Current Approaches on Viral Infection: Proteomics and Functional Validations

Zheng, Jie; Tan, Boon Huan; Sugrue, Richard; Tang, Kai

With the advent of high throughput short-gun proteomics, liquid chromatography coupled with tandem MS (LC-MS/MS) could characterize numerous fragment ion spectra and thus is able to identify large number of peptide sequences. In addition, accessing highlighted post translational modifications (PTMs) and protein quantifications by stable isotope labeling or label-free analysis have incorporated into the multi-functional search engines, enriching the mass spectrometric data. Presently, several proteomic search engines are available from distinctive searching algorithms and compatible with mass spectrometric data, including some traditional ones, e.g., Mascot, SEQUEST, X!TANDEM, as well as some new search engine, e.g., ProluCID, etc. Mascot incorporates peptide mass fingerprint, sequence query, and MS/MS ions search. It is a possibility-based scoring engine by calculating the observed match between experimental data and theoretical sequence data (Perkins et al., 1999) . For instance, Lai et al. (2007) employed Mascot to analyzed the nano LC-MS/MS data and identified up-regulated proteins in human promonocyte cells stably expressing SARS CoV 3C-like protease. Although Mascot provides high throughput protein identification dependent on possibility rankings, there exist some limitations associated with searching Frontiers in Microbiology | Virology non-independent dataset, and non-statistical validation of atypical sequence entries. SEQUEST could correlate ion fragmentations in the processed tandem mass spectrometric data with their corresponding amino acid sequences in the FASTA database files. There are generally four steps of this search algorithm, including tandem mass spectra reduction, matching spectra with amino acids, generating high-ranked sequences, and correlating with protein identification (Eng et al., 1994; Yates, 1998) . SEQUEST also is able to search several covalent modification-bearing peptides by matching the nascent tandem mass spectra (Yates et al., 1995) . It had been applied to identify secretome of human monocytederived macrophages after HIV-1-infection (Ciborowski et al., 2007) . X!TANDEM is an open-source platform for proteomic researchers to efficiently process MS/MS data (Craig and Beavis, 2004) . Its analysis on a mixture of peptides is based on one axiom: for each detectable protein in the original protein mixture, there will be at least one good tryptic peptide match within a designed scope. In the first step of the analysis, a smaller set of protein sequences is generated from the original protein database by thoroughly filtering with the designed scope that set as small as possible. From this step onward, the subsequent searches are within this refined protein sequences, thus improving the efficiency and saving the overall search time. In the second step of analysis, a bigger scope is set to perform multiple comparisons of the spectra with those refined protein sequences in respect to the different peptide modifications, number of missed cleavages, and non-specific hydrolysis, etc (Craig and Beavis, 2003) . Trans-Proteomic Pipeline (TPP) was utilized to statistically analyze the viral and host proteins in purified RSV (Radhakrishnan et al., 2010) . And global proteome machine (gpm) is a well-established open-source search engine based on TANDEM 3 . Recently, Xu et al. developed a new search engine ProluCID, which is based on binomial probability preliminary scoring scheme to filter candidate peptides for furth