SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Porins form channels in the mycolic acid layer of mycobacteria and thereby control access of hydrophilic molecules to the cell. We purified a 100 kDa protein from Mycobacterium smegmatis and demonstrated its channel-forming activity by reconstitution in planar lipid bilayers. The mspA gene encodes a mature protein of 184 amino acids and an N-terminal signal sequence. MALDI mass spectrometry of the purified porin revealed a mass of 19 406 Da, in agreement with the predicted mass of mature MspA. Dissociation of the porin by boiling in 80% dimethyl sulphoxide yielded the MspA monomer, which did not form channels any more. Escherichia coli cells expressing the mspA gene produced the MspA monomer and a 100 kDa protein, which had the same channel-forming activity as whole-cell extracts of M. smegmatis with organic solvents. These proteins were specifically detected by a polyclonal antiserum that was raised to purified MspA of M. smegmatis. These results demonstrate that the mspA gene encodes a protein of M. smegmatis, which assembles to an extremely stable oligomer with high channel-forming activity. Database searches did not reveal significant similarities to any other known protein. Southern blots showed that the chromosomes of fast-growing mycobacterial species contain homologous sequences to mspA, whereas no hybridization could be detected with DNA from slow growing mycobacteria. These results suggest that MspA is the prototype of a new class of channel-forming proteins.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The mycobacterial cell envelope is extremely hydrophobic and forms an exceptionally strong permeability barrier rendering mycobacteria naturally resistant to a wide variety of antimicrobial agents. This is because of the large amounts of long fatty acids (C60 to C90), mycolic acids, in the mycobacterial cell wall and its unique structure (Brennan and Nikaido, 1995). There is convincing evidence in support of the model that most of the mycobacterial cell wall lipids are organized in a bilayer forming an outer membrane of exceptional thickness. (i) Electron microscopy pictures show a freeze-fracture plane within the cell wall. The transparent layer of the cell wall is twice as thick as that in the plasma membrane, as predicted for an asymmetric bilayer with one leaflet composed of mycolic acids (Daffé and Draper, 1998). (ii) X-ray diffraction studies of purified mycobacterial cell wall showed that the hydrocarbon chains of the cell wall lipids are arranged in a highly ordered structure and predominantly in a direction perpendicular to the cell wall surface (Nikaido et al., 1993). (iii) Investigation of mycobacterial cell wall lipids by differential scanning calorimetry and electron spin resonance techniques suggested that an asymmetric bilayer structure of extremely low fluidity exists in mycobacteria (Liu et al., 1995; 1996).

It was not clear how hydrophilic molecules could pass through this hydrophobic mycolic acid layer until channel-forming proteins functionally similar to the well-known porins of Gram-negative bacteria were found in Mycobacterium chelonae (Trias et al., 1992) and Mycobacterium smegmatis (Trias and Benz, 1994). The number and the channel properties of porins determine the permeability of the cell for hydrophilic molecules, which is 100- to 1000-fold lower for mycobacteria than for E. coli (Jarlier and Nikaido, 1990). The availability of structural information about mycobacterial porins would be important for our understanding of how the accessibility of hydrophilic molecules into the mycobacterial cell is controlled. Despite this fact, there is only one sequence of a mycobacterial protein known (OmpATb) that has channel-forming activity in vitro (Senaratne et al., 1998).

We have purified the MspA porin from M. smegmatis (Mycobacterium smegmatisporin A) and cloned its gene mspA. Biochemical and sequence data show that this mycobacterial porin is a channel-forming protein with properties different from those of OmpATb and from porins of Gram-negative bacteria.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Purification of a porin from M. smegmatis

Three detergents were tested for their abilities to solubilize porins from whole cells of M. smegmatis. Extracts with the anionic detergent dodecyl sulphate (SDS), the zwitterionic N,N-dimethyldodecylamine-N-oxide (LDAO) and the non-ionic isotridecylpoly (ethyleneglycolether) (Genapol) in lipid bilayer experiments caused an increase in the membrane conductance by many orders of magnitude. The conductance of more than 30% of the single channels recorded with the Genapol extract was 4.6 nS (Fig. 1A), which is in agreement with previously published results (Trias and Benz, 1994). The single-channel recordings of cell extracts with 1% SDS revealed a conductance shift to 5.2 nS (data not shown). The pores of the LDAO extract reconstituted very slowly into the lipid bilayer and caused noisy signals, but had the same distribution of conductance steps as the Genapol extract (data not shown). These results indicated that the properties and/or the solubilization of the porin from M. smegmatis are influenced to some extent by the type of detergent used for extraction.

image

Figure 1. . Probability (P ) of conductance steps (G) observed with membranes formed of diphytanoyl phosphatidylcholine/phosphatidylserine in the presence of porin from M. smegmatis. Protein solutions were added to both sides of the membranes. Data were collected from at least three different membranes. A. Extract of M. smegmatis with 1% Genapol. A total of 129 single-channel events was analysed. The average single-channel conductance was 2.3 nS for the lefthand maximum (35 single-channel events) and 4.6 nS for the righthand maximum (63 single-channel events). B. Extract of M. smegmatis with methylenechloride/methanol. A total of 136 single-channel events was analysed. The average single-channel conductance was 2.3 nS for the lefthand maximum (90 single-channel events) and 4.6 nS for the righthand maximum (16 single-channel events). C. Porin purified from M. smegmatis. The protein concentration in the cuvette was 2 ng ml−1. A total of 99 single-channel events was analysed. The average single-channel conductance was 2.3 nS for the lefthand maximum (14 single-channel events) and 4.6 nS for the righthand maximum (71 single-channel events). D. Recombinant MspA purified as a 100 kDa protein complex after expression in E. coli. The protein concentration in the cuvette was 20 ng ml−1. A total of 163 single-channel events was analysed. The average single-channel conductance was 2.3 nS for the lefthand maximum (110 single-channel events) and 4.6 nS for the righthand maximum (34 single-channel events).

Download figure to PowerPoint

The current through reconstituted porins from crude cell extracts of M. smegmatis with Genapol was very noisy when the electrolyte solution bathing the bilayer was buffered at a pH of 8.0 or 5.4, whereas no noise was detected at pH 6.0 (data not shown). Consequently, all further experiments with planar lipid bilayers were performed at pH 6.0.

