The C‐terminus is essential for the stability of the mycobacterial channel protein MspA

Abstract

Outer membrane proteins perform essential functions in uptake and secretion processes in bacteria. MspA is an octameric channel protein in the outer membrane of Mycobacterium smegmatis and is structurally distinct from any other known outer membrane protein. MspA is the founding member of a family with more than 3000 homologs and is one of the most widely used proteins in nanotechnological applications due to its advantageous pore structure and extraordinary stability. While a conserved C‐terminal signal sequence is essential for folding and protein assembly in the outer membrane of Gram‐negative bacteria, the molecular determinants of these processes are unknown for MspA. In this study, we show that mutation and deletion of methionine 183 in the highly conserved C‐terminus of MspA and mutation of the conserved tryptophan 40 lead to a complete loss of protein in heat extracts of M. smegmatis. Swapping these residues partially restores the heat stability of MspA indicating that methionine 183 and tryptophan 40 form a conserved sulfur‐π electron interaction, which stabilizes the MspA monomer. Flow cytometry showed that all MspA mutants are surface‐accessible demonstrating that oligomerization and membrane integration in M. smegmatis are not affected. Thus, the conserved C‐terminus of MspA is essential for its thermal stability, but it is not required for protein assembly in its native membrane, indicating that this process is mediated by a mechanism distinct from that in Gram‐negative bacteria. These findings will benefit the rational design of MspA‐like pores to tailor their properties in current and future applications.

1. INTRODUCTION

The cell envelope of diderm bacteria contains two membranes. The outer membrane is the main permeability barrier providing protection against toxic compounds (Delcour, 2009) and stability against mechanical stress (Rojas et al., 2018). This membrane is functionalized by proteins (OMPs) mediating nutrient uptake, secretion of small molecules and proteins, motility, DNA transfer, surface interactions and many other functions (Koebnik et al., 2000; Nikaido, ²⁰⁰³). Almost all outer membrane proteins (OMPs) form a membrane‐spanning β‐barrel consisting of 8 to 36 amphipathic β‐strands (Koebnik et al., 2000; Lauber et al., ²⁰¹⁸; Schulz, ²⁰⁰⁰). The β‐barrels of OMPs with transport functions form water‐filled channels (Koebnik et al., 2000; Schulz, ²⁰⁰⁰). OMPs are highly resistant to chemical and thermal denaturation due to the intrinsic stability of their β‐barrel in contrast to the α‐helical transmembrane proteins in bacterial inner membranes (Restrepo‐Perez et al., 2019). OMPs are exported across the inner membrane by the Sec machinery as unfolded polypeptides, which are guided through the periplasm by the chaperones SurA and Skp (Wang et al., 2021) and are then folded and integrated into the outer membrane by the β‐barrel assembly machinery (Bam) (Doyle & Bernstein, 2022; Rollauer et al., ²⁰¹⁵). A major determinant of OMP folding is a conserved C‐terminal motif called β‐signal at the last β‐strand (Doyle & Bernstein, 2019; Struyve et al., ¹⁹⁹¹). The β‐signal has a canonical sequence of X‐Z‐X‐Z‐X‐Z‐Tyr‐Z‐Trp/Phe, where X is hydrophobic and Z is any residue (Kim et al., 2012; Struyve et al., ¹⁹⁹¹), and is found in the vast majority of OMPs. During OMP assembly the β‐signal interacts with chaperones  and makes important contacts with BamA, the central component of the Bam machinery, to orient the polypeptide and facilitate folding of the β‐barrel domain (Doyle & Bernstein, 2019; Doyle & Bernstein, ²⁰²²; Noinaj et al., ²⁰¹³; Tomasek et al., ²⁰²⁰; Voulhoux et al., ²⁰⁰³).

