OUP user menu

Mycobacterium tuberculosis mammalian cell entry operon (mce) homologs in Mycobacterium other than tuberculosis (MOTT)

Yoseph Haile, Dominique A. Caugant, Gunnar Bjune, Harald G. Wiker
DOI: http://dx.doi.org/10.1111/j.1574-695X.2002.tb00581.x 125-132 First published online: 1 June 2002


The cloned mammalian cell entry gene mce1a from Mycobacterium tuberculosis confers to non-pathogenic Escherichia coli the ability to invade and survive inside macrophages and HeLa cells. The aim of this work was to search for and characterize homologs of the four M. tuberculosis mammalian cell entry operons (mce1, mce2, mce3 and mce4) in mycobacteria other than tuberculosis (MOTT). The dot-blot and polymerase chain reaction (PCR) experiments performed on 24 clinical isolates representing 20 different mycobacterial species indicated that the mce operons were widely distributed throughout the genus Mycobacterium. BLAST search results showed the presence of mce1, mce2 and mce4 homologs in Mycobacterium bovis, Mycobacterium avium and Mycobacterium smegmatis. A homologous region for the mce3 operon was also found in M. avium and M. smegmatis. DNA and protein alignments were done to compare the M. tuberculosis mce operons and the deduced M. bovis, M. avium, and M. smegmatis homologs. The deduced proteins of M. bovis mce1, mce2 and mce4 operons had 99.6–100% homology with the respective M. tuberculosis mce proteins (MTmce). The similarity between M. avium mce proteins and the individual M. tuberculosis homologs ranged from 56.2 to 85.5%. The alignment results between M. smegmatis mce proteins and the respective MTmce proteins ranged from 58.5% to 68.5%. Primer sets were designed from the M. tuberculosis mce4a gene for amplification of 379-bp fragments. Amplification was successful in 14 strains representing 11 different mycobacterial species. The PCR fragments were sequenced from 10 strains representing eight species. Alignment of the sequenced PCR products showed that mce4a homologs are highly conserved in the genus Mycobacterium. In conclusions, the four mce operons in different mycobacterial species are generally organized in the same manner. The phylogenetic tree comparing the different mce operons showed that the mce1 operon was closely related to the mce2 operon and mce3 diverged from the other operons. The wide distribution of the mce operons in pathogenic and non-pathogenic mycobacteria implicates that the presence of these putative virulence genes is not an indicator for the pathogenicity of the bacilli. Instead, the pathogenicity of these factors might be determined by their expression.

  • Mammalian cell entry
  • Virulence
  • Mycobacterium tuberculosis
  • Mycobacterium bovis
  • Mycobacterium avium
  • Mycobacterium smegmatis
  • Mycobacterium

1 Introduction

The genus Mycobacterium constitutes a large group of facultative and obligate pathogens, of which many members, in addition to Mycobacterium tuberculosis, cause disease in humans. The decline in the incidence of tuberculosis in the industrialized world and the development of sophisticated microbiological methods have increased the understanding of the clinical importance of atypical mycobacteria. Humans often become naturally sensitized to one or more of these mycobacterial species [1]. Colonization and disease due to atypical mycobacteria may be difficult to differentiate, but epidemiological studies have shown increasing prevalence of pulmonary tuberculosis due to Mycobacterium bovis, Mycobacterium kansasii, Mycobacterium intracellulare and Mycobacterium avium, particularly in immunocompromised individuals [24]. Even though their virulence is relatively low, Mycobacterium scrofulaceum, Mycobacterium xenopi, Mycobacterium simiae, Mycobacterium szulgai and several other mycobacterial species are occasionally isolated from clinical specimens [5].

Since mycobacteria are facultative pathogens and assumed to selectively express specific genes inside the host macrophage, much attention has been directed towards the characterization and identification of virulence genes that are important for the entry and persistence in the host. Previously, expression of a cloned DNA fragment from M. tuberculosis in non-pathogenic Escherichia coli strain was shown to confer the ability to invade and survive within macrophages and human HeLa cells, and it was named the mammalian cell entry (mce) gene [6].