Proteins from Genapol extracts of M. smegmatis were concentrated by a two-phase precipitation method using a defined methanol–chloroform–water mixture before analysis by SDS–PAGE (Wessel and Flügge, 1984). We detected channel-forming activity not only in the protein pellet but also in the organic phase, indicating that the porin is not precipitated quantitatively from dilute solutions by this method. This is in contrast to the behaviour of most other proteins (Wessel and Flügge, 1984), but agrees with the results obtained for the porin of C. glutamicum (Lichtinger et al., 1998). Based on this observation, whole cells of M. smegmatis were extracted with dichloromethane/methanol. The extract showed a rapid reconstitution of channels in planar lipid bilayers, demonstrating that a significant amount of the porin is dissolved in the organic phase and still possesses a channel-forming structure. More than 60% of these channels had a conductance of 2.3 nS (Fig. 1B), whereas only 10% of the channels had a conductance of around 4.6 nS, which was the main peak recorded with the Genapol extract. A part of the channel-forming activity could be precipitated with ether. Proteinase K treatment eliminated the channel-forming activity, which demonstrated that the channels were formed by proteins. These proteins caused the same conductance steps in the lipid bilayer as the dichloromethane/methanol extract (data not shown). A protein with an apparent molecular mass of about 100 kDa was purified by preparative gel electrophoresis (Fig. 2, lane 3). Analysis of 100 ng of this protein in a denaturing polyacrylamide gel showed a single band after silver staining (Fig. 2, lane 4). The addition of 10 ng of protein to a planar lipid bilayer resulted in a fast reconstitution of channels (Fig. 3A). Most of these channels exhibited a conductance of 4.6 nS, indicating an equilibrium between small and large channels depending on the treatment of the protein (Fig. 1C). To test whether the purified 100 kDa protein is contaminated with other proteins, 10 μg was electrophoresed in a denaturing polyacrylamide gel. Staining with Coomassie blue resulted in a single band at 100 kDa (Fig. 4, lane 1). Although this indicates that the 100 kDa protein was purified to near homogeneity, this preparation could be contaminated by a small amount of a highly active porin that might be invisible even on silver-stained polyacrylamide gels. Therefore, twofold serial dilutions of the protein solution were made. Significant channel activity with conductance steps identical to those shown in 1Fig. 1C was observed down to a protein concentration of 80 pg ml−1. As the channel-forming activity of porins in planar lipid bilayers is usually detected in a concentration range of ng ml−1 to μg ml−1 (Benz and Bauer, 1988), this experiment indicates that the purified 100 kDa protein is probably identical to a porin of M. smegmatis.

image

Figure 2. . Isolation of a porin from M. smegmatis by extraction with methylenechloride/methanol. Proteins were separated on an 8% SDS–polyacrylamide gel, which was stained with silver (Morrissey, 1981). Lanes: M, molecular mass marker; 1, methylenechloride/methanol extract; 2, supernatant of the methylenechloride/methanol extract after ether precipitation; 3, pellet of the methylenechloride/methanol extract after ether precipitation; 4, 100 ng of gel-eluted porin from M. smegmatis ; 5–8, gel-eluted proteins of the four bands of lane 2 above 100 kDa with decreasing molecular masses. All samples were incubated at 37°C for 30 min before loading the gel. Boiling of the samples for 2 min did not change the electrophoretic mobility of any protein (data not shown).

Download figure to PowerPoint

image

Figure 3. . A. Single-channel recordings of a diphytanoyl phosphatidylcholine/phosphatidylserine membrane in the presence of 2 ng ml−1 purified porin from M. smegmatis. Mainly, conductance steps of 4.6 nS were observed. B. Record of a 2.3 nS channel of a porin from M. smegmatis switching between open and partially closed conformations.

Download figure to PowerPoint

image

Figure 4. . Denaturation of a porin from M. smegmatis. Proteins were separated on a 10% SDS–polyacrylamide gel with a buffer containing tricine as described by Schägger and Jagow (1987). The gel was stained with Coomassie blue. Lanes: M, molecular mass marker (200, 116, 97, 66, 55, 36.5, 31, 21.5, 14.4, 6 kDa); 1, 9.6 μg of purified porin; 2, 9.6 μg of purified porin after boiling in 80% DMSO and acetone precipitation. The samples were incubated at 37°C for 30 min before loading the gel.

Download figure to PowerPoint

Cloning and sequence analysis of the mspA gene

The proteins of the ether-precipitated dichloromethane/methanol extracts were separated by SDS–PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane. The protein band at 100 kDa was excised from the membrane. Edman degradation yielded an N-terminal amino acid sequence of 42 amino acids (Fig. 5A). A single oligonucleotide (MP01) was synthesized with a sequence derived from a region of the N-terminus rich in amino acids with few codons. As codon usage in mycobacteria is strongly biased, we chose guanine or cytosine at the third position of each codon. Amplification of a genomic library of M. smegmatis with this porin-specific primer MP01 in combination with the primer KS19 binding next to the polycloning site of plasmid pOLYG (Garbe et al., 1994) yielded a DNA fragment of 1300 bp. This PCR product was labelled with digoxigenin and used to identify E. coli colonies of the same genomic library of M. smegmatis. Three clones hybridized with the porin-specific probe from which the plasmids pPOR6, 7 and 10 were isolated. The plasmid pPOR6 contained an insert of about 3 kbp of chromosomal M. smegmatis DNA and was chosen for further characterization.

image

Figure 5. . A. MspA porin of M. smegmatis. The putative signal peptide (SP) is as depicted as a grey box. Amino acid sequences of the N-terminus (NT) and the C-terminus (CT) of the mature protein were obtained by sequencing of the purified protein or of 7 tryptic fragments are shown below. The accession number to EMBL Nucleotide Sequence Database for the mspA gene is AJ001442. B. Hydrophobicity profile of MspA. Hydrophobicity values for each amino acid were taken from the consensus scale of Eisenberg (Sweet and Eisenberg, 1983). Hydrophobicity indices were calculated using the algorithm of Vogel and Jähnig: Hi = (hI–4 + hI–2 + hi + hi+2 + hi+4)/5 as described previously (Rauch and Moran, 1994). Hydrophobic values are positive, and hydrophilic ones are negative. Amphiphilic β-strands are the secondary structure elements of porins of Gram-negative bacteria and possess alternating hydrophobic and hydrophilic amino acids, which can be recognized in a hydrophobicity profile.

Download figure to PowerPoint

Both strands of 700 bp of the genomic DNA insert within plasmid pPOR6 were sequenced at least twice by primer walking using an automated DNA sequencer or 33P-terminator sequencing. The nucleotide sequence contained the sequence of MP01 without any deviation and an open reading frame (ORF) of 636 bp encoding a protein of 211 amino acids (accession number to the EMBL nucleotide sequence database: AJ001442.). This ORF has three potential ATG or GTG translation start sites. Two of them are preceded by a sequence similar to E. coli ribosome binding sites. Five out of six nucleotides of these potential ribosome binding sites are complementary to 16S rRNA of M. smegmatis. We named this gene mspA (Mycobacterium smegmatisporin A). The guanine plus cytosine (G + C) content in the M. smegmatis mspA gene is 66%, which is close to the 67.3% determined from all M. smegmatis genes sequenced to date (Codon Usage Database, GenBank release 108; Nakamura et al., 1997).

The N-terminal amino acid sequence of the porin from M. smegmatis was identical to residues 28–71 of the derived amino acid sequence of MspA, indicating that the protein was processed. The putative signal peptide of 27 amino acids exhibited the characteristics of a standard signal sequence (Pugsley, 1993): a basic N-terminal region of 6 amino acids (one lysine, one arginine residue, no acidic residue), a central region of high hydrophobicity of 13 amino acids, as evidenced by a hydrophobicity profile of MspA based on Kyte–Doolittle assignments (Kyte and Doolittle, 1982) and a C-terminal domain of 8 amino acids containing a glycine residue.