While the cell envelope of mycobacteria also contains two membranes (Hoffmann et al., 2008), the architecture and lipid composition of mycobacterial outer membranes are very different from that of Gram‐negative bacteria (Jackson et al., 2021; Niederweis et al., ²⁰¹⁰). Uptake of hydrophilic solutes such as glucose, amino acids and ions across the outer membrane is mediated by pore‐forming proteins such as MspA, the major porin of Mycobacterium smegmatis (Stahl et al., 2001, 2005; Wolschendorf et al., ²⁰⁰⁷). MspA is a membrane‐spanning β‐barrel protein (Faller et al., 2004) similar to porins of Gram‐negative bacteria, but its structure is distinct. Eight identical MspA monomers form a tightly interconnected octameric β‐barrel consisting of 16 β‐strands (Faller et al., 2004), in contrast to porins of Gram‐negative bacteria that are trimeric or monomeric (Hermansen et al., 2022). The MspA β‐barrel is nearly circular, has a goblet‐like shape and is almost twice as long as OMPs of Gram‐negative bacteria. Furthermore, MspA is an extremely robust porin. It is resistant to proteolysis by trypsin and proteinase K, can withstand high temperatures, exposure to a broad pH range, incubation with strong anionic detergents and denaturants without loss of its channel‐forming properties (Heinz, Engelhardt, & Niederweis, 2003).

Importantly, MspA is one of the most widely used proteins in nanopore technologies due to its advantageous pore structure and its extreme stability towards thermal and chemical denaturation. MspA with an engineered constriction zone is capable of translocating DNA (Butler et al., 2008), discriminating all four nucleotides and their modifications (Derrington et al., 2010; Laszlo et al., ²⁰¹³; Manrao et al., ²⁰¹¹) and was the first protein enabling DNA sequencing with nanopores (Manrao et al., 2011). Subsequently, MspA was used in many nanotechnology applications including sequencing of RNA (Wang, Guan, et al., 2021; Zhang et al., ²⁰²⁰), detection of small molecules (Jia et al., 2022; Zhang, Dou, et al., ²⁰²¹), sub‐angstrom distance measurements in motor proteins (Laszlo et al., 2017), direct reads of nucleic acid barcodes (Yan et al., 2022), catalyzing chemical reactions as a nanoreactor (Zhang et al., 2021), detection of amino acid substitutions in peptides (Ringer et al., 2007), and sequencing of polypeptides (Brinkerhoff et al., 2021; Yan et al., ²⁰²¹). These examples highlight the enormous potential of MspA in nanotechnology. Furthermore, MspA is the founding member of a large protein family with more than 3000 homologs in over 150 species (Pfam: PF09203), and is the only protein in this family with a known structure (Faller et al., 2004).

The large interest in MspA in nanotechnology and its role as the best studied member of a novel protein family provide ample reasons to examine the unknown molecular mechanisms of membrane integration and thermal stability of MspA. In this study we show that modifications of the highly conserved C‐terminus drastically increases the susceptibility of MspA to denaturation. Mutational analysis identified a critical tryptophan‐methionine interaction as a key determinant of the thermal stability of MspA. This finding paves the way for a better molecular understanding and engineering of MspA which constitutes the founding member of an important class of pore proteins.

2. RESULTS

2.1. A C‐terminal hexa‐histidine affinity tag reduces MspA stability

Purification of MspA relies on its thermal stability at 100°C by detergent extraction from M. smegmatis cells (Heinz & Niederweis, 2000). This procedure denatures most other proteins, and is followed by two subsequent chromatography steps (Heinz, Roth, & Niederweis, 2003). In an attempt to shorten the purification and to improve yield and purity of MspA preparations, we designed the plasmid pML421 encoding MspA with a C‐terminal His₆‐tag (Figure 1a and Table S2). However, extraction of MspA‐His₆ in the presence of n‐octylpolyoxyethylene at 100°C produced only small quantities of octameric protein, which were detectable only in Western blots (Figure 1b). We then tried to determine the optimal temperature for the extraction of MspA‐His₆ from the outer membrane of M. smegmatis in the presence of detergents (Figure 1b, c). Almost no octameric MspA‐His₆ protein was detected at any temperature when SDS was used for extraction, while large quantities of monomeric MspA‐His₆ were observed in both octylpolyoxyethylene and sodium dodecyl sulfate (SDS) extracts (Figure 1b, c) indicating a reduced stability of MspA‐His₆ compared to wt MspA (Heinz & Niederweis, 2000; Heinz, Roth, & Niederweis, ²⁰⁰³; Niederweis et al., ¹⁹⁹⁹). These results indicate that the C‐terminus plays a role in stability of the MspA octamer.

