Awardee Abstracts

Ten-eleven translocation (TET) enzymes oxidize 5-methylcytosine in DNA, leading to DNA demethylation. TET loss-of-function is commonly observed in many types of cancers. Mice with profound, acute TET deficiency induced by deletion of all Tet genes (Tet iTKO mice) show rapid onset of uncontrolled malignant myeloid expansion, and the expanded myeloid cells from Tet iTKO mice display increased levels of DNA damage and inflammation compared to control myeloid cells. Using Tet iTKO hematopoietic stem and progenitor cells (HSPC), we performed an in vivo CRISPR knockout screen of genes encoding RNA and DNA sensors, as well as regulators of RNA-DNA hybrids (known as R-loops), to find how loss of genes related to inflammatory response would affect the proliferation of TET deficient cells. We observed that loss of enzymes that prevent the accumulation of R-loops impaired the malignant myeloid expansion of Tet iTKO cells. Expanded myeloid cells from Tet iTKO mice display increased levels of R-loops compared to control myeloid cells, and in Tet iTKO cells, loss of RNaseH1 and RNaseH2, the two enzymes in the genome that degrade the RNA strand in R-loops, compromised the proliferation and stem cell activity of Tet iTKO HSPC, while further increasing the levels of R-loops and DNA damage compared to Tet iTKO cells without RNaseH loss. These results highlight R-loops as a vulnerability of TET-deficient cancer cells. Given that increased R-loops is a feature of many cancers, our results suggest that ramping up R-loops’ accumulation in cancer cells might be a new opportunity for cancer treatment.

In cancer and chronic infection CD8 T cell exhaustion is hallmarked by expression of inhibitory surface receptors such as PD-1, TIM-3, and LAG-3. However, inhibitory receptor focus has been limited to surface proteins and an expansion into different classes of inhibitory molecules is needed. Here, we identifiy the lipid metabolite phosphatidylserine (PS) as a regulator of T cell exhaustion. PS is localized to the inner plasma membrane of cells, but externalized to the outer membrane during cell death where it has potent immunosuppressive functions. However, beyond apoptosis, the function of PS is unclear. We find that live, exhausted, antigen-specific CD8 T cells externalize PS in vivo during chronic LCMV infection and in tumors. PD-1+ Tcf1+ stem-like and Tim-3+ effector cells exposed PS. Furthermore, transcriptomic and lipidomic analyses identified increased PS metabolism during exhaustion. To evaluate the checkpoint potential of exposed PS we treated chronically infected mice with the PS targeting antibody (aPS) (mch1N11) and found it significantly expanded the CD8 T cell immune response. PD-1+ Tcf1+ stem-like CD8 T cells downregulated quiescence-associated modules and increased their proliferation. Mechanistically, exposed PS on CD8 T cells contributed to an immunosuppressive environment by inhibiting antigen presenting cells. Combination of aPS with aPD-L1 significantly increased the CD8 T cell effector response and improved viral control, emphasizing the synergy of treatments. These data demonstrate that PS externalization occurs on live CD8 T cells in vivo and describe its function as a ‘non-classical’ inhibitory molecule during exhausted.

Chimeric antigen receptors (CARs) are recombinant proteins that can be introduced into immune cells to endow antigen-targeting specificity to redirect cellular immune effector functions. The antigen binding moiety is comprised of an antibody derived single chain variable fragment (scFv) consisting of variable heavy (VH) and light chains (VL) connected with a linker peptide. The Whitlow linker (WL) is a synthetic linker frequently incorporated into scFvs engineered for use in CAR T cells. We developed proprietary monoclonal antibodies that demonstrate high sensitivity and specificity to bind scFvs containing this WL linker sequence. In this seminar, I will share our characterization of these novel anti-WL antibodies, and a vision for their use in CAR T-cell translational research and therapeutic applications.

Niche partitioning is one of the mechanisms that promotes diversity of the human gut microbiome. However, the molecular basis of resource specialization and niche separation in the gut remain poorly understood. Here we profiled the fructan consumption capabilities of several representative Bacteroides species in the gut and found that there is resource partitioning dictated by fructan length. Interestingly, preference for either short or long fructans could be predicted by the size of the TonB-dependent membrane transporter that bacteria use to import these molecules intracellularly. In particular, we found that bacteria with short transporters encode less enzymes for fructan degradation and prefer short fructans, while this pattern is reversed in bacteria with large transporters. Strikingly, shortening a large transporter by deleting its plug domain did not lead to a loss-of-function phenotype but instead reversed the original preference of the transporter for long over short fructans. Such differences in the molecular structure of the transporters predict pairwise competition outcomes in fructans of different length and therefore could inform the development of targeted prebiotic interventions using dietary fiber with varying lengths aimed at enriching specific groups of microbes in the gut.

