Elucidation of the functional metal binding profile of a Cd(II)/Pb(II) sensor CmtR(Sc) from Streptomyces coelicolor

Abstract

Metal homeostasis and resistance in bacteria is maintained by a panel of metal-sensing transcriptional regulators that collectively control transition metal availability and mediate resistance to heavy metal xenobiotics, including As(III), Cd(II), Pb(II), and Hg(II). The ArsR family constitutes a superfamily of metal sensors that appear to conform to the same winged helical, homodimeric fold, that collectively "sense" a wide array of beneficial metal ions and heavy metal pollutants. The genomes of many actinomycetes, including the soil dwelling bacterium Streptomyces coelicolor and the human pathogen Mycobacterium tuberculosis, encode over ten ArsR family regulators, most of unknown function. Here, we present the characterization of a homologue of M. tuberculosis CmtR (CmtR(Mtb)) from S. coelicolor, denoted CmtR(Sc). We show that CmtR(Sc), in contrast to CmtR(Mtb), binds two monomer mol equivalents of Pb(II) or Cd(II) to form two pairs of sulfur-rich coordination complexes per dimer. Metal site 1 conforms exactly to the alpha4C site previously characterized in CmtR(Mtb) while metal site 2 is coordinated by a C-terminal vicinal thiolate pair, Cys110 and Cys111. Biological assays reveal that only Cd(II) and, to a lesser extent, Pb(II) mediate transcriptional derepression in the heterologous host Mycobacterium smegmatis in a way that requires metal site 1. In contrast, mutagenesis of metal site 2 ligands Cys110 or Cys111 significantly reduces Cd(II) responsiveness, with no detectable effect on Pb(II) sensing. The implications of these findings on the ability to predict metal specificity and function from metal-site signatures in the primary structure of ArsR family proteins are discussed.

Figure 1. Genomic region of CmtR homologs in S. coelicolor, solution structure and multiple sequence alignment of M. tuberculosis CmtR

(A) Genomic region around S. coelicolor CmtR^Sc (SCO0875) and SCO3522; each gene is separated by a single TGA termination codon (*) from homologous downstream genes SCO0874 and SCO3521 that encode putative CDF-family heavy metal transporters. The surrounding genomic regions are completely unrelated in the two loci. (B) Ribbon diagram of the solution structure of the M. tuberculosis CmtR-Cd^II complex (2jsc) with individual protomers shaded slate and violet and the two symmetry-related Cd^II ions colored green (8). The secondary structural units are labeled, with the side-chains of C24 and the Cd^II-coordinating residues, C57, C61 and C102' (prime designation, opposite subunit) highlighted in stick. The most C-terminal residue in the structural model of each protomer is R106. (C) Multiple sequence alignment of CmtR^Mtb, CmtR^Sc and the product of SCO3522. Cd^II-coordinating residues in CmtR^Mtb are highlighted in bold, with other residues as in panel B.

Figure 2. Analytical sedimentation equilibrium ultracentrifugation of CmtR^Sc and C110G/C111S CmtR^Sc

(A) 5.03 μM monomer wild-type CmtR^Sc. (B) 8.0 μM monomer C110G/C111S CmtR^Sc. Filled symbols in upper panels represent an overlay of data collected during the last seven scans and indicate that equilibrium had been reached. The solid line represents the global simultaneous fit for a single ideal species model using Ultrascan. For wild-type CmtR^Sc, the fitted M_w is 23,820 Da (theoretical dimer M_w = 24,378 Da), variance = 1.0689 e⁻⁴. For C110G/C111S CmtR^Sc, the fitted M_w is 25,690 Da (theoretical dimer M_w = 24,252 Da), variance = 5.8026 e⁻⁵. Conditions: 10 mM Hepes, 0.4 M NaCl, and 0.1 mM EDTA at pH 7.0, 25.0 °C, and 20,000 rpm rotor speed.

Figure 3. Cd^II titrations of wild-type and C110G/C111S CmtR^Sc

(A) Apoprotein subtracted difference spectrum of wild-type CmtR^Sc (50.7 μM monomer). (B) Apoprotein subtracted difference spectrum of C110G/C111S CmtR^Sc (44.8 μM monomer). CmtR^Sc variants were titrated anaerobically with increasing concentrations of Cd^II. Inset, Cd^II binding isotherm plotted as change in A₂₄₀ vs. [CmtR^Sc variant monomer]. (C) Cd^II-EDTA competition binding isotherm in which 20.9 μM wild-type CmtR^Sc was titrated with Cd^II in the presence of 227 μM EDTA. Conditions: 10 mM Hepes, 0.4 M NaCl, pH 7.0, 25°C.

