes of rare diseaseStephanie DiTroia, Melanie O’Leary, Eva Martínez, Ashana Neale, Siwaar Abouhala, Lynn Pais, Emily O’Heir, Ikeoluwa Osei-Owusu, Moriel Singer-Berk, Kathryn Russell, Carmen Glaze, Grace VanNoy, Brian Mangilog, Gulalai Shah, Jillian Serrano, Gabrielle Lemire, Vijay Ganesh, Sarah Stenton, Mutaz Amin, Kayla Socarras, Mugdha Singh, Stacey Hall, Katie Larsson, Daniel Marten, Michael Wilson, Hana Snow, Benjamin Blankenmeister, Jialan Ma, Ben Weisburd, Alba Sanchis-Juan, Harrison Brand, Emily Groopman, Alysia Lovgren, Clara Williamson, Marissa Hollyer, Eleina England, Eleanor Seaby, Katherine Chao, Julia Goodrich, Samantha Baxter, Monica Wojcik, Christina Austin-Tse, Daniel MacArthur, Michael Talkowski, Anne O'Donnell-Luria*, Heidi Rehm*The majority of patients with rare, suspected monogenic conditions do not have a molecular diagnosis. As management and treatments for rare diseases improve, the implications of a diagnosis broaden and keep patients motivated to seek an answer. Common barriers to obtaining a molecular diagnosis include exhaustion of clinically available testing, difficulty accessing genetics care, and insurance coverage issues. The Rare Genomes Project (RGP) is addressing these problems by partnering directly with families and advocacy groups to increase the reach of rare disease genomic research through a remote study design. In the first 6 years, RGP performed short-read genome sequencing on 1,940 samples from 798 families, and returned results to 150 families. The experience of RGP highlights challenges in rare disease diagnosis in our current healthcare system and demonstrates the critical role of research programs in building the knowledge base and advancing methodologies to improve the diagnostic yield for rare disease.
25
6prissom@broadinstitute.orgLeveraging protein language models to predict loss- and gain-of-function effects in ion channel variantsFrancesca Rissom, Jordan Safer, Paulo Yanez Sarmiento, Andreas Brunklaus, Damian Balaz,
Roberta Castelli, Connor Coley, Christel Depienne, Alfred George, Erkin Kurganov, Carla Marini, Rikke Steensbjerre Møller, Anna Moroni, Jen Pan, Bina Santoro, Bernhard Renard, Sumaiya Iqbal *, Henrike Heyne *; (* co-supervision)
Voltage-gated ion channels are pivotal in the evolution of rapid signal propagation within organisms and are targeted for treating diseases like epilepsy and cardiac arrhythmias. However, the impact of many genetic variants in ion channel genes on the function of the channels remains uncertain as experiments are resource-intensive. Here, we investigate the use of Protein Language Models (PLMs) to predict the functional effects of genetic missense variants across a wide range of voltage-gated ion channels including Nav, Cav, Kv, Kir, and HCN channels.

We curate a comprehensive benchmark dataset comprising functional labels for over 50 genes from electrophysiology experiments and clinical data. Additionally, we gather a wide array of protein features, including structural characteristics, protein-protein-interaction sites, protein stability metrics, and post-translational modification sites, to provide comprehensive meta-information for the variants.

We assess the performance of PLMs in predicting loss- and gain-of-function mutations in ion channels, employing two approaches: training classical machine learning models on PLM-embeddings and fine-tuning PLMs directly. We compare these to classical machine learning techniques on the collected protein features, assessing performance for specific variant groups of shared characteristics (e.g. protein-protein interaction sites). Additionally, we validate our models on large-scale clinical and functional datasets for robustness.