FIGURE 1.

Extraction of MspA‐His₆ from M. smegmatis. (a) The structure of the MspA octamer and its membrane topology. Model of the MspA octamer is shown (PDB accession: 1UUN). A single monomer is depicted in black. The C‐terminal β12 strand is highlighted in red. The amino acids examined in this work are M183 (red), W181 (green), and W40 (magenta) are shown in a single monomer. The W21 and F99 residues (blue) constitute the boundaries of the membrane‐embedded part of MspA (Mahfoud et al., 2006). The image was prepared using the Chimera 12 software. (b/c) Extraction of MspA‐His₆ from the porin mutant M. smegmatis ML16 with the detergents 0.5% (v/v) OPOE (b) and 1% (v/v) SDS (c) at the indicated temperatures. 5 μg of cell lysate proteins were loaded on a 10% polyacrylamide gel. Proteins were transferred onto a PVDF membrane and stained with a mouse αHis‐tag monoclonal antibody. A molecular weight marker with protein masses in kDa was loaded in both gels. O, oligomer; M, monomer. (d) Sequence logo plot of the MspA C‐terminus. The logo plot of 254 MspA homologs was prepared using the MspA sequence of M. smegmatis (UniProtKB accession#: A0QR29) as the query sequence and the WebLogo 2 server as shown in Figure S1. Highly conserved sections of MspA homologs comprising residues 30–40 and 174–184 are shown here. The amino acids examined in this work are labeled with red numbers (W40, M181, W183). The color code for amino acids is: hydrophobic: black; polar: green; acidic: red; basic: blue; neutral: purple.

2.2. The C‐terminus of MspA is highly conserved

To examine the unexpected detrimental effect of a C‐terminal histidine tag on MspA we analyzed the sequences of all MspA homologs in the Mycobacteriaceae family. Sequences covering <80% of the mature protein were removed. The remaining 254 sequences were aligned and visualized in a logo plot (Figure S1). The sequence alignment revealed that the majority of the β‐strands of MspA and the C‐terminus of MspA are highly conserved, while considerable sequence variability was observed in the periplasmic loop including the constriction zone and surface‐accessible loops (Figures 1d and S1). Notably, tryptophan 181 and methionine 183 at positions −4 and −2 (from the C‐terminus), respectively, are highly conserved (Figure 1d). This is reminiscent of the highly conserved −3 and −1 aromatic residues of E. coli outer membrane proteins (Tak et al., 2021). The C‐terminal β‐strand of the MspA monomer (Figure 1a) containing tryptophan 181 and methionine 183 is located within the outer membrane of M. smegmatis as indicated by MspA topology studies (Mahfoud et al., 2006).

2.3. Deletion of methionine 183 eliminates MspA protein in heat extracts of M. smegmatis

Based on the detrimental effect of the C‐terminal His₆ tag on MspA, we hypothesized that the conserved C‐terminus of MspA plays a role in protein stability. To examine this hypothesis we deleted C‐terminal amino acids in a sequential manner starting from N184 towards the N terminus of MspA. Due to its technological relevance we used the MspA M2 variant for our experiments (Butler et al., 2008). MspA M2 is capable of translocating DNA and is widely used in nanopore DNA sequencing (Derrington et al., 2010; Manrao et al., ²⁰¹¹; Manrao et al., ²⁰¹²). The overall structure of MspA M2 is likely very similar to that of wt MspA (Faller et al., 2004), since the mutated residues are not known to be involved in critical interactions and are either located at the surface of the rim domain (D134R/E139K) or in the highly variable constriction zone (D90N/D91N/D93N). The only exception is the D118R mutation, which is located inside the β‐barrel structure facing the water‐filled channel interior and does not affect the thermostability of MspA (Butler et al., 2008). The mspA genes were expressed in the porin mutant M. smegmatis ML712, which lacks all four msp genes and produces the outer membrane channel protein CpnT to enable survival of the strain in the absence of any Msp pores (Pavlenok & Niederweis, 2016). As expected our selective extraction procedure by boiling with the detergent OPOE did not show any MspA protein in the porin‐deletion mutant ML712 in contrast to wt M. smegmatis (Figure 2). It should be noted that the reduced negative net charge of MspA M2 results in a lower electrophoretic mobility than wt MspA (Figure 2) in agreement with previous results (Butler et al., 2008). Deletion of N184 slightly reduced MspA protein levels (Figure 2). In contrast, the additional deletion of M183 resulted in a complete loss of MspA protein in detergent extracts of M. smegmatis (Figure 2). No protein was detected in the MspA mutant lacking the last three C‐terminal residues (182–184) (Figure 2).

