Keith R. Jerome, MD, PhD, is a physician and medical researcher on the faculty at the University of Washington and the Fred Hutchinson Cancer Research Center. His research focuses on the biology of chronic viral infections. He has published extensively on pathogen-host interactions and immune evasion by herpesviruses and is now pioneering the use of gene editing and gene therapy as potentially curative therapies for HIV, hepatitis B, human papillomavirus and herpesvirus infections. He has served as co-principal investigator and NIH contact investigator for defeatHIV, one of the original three Martin Delaney Collaboratories working toward a cure for HIV disease.

In addition to his basic science research efforts, Dr. Jerome leads the diagnostic virology program at the University of Washington. Under his guidance the program has designed and implemented molecular testing assays for a wide variety of human viruses, including SARS-CoV-2, hepatitis B and C, enterovirus, BK virus and cytomegalovirus. The laboratory provides diagnostic support for stem cell transplant and other patients in the Pacific Northwest and throughout the country via its reference testing services.

Dr. Jerome earned a bachelor of science from Georgetown College before earning a PhD in microbiology and immunology and his MD from Duke University.

Title: Toll-Like Receptor Adaptors TRIF and MyD88: Coordinators of Macrophage Responses to Pathogenic Betacoronavirus

Abstract:

Disease severity increases during highly pathogenic betacoronavirus (b-CoV) clearance, suggesting an immune response that is protective and pathogenic. CoV infections in human and mice are associated with M1-like macrophage infiltration to infected tissues, including lungs (SARS-CoV-2 and murine CoVs, MHVs) and the brain (MHV). However, the role of innate immune receptors, specifically toll-like receptors (TLRs), and their adaptors in macrophage sensing of b-CoVs remains unclear. MHVs are well-established prototypes for modeling b-CoV pathogenesis in the lungs and CNS. Using two MHV strains, the neurovirulent MHV-JHM and the neuroattenuated MHV-A59, we investigated the role of TLR adaptors TIR-domain-containing adapter-inducing interferon-β (TRIF) and myeloid differentiation factor 88 (MyD88). Wild-type or adaptor-deficient cells were infected with either MHV strain alone or before TLR ligand stimulation. Plaque assays, immunoblotting, and enzyme-linked immunosorbent assays (ELISAs) were used to evaluate virus production, intracellular protein expression, and cytokine secretion, respectively. Replication and protein co-localization was visualized via immunofluorescence. Unexpectedly, CoV replication was suppressed in TRIF-deficient cells while expression of the RNA sensor melanoma differentiation-associated protein 5 (MDA5) was increased in TRIF- or MyD88-deficient cells. The ubiquitin-like interferon-stimulated gene 15 (ISG15) was induced by MHV and co-localized with MHV nucleocapsid. In adaptor-deficient cells, ISG15 was upregulated compared to wild-type cells, suggesting a regulatory role of TLR adaptors in ISG15 expression. Our findings demonstrate that macrophage inflammatory responses are MyD88-dependent, while TRIF negatively regulates CoV-induced macrophage antiviral responses. Defining the signaling pathways that govern macrophage responses to b-CoVs will be key for developing novel therapeutic strategies to counteract macrophage-driven hyperinflammation.

Title: AURKA inhibition as single agent or combination therapy in EGFR inhibitor-resistant head and neck squamous cell carcinoma (HNSCC)

Abstract:

Epidermal growth factor receptor (EGFR) is frequently overexpressed in head and neck squamous cell carcinomas (HNSCCs), with the EGFR inhibitor cetuximab being FDA-approved as a targeted agent for HNSCCs. Even though most patients initially respond, almost all develop resistance to EGFR inhibition. HNSCCs also express high levels of Aurora kinase A (AURKA), most studied as a kinase important for mitotic entry and progression. AURKA is overexpressed in many cancers, where it acquires new cancer-promoting activities, including promotion of resistance to EGFR inhibitors (EGFRis). Here, we have evaluated whether and how AURKA modulates EGFR signaling in HNSCCs, and identified potential AURKA therapeutic combinations to overcome EGFRi resistance in HNSCCs. Using viability and clonogenic assays, we have found that the AURKAi VIC-1911 synergizes with the EGFRis afatinib and erlotinib to inhibit growth of EGFR-sensitive and -resistant FaDu and Cal27 cell models. To elucidate the underlying mechanisms for AURKA and EGFR inhibitors synergism, we assessed core EGFR and AURKA signaling after 24 hours of drug treatment in vitro. VIC-1911 had little to no effect on expression or activation of EGFR, ERK, or AKT, indicating no significant input of AURKA activity on these core processes. However, this treatment significantly induced expression of total AURKA and TPX2, reflecting a compensatory response to enhance availability of active AURKA. RNAseq analysis of the resistant models indicated a strong EMT signature in both resistant models. To further investigate alternative approaches to overcome EGFR resistance, we combined inhibition of AURKA with inhibition of WEE1 kinase, a regulator of the G2/M checkpoint. Although Loewe synergy indicated that synergy between the AURKA inhibitor VIC-1911 and the WEE1 inhibitor adavosertib was reduced in the EGFR inhibitor resistant models, clonogenic assays demonstrated that this combination effectively inhibits growth of both models in comparison with single drugs. We further evaluated the effect of the VIC1911 and adavosertib drug combination in vivo, using Fadu parental and Fadu^AfaR HNSCC xenograft models, this combination showed to be extremely effective in both the EGFRi-sensitive and -resistant setting when compared to single drugs. These results demonstrate the role of AURKA in promoting EGFR resistance in HNSCC and reveal therapeutic approaches that can overcome the growth of HNSCCs resistant to EGFR inhibition.

