The Mycobacterium avium subsp. paratuberculosis MAP3464 gene encodes an oxidoreductase involved in invasion of bovine epithelial cells through the activation of host cell Cdc42

Abstract

Mycobacterium avium subsp. paratuberculosis infection of cattle takes place through the intestinal mucosa. To identify M. avium subsp. paratuberculosis genes associated with the invasion of bovine epithelial cells in vitro, we screened a library of transposon mutants. Several mutants of M. avium subsp. paratuberculosis were identified which invaded Madin-Darby bovine kidney (MDBK) epithelial cells less efficiently than wild-type (wt) M. avium subsp. paratuberculosis. The deltaOx mutant had the transposon located in the MAP3464 gene, a putative oxidoreductase gene whose expression is upregulated upon bacterial contact with MDBK cells. Complete restoration of invasion comparable to that for the wt bacterium was achieved by introducing a copy of the complete oxidoreductase operon into the deltaOx mutant. Immunoprecipitation and Western blot analysis indicated that wt M. avium subsp. paratuberculosis activates Cdc42 and RhoA pathways of internalization 15 and 60 min after infection of the host cell, respectively. The deltaOx mutant, however, failed to activate the Cdc42 pathway. To determine whether an M. avium subsp. paratuberculosis protein delivered to the host cell mediates the entry of the wt bacterium by activation of the Cdc42 pathway, affinity precipitation of active Cdc42 from MDBK-infected cells followed by mass spectrometry was carried out. We identified a 17-amino-acid bacterial peptide associated with the Cdc42 of cells infected with wt M. avium subsp. paratuberculosis but not with the deltaOx mutant. The sequence of the peptide matches MAP3985c, a hypothetical protein, possibly functioning as a putative Cdc42 effector. These findings reveal a novel signaling pathway activated during M. avium subsp. paratuberculosis entry that links the product of MAP3464 gene to activation of Cdc42 in the host cell.

FIG. 1.

Invasion efficiency of wt M. avium subsp. paratuberculosis and transposon mutants 4 h after infection of MDBK cells. MDBK cells were infected at a MOI of 10 and incubated at 37°C for 4 h. The number of intracellular bacteria was calculated as the fraction of the inoculated bacteria that was recovered from the cell lysate. ΔOx mutant, MAP3464; Δ41 mutant, MAP3212; Δ42 mutant, MAP3607; Δ32 mutant, MAP0941; Δ37 mutant, MAP2808. Values represent the means of three experiments ± standard deviations. *, P value of <0.05 compared with the invasion percentage for the wt bacteria.

FIG. 2.

The ΔOx mutant has the transposon inserted in the MAP3464 gene. (A) PCR from genomic DNA using specific primers for the MAP3464 and 16S RNA genes. Lanes: MW, molecular weight marker; 1, ΔOx mutant; 2 and 3, wt M. avium subsp. paratuberculosis; 4, PCR negative control. (B) Organization of the oxidoreductase operon in M. tuberculosis and M. avium subsp. paratuberculosis genomes.

FIG. 3.

Epithelial cell contact triggers upregulation of MAP3464 gene expression. (A) The expression of the MAP3464 and 16S genes in wt M. avium subsp. paratuberculosis upon contact with MDBK cells was examined by semiquantitative RT-PCR. After 0, 15, 30, and 60 min of infection, RNA was extracted from extracellular bacteria and cDNA was synthesized as described in Materials and Methods (lanes 1, 2, 3, and 4). For RT-PCR, the MAP3464 and 16S RNA genes were amplified from 1 μl of cDNA by use of specific primers. (B) Upregulation of MAP3464 gene expression (n-fold) upon 15, 30, and 60 min of infection with wt M. avium subsp. paratuberculosis as determined by real-time RT-PCR. The data represent the average of three independent experiments ± standard deviation. The levels of MAP3464 gene expression at 30 and 60 min after infection were comparable. (C) The expression of the MAP3464 and 16S genes in wt M. avium subsp. paratuberculosis and the ΔOx mutant upon 1 h of contact with MDBK cells was examined by semiquantitative RT-PCR. The expression of MAP3464 increased in wt M. avium subsp. paratuberculosis after 1 h of infection (lane 2) compared with what was seen for bacteria incubated in HBSS at 37°C for 1 h (lane 1). As shown in lanes 3 and 4, MAP3464 was not expressed in the ΔOx mutant in the tested conditions. The 16S RNA gene was used as a control because it is constitutively expressed. MW, molecular weight marker.