The analysis of the complete sequence of the M. tuberculosis H37Rv genome revealed the presence of four homologous mammalian cell entry operons. The genes within each of the four operons (mce1, mce2, mce3 and mce4) are organized in the same way. Each operon consists of eight genes encoding two integral membrane proteins at the 5′-end, followed by six genes encoding potentially exported proteins thought to be important for the entry and survival of the pathogen in mammalian cells [7].

Functional and structural analysis of these genes have shown the presence of a mce1 homolog in Mycobacterium leprae, and an extensive sequence homology between mce1a and mce2a genes of M. tuberculosis[8]. In addition, the expression of the six mce1 genes in vitro was shown by anti-peptide antibodies and reverse transcriptase-polymerase chain reaction (RT-PCR) [9]. PCR and sequence analysis showed the presence of a mce1 homolog in members of the M. tuberculosis complex, and of the M. avium complex including M. avium, M. scrofulaceum and M. intracellulare[10]. Furthermore, the deletion of most of the mce3 operon in M. bovis has been documented [11].

In this study, we searched for mce1, mce2, mce3 and mce4 homologs in the M. avium, M. smegmatis and M. bovis genome sequences and used PCR and dot-blot techniques to assess the occurrence of mce operons in different mycobacterial species.

2 Materials and methods

2.1 Bacterial strains

A total of 24 clinical isolates of mycobacteria from the National Reference Laboratory for Mycobacteria, Norwegian Institute of Public Health, Oslo, Norway, was analyzed (Table 1). Species identification was performed using probes specific for M. tuberculosis, M. avium and M. gordonae (AccuProbe, GenProbe Inc., San Diego, CA, USA). Strains negative by this method were identified by sequencing of a fragment of the 16S rRNA gene amplified by PCR [12].

View this table:
Table 1

Species of mycobacterial strains tested in this study and their origin

Strain No.SpeciesSourceSite of isolation
229/99M. aviumhumanexpectorat
171/98M. bovishumanbronchial alveolar lavage
144/99M. celatumhumanexpectorat
400/97M. chelonaehumanunknown
156/97M. confluentisunknownzoological feces
436/99M. fortuitumhumanexpectorat
446/99M. fortuitumhumanexpectorat
60/98M. gastrihumanexpectorat
169/98M. gastrihumanabscess from peritoneum
1394/00M. gordonaehumanexpectorat
66/99M. interjectumhumanexpectorat
343/98M. interjectumhumanexpectorat
238/98M. mucogenicumhumanpleural fluid
155/97M. nonchromogenicumcattle
250/99M. obuensehumanunknown
186/98M. scrofulaceumhumanexpectorat
300/99M. shimoideihumanunknown
183/98M. simiaehumanexpectorat
6841/99M. smegmatislaboratory
103/95M. szulgaiswimming pool water
320/96M. szulgaihumanlung
338/95M. terraehumanexpectorat
286/99M. tuberculosishumanurine
229/98M. xenopihumanexpectorat

2.2 DNA isolation and dot-blot technique

Chromosomal DNA of the isolates was prepared as described previously [13]. 5 µl with 50 ng of chromosomal DNA was dotted on a nylon membrane and air-dried. After cross-linking using UV light, prehybridization with 80 µl hybridization solution (5×SSC, 10 mg ml−1 blocking reagent (supplied), 0.1% sarcosyl and 0.02% SDS) was performed at 60°C for 2 h. Then, 25 µl digoxigenin-labeled probe was added in a fresh hybridization solution, denatured by boiling at 100°C for 10 min, and the membrane hybridized overnight at 60°C. After washing with 2×SSC, 0.1% SDS twice at room temperature for 5 min, with 0.1×SSC, 0.1% SDS at 68°C for 20 min with constant agitation, the membrane was incubated further for 15 min at the same temperature. After incubation with the supplied blocking reagent at a 1:200 dilution in buffer one (0.15 M NaCl, 0.1 M Tris base, pH 7.5) for 30 min, and washing with buffer one for 1 min, the antibody conjugated with alkaline phosphatase was added, and the blot was incubated for 30 min at room temperature. After washing with 80 ml buffer two (0.1 M NaCl, 0.05 M MgCl2, and 0.1 M Tris base, pH 9.5) for 2 min, 400 µl BCIP/NBT solution mixed with 20 ml buffer two was added, and the membrane was incubated at room temperature in the dark for 1 day without shaking. The reaction was stopped by adding 300 ml buffer three (0.6 mM EDTA, 6 mM Tris base, pH 8.0). Probe labeling followed the protocol for digoxigenin PCR-oligolabeling and detection systems (Roche Molecular Biosystems, GmbH, Germany). The probes were generated by purification of PCR products using the primer pairs described in Table 2. They were designed from the M. tuberculosis sequence.