Purified porin was digested with trypsin to obtain internal peptide sequences. After separation by high-performance liquid chromatography (HPLC), the N-terminal amino acid sequences of seven peptides were determined. All sequences were found in the amino acid sequence derived from the DNA sequence of mspA. Five tryptic fragments were located in the N-terminus (NT), whereas two define the C-terminal end of mature MspA (Fig. 5A). MALDI mass spectrum of purified porin revealed a mass of 19 406 Da (Fig. 6), which is in agreement with the predicted molecular mass of 19 404 Da. Both results confirm the DNA-derived amino acid sequence of MspA and the cleavage site of the signal peptide. It also demonstrates that MspA is not covalently modified after translation. An overall hydrophobicity index of 0.02, an isoelectric point of 4.04 and a net charge of −11.7 is calculated for the derived amino acid sequence of mature MspA (Lasergene program package, dnastar). Thus, the porin from M. smegmatis is an acidic protein composed of 184 amino acids with a balanced distribution of hydrophilic and hydrophobic amino acids.

image

Figure 6. . MALDI mass spectrum of purified MspA. No signals were observed with mass-to-charge ratios above 20 000.

Download figure to PowerPoint

At present, two types of membrane protein structures are known: α-helical proteins from cytoplasmic membranes and β-barrel proteins from outer membranes of Gram-negative bacteria. Sequence analysis of mature MspA at the PredictProtein server (http://www.embl-heidelberg.de/predictprotein/predictprotein.html) did not reveal any transmembrane helices (Rost et al., 1995). As the β-strands of porins line a water-filled channel, they must be amphipathic and are therefore defined by alternating hydrophobic and hydrophilic residues. Mean hydrophobicities on each side of a putative β-strand of mature MspA were calculated using a consensus scale (Sweet and Eisenberg, 1983) and the algorithm of Vogel and Jähnig as described previously (Rauch and Moran, 1994). The hydrophobicity profile is displayed in 5Fig. 5B and reveals that many amino acids of MspA are compatible with amphipathic β-strand conformation.

The amino acid sequence of mature MspA showed no significant matches to any other sequence in a search in the non-redundant database (342514 sequences, 12/98, National Center for Biotechnology Information) and in the SWISSPROT data bank (release 36, 07/98) using the Basic Local Alignment Search Tool (blastP 2.0; Altschul et al., 1997) and default parameters.

Biochemical analysis of the MspA porin

However, two questions arise from these results. First, the porin was purified as a 100 kDa protein from a denaturing gel, but exhibits a mass of 19.4 kDa in the mass spectrum. Thus, it is not clear whether the porin was denatured during sample preparation for mass spectrometry and/or laser desorption or whether it has an unusually low apparent electrophoretic mobility in denaturing gels. Secondly, the mass spectrum shows several peaks with a mass-to-charge ratio below 10 000. These peaks could arise from fragmented MspA or could result from other small proteins that might be strongly associated with MspA.

To analyse whether it is possible to release a 20 kDa protein from the purified porin of M. smegmatis, several denaturing agents were tested. While the 100 kDa protein changed neither its electrophoretic mobility nor its channel-forming activity after boiling in 3% SDS with or without β-mercaptoethanol for 10 min or treatment with 8 M urea or 6 M guanidium hydrochloride, it could be converted into a single protein of 20 kDa after boiling in 80% dimethyl sulphoxide (DMSO) for 15 min (Fig. 4, lane 2). N-terminal sequence analysis after blotting of this 20 kDa protein on a PVDF membrane revealed 17 amino acids that were identical to that of mature MspA. These results show that the porin from M. smegmatis is an extremely stable protein composed of several MspA subunits. No channel-forming activity of the MspA monomer could be detected with lipid bilayer measurements, indicating that only the oligomeric structure of MspA is capable of reconstituting as a channel into lipid membranes.

The volume of a dichloromethane/methanol extract of whole cells of M. smegmatis (Fig. 2, lane 1) was reduced 18-fold by evaporation. Analysis in a denaturing polyacrylamide gel showed several bands with molecular masses above 100 kDa (Fig. 2, lane 2). Four of these bands were excised from the gel and extracted with Genapol. All extracts caused a rapid conductance increase in a lipid bilayer with mainly 4.6 nS channels, as was the case with the protein purified from the ether pellet. In addition, SDS–PAGE of these extracts revealed a protein band of the same molecular mass as the purified porin (Fig. 2, lanes 4–7). It is concluded that the four excised bands represented different oligomers of the porin that are stable in a denaturing polyacrylamide gel.

Expression of the mspA gene in E. coli

To analyse whether MspA alone is able to form channels in lipid membranes, a codon usage-modified mspA gene encoding the mature protein without the signal sequence was cloned downstream of a T7 promoter to give the plasmid pMN501. Induction of E. coli BL21(DE3)/pMN501 with IPTG led to the expression of proteins at 20 kDa and 100 kDa compared with E. coli BL21(DE3) transformed with the control plasmid pET24+ (Fig. 7). Cells of both strains were lysed by boiling in water. Loading buffer containing SDS was added, and the supernatants were boiled to dissociate the trimers of E. coli porins, which have an electrophoretic mobility of about 90 kDa in a denaturing polyacrylamide gel and could interfere with the detection of channel-forming activity in recombinant MspA. In addition, the preparative gel was stained with Coomassie blue to excise recombinant MspA precisely in its monomeric and oligomeric forms and to minimize contamination from other E. coli proteins. These recombinant proteins were eluted from the polyacrylamide gel and had the same electrophoretic mobilities as native and denatured MspA from M. smegmatis (Fig. 7, lanes 5 and 6). In contrast, only traces of protein or no protein could be detected after excision of the molecular mass regions of 20 kDa or 100 kDa, respectively, from extracts of E. coli BL21(DE3)/pET24+ that did not express the mspA gene (Fig. 7, lanes 2 and 3).

image

Figure 7. . Expression of the mspA gene in E. coli and purification of recombinant MspA. All samples were incubated at 37°C for 30 min before loading the gel. A. Proteins were separated on a 10% SDS–polyacrylamide gel with a buffer containing tricine as described by Schägger and Jagow (1987). The gel was stained with silver (Morrissey, 1981). Lanes: M, molecular mass marker (200, 116, 97, 66, 55, 36.5, 31, 21.5, 14.4, 6 kDa); 1, 1 μg of total protein from E. coli BL21(DE3)/pET24(+) after induction; 2, about 400 ng of gel-eluted proteins from lane 1 with a molecular mass of about 20 kDa; 3, about 400 ng of gel-eluted proteins from lane 1 with a molecular mass of about 100 kDa; 4, 1 μg of total protein from E. coli BL21(DE3)/pMN501 after induction; 5, about 400 ng of gel-eluted proteins from lane 4 with a molecular mass of about 20 kDa; 6, about 400 ng of gel-eluted proteins from lane 4 with a molecular mass of about 100 kDa; 7, 800 ng of purified porin from M. smegmatis ; 8, 800 ng of purified porin from M. smegmatis denatured by boiling in 80% DMSO and by precipitation with acetone. B. The same samples were separated on a 10% SDS–polyacrylamide gel as described above, blotted onto a PVDF membrane and detected with rabbit antiserum to purified MspA.

Download figure to PowerPoint

A polyclonal rabbit serum raised to purified porin from M. smegmatis specifically recognized the native porin at 100 kDa and the denatured MspA monomer at 20 kDa, as shown in an immunoblot (Fig. 7B, lanes 7 and 8). After the induction of MspA expression, this serum detected two proteins in the crude extract of E. coli BL21(DE3)/pMN501 that had the same electrophoretic mobilities as the gel-purified proteins at 20 kDa and 100 kDa. These results indicated that the gel-purified proteins were identical to the MspA monomer and an oligomeric form of MspA.

