Identifying the Gene Regulatory Network of the Starvation-Induced Transcriptional Activator Nla28

Abstract

Myxococcus xanthus copes with starvation by producing fruiting bodies filled with dormant and stress-resistant spores. Here, we aimed to better define the gene regulatory network associated with Nla28, a transcriptional activator/enhancer binding protein (EBP) and a key regulator of the early starvation response. Previous work showed that Nla28 directly regulates EBP genes that are important for fruiting body development. However, the Nla28 regulatory network is likely to be much larger because hundreds of starvation-induced genes are downregulated in a nla28 mutant strain. To identify candidates for direct Nla28-mediated transcription, we analyzed the downregulated genes using a bioinformatics approach. Nine potential Nla28 target promoters (29 genes) were discovered. The results of in vitro promoter binding assays, coupled with in vitro and in vivo mutational analyses, suggested that the nine promoters along with three previously identified EBP gene promoters were indeed in vivo targets of Nla28. These results also suggested that Nla28 used tandem, imperfect repeats of an 8-bp sequence for promoter binding. Interestingly, eight of the new Nla28 target promoters were predicted to be intragenic. Based on mutational analyses, the newly identified Nla28 target loci contained at least one gene that was important for starvation-induced development. Most of these loci contained genes predicted to be involved in metabolic or defense-related functions. Using the consensus Nla28 binding sequence, bioinformatics, and expression profiling, 58 additional promoters and 102 genes were tagged as potential Nla28 targets. Among these putative Nla28 targets, functions, such as regulatory, metabolic, and cell envelope biogenesis, were assigned to many genes. IMPORTANCE In bacteria, starvation leads to profound changes in behavior and physiology. Some of these changes have economic and health implications because the starvation response has been linked to the formation of biofilms, virulence, and antibiotic resistance. To better understand how starvation contributes to changes in bacterial physiology and resistance, we identified the putative starvation-induced gene regulatory network associated with Nla28, a transcriptional activator from the bacterium Myxoccocus xanthus. We determined the mechanism by which starvation-responsive genes were activated by Nla28 and showed that several of the genes were important for the formation of a highly resistant cell type.

Keywords: biofilms; enhancer-binding proteins; fruiting body development; transcriptional activators; σ54 promoters.

FIG 1

Locations of characterized developmental promoter targets of Nla28. Genes are represented by black (Nla28 targets) or gray (other loci) arrows oriented in the direction of transcription. Dashed arrows represent relatively large genes that are not drawn to scale. Bent arrows represent known or putative σ⁵⁴ promoters. Blue asterisks denote Nla28 half-binding sites analyzed here. Red asterisks denote additional putative Nla28 half-binding sites identified in the promoter regions.

FIG 2

Quantitative real-time PCR measurement of the developmental expression levels and patterns of Nla28 target genes in wild-type and nla28 mutant strains. The developmental mRNA levels of nine Nla28 target genes in wild-type (WT) and nla28 mutant (nla28−) strains were determined using qPCR. Wild-type and nla28 mutant cells were harvested at 0, 1, 2, 8, 12, and 24 h of development for RNA isolation and qPCR analysis. N = 3 technical replicates of pooled RNA samples at each time point. Error bars are standard deviations of the means. The data were analyzed using two-way analysis of variance (ANOVA) and Tukey’s multiple comparisons post hoc tests; ***, P < 0.001; **, P < 0.01; *, P < 0.05 for mRNA levels in nla28 mutant versus wild-type cells.

FIG 3

Electrophoretic mobility shift assays with Nla28-DBD and fragments of putative target promoters. (A) Purified Nla28-DBD binds to fragments of the 9 newly identified target promoters. EMSAs performed with purified Nla28-DBD and a pilA, MXAN881, MXAN989, MXAN2511, MXAN5040, MXAN6732, MXAN7147, MXAN7280, nla28 (positive control), dev (negative control) or mrpB promoter fragment containing putative Nla28 binding sites. Binding reactions were performed with (+) or without (−) 2 μM purified Nla28-DBD and 5 ng of Cy5 5′ end-labeled promoter fragments in a total volume of 10 μL. Similar 8-bp sequences (putative Nla28 half-binding sites) were found in all promoter fragments that were positive for Nla28 binding. (B) Nucleotide frequency at each position of the 8-bp sequence. The consensus 8-bp sequence or consensus Nla28 half-binding site derived from the frequency data was CT(C/G)CG(C/G)AG.

FIG 4

Electrophoretic mobility shift assays with Nla28-DBD and wild-type or mutant fragments of the actB and nla28 promoters. (A) Two 8-bp sequences, half-site 1 (HS1) and half-site 2 (HS2), that closely matched the consensus Nla28 halfsite-binding were identified in the wild-type (WT) actB promoter fragment. Underlined sequences represent nucleotides that matched the consensus Nla28 half-binding site. actB promoter fragments (actB HS1M and HS2M) carrying mutations in these putative Nla28 half-binding sites were generated for in vitro Nla28-DBD binding analysis. (B) EMSAs were performed with (+) or without (−) 2 μM purified Nla28-DBD and 5 ng of a Cy5 5′ end-labeled actB promoter fragment containing two WT Nla28 half-binding sites, actB HS1M or actB HS2M, in a total volume of 10 μL. (C) Three 8-bp sequences (HS1, HS2, and HS3) that closely matched the consensus Nla28 half-binding site were identified in the wild-type nla28 promoter fragment. Underlined sequences represent nucleotides that matched the consensus Nla28 half-binding site. nla28 promoter fragments (nla28 HS1M, HS2M, and HS3M) carrying mutations in these putative Nla28 half-binding sites were generated for in vitro Nla28-DBD binding analysis. (D) EMSAs were performed with (+) or without (−) 2 μM purified Nla28-DBD and 5 ng of Cy5 5′ end-labeled nla2B promoter fragment containing three WT Nla28 half-binding sites, nla28 HS1M, nla28 HS2M, or nla28 HS3M, in a total volume of 10 μL.

FIG 5

In vivo activities of wild-type and mutant promoter targets of Nla28. (A, C, E, and G) nla28, actB, pilA, and MXAN5040 promoter fragments with a 2-bp substitution (mutation 1), 4-bp substitution (mutation 2), 6-bp substitution (mutation 3), or 8-bp substitution (mutation 4) in Nla28 half-site 2 were generated by site-directed mutagenesis. Substituted nucleotides are shown in red. (B, D, F, and H) Wild-type and derivatives of the nla28, actB, pilA, and MXAN5040 promoters carrying mutation 1, mutation 2, mutation 3, or mutation 4 were cloned into a lacZ expression vector and transferred to the wild-type M. xanthus strain DK1622. In vivo activities of wild-type and mutant promoters were inferred from β-galactosidase-specific activities (defined as nanomoles of ONP produced per minute per milligram of protein) at various time points (0, 2, 6, 8, 12, 24, and 48 h) during development. N = 3 biological replicates at each of the indicated time points. Error bars are standard deviations of the means.

FIG 6

Functional category and subcategory distributions of Nla28 targets based on the clusters of orthologous groups of proteins (COGs) database. (A) The two-layer pie chart shows COG functional categories, subcategories, and their relative abundances of confirmed Nla28 targets in this study. (B) The two-layer pie chart shows COG functional categories, subcategories, and their relative abundances of putative Nla28 targets in this study. The inner layer of pie charts represents categories. The outer layer represents subcategories. The subcategory of no mapping indicates hypothetical proteins without predicted orthologs in the COG database.