A mycobacterium ESX-1-secreted virulence factor with unique requirements for export

Abstract

Specialized secretion systems of pathogenic bacteria commonly transport multiple effectors that act in concert to control and exploit the host cell as a replication-permissive niche. Both the Mycobacterium marinum and the Mycobacterium tuberculosis genomes contain an extended region of difference 1 (extRD1) locus that encodes one such pathway, the early secretory antigenic target 6 (ESAT-6) system 1 (ESX-1) secretion apparatus. ESX-1 is required for virulence and for secretion of the proteins ESAT-6, culture filtrate protein 10 (CFP-10), and EspA. Here, we show that both Rv3881c and its M. marinum homolog, Mh3881c, are secreted proteins, and disruption of RD1 in either organism blocks secretion. We have renamed the Rv3881c/Mh3881c gene espB for ESX-1 substrate protein B. Secretion of M. marinum EspB (EspBM) requires both the Mh3879c and Mh3871 genes within RD1, while CFP-10 secretion is not affected by disruption of Mh3879c. In contrast, disruption of Mh3866 or Mh3867 within the extRD1 locus prevents CFP-10 secretion without effect on EspBM. Mutants that fail to secrete only EspBM or only CFP-10 are less attenuated in macrophages than mutants failing to secrete both substrates. EspBM physically interacts with Mh3879c; the M. tuberculosis homolog, EspBT, physically interacts with Rv3879c; and mutants of EspBM that fail to bind Mh3879c fail to be secreted. We also found interaction between Rv3879c and Rv3871, a component of the ESX-1 machine, suggesting a mechanism for the secretion of EspB. The results establish EspB as a substrate of ESX-1 that is required for virulence and growth in macrophages and suggests that the contribution of ESX-1 to virulence may arise from the secretion of multiple independent substrates.

Figure 1. Either EspB_T or EspB_M Is Sufficient to Restore Growth to the espB_M::tn Mutant in Mouse BMDMs

BMDMs were infected at a multiplicity of infection of 1. Colony-forming units (cfu) were determined by lysing infected monolayers and plating lysates at indicated time points. Data were combined for three experiments, with growth indicated as fold increase compared to the initial level of infection for each strain. (A) Wild type (Wt) and espB_M::tn each contain the empty non-integrating plasmid pLYG206; p(espB_M) and p(espB_T) are plasmids for expression of EspB_M and EspB_T, respectively; and p(espB_M-Mh3880c) and p(espB_T-Rv3880c) encode the second gene of the operon as well as espB for each species. (B) Wild type (Wt) and espB_M::tn each were transformed with the empty integrating plasmid and the indicated genes were integrated into the attB of the espB_M::tn mutant. espB_M::tn + p(Mh3883c-Mh3880c) expresses the locus Mh3883-Mh3880c on a non-integrating plasmid. Strains differed significantly by one-way ANOVA in (A) after 24 h, 48 h, 72 h, and 96 h and in (B) after 16 h, 43 h, 75 h, and 92 h (p < 0.001 for each).

Figure 2. Requirements for EspB Secretion in M. marinum and M. tuberculosis

Cell lysates (CL) and culture filtrates (CF) were prepared from the indicated strains as described in Materials and Methods. Proteins were separated by SDS-PAGE, and the indicated proteins were detected by western blot as described in Materials and Methods. (A) 60 μg of total CL and 30 μg of total CF of M. tuberculosis Erdman and each of the M. marinum strains were loaded in each well. (B) 60 μg of total CL and 30 μg of total CF were loaded in each well. (C) Cultures of each strain were grown in 7H9 to an OD of 0.5 and then inoculated into Sauton's medium at an OD of 0.5 and grown for 36 h. Therefore, the samples of each strain are normalized by OD readings. Of the total CL and CF fractions collected for each strain from one experiment, 3% of the CL was loaded in each lane and 15% of the total CF was loaded into each lane. The results shown are representative of the results obtained in four replications of this experiment. (D) 30 μg of total CL and 30 μg of total CF of each M. tuberculosis Erdman strain were loaded in each well.

Figure 3. Growth of M. marinum Secretion Mutants in BMDMs

BMDMs were infected with M. marinum strains as described in Materials and Methods at a multiplicity of infection of 1, and growth of bacteria was monitored over time as in Figure 1. Data are summarized from three independent experiments. Strains differed significantly by one-way ANOVA after 24 h, 48 h, 72 h, 96 h, and 120 h (p < 0.001 for each).

Figure 4. Bacterial Two-Hybrid Analysis of Interaction of EspB_T with Proteins of extRD1

The target plasmid containing EspB_T fused to the RNA polymerase alpha subunit was co-transformed with each of the bait plasmids containing the indicated extRD1 proteins fused to the lambda repressor into the reporter validation strain. Shown is the ratio of growth of the co-transformants on selective (+5 mM 3AT) versus non-selective plates. The experiment depicted is representative of three independent determinations.

Figure 5. GST Pulldown Analysis of EspB and Rv3879 Interactions

(A) Agarose beads with immobilized GST, GST-SHP1, GST-syntaxin2, GST-Mh3879c, and GST-EspB_M were incubated with lysates of E. coli that express V5-Mh3879c and V5-EspB_M. (B) Agarose beads with immobilized GST, GST-SHP1, GST-syntaxin2, GST-Rv3879c, or GST-EspB_T were incubated with lysates of E. coli that express V5-Rv3871, V5-Rv3879c, and V5-EspB_T. Proteins from cell lysates retained on the beads after washing were separated by SDS-PAGE and detected by western blotting with an antibody against V5. To the right of each set of pulldowns, 0.1% of the input E. coli lysate was analyzed. EspB_M physically interacts with Mh3879c, EspB_T physically interacts with Rv3879c, and Rv3879c also interacts with Rv3871.

Figure 6. Deletion Analysis of EspB_M Secretion and Interaction with Mh3879c

(A) The M. marinum espB_M::tn mutant was transformed with a non-integrating plasmid expressing N-terminally V5-tagged EspB_M full length, EspB_M Δ(2–31), EspB_M Δ(264–271), or EspB_M Δ(400–454). Cell lysates (CL) and culture filtrates (CF) were prepared from the indicated strains as described in Materials and Methods. The samples of each strain are normalized by OD readings. Of the total CL and CF fractions collected for each strain from one experiment, 3% of the CL was loaded in each lane and 15% of the total CF was loaded into each lane, separated by SDS-PAGE and detected by western blotting with an antibody against V5. (B) Agarose beads with immobilized GST-Mh3879c were incubated with E. coli lysates expressing V5-tagged EspB_M full length, EspB_M Δ(2–31), EspB_M Δ(264–271), or EspB_M Δ(400–454). The input lysate (0.1%) and the material from the cell lysates that bound to the beads was run on SDS-PAGE and detected by western blotting with an antibody against V5.

Figure 7. Model for EspB Secretion

Depicted are the core ESX-1 components Rv3870, Rv3871, and Rv3877, as well as the ESAT-6/CFP-10 and EspB/Rv3879c complexes. Both cytosolic complexes require interaction with Rv3871 for substrate secretion.