After cell lysis and removal of cell debris by centrifugation, the supernatant of E. coli BL21(DE3)/pMN501 contained more than 50% of recombinant MspA monomer, as estimated from a Coomassie-stained denaturing polyacrylamide gel, whereas no MspA oligomers could be detected at 100 kDa. This indicated that this MspA oligomer is not soluble in water, which is the expected result for integral membrane proteins. To analyse whether the MspA monomer is water soluble, the supernatant was centrifuged at 100 000 × g for 1 h. Almost all of the MspA monomer was found in the supernatant after ultracentrifugation, indicating that the monomer is indeed soluble in water (data not shown).

The detergent-solubilized recombinant MspA oligomer showed a high channel-forming activity after reconstitution in planar lipid membranes with a similar distribution of single-channel conductances to the native porin after extraction with organic solvents (Fig. 1D). This channel-forming activity was observed down to a protein concentration of 10 ng ml−1. Rarely, channels with conductance values of about 0.5 nS and 1.0 nS were detected, which were probably caused by a minor contamination with porins from E. coli (Fig. 1D). However, more than 20% of the channels recorded with purified MspA oligomer had conductance values of 4.6 nS and above, which were never observed with porins from E. coli. In contrast, no significant reconstitution of channels larger than 1 nS could be detected after addition of the proteins at 20 kDa and 100 kDa eluted from extracts of E. coli that did not express the mspA gene. The recombinant MspA monomer did not form channels in diphytanoyl phosphatidylcholine bilayers, in agreement with the result obtained with the denatured MspA monomer from M. smegmatis. These experiments demonstrate that expression of the mspA gene is sufficient for the formation of MspA oligomers with high channel-forming activity.

Channel properties of the MspA porin

The channel conductance of purified MspA porin was influenced considerably by the salt composition of the solution bathing the lipid bilayer in reconstitution experiments (Table 1). Cations moved through the MspA pore as in the aqueous environment, as demonstrated by the linear increase in pore conductivity with the specific conductance of the different salts in water. This suggests that MspA forms a wide water-filled channel. Exchange of K+ by the less mobile Tris+ reduced the single-channel conductance by a factor of 20, whereas replacement of Cl by the less mobile acetate anion had almost no effect. This indicated a selectivity of MspA for cations, in agreement with experiments with crude cell wall extracts from M. smegmatis (Trias and Benz, 1994). To quantify the cation selectivity of MspA, zero-current potentials of membranes containing several thousand MspA porins were measured in the presence of salt gradients. A 2.4-fold KCl gradient resulted in a potential of 16.0 ± 0.15 mV, which was positive at the more-dilute side. Using the Goldman–Hodgkin–Katz equation (Benz et al., 1979), this is equivalent to a permeability ratio of cations over anions (Pc/Pa) of 6.6 ± 0.3. A 2.4-fold NaCl gradient resulted in a permeability ratio of cations over anions (Pc/Pa) of 6.3 ± 0.5. The Pc/Pa ratios were consistent with a strong cation selectivity of the MspA porin, as indicated by the single-channel experiments. As these properties are very similar to those of the channels of cell wall extracts of M. smegmatis (Trias and Benz, 1994), it is concluded that MspA is the dominant channel in these extracts.

Table 1.  . Average single-channel conductances (G) of the MspA porin in 1 M solutions of different salts. MspA was added to both sides of the membranes to a concentration of 30 pg ml−1. A 10-fold higher MspA concentration was used to measure the channel conductivity in the presence of 1 M Tris-Cl. Salt solutions were buffered with 10 mM 2-(N-morpholino)ethanesulphonic acid (MES) at pH 6.0. The pH values of solutions of Tris-Cl and KAc were 8.0 and 7.3 respectively. Data were collected from at least three different membranes. Average single-channel conductances were calculated from at least 50 events.a. From Trias and Benz (1994).Thumbnail image of

Reconstitution of cell wall channels from M. smegmatis occurred almost exclusively in a step-like fashion, as shown in 3Fig. 3A. Only a few MspA channels were observed switching between open and closed states (Fig. 3B). Switching channels in the open state had a conductance of 2.3 nS, which was reduced to about 40% in the partially closed state (Fig. 3B). Transitions between both types of channels were observed, which indicated that two different configurations of the same channel were reconstituted rather than two different channel species. The switching of the MspA pore occurred with a slow time constant compared with the porin from M. chelonae (Trias and Benz, 1993).

Occurrence of the mspA gene in mycobacteria

To find out whether the mspA gene exists in other mycobacteria, we examined its presence in different species of mycobacteria. A probe derived from the mspA gene hybridized to the genomes of M. smegmatis, M. fortuitum and M. chelonae. None of the slow-growing mycobacterial strains tested hybridized with this probe (Fig. 8). Thus, the mspA gene appears to be specific for fast-growing mycobacteria. The mspA probe hybridized to three fragments of the BamHI-digested chromosomal DNA of M. smegmatis. Thus, the M. smegmatis chromosome may contain several copies of the porin gene, as there are no BamHI restriction sites in the sequenced DNA fragment containing the mspA gene.

image

Figure 8. . Presence of the mspA gene in different species of mycobacteria. Chromosomal DNA (1 μg) of each strain was digested to completion with BamHI and analysed by Southern blotting with a digoxigenin-labelled probe corresponding to the mspA gene. Lanes: 1 and 12, molecular mass markers; 2, M. tuberculosis H37Rv; 3, M. bovis ; 4, M. africanum ; 5, M. microti ; 6, M. avium ; 7, M. intracellulare ; 8, M. kansasii ; 9, M. smegmatis mc2155; 10, M. fortuitum ; 11, M. chelonae.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The existence of porins in Gram-positive bacteria has been demonstrated for several mycobacteria. In this study, we have cloned the mspA gene, which is shown to encode a porin from M. smegmatis.

A 100 kDa protein with high channel-forming activity was extracted from whole cells of M. smegmatis by different detergents and organic solvents and purified by preparative gel electrophoresis. The same single-channel recordings as with whole-cell extracts were obtained with less than 80 pg ml−1 of the purified 100 kDa protein, indicating that channel proteins were not a minor component in this preparation. The MspA porin forms a water-filled channel, which favours the permeation of cations, as demonstrated by the strong influence of the cation size on its pore conductance and by selectivity measurements. A similar cation selectivity has been found for detergent extracts of the cell wall of M. smegmatis (Trias and Benz, 1994). These results suggest that MspA is the dominant porin in cell extracts of M. smegmatis. However, it might be possible that other porins with very similar single-channel conductances and ion selectivities are expressed in M. smegmatis.