View this table:
Table 2

Primers used for PCR and generation of probes

1mce1aforward 225–2455′-CAC CAT CTC GGA GGT CAC ACG-3′
backward 520–5015′-CGC TCA GGG TCA GGT TCA GC-3′
2mce2aforward 117–1365′-ATT CAC GCC GAA GAC CGA GC-3′
backward 430–4115′-CCG ACC CCA CAT CAA TCA CG-3′
3mce3aforward 529–5485′-GGG CAA GGT GAA GGC ATA GG-3′
backward 883–8645′-GGA AGC AGG TGT ATT CGG GG-3′
4mce4aforward 190–2105′-GGC ATC CAG GTA GGC AAG GTC-3′
backward 569–5505′-CGG GTC AGC GTG TTC AGT CC-3′

2.3 PCR

PCR was performed in a final volume of 25 µl containing 2 ng of chromosomal DNA as template, 2 µM of each primer (Table 2), dNTPs (0.2 mM each) (Roche Molecular Biosystems, Foster City, CA, USA) and 1.25 U Taq DNA polymerase (Roche Molecular Biosystems, Branchburg, NJ, USA). The PCR was carried out for an initial 94°C for 4 min, followed by 30 cycles (94°C for 1 min, 60°C for 1 min, and 72°C for 30 s) and, finally the extension time continued for 7 min at 72°C. A 5-µl portion of the PCR products was subjected to electrophoresis in 1% agarose gel in 1×TBE buffer for 90 min. The gel was stained with ethidium bromide for 10 min. The presence of amplified product was visualized under UV light.

2.4 Sequencing the amplified products

The PCR products were purified for sequencing with exonuclease I and shrimp alkaline phosphatase according to the manufacturer's instruction (Amersham Life Science, Cleveland, OH, USA). Sequencing was undertaken using the ABI Prism Big Dye terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems, Oslo, Norway). 3 µl 5×TBE (Tris-borate-EDTA) buffer, 2 µl Big-dye terminator mix, 3 µl PCR product and 1.6 µM of sequencing primers Mce4aFP and Mce4aBP in separate tubes were mixed together and the final volume adjusted to 20 µl by adding deionized H2O. The sequencing reaction was run in the PCR machine using the following program: 5 min at 94°C, followed by 10 s at 96°C, 5 s at 50°C and 4 min at 60°C for 25 cycles. Then the contents of the tubes were spun down by centrifugation. The sequenced products were washed and denatured according to the manufacturer's protocol (Applied Biosystems). The samples were loaded in 0.5% polyacrylamide long ranger gel (BMA Rockland, ME, USA) and the electrophoresis run for 7 h on an ABI Prism 377 sequencer (Applied Biosystems). DNA sequences were determined on both DNA strands. Sequences were assembled using DNA Sequence Assembly Software version 2.0 (Applied Biosystems).