Figure 4. Pb^II titrations of CmtR^Sc and C110G/C111S CmtR^Sc

(A) Apoprotein subtracted difference spectrum of wild-type CmtR^Sc (50.8 μM). (B) Apoprotein subtracted difference spectrum of C110G/C111S CmtR^Sc (53.7 μM). CmtR^Sc variants were titrated anaerobically with increasing concentrations of Pb^II. Inset, Pb^II binding isotherm plotted as change in A₃₃₃ vs. [CmtR^Sc variant monomer] Conditions: 10 mM Bis-Tris, 0.4 M NaCl, pH 7.0, 25.0 °C.

Figure 5. Zn^II titrations of CmtR^Sc and C110G/C111S CmtR ^Sc using magfura-2 as an indicator

(A) 30 μM CmtR^Sc and 14.7 μM magfura-2 and (B) 30 μM C110G/C111S CmtR^Sc and 15 μM magfura-2 were present. The empty circles represent A₃₂₅ and the filled squares represent A₃₆₆. The solid line represents a global non-linear least square fit to a model that incorporates the stepwise binding of two Zn^II (defined by K_PZn and K_PZn2) to a CmtR^Sc monomer using Dynafit. The titration for C110G/C111S CmtR^Sc was fitted using one Zn^II to protein monomer binding model. The following parameters were obtained for wild-type CmtR^Sc: K_PZn = 5.3 (±1.8) × 108 M⁻¹ (a lower limit under these conditions), K_PZn2 = 6.7 (±1.4) × 10⁸ M⁻¹. For C110G/C111S CmtR^Sc, K_PZn = 5.5 (±1.2) × 10⁷ M⁻¹. Conditions: 10 mM Hepes, 0.4 M NaCl at pH 7.0 and 25.0 °C.

Figure 6. CmtR^Sc responds to Cd^II and Pb^II in an actinomycete host

(A) β-galactosidase activity measured in M. smegmatis mc²155 containing cmtR^Sc and its operator-promoter region fused to lacZ following growth in LB medium with no metal supplement or with maximum permissive concentrations of Zn^II (100 μM), Co^II (200 μM), Ni^II (500 μM), Cd^II (7.5 μM), Cu^II (500 μM), Pb^II (3.75 μM), As^III(20 μM), or Hg^II (0.025 μM). (B) β-galactosidase activity in cells containing wild-type CmtR^Sc (WT) or the stop codon derivative (R16*) following growth in LB medium with no metal supplement (black) or maximum permissive concentrations of Cd^II (gray) or Pb^II (white).

Figure 7. Metal sensing ligands of CmtR^Sc

(A) β-galactosidase activity measured in M. smegmatis mc²155 containing wild-type CmtR^Sc (WT) or derivatives with indicated cysteine to serine codon substitutions, following growth in LB medium with no metal supplement (black) or maximum permissive concentrations of Cd^II (gray) or Pb^II (white). (B) β-galactosidase activity in cells containing wild-type CmtR^Sc (black circles) or the C110S (gray squares) and C111S derivatives (gray diamonds) grown in LB with up to inhibitory concentrations of Cd^II or Pb^II. Data points represent the mean (± SE) for three independent experiments, each performed in triplicate.

Figure 8. Metal sensing ligands of CmtR^Sc as determined on a chemically defined minimal medium

Top panel, Cell viability of M. smegmatis mc²155 containing wild-type CmtR^Sc (black circles), C110S (gray squares) and C111S derivatives (gray diamonds) grown on minimal Sauton medium (containing 2.9 mM phosphate) as a function of total added Cd^II (left) or Pb^II (right). Lower panel, β-galactosidase activity in cells containing wild-type CmtR^Sc (black circles), C110S (gray squares) or C111S CmtR^Sc (gray diamonds) grown in minimal media up to inhibitory concentrations of Cd^II or Pb^II. Data points represent the mean (± SE) for three independent experiments, each performed in triplicate.