Identifying Divergent HAMP Domains and Poly-HAMP Chains

Abstract

Meena et al. (1Meena N. Kaur H. Mondal A.K. J. Biol. Chem. 2010; 285: 12121-12132Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar) report the mechanistic investigation of poly-HAMP domains from Debaryomyces hansenii Nik1 (DhNik1), a soluble hybrid histidine kinase. A standard SMART search (http://smart.embl-heidelberg.de/) predicts hNik1 to contain 5 successive HAMP (histidine kinase, adenylyl cyclase, methyl-accepting chemotaxis protein, and phosphatase) domains, each separated by a 40-residue linker. We wish to make the authors and general community aware that a SMART or BLAST search (http://blast.ncbi.nlm.nih.gov/) is usually incapable of correctly identifying all HAMP domains in a protein sequence. In the case of DhNik1, a more detailed analysis reveals that the N-terminal region actually contains 9 successive HAMP domains, instead of 5 (Fig. 1). In light of this information, the domain deletion and yeast 2-hybrid results reported by Meena et al. (1Meena N. Kaur H. Mondal A.K. J. Biol. Chem. 2010; 285: 12121-12132Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar) should be reassessed. In particular, the choice of inappropriate domain boundaries in these experiments calls into question the proposed model of side-on HAMP-HAMP interactions. The DhNik1 HAMP domains belong to a recently identified subset of divergent HAMP domains (2Dunin-Horkawicz S. Lupas A.N. J. Mol. Biol. 2010; 397: 1156-1174Crossref PubMed Scopus (70) Google Scholar) that form repeating poly-HAMP units. Most divergent HAMP domains have been identified using sensitive HMM-HMM sequence alignment methods (2Dunin-Horkawicz S. Lupas A.N. J. Mol. Biol. 2010; 397: 1156-1174Crossref PubMed Scopus (70) Google Scholar). However, the structure of a 3-unit poly-HAMP chain from the soluble receptor Aer2 reveals the interwoven and linear arrangement of successive HAMP domains within poly-HAMP chains and thereby provides a useful template for the interpretation of sequence alignments (3Airola M.V. Watts K.J. Bilwes A.M. Crane B.R. Structure. 2010; 18: 436-448Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar). As this new knowledge is very recent and was not available to the authors at the time of their study, it is not surprising that the divergent HAMP domains of DhNik1 were misassigned. To avoid future discrepancies we urge a careful inspection of any protein sequence predicted to contain a HAMP domain. A defining feature of divergent HAMP domains is 2 glycine residues located at the end of the N-terminal helix AS1 and start of the C-terminal helix AS2. These residues allow the close association of neighboring domains in a poly-HAMP chain (3Airola M.V. Watts K.J. Bilwes A.M. Crane B.R. Structure. 2010; 18: 436-448Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar) and can be used as a marker, in conjunction with other conserved motifs (Fig. 1), to identify divergent HAMP domains by manual sequence inspection. Reply to Airola et al.: Linker or a HAMP?Journal of Biological ChemistryVol. 285Issue 23PreviewThis is a response to a letter by Airola et al. (1) Full-Text PDF Open Access