2.5 Sequences analysis

Known DNA sequences from databases were obtained from the Sanger Center for genomic sequences; http://www.sanger.ac.uk/l and Institute for genomic research; http://www.tigr.org. Assessment of mce operons in different mycobacterial species was performed using the BLASTN search tool [14]. The DNA-Star Megalign program was used for comparing DNA and protein sequences. Multiple alignments were done using the Clustal method and pairwise alignments were done by the Martinez/Needleman–Wunsch method [15,16] using a minimum match of 9; gap penalty 1.10; and gap-length penalty 0.33 as parameters. Protein sequences were compared by the Lipman and Pearson method [17] using the following parameters: K-tuple 2; gap penalty 4; and gap-length penalty 12.

3 Results

The presence of M. tuberculosis mce1, mce2, mce3 and mce4 operons in 24 mycobacterial isolates (Table 1) were assessed by dot blots. Four digoxigenin-labeled probes from M. tuberculosis mce1a, mce2a, mce3a and mce4a genes were prepared by PCR and hybridized with the different mycobacterial isolates. All the tested isolates gave positive signals by dot blots (data not shown). The intensities of the dots varied from species to species. In addition, hybridizations of the mce1a probe with mce2a PCR product and vice versa gave strong positive signals. Since the different mce operons share common DNA sequences, cross-hybridization among the different mce operons makes it difficult to determine the representation of the individual operons by the dot-blot technique.

To obtain more specific information about individual mce operons in different mycobacterial species, DNA sequence homology searches were done. The search results revealed the presence of M. tuberculosis mce1, mce2 and mce4 homologs in M. bovis, M. avium and M. smegmatis (Table 3). Homologues for mce3 were also found in M. avium and M. smegmatis (Table 3). DNA sequence alignments of M. bovis mce operons (MBmce) with the M. tuberculosis mce operons (MTmce) showed that MBmce1, MBmce2 and MBmce4 homologs have the same structural organization. In addition, the sizes of the homologous proteins were similar. The predicted MAmce1 genes were always shorter than the MTmce1 genes except for the M. avium mce1b (MAmce1b) gene (not shown).

View this table:
Table 3

Alignments of deduced M. bovis (MBmce), M. avium (MAmce) and M. smegmatis (MSmce) proteins with mce homologs from M. tuberculosis, MTmce1–4

OperonSimilarity index (%)
MBmce1 vs. MTmce199.6–100
MBmce2 vs. MTmce298.6–100
MBmce4 vs. MTmce4100
MAmce1 vs. MTmce169.1–80.5
MAmce2 vs. MTmce276.4–82.4
MAmce3 vs. MTmce356.2–66.2
MAmce4 vs. MTmce480.5–85.5
MSmce1 vs. MTmce158.9–67.6
MSmce2a vs. MTmce2a62.8
MSmce3 vs. MTmce358.5–67.7
MSmce4 vs. MTmce465.4–68.5
  • The numbers represent the maximum and minimum percent similarity using the Lipman and Pearson method [17] for pairwise protein alignments.

Homologies between MTmce1 and MAmce1 proteins, ranged from 69.1% for MAmce1a to 80.5% for MAmce1d (Table 3). Comparison of MT/MBmce1a with MAmce1a, MLmce1a and MSmce1a proteins revealed deletions in the latter species corresponding to amino acids 353–363 in MT/MBmce1a with variable lengths in the flanking regions (Fig. 1). The length of the deleted segments ranged from 12 on MLmce1a up to 48 on MSmce1a. Structural analysis of the MTmce1a protein by the DNA-Star program Protean showed that the deleted segments were hydrophobic. Prediction of MAmce2 proteins (MAmce2a–f) by EditSeq DNA-Star program from the M. avium genome showed the presence of all six proteins with a comparable size, except for MAmce2b which had 66 extra amino acids at the C-terminus. To investigate the presence of these extra nucleotides in the M. tuberculosis genome, the MAmce2b gene was aligned with MTmce2b. Non-coding sequences corresponding to the MAmce2b gene were found down stream of MTmce2b gene (not shown). This result showed that nonsense mutation at nucleotide position 828 on the MTmce2b gene might be responsible for its termination. Alignments of MAmce2a–f with MTmce2a–f showed similarities from 76.4 up to 82.4% (Table 3). Assessment of the MSmce2 operon showed that all genes except the MSmce2a gene were deleted (Table 3).