Dissociation of the 100 kDa protein by boiling in 80% DMSO yielded the MspA monomer, which did not form channels any more. This demonstrates that the MspA channel is composed of subunits, as is the case for other porins from Gram-positive bacteria (Lichtinger et al., 1998; Riess et al., 1998). E. coli cells expressing the mspA gene produced a 100 kDa protein, which was purified by preparative gel electrophoresis. In contrast to the MspA monomer, the recombinant MspA oligomer was not soluble in water and had the same channel-forming activity as whole-cell extracts of M. smegmatis with organic solvents. The observed channels were easily distinguishable from the small amount of E. coli porins, which contaminated this preparation and had conductance values of about 1 nS, in agreement with published data for OmpF and OmpC, the main porins of E. coli (Benz and Bauer, 1988). More than 50% of the channels observed with purified recombinant MspA oligomer had conductance values of about 2.3 nS, and more than 20% of the conductance steps were greater than 4.6 nS. This experiment proved that the mspA gene encodes a porin from M. smegmatis. In addition, a polyclonal antiserum specifically detected MspA in its monomeric and oligomeric form in both M. smegmatis and E. coli. These experiments indicated that, even in E. coli, expression of the mspA gene led to the formation of an oligomeric channel protein resistant to denaturation by heat and detergents and suggested a structure similar to the native MspA pore. However, less than 1% of the recombinant MspA oligomer seemed to be folded correctly, as estimated from the minimal protein concentration of 10 ng ml−1 that was necessary to detect channels in lipid bilayer experiments. Presumably, the MspA oligomer aggregated in E. coli, as observed for many porins from Gram-negative bacteria that are expressed heterologously without signal sequence (Schmid et al., 1996).

MspA can be dissolved in organic solvents, which indicates that it possesses a hydrophobic surface, as expected for an integral membrane protein. Even boiling in 3% SDS neither dissociated the MspA oligomer nor destroyed its channel-forming activity. This demonstrates that MspA is an extraordinarily stable protein, even when compared with porins of Gram-negative bacteria, which are also stable proteins (Cowan et al., 1992) but are completely denatured after boiling in 0.1% SDS.

In lipid bilayer experiments, large current fluctuations of the MspA channels were observed at pH 5.4 and pH 8, whereas at pH 6, more than 100 channels reconstituted into the membrane without a significant increase in current noise. This might indicate an equilibrium of different pH-dependent substrates of the MspA channel, as is the case for porins of Gram-negative bacteria (Todt et al., 1992; Liu and Delcour, 1998). Different conductance values with maxima at 2.3 nS and 4.6 nS might be caused by a simultaneous reconstitution of MspA channels into the membrane or by the existence of different MspA conformations.

The size of mature MspA with 184 amino acids is surprisingly small compared with other general diffusion porins, which consist of at least 301 amino acids (porin from Rhodobacter capsulatus ). One monomer of a porin from Gram-negative bacteria needs 16 amphipathic β-sheets to form a β-barrel-like channel. It is difficult to imagine how such a structure spanning the mycobacterial outer membrane, which is twice as thick as that from E. coli (Brennan and Nikaido, 1995), can be built with about half of the amino acids. It should be noted that MspA effectively reconstitutes itself into planar lipid membranes with a thickness of about 5 nm, which is about half as thick as the mycolic acid layer. It is not known whether the channel-forming structure of MspA is the same in both membranes. Although our experiments clearly demonstrate that MspA is oligomeric in its active form, additional experiments are necessary to assess the oligomerization state of the MspA channel in vitro and in vivo. The result that oligomers of MspA up to 220 kDa can be detected in denaturing polyacrylamide gels and also reconstitute as channels in lipid bilayers indicates that MspA is able to form oligomers of many MspA monomers. This oligomerization occurs after a roughly 20-fold increase in MspA concentration and indicates an equilibrium between different MspA oligomers. Switching channels might give a hint about the number of channel-forming subunits that form a pore, but these have not been observed in previous experiments with the porin from M. smegmatis (Trias and Benz, 1994). In this study, we observed switching channels of MspA with a closed state exhibiting about 40% of the conductance of the same channel in the open state (Fig. 3B). This is consistent with the existence of several channel-forming subunits in an MspA pore, but a more detailed analysis will be necessary to determine their number.

Sequence analysis of mature MspA did not reveal membrane-spanning hydrophobic α-helices, but showed that a large part of the protein is compatible with the formation of amphipathic β-strands. This might be a common property of all outer membrane proteins, as a hydrophobic α-helix would act as a stop–transfer sequence blocking their transfer across the cytoplasmic membrane (Pugsley, 1993).

Database searches did not reveal significant similarities to any other known protein. In addition, MspA has no phenylalanine at the C-terminus, which is a highly conserved residue among outer membrane proteins and essential for their efficient membrane incorporation (Struyvéet al., 1991). Thus, MspA does not seem to be related to porins from Gram-negative bacteria, although it should be noted that sequence variation among porins from Gram-negative bacteria can be considerable (Welte et al., 1991). There were also no significant similarities of MspA to the translated genome sequence of M. tuberculosis, especially not to OmpATb, which has recently been described as a porin-like protein. Thus, MspA might be the prototype of a new class of channel-forming proteins in the outer membranes of bacteria.

Mukhopadhyay et al. (1997) isolated cell walls from M. smegmatis and demonstrated channel-forming activity in fractions enriched in a 40 kDa protein. The N-terminal sequence determined from one of these protein fractions did not show any similarity to the sequence of MspA. In preparative SDS–polyacrylamide gels loaded with whole-cell extracts of M. smegmatis, we detected high channel-forming activity at 100 kDa as a result of MspA and low activity down to 20 kDa, which might be caused by minor amounts of non-denatured MspA monomer and different MspA oligomers or of another porin. Another possibility is the existence of channels with low activity in the cell wall of M. smegmatis, such as OmpATb of M. tuberculosis (Senaratne et al., 1998). The small channel-forming activity observed by Mukhopadhyay et al. (1997) appears to be consistent with both explanations.

Three fragments of chromosomal DNA of M. smegmatis with homology to mspA containing DNA were detected in Southern blot experiments. This situation seems to be similar to that of E. coli and other Gram-negative bacteria, which have several porin genes including porin genes that are only expressed in pseudorevertants of porin-deficient mutants (Hindahl et al., 1984). However, our data do not allow us to distinguish whether the chromosome of M. smegmatis contains three copies of the mspA gene or three homologous porin genes. Expression of other porins in minor quantities could result in the formation of mixed oligomers. Assuming that other porins produce channels with different conductances, as observed for all porins so far, they should, in principle, be detectable by single-channel analysis. The conductances of the channels produced by recombinant MspA and that produced by methylenechloride/methanol-solubilized native MspA did not show differences greater than 0.5 nS. However, it cannot be excluded that mixed porin oligomers are present in the cell wall of M. smegmatis in minor quantities.

