Identification of Somatic Structural Variants in Solid Tumors by Optical Genome Mapping.

Genomic structural variants comprise a significant fraction of somatic mutations driving cancer onset and progression. However, such variants are not readily revealed by standard next-generation sequencing. Optical genome mapping (OGM) surpasses short-read sequencing in detecting large (>500 bp) and complex structural variants (SVs) but requires isolation of ultra-high-molecular-weight DNA from the tissue of interest. We have successfully applied a protocol involving a paramagnetic nanobind disc to a wide range of solid tumors. Using as little as 6.5 mg of input tumor tissue, we show successful extraction of high-molecular-weight genomic DNA that provides a high genomic map rate and effective coverage by optical mapping. We demonstrate the system's utility in identifying somatic SVs affecting functional and cancer-related genes for each sample. Duplicate/triplicate analysis of select samples shows intra-sample reliability but also intra-sample heterogeneity. We also demonstrate that simply filtering SVs based on a GRCh38 human control database provides high positive and negative predictive values for true somatic variants. Our results indicate that the solid tissue DNA extraction protocol, OGM and SV analysis can be applied to a wide variety of solid tumors to capture SVs across the entire genome with functional importance in cancer prognosis and treatment.

DNA quality assessment for array CGH by isothermal whole genome amplification.

BACKGROUND: Array Comparative Genomic Hybridization (array CGH) is increasingly applied on DNA obtained from formalin-fixed paraffin-embedded (FFPE) tissue, but in a proportion of cases this type of DNA is unsuitable. Due to the high experimental costs of array CGH and unreliable methods for DNA quality testing, better prediction methods are needed. The aim of this study was to accurately determine the quality of FFPE DNA input in order to predict quality of array CGH outcome. MATERIAL AND METHODS: DNA quality was assessed by isothermal amplification and compared to array CGH quality on 59 FFPE gastric cancer samples, one FFPE colorectal cancer sample, two FFPE normal uvula samples, one fresh frozen and six FFPE HNSCC samples. Gastric cancer DNA was also quality tested by beta-globin PCR. RESULTS: Accurate prediction of DNA quality using the isothermal amplification was observed in the colorectal carcinoma, HNSCC and uvula samples. In gastric cancer samples, the isothermal amplification was a more accurate method for selecting good quality DNA for array CGH compared to using PCR product lengths. The isothermal amplification product was used for array CGH and compared to the results achieved using non-amplified DNA in four of the samples. DNAs before and after amplification yielded the same segmentation patterns of chromosomal copy number changes for both the fresh DNA sample and the FFPE samples. CONCLUSION: The efficiency of isothermal DNA amplification is a reliable predictor for array CGH quality. The amplification product itself can be used for array CGH, even starting with FFPE derived DNA samples.

Dual mechanism of signal transducer and activator of transcription 5 activation by the insulin receptor.

Insulin stimulates signal transducer and activator of transcription 5 (Stat5) activation in insulin receptor (IR)-overexpressing cell lines and in insulin target tissues of mice. Stat5b and insulin receptor substrate 1 (IRS-1) interact with the same autophosphorylation site in the IR [phosphotyrosine (pY) 972] in yeast two-hybrid assays, and the IR phosphorylates Stat5b in vitro. These data suggest that Stat5 proteins might be recruited to, and phosphorylated by, the activated IR in vivo. Nevertheless, insulin activates Janus kinases (JAKs) in IR-overexpressing cell lines and in insulin target tissues. To determine whether Stat5 proteins must be recruited to the pY972LSA motif in the IR for insulin-stimulated activation in mammalian cells, we generated and tested a series of IR mutants. The L973R/A975D mutation abolishes the ability of the IR to induce Stat5 activation, whereas IRS-1 phosphorylation is unaffected. In contrast, the N969A/P970A mutation in the IR has no effect on Stat5 activation but significantly reduces IRS-1 phosphorylation. In coimmunoprecipitation assays, insulin-stimulated Stat5 activation correlates with Stat5 recruitment to the IR. We also find that insulin stimulates tyrosine phosphorylation of JAKs that are constitutively associated with the IR. Expression of dominant-negative (DN) JAKs, the JAK inhibitor suppressor of cytokine signaling 1, or pretreatment with the JAK inhibitor, AG490, reduces, but does not eliminate, insulin-induced Stat5 activation. Expression of the appropriate pair of DN JAKs in each of the singly JAK-deficient cell lines further establishes a component of insulin-stimulated Stat5 activation that is JAK independent. This likely represents phosphorylation of Stat5 proteins by the IR, as we find that IR kinase domain phosphorylates Stat5b in vitro on Y699 as efficiently as JAK2. Increasing the concentration of Stat5 proteins in cells favors the direct phosphorylation of Stat5 by the IR kinase where the DN-JAK inhibition of insulin-stimulated Stat5 activation becomes insignificant. At physiological levels of Stat5 however, we propose that JAKs and the IR both contribute to the insulin-induced phosphorylation of Stat5.

Discrimination between signaling pathways in regulation of specific gene expression by insulin and growth hormone in hepatocytes.

Insulin and GH can activate common signaling elements in many tissues and cell lines. We investigated the possibility of overlap in signaling pathways activated by insulin and GH in a key target cell, the hepatocyte. In primary cultures of rat hepatocytes, GH caused a dose- and time-dependent increase in tyrosine phosphorylation of signal transducer and activator of transcription 5. This was accompanied by the induction of the mRNA encoding suppressor of cytokine signaling 2. Neither of these effects took place in companion hepatocytes challenged with insulin. By contrast, insulin caused a rapid and sustained phosphorylation of protein kinase B, accompanied by a massive induction of the mRNA encoding glucokinase. GH had no detectable effect on phosphorylation of protein kinase B or level of glucokinase mRNA. Insulin also elicited brief hyperphosphorylation of ERK1 and 2, an effect not seen in GH-stimulated hepatocytes. Thus, there was a clear demarcation of signaling events triggered in hepatocytes by insulin and GH, and this was accompanied by hormone-specific responses with respect to the induction of gene expression. Additionally, the current results show that signal transducer and activator of transcription 5 activation is neither necessary nor sufficient for the insulin-dependent induction of hepatic glucokinase.