A Moraxella catarrhalis two-component signal transduction system necessary for growth in liquid media affects production of two lysozyme inhibitors

Abstract

There are a paucity of data concerning gene products that could contribute to the ability of Moraxella catarrhalis to colonize the human nasopharynx. Inactivation of a gene (mesR) encoding a predicted response regulator of a two-component signal transduction system in M. catarrhalis yielded a mutant unable to grow in liquid media. This mesR mutant also exhibited increased sensitivity to certain stressors, including polymyxin B, SDS, and hydrogen peroxide. Inactivation of the gene (mesS) encoding the predicted cognate sensor (histidine) kinase yielded a mutant with the same inability to grow in liquid media as the mesR mutant. DNA microarray and real-time reverse transcriptase PCR analyses indicated that several genes previously shown to be involved in the ability of M. catarrhalis to persist in the chinchilla nasopharynx were upregulated in the mesR mutant. Two other open reading frames upregulated in the mesR mutant were shown to encode small proteins (LipA and LipB) that had amino acid sequence homology to bacterial adhesins and structural homology to bacterial lysozyme inhibitors. Inactivation of both lipA and lipB did not affect the ability of M. catarrhalis O35E to attach to a human bronchial epithelial cell line in vitro. Purified recombinant LipA and LipB fusion proteins were each shown to inhibit human lysozyme activity in vitro and in saliva. A lipA lipB deletion mutant was more sensitive than the wild-type parent strain to killing by human lysozyme in the presence of human apolactoferrin. This is the first report of the production of lysozyme inhibitors by M. catarrhalis.

FIG 1

Characterization of wild-type M. catarrhalis ETSU-9 and the mesR, mesS, and complemented mutants of M. catarrhalis ETSU-9. (A) The mesRS locus in wild-type ETSU-9; (B) the mesR mutant ETSU-9ΔmesR; (C) the mesS mutant ETSU-9ΔmesS. (D) Western blot analysis of MesR and CopB protein production by wild-type ETSU-9 (lane 1), ETSU-9ΔmesR (lane 2), ETSU-9ΔmesR(pSNJ224) (lane 3), and ETSU-9ΔmesR(pWW115) (lane 4). (E) Western blot analysis of MesS and CopB protein production by wild-type ETSU-9 (lane 1), ETSU-9ΔmesS (lane 2), ETSU-9ΔmesS(pSNJ225) (lane 3), and ETSU-9ΔmesS(pWW115) (lane 4). Detection of the CopB outer membrane protein (22) was used as a loading control to ensure that equivalent amounts of whole-cell lysates were present in each lane. Size position markers (in kDa) are present on the left side of panels D and E. (F) RT-PCR-based analysis of transcriptional linkage between mesR and mesS. Primers 724 F and 725 R were designed to allow amplification of a 233-bp product that spans from mesR to mesS. Primers 990 F and 990 R bind inside the oppA ORF (105), were designed to allow amplification of a 396-bp product, and were used here as a positive control. Lanes 1 and 4, samples from RT-PCR with wild-type ETSU-9 RNA as the template and RT added; lanes 2 and 5, samples from RT-PCR with wild-type ETSU-9 RNA as the template and no RT; lanes 3 and 6, samples from RT-PCR with wild-type ETSU-9 chromosomal DNA as the template. Size position markers (in 100-bp increments) are presented on the left side.

FIG 2

Growth characteristics of wild-type M. catarrhalis ETSU-9 and mesR, mesS, and complemented mutants of M. catarrhalis ETSU-9. (A) Broth growth curves for wild-type ETSU-9 (filled diamonds), ETSU-9ΔmesR (open squares), ETSU-9ΔmesR(pSNJ224) (open circles), and ETSU-9ΔmesR(pWW115) (filled triangles). (B) Broth growth curves for wild-type ETSU-9 (filled diamonds), ETSU-9ΔmesS (open squares), ETSU-9ΔmesS(pSNJ225) (open circles), and ETSU-9ΔmesS(pWW115) (filled triangles). (C) Inhibition of the growth of wild-type ETSU-9, the mesR mutant, and complemented mutants on BHI agar by different agents, including SDS, polymyxin B, and hydrogen peroxide; data were analyzed by using a 2-way analysis of variance with a Sidak correction (**, P < 0.0022; ***, P < 0.0009; ****, P < 0.001). All growth data are the results of two independent experiments.

FIG 3

Comparison of DNA microarray and real-time RT-PCR (quantitative RT-PCR [qRT-PCR]) data for selected M. catarrhalis ETSU-9 genes whose expression was affected by inactivation of mesR. (A) Expression levels of 21 selected genes in ETSU-9ΔmesR cells relative to those in wild-type ETSU-9 cells were measured by DNA microarrays (black bars) or quantitative RT-PCR analysis (white bars), as described in Materials and Methods. These data are the means of the results from three independent DNA microarray experiments and three independent quantitative RT-PCR experiments. (B) Plot of the correlation between the log₂ values from the DNA microarray and quantitative RT-PCR analyses. The diagonal line represents the power trend line.