Title: Isogenic human cortical organoids reveal genotype-phenotype distinctions and a novel therapeutic target for Spastic Paraplegia Type 4

Abstract:

Hereditary Spastic Paraplegia (HSP) is a neurodegenerative disorder marked by lower limb spasticity, leading to severe gait impairment. The hallmark pathology of HSP is the axonal swelling and degeneration of the corticospinal tract, projections of upper motor neurons in the motor cortex. Spastic Paraplegia Type 4 (SPG4) is caused by mutations in the SPAST gene encoding a protein named spastin. It is an autosomal dominant form of HSP and accounts for nearly half of HSP cases. Despite the range of SPAST mutations, including missense, nonsense, frameshift, and truncation variants, the clinical presentation is generally consistent. However, recent studies have identified genotype-phenotype correlations in large patient cohorts, though these phenotypes have yet to be fully recapitulated in experimental models exploring their molecular mechanisms. Here, we sought to bridge this gap by utilizing CRISPR-Cas9 to generate isogenic lines of human induced pluripotent stem cells, harboring distinct SPAST mutations. They were differentiated into cortical organoids enriched with upper motor neurons, providing a humanized platform to study SPG4 pathogenesis. Across different developmental stages, these organoids revealed clear genotype-phenotype differences, shedding light on disease mechanisms that had previously been elusive. Notably, we identified aberrant activation of histone deacetylase 6 (HDAC6) as a key contributor of SPG4 pathophysiology, highlighting HDAC6 as a potential therapeutic target. Moreover, we extended our findings by validating the therapeutic efficacy of Tubastatin A, a specific HDAC6 inhibitor in SPG4 transgenic mice, further supporting its potential for clinical application. Overall, our study offers new insights into SPG4 mechanisms, paving the way for HDAC6-targeted therapies. By connecting genotype-phenotype correlations with underlying mechanisms, we establish a foundation for personalized medicine, demonstrated through our innovative organoid models.

Title: Mechanisms supporting spinal locomotor rhythm generation in adult mouse

Abstract:

Walking involves circuitry spanning the central nervous system. The brain initiates locomotor activity and sends “go” signals via the brainstem to the spinal locomotor central pattern generator (CPG). Locomotor timing is encoded in this CPG by rhythmogenic interneurons (INs), such as Shox2 INs in mice. Our understanding of how these neurons function comes mostly from neonatal studies, but we must examine how these neurons function in adult mice, which are capable of weight-supported stepping. Our overarching hypothesis is that Shox2 INs from adult mice have enhanced intrinsic properties compared to the neonate and receive monosynaptic input from a locomotor-initiating brainstem region. We first used whole-cell patch clamp to characterize the intrinsic conductances displayed by Shox2 INs. We found that adult Shox2 INs have stronger persistent inward currents, M-type K+ currents, and slow afterhyperpolarization compared to neonates, suggesting that adult Shox2 INs employ these currents to maintain rhythmic firing. We then asked whether Shox2 INs are directly contacted by the brainstem, as this could provide the theorized tonic drive from brainstem to rhythmogenic CPG. We performed anterograde viral tracing targeting the lateral paragigantocellular nucleus (LPGi), a region containing key locomotor-initiating reticulospinal neurons. Our immunohistochemistry on lumbar spinal slices showed excitatory LPGi synaptic puncta in apposition to Shox2 INs, and RNAscope on brainstem slices confirmed labeling of excitatory LPGi neurons. Finally, we performed optogenetic experiments with whole-cell patch clamp in lumbar spinal slices after injecting channelrhodopsin into the LPGi. These data are the first to show a monosynaptic connection between the LPGi and genetically-identified spinal INs within locomotor circuitry. Overall, these findings highlight putative therapeutic targets, both intrinsic and synaptic, after injury or disease that results in paralysis.

