Title:	Functional assignments of short-chain dehydrogenases/reductases (SDR)
DNr:	SNIC 001/08-93
Project Type:	SNIC Medium Compute
Principal Investigator:	Yvonne Kallberg <yvonne.kallberg@ki.se>
Duration:	2008-06-04 – 2009-06-30
Classification:	152400
Keywords:	short-chain dehydrogenases/reductases, hidden Markov models, bioinformatics

Abstract

The super-family of short-chain dehydrogenases/reductases (SDR) consists of close to 20 000 enzymes, displaying a large spectrum of functions. The term SDR was coined by us in 1991 and the enzyme family has been shown to be present in all domains of life, from primitive bacteria to higher eukaryotes The three-dimensional structure is a single-domain globular Rossmann-related fold consisting of a 7-stranded beta sheet sandwiched between 3 alpha helices on each side. The SDR family shows early divergence where the majority of its members only have low pairwise sequence identity, but share common sequence motifs that define the cofactor binding site (TGxxxGxG), and the catalytic tetrad (N-S-Y-K). In humans, there exist 135 SDR forms, corresponding to at least 72 different genes, of which several are known to exhibit crucial functions, e.g. in steroid and prostaglandin metabolism. SDRs also play important roles in the metabolism of xenobiotics, drugs and carcinogens. Of the 38 human SDRs that are listed in the well-annotated database Swissprot, 24 enzymes are associated with diseases in the OMIM database (Online Mendelian Inheritance in Man). Thus, many (or even all) of these enzymes are medically important. However, the function is still unknown for about half of the human SDR enzymes. As this field continues to grow, the need for a sub-grouping is essential for functional assignments and reference purposes. The SDRs constitute a large super-family of ancient origin, with most members equidistantly related at the 20-30% residue identity level. Consequently, there are no natural hierarchical relationships to rely on for the functional assignments. Hidden Markov models (HMMs) have successfully been used in protein family characterization, and are today one of the standard techniques when annotating new sequences. This technique is used in our functional categorization of all SDR members, where each SDR family will have an HMM and this set of resulting HMMs forms the basis for a sustainable sub-classification and functional assignment scheme for the whole SDR super-family.