Using Proteomics to Identify Viral microRNA-Regulated Genes
Proteomic screening for KSHV miRNA targets: To observe proteins repressed by KSHV miRNAs, Ziegelbauer’s team transfected human endothelial cells with control or KSHV miRNAs, and then labeled with cells with amino acids containing either medium-heavy (M) or heavy (H) isotopes. They combined cells from both conditions and used mass spectrometry to measure the abundance of isotope-labeled amino acids. Green proteins symbolize proteins that were translated before the amino acid labeling and/or do not contain stable isotope-labeled amino acids
Kaposi sarcoma is a soft tissue malignancy that affects the skin, the mucous membranes, the lymph nodes and other organs of individuals with compromised immune systems. It is caused by infection with human herpesvirus-8 also known as Kaposi sarcoma-associated herpesvirus or KSHV. The herpesvirus family is unique in that it is the only viral family currently known to express multiple microRNAs (miRNAs); KSHV produces 12 pre-miRNAs, which are processed into at least 25 mature miRNAs. While their functions are not well understood, these miRNAs may be a way for the virus to alter the host immune response without producing proteins that could be recognized and targeted by the immune system. Joseph Ziegelbauer, Ph.D., in CCR’s HIV and AIDS Malignancy Branch, and his colleagues set out to identify human targets of KSHV miRNAs and to understand their functional importance.
Most studies of miRNA targets investigate changes to the RNA, but these studies may miss genes regulated at the protein level. To observe proteins repressed by the KSHV miRNAs, the researchers used a novel method called pulsed stable isotope labeling of amino acids in cell culture (pSILAC) coupled with tandem mass spectrometry (MS). In this assay, they expressed a collection of 16 KSHV miRNAs or a control miRNA in human endothelial cells then introduced amino acids containing stable medium-heavy or heavy isotopes, respectively, into the cell media to label newly-made proteins. The investigators pooled the samples, extracted the proteins, and analyzed them with MS, which can detect the isotope-labeled amino acids. They identified 1,276 proteins that met stringent detection criteria. The 13 most highly repressed proteins included thrombospondin-1, a previously recognized target of KSHV miRNAs, suggesting pSILAC can identify authentic miRNA targets.
miRNAs tend to target sequences in the 3’ untranslated regions (3’UTRs) of mRNA transcripts. Ziegelbauer and colleagues used TargetScan, a bioinformatics program, to find sequences matching the KSHV miRNAs in the 3’UTRs of the transcripts of proteins identified using pSILAC. Their analysis divided the genes into two sets: 847 with at least one 3’UTR matching sequence and 424 without a matching sequence. The strongly repressed proteins were more likely to have at least one predicted 3’UTR matching site. Interestingly, over half of the proteins that were not repressed by KSHV miRNAs had at least one 3’UTR matching site, suggesting a 60 percent false positive rate from using matching sequences alone to identify miRNA targets. But those pSILAC-identified proteins with 3’UTR matching sites tended to have lower expression in the presence of KSHV miRNAs as did proteins with multiple matching sites, suggesting they are bone fide targets.
Because the proteins identified by pSILAC may include direct as well as indirect targets of KSHV miRNAs, the researchers tested whether six of the highly repressed proteins (GRB2, ROCK2, STAT3, HMGCS1, TSPAN3, and AKAP9) were direct targets of KSHV miRNAs. They used fluorescent reporter genes each containing the 3’UTR from the transcript of one of the six proteins and found that the fluorescence from all six reporter genes was suppressed by at least one KSHV miRNA. After mapping the specific sites targeted by the miRNAs for two genes, the investigators mutated these sites, which relieved the miRNA-mediated repression. These results indicate that all six proteins are direct targets of KSHV miRNAs.
To verify that the miRNAs can reduce the endogenous protein levels of GRB2, ROCK2, STAT3 (α and β), and HMGCS1, the scientists expressed the 16 KSHV miRNAs in primary endothelial cells and used quantitative Western blotting to measure protein changes. All five proteins were inhibited by at least one miRNA, and expression of four of the five was repressed by multiple miRNAs. In general, the miRNAs that repressed the 3’UTR reporter of a gene also inhibited the gene’s protein expression. The researchers also wanted to verify that pSILAC-identified proteins are repressed by KSHV infection. In virus infected endothelial cells, they observed significant decreases in the levels of the same five proteins, particularly HMGCS1, validating these targets.
The data generated by the authors also can be used to address the question of whether miRNA-mediated changes in gene expression are mostly seen at the mRNA or protein level. The investigators took cells from their pSILAC study and used microarray analysis to look at mRNA changes. Comparing the protein and mRNA changes, they found that for the six newly-identified targets, the protein changes were stronger than the alterations in mRNA. These results suggest that examining protein expression can indeed find targets missed by studying mRNA alone.
Finally, the researchers wanted to determine the functional relevance of KSHV miRNA-mediated gene expression changes. They found that repression of ROCK2 and STAT3 by the miRNAs prevented lysophosphatidic acid-induced expression of ICAM1, which recruits white blood cells to sites of inflammation. KSHV infection alone also strongly repressed ICAM1 levels, suggesting that the miRNAs could enhance viral immune system evasion. In addition to identifying proteins downregulated by the KSHV miRNAs, the investigators observed enhanced expression of HIF1α and HMOX1, two genes known to promote angiogenesis, which is a hallmark of Kaposi sarcoma. The scientists showed that their increased expression was due at least in part to miRNA-mediated repression of the ubiquitin ligase RBX1 and the transcription factor BACH1, respectively.
Looking more broadly at the pathways regulated by proteins highly repressed by the KSHV miRNAs, the researchers found that they are involved in translation, the cytoskeleton, the cell cycle, chromatin modification, and angiogenesis. These results suggest that KSHV miRNAs target important cell functions and are likely to be critical for KSHV pathogenicity.Summary Posted: Sun, 09/01/2013
Gallaher AM, Das S, Xiao Z, Andresson T, Kieffer-Kwon P, Happel C, Ziegelbauer J. Proteomic Screening of Human Targets of Viral microRNAs Reveals Functions Associated with Immune Evasion and Angiogenesis. PLoS Pathogens. September 5, 2013 PubMed Link