Title: Neuronal identity control at the resolution of a single transcription factor isoform

Abstract:

Although the recent emergence of extensive transcriptional, morphological, and connectome information at the resolution of individual neurons and their synapses has enabled investigators to define and categorize neuronal cell types with a new level of precision, these newly uncovered complexities in identity and the sheer numbers of cell types highlight a dearth of understanding in how unique cell identities are established during development. In post-mitotic cells, “terminal selectors” are transcription factors hypothesized to determine the final identity characteristics—but which transcription factors function as terminal selectors and the level of control they exert over different terminal characteristics is not well known. Here for the first time, we establish a novel role for the transcription factor broad as a terminal selector in Drosophila melanogaster. Employing an in-depth characterization of terminal characteristics, we find the expression of a single isoform of broad (broad-z4) serves as the switch between two visual projection neuron identities and is sufficient to induce a comprehensive change in identity—changing gene expression, morphology, and functional connectivity—when perturbed. Only changing the expression of broad in lobula plate lobula columnar 2 (LPLC2) cells, change their morphological patterns, gene expression profiles, and synaptic partners to resemble lobula plate lobula columnar 1 (LPLC1) cells. For the first time, our work identifies a single isoform as the smallest unit underlying an identity switch, which may emerge as a conserved strategy replicated across developmental programs, and this finding has major implications for how we may be able to repair or engineer connectivity to overcome human neurodevelopmental and neuropsychiatric disorders that disrupt proper neuronal wiring patterns.

Title: Influenza virus-induced type I interferons disrupt alveolar epithelial repair and tight junction integrity in the developing lung

Abstract:

Infants are remarkably susceptible to influenza (IV) infection. Type I interferons (IFN-I) limit IV replication in adults. However, IV-infected murine neonates lacking a functional IFN-I receptor (IFNAR-/-) have significantly improved survival and reduced lung pathology relative to wild-type (WT) neonates, indicating IFN-I toxicity in neonates. We hypothesized that IV-induced IFN-I signaling in type II alveolar epithelial cells (AT2s) contributes to viral pathogenesis in neonates due to their immunoregulatory and surfactant producing role in the lung. Moreover, AT2s are the primary cell type of IV infection and initiate host response to IV. To investigate the role of IFN signaling in AT2s, we performed RNA-Seq on AT2s purified from IV-infected IFNAR-/- and WT neonates and adults at 1, 2, and 3-days post infection (dpi). Pairwise comparisons highlighted differentially expressed genes (DEGs) dependent on genotype, timepoint, and age. Of note age was the major driver of transcriptional differences in AT2s across genotypes. Within neonatal comparisons, infected IFNAR-/- and WT neonates at 2dpi had the highest number of DEGs. Subsequent analysis of 2dpi showed significant upregulation of pathways related to organ development in IFNAR-/- neonates compared to WT. Common genes among these pathways included aldh1a2, sulf1, and sfn, genes related to alveolar regeneration and lung repair. Therefore, we hypothesized IV-infected IFNAR-/- neonates possess improved barrier integrity to support essential repair following infection. IV-infected neonates were harvested at 6dpi and whole lungs were processed for immunofluorescence imaging. WT neonatal lungs showed a significant reduction in the expression of the tight junction molecule occludin relative to IFNAR-/- neonates at 6dpi, a time of peak mortality. This suggests that early transcriptional differences in neonatal IFNAR-/- AT2s contribute to enhanced barrier integrity, mitigating later IV-induced lung damage and pathology.

Title: LncRNA SChLAP1 promotes cancer cell progression and invasion via its distinct structural domains and conserved regions

Abstract:

Long non-coding RNAs (lncRNAs) emerged as key players in various biological processes and disease progression. Despite their functional significance and therapeutic potential, lncRNAs’ mechanisms of action remain understudied. One such lncRNA is the Second Chromosome Locus Associated with Prostate-1 (SChLAP1). SChLAP1 is overexpressed in a subset of malignant prostate cancer and is related to poor patient outcomes, including metastasis and high mortality rate. In this study, we demonstrated that SChLAP1 possesses distinct structural domains and evolutionarily conserved regions that may contribute to its function. We determined the secondary structure of SChLAP1 using chemical probing methods combined with mutational profiling. Our in vitro secondary structural model revealed that SChLAP1 consists of two, highly structured domains. Sequence conservation analysis showed that these highly structured domains are also evolutionarily conserved. Overexpression of the domains led to a notable increase in cancer cell proliferation and invasion, proving their functional significance on the oncogenicity of SChLAP1. In conclusion, we discovered functionally important, independent domains with well-defined structures of SChLAP1. This will serve as a guide to explore the detailed molecular mechanisms by which SChLAP1 promotes aggressive prostate cancer, ultimately contributing to the development of SChLAP1 as a novel therapeutic target.