FIG. 4.

Invasion of MDBK cells by the wt bacterium, the ΔOx mutant, and the complemented ΔOx mutant containing a complete copy of the oxidoreductase operon (Ox Com.). Cells were infected with bacteria at a MOI of 10, and the number of intracellular bacteria was determined at 4 h after infection as described in Materials and Methods. Values are the means of three experiments ± standard errors of the mean. P values were <0.05 for the comparison between the Ox mutant and either wt or the complemented ΔOx mutant.

FIG. 5.

Effect of the oxidoreductase mutation on presentation of proteins in the cell surface. MDBK cells were infected at 37°C with wt M. avium subsp. paratuberculosis or with the ΔOx mutant. Extracellular bacteria were collected after 1 h of infection and surface-labeled using biotinamidohexanoic acid N-hydroxysuccinimide ester (20 mg/ml). Labeled bacteria were lysed, and 0.5-μg portions of the cell lysates were analyzed on a 12% SDS-PAGE gel. Gels were transferred to nitrocellulose membranes and probed with HRP-conjugated streptavidin. Lane 2 shows a biotinylated protein identified in extracellular wt M. avium subsp. paratuberculosis upon contact with MDBK cells that was not present in the ΔOx mutant (lane 4). wt M. avium subsp. paratuberculosis and the ΔOx mutant incubated in HBSS for 1 h at 37°C were used as controls (lanes 1 and 3).

FIG. 6.

Cdc42 and RhoA transduction signals are activated during M. avium subsp. paratuberculosis infection. Lysates of MDBK cells infected with wt M. avium subsp. paratuberculosis or the ΔOx mutant at different time points were immunoprecipitated (IP) with α-phosphotyrosine (αTyr), α-phosphothreonine (αThr), or α-Cdc42 antibodies. Uninfected cells (U) were used as controls. The α-phosphotyrosine or α-phosphothreonine immunoprecipitates were examined by immunoblotting with α-Cdc42 (A) and α-RhoA (B) antibodies. Similar quantities of proteins were confirmed by immunoprecipitation and immunoblotting with α-Cdc42 antibody (C). α-, anti-.

FIG. 7.

MAP3985c binding to active Cdc42. MDBK cells were infected with wt M. avium subsp. paratuberculosis or with the ΔOx mutant for 30 min at 37°C. Cell lysates were then incubated with 20 μg of the PDB of Pak1 and a SwellGel-immobilized glutathione disc (GST-Pak1-PDB) at 4°C for 1 h. To remove unbound proteins, the resin was washed three times and the samples were eluted in 50 μl of 2× SDS sample buffer without β-mercaptoethanol. The eluted pull-down samples were analyzed on a 12% SDS-PAGE gel. The arrows point to the two bands that were excised from each lane of the gel and analyzed by mass spectrometry, because of prominence and because it correspond to the Cdc42 molecular weight. Lanes: MW, molecular weight marker; 1, uninfected cell lysates treated with 0.1 mM GTPγ; 2, lysates of uninfected cells treated with GDP; 3, lysates of MDBK cells infected with wt M. avium subsp. paratuberculosis; 4, lysates of MDBK cells infected with the ΔOx mutant. After sequencing of the 45-kDa band (approximately the size of the Cdc42 protein), we detected MAP3985c peptide in lane 3 but not in lanes 1, 2, and 4.