Figure 1

Alignment of the mycobacterial Mce1a proteins. Partial protein sequence starting at position 106 of M. tuberculosis MTmce1a was aligned with the deduced sequences from M. bovis (MBmce1a), M. avium (MAmce1a), M. smegmatis (MSmce1a), and M. leprae (MLmce1a) using the Clustal method. Identical amino acids are indicated by a hyphen. Deleted regions are indicated by dots.

Homologues of the mce3 operon were found in M. avium and M. smegmatis (Table 3). Both the MAmce3 and MSmce3 operons consisted of six genes with the same organization as MTmce3. The MTmce3 proteins were aligned with the corresponding MAmce3 and MSmce3 proteins using the Lipman and Pearson method. Comparative alignments between the different MTmce operons and the corresponding MAmce and MSmce operons demonstrated that the mce3 operon in M. avium and M. smegmatis were distantly related to the MTmce3 operon.

The DNA and protein sequences of the MTmce4 genes with the predicted MAmce4 showed from 80.5–85.5% homology (Table 3). MAmce4a–f genes had comparable sizes with MTmce4a–f homologs. The MSmce4 proteins had almost the same size as the MTmce4 proteins. Pairwise alignments showed that the MSmce4 proteins are quite diverged from the MTmce4 and MAmce4 proteins (Table 3).

To exploit the conservation of mce4 operons in different mycobacterial species, we used the primer set designed for a portion of the MTmce4a gene to determine the presence of genes in other species. Among the 24 mycobacterial strains (Table 1), this primer set successfully amplified a product of the approximate predicted size (379 bp) in 14 strains, representing 11 mycobacteria other than tuberculosis (MOTT) species: M. gastri, M. interjectum, M. szulgai, M. celatum, M. bovis, M. simiae, M. scrofulaceum, M. xenopi, M. avium, M. terrae, and M. obuense, and one M. tuberculosis strain (Fig. 2). Despite the presence of 65% homology between MSmce4a and MTmce4a proteins, no PCR amplification of the MSmce4a gene was obtained for the M. smegmatis strain because of 14/20 mismatches in the sequence of the reverse primer. In several strains, such as M. gordonae, amplification generated bands at a different position or there were multiple bands (for example M. shimoidei). Thus, the strains were excluded from further analysis.

Figure 2

The primer set designed for the amplification of MTmce4a 379-bp PCR products was used to assess the presence of mce4a gene from different mycobacterial species. Lane 1, molecular mass marker; lane 2, M. gastri; lane 3, M. interjectum; lane 4, M. szulgai; lane 5, M. celatum; lane 6, M. noncromogenicum; lane 7, M. confluentis; lane 8, M. gastri; lane 9, M. bovis; lane 10, M. simiae; lane 11, M. scrofulaceum; lane 12, M. xenopi; lane 13, M. avium; lane 14, M. mucogenicum; lane 15, M. obuense; lane 16, M. shimoidei; lane 17, M. szulgai; lane 18, M. terrae; lane 19, M. interjectum; lane 20, M. chelonae; lane 21, M. fortuitum; lane 22, M. fortuitum; lane 23, M. gordonae; lane 24, M. smegmatis; lane 25, M. tuberculosis, and lane 26, negative control.

