Mycobacterium tuberculosis expresses a novel pH-dependent divalent cation transporter belonging to the Nramp family

Abstract

Mammalian natural resistance-associated macrophage protein (Nramp) homologues are important determinants of susceptibility to infection by diverse intracellular pathogens including mycobacteria. Eukaryotic Nramp homologues transport divalent cations such as Fe(2+), Mn(2+), Zn(2+), and Cu(2+). Mycobacterium tuberculosis and Mycobacterium bovis (bacillus Calmette-Guérin [BCG]) also encode an Nramp homologue (Mramp). RNA encoding Mramp induces approximately 20-fold increases in (65)Zn(2+) and (55)Fe(2+) uptake when injected into Xenopus laevis oocytes. Transport is dependent on acidic extracellular pH and is maximal between pH 5.5 and 6.5. Mramp-mediated (65)Zn(2+) and (55)Fe(2+) transport is abolished by an excess of Mn(2+) and Cu(2+), confirming that Mramp interacts with a broad range of divalent transition metal cations. Using semiquantitative reverse transcription PCR, we show that Mramp mRNA levels in M. tuberculosis are upregulated in response to increases in ambient Fe(2+) and Cu(2+) between <1 and 5 microM concentrations and that this upregulation occurs in parallel with mRNA for y39, a putative metal-transporting P-type ATPase. Using a quantitative ratiometric PCR technique, we demonstrate a fourfold decrease in Mramp/y39 mRNA ratios from organisms grown in 5-70 microM Cu(2+). M. bovis BCG cultured axenically and within THP-1 cells also expresses mRNA encoding Mramp. Mramp exemplifies a novel prokaryotic class of metal ion transporter. Within phagosomes, Mramp and Nramp1 may compete for the same divalent cations, with implications for intracellular survival of mycobacteria.

Figure 1

Putative distribution of membrane-spanning segments of Mramp to illustrate alignment of conserved residues with homologues. This sketch has been derived from detailed hydropathy analyses performed by Cellier et al. (reference 31). The precise number of membrane-spanning segments and topology of the COOH-terminal region are still uncertain. Mramp sequence was analyzed using the Kyte-Doolittle algorithm (window size of 16 amino acids). The highly conserved distribution of thermodynamically unfavored charged residues within transmembrane segments (M), which is a feature of eukaryotic Nramp homologues (reference 31), is also discernible in Mramp and other prokaryotic homologues (boxed residues). The positions of two adjacent glycine residues in M4 (region c, marked *) are known to be functionally important; in murine Nramp1, a G169D mutation is associated with the pathogen-susceptible phenotype (reference 3), while in Nramp2, a mutated neighboring glycine in the mk mouse (G185R; reference 36) and Belgrade (b) rat (reference 11) causes microcytic (iron deficiency) anemia, probably through steric and charge effects. Mramp contains a consensus transport motif (CTM) resembling the “EAA” box or “binding protein–dependent transport system's inner membrane component signature” (region f) (see reference 31). This motif [E,Q] [S,T,A]×2, 3X,G, [L,I,V,M,F,Y,A], 4X, [F,L,I,V], [P,K] is found in numerous bacterial periplasmic permeases and some eukaryotic multisubunit transporters (reference 31). Mutational analysis of three residues in this region (marked * in region f) in murine Nramp2 has demonstrated the functional importance of the first two residues (reference 10). Other highly conserved residues are shown in bold. DCT1 sequence = rat Nramp2; S. cerevisiae sequence = smf1. Sequence data are available from EMBL/GenBank/DDBJ under accession nos. U15184 (M. leprae), Z99106 (B. subtilis), U00096 (E. coli), U15929 (Smf1), AF008439 (DCT1), and L32185 (Nramp1, human).