FIGURE 2.

Analysis of C‐terminal MspA mutants in detergent extracts of M. smegmatis. (a) Coomassie‐stained 10% polyacrylamide gel with high‐temperature detergent extracts of the porin mutant M. smegmatis ML712 containing the backbone plasmid pMS2 or plasmids encoding wt MspA, MspA M2, MspA M2 Δ184, MspA M2 Δ183‐184, MspA M2 Δ182‐184, MspA M2 M183D, and MspA M2 M183A. 2 μg of the protein extracts were loaded in all lanes. (b) Western blot of the samples shown in A. Proteins were transferred onto a PVDF membrane and stained with rabbit MspA antiserum #813. 50 ng of protein extract was loaded in all lanes. The molecular masses of the marker are indicated in kDa.

To examine whether heat extraction using octylpolyoxyethylene extracts only a fraction of mutant MspA protein, we examined whole cell lysates of M. smegmatis. We confirmed that similar protein levels were extracted with SDS in all samples as shown by the RNA polymerase loading control (Figure S2). Western blots of the same whole cell lysates using an MspA antiserum confirmed a strongly reduced level of oligomeric MspA for the MspA M2 Δ183‐184 mutant compared to the parent MspA M2 protein indicating that octylpolyoxyethylene heat extraction reflects the amount of oligomeric MspA in the outer membrane of M. smegmatis. Interestingly, these experiments also revealed increased monomer levels for the MspA M2 Δ183‐184 mutant compared to the parent MspA M2 protein (Figure S2B) indicating that the deletion of the last two amino acids impaired the protein stability similar to that observed for histidine‐tagged MspA.

Next, we mutated methionine 183 to alanine to examine the role of this amino acid for MspA protein stability. Unexpectedly, the protein levels of MspA M2 M183A were restored to MspA M2 levels (Figure 2). Since alanine and methionine are both hydrophobic amino acids with similar hydrophobicities (Kyte & Doolittle, 1982), we reasoned that alanine at position 183 might function similarly to methionine. Therefore, we mutated methionine 183 to the negatively charged aspartate. The resulting MspA M2 M183D protein was not detected in the polyacrylamide gel and barely detectable in the Western blot (~2% of MspA M2 level in Figure 2b and S2). Taken together these results suggest that M183 is involved in a hydrophobic interaction that is important for MspA protein stability.

2.4. Tryptophan 40 is an important residue in the MspA protein

Early analyses of protein structures indicated interactions between methionine and aromatic amino acids (Morgan et al., 1978; Morgan & McAdon, ¹⁹⁸⁰). We identified two tryptophan residues at positions 40 and 181 in close proximity to methionine 183 (Figure 3a). The distances between W40 and W181 to M183 of 5.4 and 6.9 Å, respectively, are within the range of previously observed methionine‐tryptophan interactions (Morgan et al., 1978; Morgan & McAdon, ¹⁹⁸⁰). We then mutated W40 and W181 to serines to examine whether these residues play a role in MspA protein stability. While we did not observe any effect of the W181S mutation, the MspA M2 W40S protein was not detectable in high‐temperature detergent extracts of the porin deletion M. smegmatis ML712 (Figure 3b, c), indicating that the tryptophan 40 is important for protein stability and/or oligomerization of MspA in a similar manner as the methionine 183.