Merkel cell carcinoma (MCC) is a deadly skin cancer that is caused by Merkel cell polyomavirus (MCV), a small, circular dsDNA virus. MCV infects most people during childhood and maintains a persistent lifelong asymptomatic infection. However, MCV goes on to cause cancer in a subset of people later in life. In MCC cells, the MCV genome become integrated into the cellular genome and two viral proteins are expressed: the Small and Large Tumor antigens (ST, LT), which inactivate two major tumor suppressor genes, TP53 and RB1. We previously identified that MCV transcripts expressed in MCC encode a third viral protein, ALTO, however its role in tumorigenesis was unknown. During my postdoctoral training, I discovered that ALTO protein is not expressed in MCC cells. Restoring ALTO expression with a lentiviral vector arrests cell growth by activating NF-kB signaling and downregulating LT/ST. ALTO also strongly upregulates antigen presentation and secretion of inflammatory chemokines via NF-kB. In future work, I will examine whether ALTO promotes anti-tumor immunity and responses to immunotherapy, which currently works in only half of patients. I will also determine whether ALTO can be reactivated in MCC tumors and if not how to deliver ALTO as a gene therapy in vivo.

Retrotransposons are mobile genetic elements in the human genome that are recognized as drivers of genome expansion and evolution. The Long Interspersed Element-1 (LINE-1) retrotransposon has generated over one-third of the human genome and serves as an active source of genetic diversity and human disease. Yet, how LINE-1 mobilizes within the human genome remains poorly understood. I will discuss our efforts to biochemically reconstitute the mobility mechanism by the human LINE-1 encoded enzyme with purified components. Our reconstitutions demonstrate how the LINE-1 enzyme nicks the target DNA to prime reverse transcription of the LINE-1 or SINE RNAs in vitro.

Using cryo-electron microscopy, we have obtained structures of retrotransposition intermediates with the LINE-1 enzyme engaging its native RNAs, e.g. the SINE RNAs, and target DNA to prime reverse transcription. We visualize extensive interactions with the single-stranded RNA and RNA secondary structures by five distinct domains of the LINE-1 enzyme, including sequence-specific contacts. Most surprisingly, we demonstrate an unexpected target-site requirement for DNA cleavage and reverse transcription where the enzyme recognizes an upstream single-stranded DNA to position adjacent DNA duplex in the endonuclease active site for nicking, generating a staggered DNA break with a single nick. These findings demonstrate that LINE-1’s mobility is coupled to the DNA replication within human cells.

In summary, our work provides key insights into the mechanism of ongoing transposition in the human genome and informs the engineering of retrotransposon proteins for gene therapy.

Teleost fish exhibit the widest variation in craniofacial morphologies among vertebrates, making them a unique source of information about the genetics behind naturally divergent groups, allowing for the discovery of novel gene networks and genome regulatory elements with potential clinical relevance. Using a flock of three specialist species of Cyprinodon pupfishes that have colonized the hypersaline lakes of San Salvador Island (SSI), Bahamas, with the fastest rate of craniofacial evolutionary divergence ever documented, I am studying 1) the genetics behind their extremely divergent craniofacial phenotypes and, 2) the role of those genetic changes in craniofacial development. By integrating multiple disciplines like evolutionary genomics, DNA sequencing, mRNA in-situ hybridization, microscopy, and synthetic receptor antagonists, my first results demonstrated the advantages of using this non-model adaptive radiation to reveal a novel function for galr2 (galanin receptor 2) in the evolution of divergent jaw development. Now, using tissue-specific transcriptomics across SSI pupfishes and two distant outgroup species, I have identified two more novel candidate genes important for the evolved dietary behavior of scale-eating: pycr3 (pyrroline-5-carboxylate reductase 3) and atp8a1(ATPase phospholipid transporting 8A1). Both pycr3 and atp8a1 are found differentially expressed in craniofacial tissues (i.e., cartilages and muscles) between the specialists and the generalist during development and are associated with fixed single nuclear polymorphisms within their 20 kb cis-regulatory region in the scale-eaters. My future research will shed light on how these fixed adaptive alleles control the spatial and temporal expression of novel candidate genes and uncover novel gene functions crucial for craniofacial divergence.