In contrast to fast-growing mycobacteria, only very weak hybridization signals of chromosomal DNA of M. tuberculosis or other slow-growing mycobacteria were detected. However, the absence of a porin in M. tuberculosis would be surprising, because a porin in its outer membrane is essential for physiological reasons. Apparently, the sequence similarity between the porin genes of the fast- and the slow-growing mycobacteria is below the detection limit of hybridization experiments. It has been proposed that the low permeability of the mycobacterial cell wall might be the reason for the slow growth of some mycobacterial species, which has been supported by measuring the transport kinetics for several hydrophilic nutrients, at least in certain cases (Jarlier and Nikaido, 1990). The apparent sequence divergence of porin genes of fast- and slow-growing mycobacteria might be correlated with different channel sizes of the porins and might, therefore, be responsible for different growth rates. The small single-channel conductance of the porin-like protein OmpATb of M. tuberculosis and its low activity are in contrast to the channel properties of MspA and seem to support this hypothesis (Senaratne et al., 1998). However, the permeability of M. tuberculosis to hydrophilic β-lactam compounds has been shown to be similar to that of M. smegmatis (Chambers et al., 1995). Therefore, the availability of structural information about mycobacterial porins and analysis of their expression is important for our understanding of how mycobacteria control the access of hydrophilic molecules into the cell. In the case of M. tuberculosis, this knowledge could be of paramount medical importance.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Bacterial strains, plasmids and growth conditions

The following mycobacterial strains were used in this study: M. smegmatis mc2155 (Snapper et al., 1990); M. tuberculosis H37Rv (ATTC no. 25618); M. bovis (ATTC no. 19210); M. africanum (ATTC no. 25420); M. microti (ATTC no. 19422); M. avium (ATTC no. 25291); M. intracellulare (ATTC no. 13950); M. kansasii (ATTC no. 12478); M. fortuitum (ATTC no. 6841); M. chelonae (ATTC no. 35752). Cultures of mycobacteria were grown in Middlebrook 7H9 medium (Difco Laboratories) supplemented with 0.2% glycerol, 0.05% Tween 80 and ADC enrichment (Difco Laboratories) at 37°C and 50 μg ml−1 hygromycin, where appropriate.

E. coli DH5α was used for all cloning experiments and for the construction of the genomic library of M. smegmatis. E. coli cultures were grown routinely in LB medium at 37°C and supplemented with 200 μg ml−1 hygromycin, where required.

Plasmid pBluescript (Stratagene) served as a cloning vector in E. coli. The shuttle plasmid pOLYG is a derivative of p16R (Garbe et al., 1994).

Protein purification

Extraction of the porin of M. smegmatis was performed as described previously (Lichtinger et al., 1998). Cells were resuspended in a mixture of dichloromethane and methanol (1:2) in a Teflon centrifuge tube (Nalgene) and kept at room temperature overnight while stirring with a magnet. Incubation on ice for 10 min was followed by centrifugation at 15 000 r.p.m. for 15 min at 4°C. The supernatant was transferred to another Teflon centrifuge tube, diluted 10-fold with diethylether, stirred on ice for 1 h and centrifuged for 30 min at 5000 r.p.m. and 4°C. The pellet was dissolved in 1% Genapol X-080 (Fluka). The ether phase was rejected, and the remaining yellow aqueous phase was concentrated by evaporation in a speed-vac to a volume of about 1 ml and stabilized by the addition of Genapol to a final concentration of 1%. Protein samples were incubated at 40°C for 30 min before loading on a gel, unless otherwise indicated. Analytical and preparative SDS–polyacrylamide gels were performed as described previously (Laemmli, 1970) and stained with silver using the method of Morrissey (1981).

To purify proteins, the gel was cut into two halves. One half was stained with silver to identify the protein bands of interest. These bands were excised from the second half of the gel, which remained unstained, and the proteins were eluted with 0.5% Genapol and analysed for channel-forming activity using the lipid bilayer method and by analytical SDS–PAGE. Protein concentrations were determined using bicinchoninic acid (Smith et al., 1985) and Coomassie brilliant blue G 250 (Bradford, 1976) with BSA as standard protein.

Lipid bilayer experiments

The methods used for black lipid bilayer experiments were essentially the same as those described previously (Benz et al., 1978). Characterization of the porin from C. glutamicum has shown that negative charges in the lipid bilayer provided by incorporation of phosphatidylserine are essential for rapid reconstitution of these channels (Lichtinger et al., 1998). Therefore, membranes were formed from a 1% (w/v) solution of diphytanoylphosphatidylcholine (0.8%) and phosphatidylserine (0.2%) (Avanti Polar Lipids) in n-decane, unless otherwise indicated, across a circular hole of 0.1 mm2 or 0.5 mm2 surface area between two aqueous compartments in a Teflon cell. An aqueous solution (5 ml) of 1 M KCl buffered with 10 mM 2-(N-morpholino)ethanesulphonic acid (MES), pH 6.0, were used in all experiments unless otherwise stated. The temperature was kept at 20°C throughout the experiment. Membrane current was measured after the application of a transmembrane potential of 10 mV with a pair of silver electrodes. The current was boosted by a current amplifier (427, Keithley) and recorded with a strip chart recorder. Whole-cell extracts of M. smegmatis containing the porin were added after the lipid membrane turned optically black to incident light.

Zero-current membrane potential measurements were performed as described by Benz et al. (1979). Membranes were formed from a 1% (w/v) solution of diphytanoylphosphatidylcholine in n-decane. An aqueous solution (4 ml) of a 50 mM salt solution buffered at pH 6.0 with 10 mM MES was used in each compartment of a Teflon cell. After the addition of purified MspA (c = 20 ng ml−1), the conductance of the membrane increased about 1000- to 50 000-fold within 30–60 min, corresponding to 500–25 000 MspA channels if one assumes a conductance of 4.6 nS for a single MspA channel. Then, the voltage was turned off, and the electrometer (6517, Keithley) was switched to allow measurements of membrane potentials. A 2.4-fold salt gradient was established across the membranes by adding 100 μl of a 3 M salt solution to the cis side and 100 μl of a 50 mM solution of the same salt to the other side of the membrane while stirring. The zero-current membrane potential reached its final value within about 1 h.

N-terminal amino acid sequencing

Proteins were separated by SDS–PAGE and transferred onto a polyvinylidene (PVDF) membrane (Immobilon-P; Millipore). The membrane was stained with Coomassie blue, and the protein band was cut out from the membrane and sequenced by automated Edman degradation using the HP G1005A Protein Sequencing system (Hewlett Packard). The sequencing system was operated using standard reagents, solvents and programs supplied by the manufacturer. Proteins blotted onto PVDF membrane were sequenced using routine 3.1, PVDF, after the loading of the PVDF membrane of the upper part of the sample cartridge. Peptides were sequenced using routine 3.5. Samples were loaded onto the hydrophobic half of the biphasic cartridge after 1:10 dilution with 2% TFA using the sample loading station.

For internal peptide sequences, gel slices with the protein of interest were digested with trypsin (Hellman et al., 1995), the peptides were separated by high-performance liquid chromatography (HPLC) and the N-terminal amino acid sequence was determined as described above.

Cloning and sequencing of the mspA gene

Chromosomal DNA was isolated from M. smegmatis as described previously (van Soolingen et al., 1991) and partially digested with Sau3A1. A genomic library was constructed by ligation of the DNA fragments into the BamHI-digested, dephosphorylated shuttle plasmid pOLYG (Garbe et al., 1994). The library was amplified in E. coli DH5α.

The sequence of oligonucleotide MP01 (5′-CAGCAGTGGGACACCTTCCT-3′) was derived from amino acids 19–25 of the N-terminal sequence (QQWDTFL) of MspA. Polymerase chain reaction (PCR) amplification of the genomic library of M. smegmatis with MP01 and KS19 (5′-GCCGCTCTAGAACTAGTGG-3′), a primer binding next to the polycloning site of plasmid pOLYG, led to the isolation of a specific 1.3 kbp fragment. The conditions for PCR were 5 min denaturing at 95°C, then 30 cycles of 30 s at 95°C, 30 s at 63°C, 2 min at 72°C and, finally, 7 min at 72°C. The amplified PCR fragment was purified from an agarose gel and labelled with digoxigenin as described by the manufacturer (Boehringer Mannheim).