The amplified products were successfully sequenced in 10 MOTT strains representing eight different species, and the deduced amino acid sequences (Fig. 3) exhibited more than 77% similarity, indicating that the mce4 operon is highly conserved. To clarify the relationship between the sequenced mce4a PCR products from the MOTT and the MTmce1a, MTmce2a, MTmce3a, MTmce4a, MAmce1a, MAmce2a, MAmce3a, MAmce4a, MSmce1a, MSmce2a, MSmce3a and MSmce4a in the databases, the similarities between the translated sequences were represented by a dendrogram. The tree presented three major branches. The first group consisted of all the mce4a partial sequences. The second group consisted of mce1 and mce2 sequences, and the third group of mce3 sequences (Fig. 4). The phylogenetic tree, based on the mce4a partial sequences, was not in agreement with the Runyon classification or clinical importance of the mycobacteria (Fig. 4). In addition, alignment results between the two M. szulgai and M. interjectum were identical. The mce4a fragments of M. scrofulaceum and M. obuense were also identical, suggesting that these two species might be closely related. The Mce1a, Mce2a, Mce3a and Mce4a proteins in M. tuberculosis, M. avium and M. smegmatis were also aligned. The results indicated that the Mce1a protein was always closely related to Mce2a, and the Mce3a and Mce4a proteins diverged from the Mce1a and Mce2a proteins. The MSmce1a and MSmce2a proteins were both more closely related to the MTmce2a protein than to the MTmce1a protein (Table 4).

Figure 3

Amino acid sequences deduced from the mce4a gene segments of 10 strains representing eight mycobacterial species aligned with M. tuberculosis mce1a partial protein sequence. Amino acids are noted only where different from those in M. tuberculosis; identical amino acids are shown by a hyphen.

Figure 4

Relationships among the Mce proteins of different species of mycobacteria. Deduced sequences of mce4a PCR products from two strains of M. szulgai, M. interjectum, and one strain of M. avium, M. scrofulaceum, M. gastri, M. obuense, M. celatum and M. xenopi were aligned with the corresponding partial sequences of MTmce1a, MTmce2a, MTmce3a, MTmce4a, MBmce1a, MBmce2a, MBmce4a, MAmce1a, MAmce2a, MAmce3a, MAmce4a, MLmce1a, MSmce1a, MSmce3a and MSmce4a proteins. The partial sequences, delineated as MT, MA, MB, ML and MS were obtained from the Sanger Center for genomic sequences; http://www.sanger.ac.uk/l and Institute for genomic research; http://www.tigr.org. Divergence in amino acid sequences among various proteins is indicated below.

View this table:
Table 4

Comparison of the mce1a–4a proteins from M. tuberculosis, M. bovis, M. avium and M. smegmatis

  • The numbers represent the percentage of identity of each of the MceA proteins from the different mycobacterial species with the other Mce proteins using Clustal multiple-protein alignment. Values larger than 50% identity are in bold.

4 Discussion

The aim of this work was to provide a detailed description of homologs of the four MTmce operons in MOTT using a combination of dot-blot, bioinformatic and PCR-based sequence analysis. This information gives us a new insight into the distribution of mce genes and the phylogenetic relationship among members of the genus Mycobacterium.

Previously, assessments of mce1a in MOTT by PCR have shown that it was restricted to the M. tuberculosis complex, the M. avium complex and M. scrofulaceum[10]. However, slight mismatch between the primer sets and the target sequences may give false negative results, limiting the use of PCR. To get more information concerning the distribution of mce operons we initially decided to assess this by dot blot. Using this approach, we were able to demonstrate the presence of representatives of the mce operons in all tested mycobacterial strains. The variation in the intensity of the dots from species to species may either reflect the mismatch between the mce probes and the target sequences in MOTT or may be due to the presence of cross-hybridization among the different mce operons. It has been demonstrated earlier that the different mce genes share common DNA sequences and might have originated by gene duplication [18].

The availability of the whole genome sequence of different bacterial species facilitates the identification of novel drug targets and diagnostic probes. Using the BLAST search tool, the mce1 operon (MLmce1) was identified in M. leprae[8]. The recently described reductive evolution of the M. leprae genome is responsible for the deletion or inactivation of many important genes within the MLmce2, MLmce3 and MLmce4 operons [19]. Assessment of these putative virulence genes in M. avium, M. bovis and M. smegmatis showed the presence of mce1, mce2 and mce4 operons in all tested species. However, the MSmce1a and MSmce2a proteins were both more closely related to the MTmce2a protein than to the MTmce1a protein (Table 4). Because the mce operons are assumed to have emerged from the same ancestral gene, one of the possible explanations is that the MSmce1 and MSmce2 operons duplicated after the divergence of slow-growing mycobacteria from fast-growing mycobacteria.