Figure 2

⁶⁵Zn²⁺ uptake by oocytes expressing Mramp. (A) Induction of ⁶⁵Zn²⁺ uptake by Mramp in Xenopus oocytes injected with RNA (5 ng in 50 nl) transcribed from pXmramp. The difference in ⁶⁵Zn²⁺ uptake between experimental and water (50 nl)–injected control oocytes is highly significant (P < 0.001). Inset shows a separate experiment in which ⁶⁵Zn²⁺ uptake in Mramp cRNA–injected oocytes was compared with antisense cRNA (pmarm)-injected control oocytes (P = 0.025). (B) ⁶⁵Zn²⁺ uptake in water-injected and THT1 (5 ng)–injected control oocytes. THT1 is the T. brucei hexose transporter. There was no significant difference in ⁶⁵Zn²⁺ uptake between water- and THT1-injected groups (P = 0.5). (C) 2-DOG uptake in Mramp- (5 ng), THT1- (5 ng), and water-injected oocytes. THT1 induces significantly greater 2-DOG uptake than either Mramp (P < 0.002) or water (P < 0.002). There was no significant difference in uptakes between the water- and Mramp-injected oocytes. Displayed are mean values (± SE) of uptakes (10 oocytes per experimental condition).

Figure 3

pH dependence of ⁶⁵Zn²⁺ uptake induced by Mramp. ⁶⁵Zn²⁺ uptake was assayed in Mramp-injected and control oocytes in media varying in pH (see Materials and Methods). The increase in ⁶⁵Zn²⁺ uptake (Fold) is shown for Mramp-expressing oocytes compared with controls. Uptakes at pH 5.5 (P < 0.001) and pH 6.0 (P ≤ 0.009) were significantly greater than uptakes under remaining conditions. There were no significant differences in uptake at pH 5.0, 6.5, and 7.0 (P > 0.1). Displayed are mean values (± SE) of uptakes (10 oocytes per experimental condition).

Figure 4

Substrate specificity for Mramp. (A) Abolition of ⁶⁵Zn²⁺ uptake by Mn²⁺ in Mramp RNA–injected oocytes. Increase in uptake compared with control oocytes in the absence and presence of Mn²⁺ (10 mM as MnCl₂; P = 0.01). (B) Uptake of ⁵⁵Fe²⁺ by Mramp RNA–injected oocytes compared with water-injected controls (P < 0.001). (C) Influence of divalent cation competitors on ⁵⁵Fe²⁺ (100 μM) uptake by Mramp RNA–injected oocytes. Data from one experiment. Displayed are mean values (± SE) of uptakes (10 oocytes per experimental condition).

Figure 5

Growth of M. tuberculosis H37Rv in different cationic conditions. (A) RT-PCR for Mramp and y39 carried out on total RNA isolated from M. tuberculosis H37Rv cultured in media containing different Fe²⁺ and Cu²⁺ concentrations. Equal amounts of RNA template (30 ng) were used in each reaction (3 ng for 16S RNA experiments). Lane 1, low Fe²⁺(<1 μM); lane 2, medium Fe²⁺ (4 μM); lane 3, high Fe²⁺ (48 μM); lane 4, low Cu²⁺ (<0.5 μM); lane 5, medium Cu²⁺ (5 μM); lane 6, high Cu²⁺ (70 μM). (B) Bacterial growth after 5 wk culture at each metal ion concentration. (C) Top panel, RT-PCR for Mramp on total RNA isolated from axenically cultured and intracellular BCG. Equal amounts of RNA template (135 ng) were used in each reaction. Lane 1, template from extracellular (axenically cultured) BCG; lane 2, template from extracellular BCG pretreated with RNAse A; lane 3, template from intracellular BCG; lane 4, template from intracellular BCG pretreated with RNAse A. Bottom panel, RT-PCR for rRNA (16S) using amounts of total RNA template identical to top panel. Lane 1, template from extracellular (axenically cultured) BCG; lane 2, template from extracellular BCG pretreated with RNAse A; lane 3, template from intracellular BCG; lane 4, template from intracellular BCG pretreated with RNAse A.