FIGURE 3.

Mutational analysis of tryptophans 40 and 181 in MspA. (a) Location of W40, W181, and M183 residues in the crystal structure of MspA (PDB#: 1UUN). A single MspA monomer is shown as a golden ribbon. The enlarged panel shows potential interactions of C‐terminal M183 (red) with W181 (green) in β‐strand 12 or with W40 (purple) in β‐strand 3. The distances between M183 and W40 or W181 are is 5.4 and 6.9 Å, respectively. The image was prepared with Chimera 12. (b). Coomassie‐stained 10% polyacrylamide gel with high‐temperature protein extracts of the porin mutant M. smegmatis ML712 containing the empty plasmid pMS2 or plasmids encoding wt MspA, MspA M2, MspA M2 W40S, and MspA M2 W181S protein. 2 μg of protein extract were loaded in all lanes. (c) Western blot of the samples shown in A. Proteins were transferred onto a PVDF membrane and stained with rabbit MspA antiserum #813. 50 ng of protein extract was loaded in all lanes. The molecular masses of the marker are indicated in kDa.

2.5. Amino acid swapping experiments indicate that methionine 183 interacts with tryptophan 40 in MspA

To elucidate whether W40 and M183 interact with each other, we swapped the amino acids in positions 40 and 183 and examined whether the MspA protein stability and/or outer membrane integration are restored in a manner similar as described for cross‐strand ionic interactions in β‐sheets (Huang et al., 2021). Indeed, the MspA M2 W40M/M183W mutant was detected in a Western blot (Figure 4b) at approximately 10% of wt protein level (Figure S3B), indicating that these residues interact. However, replacing the methionine‐tryptophan interaction with hydrophobic interactions in the MspA M2 W40L/M183A mutant or with ionic interactions in the MspA M2 W40R/M183D mutant did not yield any detectable protein (Figure 4).

FIGURE 4.

Protein levels of MspA M2 W40 and M183 mutants in extracts from M. smegmatis. (a) Coomassie‐stained 10% polyacrylamide gel of high‐temperature protein extracts of the porin mutant M. smegmatis ML712 containing the backbone plasmid pMS2 or plasmids encoding wt MspA, MspA M2, MspA M2 W40L/M183A, MspA M2 W40R/M183D, MspA M2 W40M/M183W protein. 2 μg of protein extract were loaded in all lanes. (b) Western blot of the samples shown in “a.” Proteins were transferred onto PVDF membrane and stained with rabbit MspA antiserum #813. 50 ng of protein extract was loaded in all lanes. The molecular masses of the marker are indicated in kDa.

To examine whether the MspA M2 W40M/M183W mutant forms functional channels, we performed lipid bilayer experiments. The MspA M2 protein readily inserted into DPhPC bilayers at very low protein concentrations (6 ng/ml) as indicated by the step‐wise increase in conductance (Figure 5a). Analysis of conductance distribution showed a predominant peak at 1.9 nS, which is in a good agreement with previously published data (Butler et al., 2008). Similar to MspA M2, the MspA M2 W40M/M183W mutant readily formed channels in bilayers at 60 ng/ml of the added extract (Figure 5c). It should be noted that MspA M2 W40M/M183W had more channel closing events as evident by more frequent downward current changes compared to MspA M2 (Figure 5a). Analysis of the conductance distribution showed a well‐defined peak at 2.9 nS for MspA M2 W40M/M183W (Figure 5d) which is greater than that of MspA M2 (Figure 5b). Taken together these results demonstrate the importance of the highly conserved tryptophan 40 and methionine 183 residues for the overall structural integrity and/or membrane insertion of MspA and indicate that these amino acids form a sulfur‐π electron interaction, which play important roles in protein folding and stabilization (Robert et al., 2006).

FIGURE 5.