FIG 4

Characterization of wild-type M. catarrhalis O35E and lipA, lipB, and lipA lipB mutants of M. catarrhalis O35E. (A) The lipA and lipB loci in wild-type O35E; (B) the lipA mutant O35EΔlipA; (C) the lipB mutant O35EΔlipB; (D) the O35EΔlipAlipB double mutant. MCORF 995 is labeled “M35-like.” (E) Western blot analysis of LipA, LipB, and CopB protein production by wild-type O35E (lane 1), O35EΔlipA (lane 2), O35EΔlipB (lane 3), and O35EΔlipAlipB (lane 4). Detection of the CopB outer membrane protein (22) was used a loading control to ensure that equivalent amounts of whole-cell lysates were present in each lane. Size position markers (in kDa) are present on the left side of panel E. (F) Attachment of wild-type O35E and O35EΔlipAlipB to 16HBE14o- cells in vitro; the attachment assay was performed as described previously (22). These results are the mean of two independent experiments (Mann-Whitney test; P = 0.85). ns, no statistically significant difference.

FIG 5

Alignment of the M. catarrhalis LipA and LipB proteins with homologous proteins. A structure-based sequence alignment of LipA, LipB, the N. meningitidis Acp adhesin (NmAcp), the two E. corrodens LecA domains (EcN, the E. corrodens amino-terminal domain; EcC, the E. corrodens carboxyl-terminal domain), the B. abortus lysozyme inhibitor PliC (PliC-Ba), and the S. Typhimurium lysozyme inhibitor PliC (PliC-St) is shown. Residue numbering of the full-length proteins is shown at the right and left. The secondary structure (106) from the known structure of S. Typhimurium PliC (65) is shown, with blue arrows depicting β strands and black lines showing loop regions; all strands are labeled, as are relevant loops. S. Typhimurium PliC residues homologous to those known to contact hen egg white lysozyme in the P. aeruginosa MliC-hen egg white lysozyme structure (66) are boxed. Also shown in boxes are residues from B. abortus PliC that contact human lysozyme in that complex structure (68). The two absolutely conserved cysteines are highlighted in yellow, and the absolutely conserved serine is highlighted in blue. Other conserved residues are highlighted in purple.

FIG 6

The M. catarrhalis LipA and LipB proteins inhibit lysozyme activity. (A) Purified recombinant MBP-LipA (closed circles), MBP-LipB (open circles), and MBP (closed squares) (1 μM each) were incubated with dried cells of M. lysodeikticus and recombinant human lysozyme (20 U) as described in Materials and Methods. Lysozyme activity was measured by determination of the decrease in the OD₆₀₀ over time. A 2-way analysis of variance with a Tukey correction for multiple comparisons indicated that the extent of lysis obtained in the presence of MBP-LipA and MBP-LipB was different (P < 0.05) from that obtained with MBP by the 15-min time point. By the 75-min time point, the extent of lysis obtained with MBP-A was different (P < 0.05) from that obtained with MBP-LipB. These are the data from two independent experiments each involving triplicate sample wells. (B) Purified recombinant MBP-LipA (open circles) and MBP (closed squares) (1 μM each) were incubated with dried cells of M. lysodeikticus and hen egg white lysozyme (40 U). A 2-way analysis of variance with a Sidak correction for multiple comparisons indicated no difference in the extent of lysis obtained with MBP-LipA and MBP at any time point during the assay. These are the data from two independent experiments each involving triplicate sample wells. (C) Inhibition of lysozyme activity in pooled normal human saliva. Dried cells of M. lysodeikticus were incubated with buffer (closed triangles), saliva (open squares), saliva and MBP (closed squares), saliva and MBP-LipA (closed circles), and saliva and MBP-LipB (open circles) as described in Materials and Methods. Data were analyzed by using a 2-way analysis of variance for multiple comparisons with a Tukey correction. By 15 min, lysis caused by saliva was more extensive than that obtained with the buffer control (P < 0.0001). By 15 min, the extent of lysis obtained in the presence of saliva with either MBP-LipA or MBP-LipB was different (P < 0.0001) from that obtained with saliva and MBP. By 15 min, the extent of lysis obtained with saliva and MBP-LipA was different (P < 0.0001) from that obtained with saliva and MBP-LipB. These are the data from two independent experiments each involving triplicate sample wells. (D) Killing of the O35E wild-type strain (black columns) and the O35E ΔlipA ΔlipB mutant (gray columns) by human lysozyme in the presence of human apolactoferrin. Data analysis with a 2-way analysis of variance for multiple comparisons with a Sidak correction indicated that the extent of killing of the mutant was different (P < 0.0001) from that of the wild-type strain. These are the data from three independent experiments each involving duplicate sample wells.

FIG 7

Expression of selected genes measured by real-time RT-PCR in an M. catarrhalis O35E mesR mutant. Expression levels of lipA, lipB, MCORF 995, MCORF 1550, and MCORF 113 in the O35E mesR mutant relative to those in wild-type O35E were determined. MCORF 1452 (gyrB) was used to normalize the amount of cDNA per sample. The results depicted are representative of those from two independent experiments.