For colony hybridization experiments, the M. smegmatis library was plated on LB-hygromycin plates, and the colonies were transferred to Nytran+ membranes (Schleicher & Schuell). The membranes were placed sequentially on 3 MM Whatman paper saturated with the following: 10% SDS (3 min); 0.5 M NaOH (5 min); 1 M Tris-HCl, pH 8.0 (3 min); 1 M Tris-HCl, pH 8.0, 1.5 M NaCl (5 min). The membranes were washed with 2 × SSC and air dried. The DNA was cross-linked to the membranes and prehybridized for 3 h at 42°C in Dig-Easy hybridization solution (Boehringer Mannheim). Hybridization was carried out in the presence of 250 ng OF digoxigenin-labelled PCR fragment at 42°C overnight. The membrane was washed twice for 5 min at room temperature with 2 × SSC, 0.1% SDS and twice for 15 min at 68°C with 0.1 × SSC, 0.1% SDS. Detection of the hybridized digoxigenin-labelled probe was performed using a Dig nucleic acid detection kit (Boehringer Mannheim).

Three colonies hybridized with the mspA probe, and the respective plasmids were purified, resulting in pPOR6, pPOR7 and pPOR10 with respective inserts of chromosomal M. smegmatis DNA of ≈3 kbp, 8 kbp and 4 kbp. Plasmid pPOR6 was chosen for further analysis. A part of this insert was sequenced using the AmpliTaq DNA polymerase, the Dye Terminator cycle sequencing kit (Perkin-Elmer) and an Applied Biosystems 310 sequencer (Perkin-Elmer) by primer walking. Both strands were sequenced at least twice. To resolve ambiguous positions, the Thermosequenase cycle sequencing kit (Amersham) with 33P-labelled terminators was used.

Mass spectrometry

Mass analysis was carried out on a Bruker Reflex II MALDI-TOF mass spectrometer. The instrument was operated in the reflector mode with 17 kV ion acceleration and 20 kV reflectron voltages. Ionization was accomplished with a 337 nm beam from a nitrogen laser with a repetition rate of 3 Hz. The signal of the detector was digitized using a transient recorder. Instrument control and data processing were performed on a Spark 5 station using programs supplied by Bruker. The resolution was  ± 2 Da. Steel targets were cleaned by sonication in 50% acetonitrile followed by washing with water and acetone. The saturated 4-hydroxy-alpha-cyanocinnamic acid matrix was freshly prepared by dissolving 5 mg ml−1 0.1% TFA/50% acetonitrile, followed by vortexing and centrifugation. The supernatant was used for sample preparation. The purified 100 kDa protein with high channel-forming activity was isolated from the methanol–chloroform phase using the precipitation method of Wessel and Flügge (1984). Samples (1 μl) were mixed with 1 μl of the saturated matrix solution, spotted on a target and air dried at room temperature. All masses were calibrated using an external calibration mixture containing cytochrome c (12 361.1 Da) and trypsinogen (32 981.9 Da).

Expression of the mspA gene in E. coli and purification of recombinant MspA

Competent cells of E. coli BL21(DE3) were transformed with the plasmids pET24+ and pMN501, which contains a transcriptional fusion of a codon usage-modified mspA gene to the T7 promoter. The sequence encoding the signal peptide was replaced by a methionine (details will be published elsewhere). The transformed cells were plated on LB agar supplemented with kanamycin (30 μg ml−1). Single colonies were inoculated into 4 ml of LB medium with ampicillin (30 μg ml−1). After incubation for 16 h at 37°C with shaking, 1 ml was used to inoculate a 2 l flask containing 1 l of LB/Kan30. Incubation was continued at 37°C with shaking until the optical density (OD) at 600 nm was 0.5. Then, IPTG was added to a final concentration of 1 mM. After further incubation for 4 h, the cells were harvested by centrifugation (5000 × g, 15 min), resuspended in H2O and lysed by boiling for 10 min. One part of the lysed cells was centrifuged to remove the cell debris (10 000 × g, 30 min). The supernatant of the cells that expressed MspA was subjected to ultracentrifugation at 100 000 × g (rotor Ti 60; Beckman) for 60 min to analyse whether MspA is soluble in water. Another part of the lysed cells was mixed with protein loading buffer to give final concentrations of 35 mM Tris-HCl (pH 7.0), 3% SDS, 1.25% mercaptoethanol, 7.5% glycerol and 0.0125% Coomassie blue. Cell debris was pelleted by centrifugation (10 000 × g, 30 min). The supernatants were boiled for 2 min and separated on a denaturing polyacrylamide gel that was stained with Coomassie blue. The supernatant of E. coli BL21(DE3)/pMN501 contained bands at 100 kDa and 20 kDa, corresponding to recombinant MspA in its monomeric and oligomeric form, which were precisely excised from the gel to minimize contamination with other E. coli proteins. The proteins were eluted from the polyacrylamide with 0.5% Genapol overnight. Proteins at 100 kDa and 20 kDa in the supernatant of E. coli BL21(DE3)/pET24+ were purified in the same way and were used as a control.

Production of polyclonal antiserum to MspA and immunoblot analysis

A polyclonal antibody was raised by the Animal Resources Services at the University of California at Davis. Female New Zealand White rabbits were injected five times subcutaneously with ≈100 μg MspA at 2 week intervals. The resulting antiserum MspA813 was used for immunoblot analysis of bacterial lysates or purified protein by standard procedures. TiterMax Gold (CytRx) was used as adjuvant.

Southern blot analysis

For Southern blot analysis, chromosomal DNA was isolated from various mycobacterial strains (van Soolingen et al., 1991), digested with BamHI, separated on a 1% agarose gel and transferred in 10 × SSC (1.5 M NaCl, 0.15 M sodium citrate) to a positively charged nylon membrane (Boehringer Mannheim). The DNA was cross-linked to the membrane, and hybridization and detection of the hybridized digoxigenin-labelled probe was performed as described above for colony hybridization experiments.

Footnotes
  1. *Zentrum für Molekulare Biologie, Universität Heidelberg, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany

  2. **Experimentelle Medizin, Universität Erlangen-Nürnberg, Schwabachanlage 10, 91054 Erlangen, Germany

