A six-amino-acid motif is a major determinant in functional evolution of HOX1 proteins

Abstract

Gene duplication and divergence is a major driver in the emergence of evolutionary novelties. How variations in amino acid sequences lead to loss of ancestral activity and functional diversification of proteins is poorly understood. We used cross-species functional analysis of Drosophila Labial and its mouse HOX1 orthologs (HOXA1, HOXB1, and HOXD1) as a paradigm to address this issue. Mouse HOX1 proteins display low (30%) sequence similarity with Drosophila Labial. However, substituting endogenous Labial with the mouse proteins revealed that HOXA1 has retained essential ancestral functions of Labial, while HOXB1 and HOXD1 have diverged. Genome-wide analysis demonstrated similar DNA-binding patterns of HOXA1 and Labial in mouse cells, while HOXB1 binds to distinct targets. Compared with HOXB1, HOXA1 shows an enrichment in co-occupancy with PBX proteins on target sites and exists in the same complex with PBX on chromatin. Functional analysis of HOXA1-HOXB1 chimeric proteins uncovered a novel six-amino-acid C-terminal motif (CTM) flanking the homeodomain that serves as a major determinant of ancestral activity. In vitro DNA-binding experiments and structural prediction show that CTM provides an important domain for interaction of HOXA1 proteins with PBX. Our findings show that small changes outside of highly conserved DNA-binding regions can lead to profound changes in protein function.

Genome-wide binding analyses of HOXB1 revealed a novel DNA binding motif associated with gene repression

“…Using replicate samples in a series of ChIP-seq experiments we reproducibly mapped 2058 binding peaks for HOXB1 in the mouse genome ( Figure 1A) [26]. Gene ontology (GO) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis of nearest neighbor genes in HOXB1 bound peaks suggest that HOXB1 downstream targets are enriched for genes and processes required for neurogenesis, neuronal processes, and behavior (Figure 2A).…”

“…Gene expression analyses of the differentiated ES cells over a time-course revealed that the transcription profile at 24 h of differentiation is similar to that of the mouse embryonic hindbrain and spinal cord. Therefore, we used cells differentiated cells for 24 h to map the genome-wide binding targets of HOXB1 using Chromatin immunoprecipitation followed by deep sequencing (ChIP-seq) with an antibody against the epitope tag [26,27].…”

“…proteins, along with several novel/unknown motifs ( Figure 1B and ref [26]). In agreement with the lack of co-occupancy of PBX and MEIS the in ChIP-seq data, we find no evidence for enrichment of PBX and MEIS binding motifs.…”

“…In a recent study, we examined this question by using ChIP-seq (Chromatin immunoprecipitation followed by deep sequencing) approaches to characterize the genome-wide binding properties of HOXA1 and HOXB1 proteins in mouse ES cells differentiated into neural fates [26]. These proteins bind to distinct and largely nonoverlapping sets of target genes.…”

“…These proteins bind to distinct and largely nonoverlapping sets of target genes. Analyses of the binding profiles demonstrated that a large number of the HOXA1 bound regions are co-bound with PBX1 and MEIS [26][27][28], which are known cofactors for many HOX proteins and other TFs [29][30][31]. In contrast, only a small subset of HOXB1 bound regions display co-occupancy with PBX and MEIS.…”