Application of computational biology in understanding emerging infectious diseases: inferring biological function for S-M complex of SARS-CoV

Bioinformatics can play a special role in studying emerging infectious diseases, where a fast characterization of the diseases is often urgently needed before they are widespread. In this chapter, we described a computational framework for fast modeling of emerging infectious diseases, including five components: (1) gene survey of the biological sequences rela ted to the emerging diseases; (2) biological problem definition; (3) computational analyses of each gene sequence, which may include molecular characterization, structure prediction, phylogenetic analysis, and regulatory motif prediction; (4) literature review; (5) inference of the biological mechanism and hypothesis for experimental testing. We selected Severe acute respiratory syndrome coronavirus (SARS-CoV) as an example to illustrate the computational framework. SARS-CoV is a most recently emerging coronavirus causing the worldwide SARS pandemic. It is a novel coronavirus, which is similar in genome organization but distantly related to previously characterized coronaviruses in gene sequences. Among the identified open ORFs, replicase ORF1ab, spike [S], envelope [E], membrane [M] and nucleopcapsid [N] are found in other known coronaviruses with a conserved genome organization. In addition to this common genome organization, this novel virus also has a number of nonstructural proteins with unknown functions. We first computationally surveyed all the open reading frames within SARS-CoV, and then predicted the cleavage sites between S1 and S2. According to the literature, M and S proteins play an important role in the processing of assembly, attachment, and fusion for coronavirus-host interaction. However, there has not been a systematically description about the mechanism of the M and S proteins. Although the cellular roles of the S and M proteins are known, it requires structural information to understand the molecular mechanisms of how these two proteins perform their functions. The protein structure of either the M protein or the S protein has not been solved. Questions remains include: (1) How S1 protein interacts with the associated human receptor? (2) How the S protein interacts with M protein? (3) What is the role of the S-M complex in virus fusion? (4) What are the other possible roles of the M protein in the virus infection besides its role in cell assembly? In this chapter, we only focused on question 2. We first performed the molecular characterization of S and M proteins. Then we predicted the threedimensional structures for the S and M proteins and the potential human receptor Naminopeptidase. We further illustrated how the S-M complex assembles based on correlated mutation analysis. In the end, hypotheses about the S-M complex was proposed based on our computational results and previous experimental results.