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The authors thank Dr W. Hillen for critical reading of the manuscript, support and helpful comments, F. Rusnak for technical assistance in N-terminal microsequencing, and Dr J. Kellermann for mass analysis. This work was supported by the Deutsche Forschungsgemeinschaft (NI 412/2-1, BE 865-91) and by grants from the Fonds der Chemischen Industrie to M.N. and R.B.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • 1
    Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., et al (1997) Gapped Blast and Psi-Blast: a new generation of protein database search programs. Nucleic Acids Res 25: 33893402.
  • 2
    Benz, R. & Bauer, K. (1988) Permeation of hydrophilic molecules through the outer membrane of Gram-negative bacteria. Review on bacterial porins. Eur J Biochem 176: 119.
  • 3
    Benz, R., Janko, K., Boos, W., Läuger, P. (1978) Formation of large, ion-permeable membrane channels by the matrix protein (porin) of Escherichia coli. Biochim Biophys Acta 511: 305319.
  • 4
    Benz, R., Janko, K., Läuger, P. (1979) Ionic selectivity of pores formed by the matrix protein (porin) of Escherichia coli. Biochim Biophys Acta 551: 238247.
  • 5
    Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248254.
  • 6
    Brennan, P.J. & Nikaido, H. (1995) The envelope of mycobacteria. Annu Rev Biochem 64: 2963.
  • 7
    Chambers, H.F., Moreau, D., Yajko, D., Miick, C., Wagner, C., Hackbarth, C., et al (1995) Can penicillins and other beta-lactam antibiotics be used to treat tuberculosis? Antimicrob Agents Chemother 39: 26202624.
  • 8
    Cowan, S.W., Schirmer, T., Rummel, G., Steiert, M., Ghosh, R., Pauptit, R.A., et al (1992) Crystal structures explain functional properties of two E. coli porins. Nature 358: 727733.
  • 9
    Daffé, M. & Draper, P. (1998) The envelope layers of mycobacteria with reference to their pathogenicity. Adv Microb Physiol 39: 131203.
  • 10
    Garbe, T.R., Barathi, J., Barnini, S., Zhang, Y., Abou-Zeid, C., Tang, D., et al (1994) Transformation of mycobacterial species using hygromycin resistance as selectable marker. Microbiology 140: 133138.
  • 11
    Hellman, U., Wernstedt, C., Gonez, J., Heldin, C.H. (1995) Improvement of an ‘In-Gel’ digestion procedure for the micropreparation of internal protein fragments for amino acid sequencing. Anal Biochem 224: 451455.
  • 12
    Hindahl, M.S., Crockford, G.W., Hancock, R.E. (1984) Outer membrane protein NmpC of Escherichia coli : pore-forming properties in black lipid bilayers. J Bacteriol 159: 10531055.
  • 13
    Jarlier, V. & Nikaido, H. (1990) Permeability barrier to hydrophilic solutes in Mycobacterium chelonei. J Bacteriol 172: 14181423.
  • 14
    Kyte, J. & Doolittle, R.F. (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105132.
  • 15
    Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680685.
  • 16
    Lichtinger, T., Burkovski, A., Niederweis, M., Kramer, R., Benz, R. (1998) Biochemical and biophysical characterization of the cell wall porin of Corynebacterium glutamicum : the channel is formed by a low molecular mass polypeptide. Biochemistry 37: 1502415032.
  • 17
    Liu, N. & Delcour, A.H. (1998) Inhibitory effect of acidic pH on OmpC porin: wild-type and mutant studies. FEBS Lett 434: 160164.
  • 18
    Liu, J., Rosenberg, E.Y., Nikaido, H. (1995) Fluidity of the lipid domain of cell wall from Mycobacterium chelonae. Proc Natl Acad Sci USA 92: 1125411258.
  • 19
    Liu, J., Barry, III, C.E., Besra, G.S., Nikaido, H. (1996) Mycolic acid structure determines the fluidity of the mycobacterial cell wall. J Biol Chem 271: 2954529551.
  • 20
    Morrissey, J.H. (1981) Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity. Anal Biochem 117: 307310.
  • 21
    Mukhopadhyay, S., Basu, D., Chakrabarti, P. (1997) Characterization of a porin from Mycobacterium smegmatis. J Bacteriol 179: 62056207.
  • 22
    Nakamura, Y., Gojobori, T., Ikemura, T. (1997) Codon usage tabulated from the international DNA sequence databases. Nucleic Acids Res 25: 244245.
  • 23
    Nikaido, H., Kim, S.H., Rosenberg, E.Y. (1993) Physical organization of lipids in the cell wall of Mycobacterium chelonae. Mol Microbiol 8: 10251030.
  • 24
    Pugsley, A.P. (1993) The complete general secretory pathway in Gram-negative bacteria. Microbiol Rev 57: 50108.
  • 25
    Rauch, G. & Moran, O. (1994) On the structure of mitochondrial porins and its homologies with bacterial porins. Biochem Biophys Res Commun 200: 908915.
  • 26
    Riess, F.G., Lichtinger, T., Cseh, R., Yassin, A.F., Schaal, K.P., Benz, R. (1998) The cell wall porin of Nocardia farcinica : biochemical identification of the channel-forming protein and biophysical characterization of the channel properties. Mol Microbiol 29: 139150.
  • 27
    Rost, B., Casadio, R., Fariselli, P., Sander, C. (1995) Transmembrane helices predicted at 95% accuracy. Protein Sci 4: 521533.
  • 28
    Schägger, H. & Von Jagow, G. (1987) Tricine-sodium dodecyl sulfate–polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166: 368379.
  • 29
    Schmid, B., Kromer, M., Schulz, G.E. (1996) Expression of porin from Rhodopseudomonas blastica in Escherichia coli inclusion bodies and folding into exact native structure. FEBS Lett 381: 111114.
  • 30
    Senaratne, R.H., Mobasheri, H., Papavinasasundaram, K.G., Jenner, P., Lea, E.J., Draper, P. (1998) Expression of a gene for a porin-like protein of the OmpA family from Mycobacterium tuberculosis H37Rv. J Bacteriol 180: 35413547.
  • 31
    Smith, P.K., Krohn, I.R., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., et al (1985) Measurement of protein using bicinchoninic acid. Anal Biochem 150: 7685.
  • 32
    Snapper, S.B., Melton, R.E., Mustafa, S., Kieser, T., Jacobs, Jr, W.R. (1990) Isolation and characterization of efficient plasmid transformation mutants of Mycobacterium smegmatis. Mol Microbiol 4: 19111919.
  • 33
    Van Soolingen, D., Hermans, P.W., De Haas, P.E., Soll, D.R., Van Embden, J.D. (1991) Occurrence and stability of insertion sequences in Mycobacterium tuberculosis complex strains: evaluation of an insertion sequence-dependent DNA polymorphism as a tool in the epidemiology of tuberculosis. J Clin Microbiol 29: 25782586.
  • 34
    Struyvé, M., Moons, M., Tommassen, J. (1991) Carboxy-terminal phenylalanine is essential for the correct assembly of a bacterial outer membrane protein. J Mol Biol 218: 141148.
  • 35
    Sweet, R.M. & Eisenberg, D. (1983) Correlation of sequence hydrophobicities measures similarity in three-dimensional protein structure. J Mol Biol 171: 479488.
  • 36
    Todt, J.C., Rocque, W.J., McGroarty, E.J. (1992) Effects of pH on bacterial porin function. Biochemistry 31: 1047110478.
  • 37
    Trias, J. & Benz, R. (1993) Characterization of the channel formed by the mycobacterial porin in lipid bilayer membranes. Demonstration of voltage gating and of negative point charges at the channel mouth. J Biol Chem 268: 62346240.
  • 38
    Trias, J. & Benz, R. (1994) Permeability of the cell wall of Mycobacterium smegmatis. Mol Microbiol 14: 283290.
  • 39
    Trias, J., Jarlier, V., Benz, R. (1992) Porins in the cell wall of mycobacteria. Science 258: 14791481.
  • 40
    Welte, W., Weiss, M.S., Nestel, U., Weckesser, J., Schiltz, E., Schulz, G.E. (1991) Prediction of the general structure of OmpF and PhoE from the sequence and structure of porin from Rhodobacter capsulatus. Orientation of porin in the membrane. Biochim Biophys Acta 1080: 271274.
  • 41
    Wessel, D. & Flügge, U.I. (1984) A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal Biochem 138: 141143.