Prediction and analysis of new HisKA-like domains
Background: Histidine kinases (HKs) are key components of many signaling pathways, particularly through their involvement in two-component systems (TCS). By utilizing autophosphorylation and phosphotransfer to response regulators (RRs), they enable organisms to adapt to their environment. Most HKs are transmembrane proteins featuring a sensing domain located outside the cell and two catalytic domains, HisKA and HATPase. The HATPase domain mediates interaction with ATP, while HisKA harbors the phosphorylatable histidine residue. HKs participate in a wide range of environmental adaptation mechanisms, including light sensing and responses to biochemical changes. Characterizing their diversity is therefore essential to better understand how cells interact with their environment. Incomplete HKs (iHKs), lacking either the HisKA or HATPase domain, have been described. Among iHKs that retain an HATPase domain, some possess a region of their sequence where an HisKA domain would be expected. These iHKs may include ``true'' HKs harboring novel, uncharacterized HisKA domains, which could fill gaps in various signaling pathways. Results: In this study, we analyzed 869 964 iHK sequences carrying an HATPase domain but lacking an HisKA domain. We generated 18 HisKA-like HMM profiles and conducted multiple meta-analyses to assess their HisKA-like characteristics. Their predicted 3D structures were found to match those of known HisKA domains. Furthermore, the genomic context of genes associated with these profiles revealed the presence of genes implicated in signal transduction pathways. Several profiles were cross-validated against curated annotations, as well as against a ``negative dataset'' composed of non-HK proteins. Lastly, using these profiles we identified 112 iHKs across 22 model organisms of known interest. Conclusions: This work describes 18 HisKA-like profiles identified in prokaryotic sequences, supported by multiple lines of evidence for their HisKA characteristics. These profiles, along with the complete methodology used to identify them, are available in the data repository https://doi.org/10.57745/Y9NOP9 and the git repository https://gitlab.in2p3.fr/louison.silly/hiska-like-domain-characterization. We encourage the integration of these profiles into genome annotation pipelines, as we believe they could facilitate the identification of novel HKs in prokaryotic regulatory pathways, both in model and non-model organisms.