Gigi C.G. Choi

The orchestrated action of genes controls complex biological phenotypes, yet the systematic discovery of gene and drug combinations that modulate these phenotypes in human cells is labor intensive and challenging to scale. Here, we created a platform for the massively parallel screening of barcoded combinatorial gene perturbations in human cells and translated these hits into effective drug combinations. This technology leverages the simplicity of the CRISPR-Cas9 system for multiplexed targeting of specific genomic loci and the versatility of combinatorial genetics en masse (CombiGEM) to rapidly assemble barcoded combinatorial genetic libraries that can be tracked with high-throughput sequencing. We applied CombiGEM-CRISPR to create a library of 23,409 barcoded dual guide-RNA (gRNA) combinations and then perform a high-throughput pooled screen to identify gene pairs that inhibited ovarian cancer cell growth when they were targeted. We validated the growth-inhibiting effects of specific gene sets, including epigenetic regulators KDM4C/BRD4 and KDM6B/BRD4, via individual assays with CRISPR-Cas-based knockouts and RNA-interference-based knockdowns. We also tested small-molecule drug pairs directed against our pairwise hits and showed that they exerted synergistic antiproliferative effects against ovarian cancer cells. We envision that the CombiGEM-CRISPR platform will be applicable to a broad range of biological settings and will accelerate the systematic identification of genetic combinations and their translation into novel drug combinations that modulate complex human disease phenotypes.

Despite significant clinical progress in cell and gene therapies, maximizing protein expression in order to enhance potency remains a major technical challenge. Here, we develop a high-throughput strategy to design, screen, and optimize 5' UTRs that enhance protein expression from a strong human cytomegalovirus (CMV) promoter. We first identify naturally occurring 5' UTRs with high translation efficiencies and use this information with in silico genetic algorithms to generate synthetic 5' UTRs. A total of ~12,000 5' UTRs are then screened using a recombinase-mediated integration strategy that greatly enhances the sensitivity of high-throughput screens by eliminating copy number and position effects that limit lentiviral approaches. Using this approach, we identify three synthetic 5' UTRs that outperform commonly used non-viral gene therapy plasmids in expressing protein payloads. In summary, we demonstrate that high-throughput screening of 5' UTR libraries with recombinase-mediated integration can identify genetic elements that enhance protein expression, which should have numerous applications for engineered cell and gene therapies.

UNLABELLED: A disintegrins and metalloproteinases with thrombospondin motifs (ADAMTS) family members have been reported dysregulated in various cancers. Through refining a loss of heterozygosity locus at 11q25 by array-CGH, we identified ADAMTS8 as a novel candidate tumor suppressor gene. Although ADAMTS8 downregulation has been reported in several tumors, its biologic function and underlying mechanism remain largely unknown. Here, we found that ADAMTS8 is broadly expressed in normal tissues but frequently downregulated or silenced by promoter methylation in common carcinoma cell lines, including nasopharyngeal, esophageal squamous cell, gastric, and colorectal carcinomas. Pharmacologic or genetic demethylation restored ADAMTS8 expression, indicating that promoter methylation mediates its silencing. Aberrant methylation of ADAMTS8 was also detected in several types of primary tumors but rarely in normal tissues. Further functional studies showed that restoring ADAMTS8 expression suppressed tumor cell clonogenicity through inducing apoptosis. ADAMTS8 as a secreted protease inhibited epidermal growth factor receptor (EGFR) signaling along with decreased levels of phosphorylated MEK and ERK. We further found that ADAMTS8 disrupted actin stress fiber organization and inhibited tumor cell motility. Thus, our data demonstrate that ADAMTS8 metalloprotease acts as a functional tumor suppressor through antagonizing EGFR-MEK-ERK signaling, in addition to its previously reported anti-angiogenesis function, and is frequently methylated in common tumors. IMPLICATIONS: This study uncovers the tumor suppressive function of ADAMTS8, one of the ADAMTS family members, and its frequent methylation in certain tumors could be developed as a potential biomarker.

Nasopharyngeal cancer (NPC) is an Epstein-Barr virus (EBV)-associated head and neck cancer prevalent in Asia. Although with reasons not fully understood, the intrinsic invasiveness of NPC is believed to be EBV-linked. Recently, EBV was found to induce STAT3 activation. Constitutive STAT3 activation correlated with advanced clinical staging in NPC. We hypothesized that STAT3 activation by EBV directly contributes to the intrinsic invasiveness of NPC cells. Phospho-STAT3-Tyr705 was detected in high percentage of NPC tumors (7/10 cases). Using a paired NPC cell line model, HONE-1 and the EBV-infected counterpart, HONE-1-EBV, we found that HONE-1-EBV expressed a higher level of phospho-STAT3-Tyr705 and was approximately 11-fold more invasive than HONE-1. In HONE-1-EBV, STAT3 siRNA targeting inhibited both spontaneous and serum-induced invasion, as well as cell growth. Conversely, activation of STAT3 (by expressing an activated STAT3 mutant, namely STAT3C) in the parental HONE-1, mimicking EBV-induced STAT3 activation, significantly enhanced its invasiveness and proliferation, which was accompanied by increased expression of markers of mesenchymal status, proliferation and anti-apoptosis. Our results demonstrated that EBV-induced STAT3 activation is responsible for NPC cell proliferation and invasion. This was further confirmed by a small molecule inhibitor of JAK/STAT3, JSI-124. JSI-124 inhibited STAT3 activation in HONE-1-EBV, with subsequent growth inhibition, induction of PARP cleavage, abrogation of anchorage-independent growth and invasion. We found that EBV-independent activation of STAT3 by a growth factor, EGF, also contributed to NPC invasion. In conclusion, EBV-induced STAT3 activation directly contributes to the intrinsic invasiveness of NPC cells and STAT3 targeting may be beneficial in treating aggressive NPC.