It has been shown that inactivation of mce1a in M. bovis BCG (Bacille Calmette-Guérin) reduced the invasion of the bacilli into macrophages [20]. The presence of surface protein on the bacilli may be relevant for the fusion and entrance of mycobacteria into host cells. In this regard, alignment of the mce1a proteins from different mycobacterial species indicated the presence of hydrophobic amino acid segments from position 327 to 377 unique to MTmce1a and MBmce1a (Fig. 1). It is not known whether the entrance and survival of M. tuberculosis and M. bovis in mammalian cells might be associated with this hydrophobic amino acid segments being deleted in MAmce1a, MLmce1a and MSmce1a (Fig. 1). This possibility needs further consideration and investigation. In addition, a recent study [21] indicated that another amino acid segment, located from position 106 to 163, was needed for the uptake of M. tuberculosis. Homologues for the MTmce3 operon were detected in M. avium and M. smegmatis, but only a small fragment was present in M. bovis as described previously [11] showing that the mce3 operon has been deleted from M. bovis. Currently, the PPD (purified protein derivative) skin test is not specific to distinguish BCG vaccination from tuberculosis and leprosy infection. It might be possible to design M. tuberculosis-specific peptides from the MTmce3 operon for PPD skin testing. PCR testing among M. tuberculosis, M. avium, M. bovis, and M. smegmatis using primers designed from the MTmce3a gene gave a positive signal for M. tuberculosis only. This should be further investigated to assess the clinical application of MTmce3a-based PCR testing in the areas with high prevalence of M. tuberculosis, M. bovis and M. avium infections.

To assess the relationships between mce genes from different mycobacterial species, we designed primer sets targeting the mce4a gene. Sequence similarity among MTmce4a, MBmce4a, MAmce4a and MSmce4a was above 65%. Thus, we expected that the mce4a gene would be suitable as a target sequence to determine the presence of mce genes in MOTT and to study the mce sequence variation within the genus Mycobacterium. Out of the 23 MOTT strains, 14 gave products of the expected size. M. chelonae, M. confluentis, M. fortuitum, M. gordonae, M. mucogenicum, M. nonchromogenicum and M. shimoidei produced multiple bands or PCR products different from the target fragment, while no product was obtained for M. smegmatis. The presence of multiple bands may be due to primers binding to multiple sites along the mce4a gene. To confirm the amplification of the right targets on MOTT, the PCR products were sequenced and aligned with the portion of published MTmce4a gene using Clustal method. The sequenced products of MTmce4a, MBmce4a and MAmce4a matched completely to the sequences in the databases. The alignment results of the different mycobacterial mce4a sequences ranged from 77 to 100%. This indicated that mce4a proteins are highly conserved in the genus Mycobacterium. The phylogenetic tree based on these data was in good agreement with the dendrogram generated previously [18]. Mce1a and mce2a genes were always closely related compared with mce3a and mce4a and this was indicated by the fact that each branch consisted of homologous operons from the different species, regardless of their pathogenicity. This finding demonstrated that the different mce operons duplicated before the divergence of these species.

Overall, our results indicate a high degree of structural homology among mycobacterial mce4a genes and that the four mce operons are present in both pathogenic and non-pathogenic mycobacteria. In contrast to a previous thought [10], these findings show that sequences from all four mce operons are present in fast-growing, clinically less-important slow-growing, and clinically important slow-growing mycobacteria. Our findings therefore indicate, that the presence of representatives of the mce operons is not an indicator of the pathogenicity of the organism. However, it is possible that the level of expression of these putative virulence genes is related to the pathogenicity of mycobacterial species.


This study was supported by the Center for International Health, University of Bergen, and Broegelmann Research Laboratory, University of Bergen. We thank Anne Klem and Torill Alvestad for assistance with DNA isolation and sequencing the PCR products.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
View Abstract