By investigating the use of pre-trained PLMs for ion channel functional effects, our study aims to deepen our understanding of their applicability for finer-grained functional prediction tasks. Moreover, we aim to contribute to the understanding of ion channel dysfunction in human disease, facilitating the development and implementation of tailored therapeutic interventions.
26
7afrahmujeeb7@gmail.comRegulatory Pathways for Advanced Therapies in BiotechAfrah MujeebAdvanced Therapy Medicinal Products (ATMPs), including gene therapies, cell therapies, and tissue-engineered products, are revolutionizing the treatment landscape for rare diseases. However, their regulatory pathways present unique challenges due to complex manufacturing, long-term safety considerations, and rigorous quality requirements. This poster explores global regulatory strategies for ATMPs, focusing on FDA's RMAT designation, EMA's PRIME scheme, and Japan's Sakigake pathway, which aim to accelerate development while ensuring patient safety. Key aspects of Chemistry, Manufacturing, and Controls (CMC), Good Manufacturing Practices (GMP), and risk-based clinical trials are highlighted alongside real-world case studies, such as Kymriah, Yescarta, and Luxturna. Special emphasis is placed on orphan designation and early regulatory engagement to streamline market access. Findings suggest that harmonized regulatory frameworks and proactive risk management can optimize approval timelines, bringing life-changing therapies to patients more efficiently.
27
8jparikh@broadinstitute.orgIdentifying germline variants that engender genetic vulnerabilitiesJulie Parikh, Sean Misek, Nicole Peiris, Angela Chen, Neha Nanda, Rameen Beroukhim, Jesse S Boehm, Francisco J. Sánchez-Rivera​Germline variants can create unique genetic dependencies in cancer cells, offering opportunities for therapeutic targeting. To identify these relationships, we integrated whole-genome sequencing with CRISPR/Cas9 gene dependency profiles from the Cancer Dependency Map. We performed a genome-wide association analysis to uncover dependency quantitative trait loci (dQTLs) and applied a computational pipeline to filter mismatched guide RNAs and reduce technical artifacts. dQTLs were classified as cis or trans depending on whether the variant and affected gene were the same or distinct. A trans-dQTL in the GEN1 locus was linked to increased dependency on MUS81, a redundant Holliday junction resolvase. We validated this synthetic lethal interaction using enAsCas12a dual knockout screens and are functionally modeling the variant using a base editing tiling library in haploid HAP1 cells. This approach supports discovery of germline variants that sensitize cancer cells to genetic perturbations
28
9tcorrido@broadinstitute.orgPrP lowering divalent siRNA for prion diseaseJuliana Gentile, Taylor Corridon, Zachary Kennedy, Fiona Serack, Nikita Kamath, Meredith Mortberg, Ken Yamada, Julia Alterman, Dimas Echeverria Moreno, Eric V. Minikel, Anastasia Khvorova, Sonia M. VallabhPrion disease is a fatal and incurable neurodegenerative disease caused by
misfolding of the prion protein (PrP), encoded by the gene PRNP. Extensive proof-of-
concept studies validate PrP lowering as a therapeutic hypothesis in prion disease.
Divalent siRNA is an oligonucleotide modality with promising properties in terms of
biodistribution, potency, and durability. We explored the utility of this modality for PrP
lowering, combined with the recently reported extended nucleic acid (exNA)
modification. In wild-type mice, a moderately potent tool divalent siRNA yielding 49%
residual PrP, increased survival time in prion-infected animals both at a pre-symptomatic
timepoint and when given at the onset of symptoms. Extensive screening in cellulo and
in naïve humanized Tg26372 mice identified a new lead compound incorporating exNA
that achieves 27% residual PrP at a 320 µg dose. In a dose response study comparing
chemical backbones, we determined that a fixed UU-3’ antisense strand tail and the
exNA modification each contribute ~7% additional PrP lowering compared to the parent
scaffold. The lead compound was highly durable with 74% residual PrP 6 months post-
dose. These data support the continued development of divalent siRNA for prion
disease.
29
10arines@broadinstitute.orgTowards finding the molecular basis of cargo recognition by the TMED9 cargo receptorFelichi Mae Arines, John Carlos Ignacio, Lindsey Ross, Juan Lorenzo Pablo, Sumaiya Iqbal, Anna GrekaThe TMED cargo receptor family traffic a diverse set of cargoes and entrap misfolded mutant proteins in proteinopathies. Pharmacological targeting of TMEDs cleared misfolded proteins in models of MUC1 disease, UMOD disease, and retinitis pigmentosa, highlighting their potential as nodal druggable targets to treat multiple rare diseases. To understand the molecular basis of TMED cargo recognition, we profiled the full cargo repertoire of TMED9, a prominent family member, using a microarray harboring >21,000 human proteins. AlphaFold3 modeling of TMED9 in complex with top interactors identified putative cargo-binding hotspots on TMED9. Mutating these residues reduced the binding of top interactors including TEX33, a poorly characterized protein. Ongoing work aims to understand the generalizability of these binding sites to other cargoes, identify sequence or structural signatures shared by TMED9 cargoes, and evaluate how these signatures influence folded vs. misfolded cargo recognition.
30
11rosslind@broadinstitute.orgTrafficking and Entrapment of Uromodulin: Elucidating the TMED9 Molecular Recognition MechanismLindsey Ross, Felichi Mae Arines, John Carlos Ignacio, Silvana Bazua-Valenti, Hyery Yoo, John Lin, Juan Lorenzo Pablo, Sumaiya Iqbal, Anna GrekaUromodulin (UMOD), the most abundant protein in human urine, is synthesized in the endoplasmic reticulum of kidney tubule epithelial cells and secreted into the urine via the secretory pathway. Mutations in UMOD result in its intracellular entrapment and accumulation in the early secretory pathway, leading to autosomal dominant tubulointerstitial kidney disease (ADTKD-UMOD) and ultimately kidney failure. We have shown that the transmembrane Emp24 domain-containing cargo receptor, TMED9, regulates UMOD trafficking and mediates mutant UMOD entrapment; however, its recognition mechanism remains unclear. Using in vitro protein-protein interaction assays, we identified the zona pellucida (ZP) domain—also implicated as the functional unit of UMOD polymerization—as critical for TMED9 binding. This finding suggests different roles for the ZP domain in intracellular trafficking and extracellular polymerization of UMOD, offering insight into protein handling in a rare kidney disease.
31
12jsafer@broadinstitute.orgGenomics 2 Proteins portal: A discovery tool to link genetic screening outputs to protein sequence and structureJordan Safer, Seulki Kwon, Surya Mani, Mikias Mohamed, Duyen Nguyen, Sumaiya IqbalWe live in the era of big biological data. The latest deep learning methods and experimental techniques have made millions of high-quality protein structures accessible to the biomedical community. At the same time, an unprecedented number of genetic variations have been accumulated in multiple databases. However, there is no readily accessible tool to connect genetic data with protein structures to help hypothesize the molecular effect of genetic variations. To fill this gap, we developed a bioinformatic method to query, retrieve, and connect genetic variations and transcripts to protein sequence annotations and structures, in a web interface called Genomics 2 Proteins portal. Using the method, we mapped over 49,000,000 human genetic variations from all protein-coding genes onto protein sequences and structures, with comprehensive annotations. Additionally, the portal includes an “Interactive Module” that allows users to upload annotations to map to the target protein’s structure.
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119