Deciphering the metabolic response of Mycobacterium tuberculosis to nitrogen stress

Abstract

A key component to the success of Mycobacterium tuberculosis as a pathogen is the ability to sense and adapt metabolically to the diverse range of conditions encountered in vivo, such as oxygen tension, environmental pH and nutrient availability. Although nitrogen is an essential nutrient for every organism, little is known about the genes and pathways responsible for nitrogen assimilation in M. tuberculosis. In this study we have used transcriptomics and chromatin immunoprecipitation and high-throughput sequencing to address this. In response to nitrogen starvation, a total of 185 genes were significantly differentially expressed (96 up-regulated and 89 down regulated; 5% genome) highlighting several significant areas of metabolic change during nitrogen limitation such as nitrate/nitrite metabolism, aspartate metabolism and changes in cell wall biosynthesis. We identify GlnR as a regulator involved in the nitrogen response, controlling the expression of at least 33 genes in response to nitrogen limitation. We identify a consensus GlnR binding site and relate its location to known transcriptional start sites. We also show that the GlnR response regulator plays a very different role in M. tuberculosis to that in non-pathogenic mycobacteria, controlling genes involved in nitric oxide detoxification and intracellular survival instead of genes involved in nitrogen scavenging.

Figure 1

Effect of nitrogen limitation on M . tuberculosis growth and gene expression. A. M . tuberculosis was grown in nitrogen‐free Sauton's medium (filled triangles), or containing 1 mM ammonium chloride (nitrogen limiting, filled diamonds), or 30 mM ammonium chloride (nitrogen excess, filled squares). B. The expression of nir B was measured by qRT‐PCR using RNA from three independent cultures, with sig A as internal control. Fold change was calculated as a ratio of the arbitrary expression units, standardised to sig A. Ct values for sig A did not change significantly over nitrogen run‐out. C. The number of genes showing greater than twofold change in differential expression (DE) over nitrogen run‐out at days 8 and 9 compared with day 7. Black bars show an increase in DE; grey bars a decrease in DE. The concentration of external ammonium concentration (mM) in the growth medium as measured by AquaQuant analysis is also shown (dashed line).

Figure 2

Ambient modules are presented in their metabolic network context, illustrating a redirection of metabolism. Squares represent enzymatic reactions and circles represent metabolites. Grey (blue online) represents negatively regulated, and dark (brown online) positively regulated reactions. No definitive claims can be made about fluxes, so arrows are only indicative and taken from the identity or context of the relevant reaction. For reference, reactions with gene numbers are given in Fig. S1.

Figure 3

Novel GlnR binding sites identified upstream of differentially expressed genes, with corresponding EMSA to confirm specific GlnR binding. EMSA were performed by incubating increasing amounts of His‐GlnR recombinant protein with labelled DNA corresponding to the promoter regions of the genes downstream of the GlnR binding site. GlnR binding was visualised in IGV. The top track represents input control DNA, the second track represents GlnR binding in nitrogen excess and the third track represents GlnR binding in nitrogen limiting conditions. Bar height is representative of fold change in gene expression in the wild type compared with the glnR  KO mutant in nitrogen limitation. Levels of gene expression are indicated in the bottom track. (A) Peak 13, Rv1386 (PE15); (B) peak 20, Rv2219A; (C) peak 17, Rv1542c (glbN); (D) peak 18, Rv1548c (PPE22). The negative control Rv1360 DNA showed no GlnR binding (Fig. S4).

Figure 4

Novel GlnR binding sites identified upstream of differentially expressed genes, with corresponding DNA enrichment verified by qPCR. DNA enrichment for an additional four novel GlnR‐binding regions was observed in nitrogen limitation, but not nitrogen excess using rate‐limiting qPCR. (A) Peak 2, Rv0260c (nna R); (B) peak 10, Rv1161 (narG); (C) peak 11, Rv1173 (fbi C); (D) peak 23, Rv2291 (sse B). Lane 1, size marker, 200 bp arrowed. Lane 2, ChIP nitrogen excess. Lane 3, nitrogen excess no antibody control. Lane 4, ChIP nitrogen limiting. Lane 5, nitrogen limiting no antibody control. Lane 6, input nitrogen excess. Lane 7, input nitrogen limiting. Lane 8, no DNA control. A negative control region showed no enrichment (Fig. S3).

Figure 5

M ycobacterium tuberculosis  GlnR consensus binding motif derived from 20 GlnR binding regions identified during nitrogen limitation. MEME generated GlnR motif from 200 bp DNA sequences surrounding the 20 peaks associated with differential gene expression.