Channel activity of the MspA M2 and MspA M2 W40M/M183W proteins. All experiments were performed in 1 M KCl, 10 mM HEPES, pH 7.4 electrolyte at 10 mV applied potential. The diameter of the diphytanoyl phosphatidylcholine (DPhPC) bilayer was approximately 1 mm. (a) Current trace of MspA M2. (b) Histogram of insertion events of MspA M2. 206 insertions from six membranes were analyzed. The predominant conductance peak is at 1.9 nS. Protein concentration of MspA M2 extract in the cuvette for A and B was 6 ng/ml. (c) Current trace of MspA M2 W40M/M183W. (d) Histogram of insertion events of MspA M2 W40M/M183W. 280 insertions from seven different membranes were analyzed. The major conductance peak is 2.7 nS. Protein concentration of MspA M2 W40M/M183W extract in the cuvette 30–60 ng/ml.

2.6. Surface detection of C‐terminal MspA mutants

To examine if the C‐terminus of MspA plays a role in protein folding or outer membrane integration rather than in protein stability we analyzed the surface accessibility of the MspA mutants in M. smegmatis by flow cytometry as shown previously for other mycobacterial outer membrane proteins (Danilchanka et al., 2014; Mitra et al., ²⁰¹⁷; Tak et al., ²⁰²¹). We were not able to use the quadruple porin deletion ML712 strain for these experiments due to its extensive clumping. Instead, in these experiments we used the triple porin deletion mutant M. smegmatis ML16, which produces MspB as the only porin at very low levels (Stephan et al., 2005) and does not clump like ML712 strain. Both flow cytometry and microscopy experiments showed no significant difference in clumping and cell morphology between wild type M. smegmatis SMR5 and ML16, whereas the ML712 strain showed a significantly higher propensity for clumping (Figure S4). Indeed, much less cell aggregates were observed in ML16 and wt SMR5 strains in comparison to ML712 (Figure S4A). Additionally, forward and side scatter plots of wt SMR5 and ML16 were nearly identical with a single population while ML712 had two distinct populations (Figure S4B).

Flow cytometry showed that all MspA mutants including MspA M2 W40M/M183W were detected on the cell surface of M. smegmatis ML16 at a similar or only slightly reduced level compared to the parent MspA M2 protein (Figure 6), indicating that folding, oligomerization to a complete pore and membrane integration is not affected. Thus, the absence of some of these mutant MspA proteins in high‐temperature detergent extracts is likely due to their reduced thermal stability.

FIGURE 6.

Detection of mutant MspA proteins on the cell surface of M. smegmatis. The porin mutant M. smegmatis ML16 with the empty plasmid pMS2 or with expression vectors encoding the parent M2 MspA protein or different MspA mutants were analyzed by flow cytometry. The bacteria were incubated with MspA antiserum followed by an anti‐rabbit Alexa 488‐labeled antiserum. The quantification of MspA‐positive cells from three independent experiments is shown. Standard deviations are shown as error bars. One‐way ANOVA analysis was performed using the SigmaPlot software. The statistical information about the ANOVA analysis is presented in Table S4.

3. DISCUSSION

3.1. A methionine‐tryptophan interaction in the C‐terminus is essential for the thermal stability of MspA

The widespread use of the channel protein MspA in nanotechnological applications benefits from its unusual stability against thermal and chemical denaturation as the protein withstands boiling in detergents, organic solvents and extreme pH (Heinz, Engelhardt, & Niederweis, 2003). The unexpected drastic reduction of the stability of MspA against thermal denaturation upon addition of a C‐terminal hexa‐histidine tag indicated that the C‐terminus might be involved in the unusual stability of MspA. Indeed deletion of the penultimate conserved methionine 183 resulted in the complete loss of protein in heat extracts of M. smegmatis, while deletion of the ultimate asparagine 184 had only minor effects. The importance of C‐terminus for MspA is reminiscent of outer membrane proteins of E. coli. For example, the mutation of the C‐terminal phenylalanine of the phosphoporin PhoE severely affected outer membrane assembly (Tak et al., 2021). This phenylalanine is conserved in outer membrane proteins of Gram‐negative bacteria and is required to correctly position the protein for folding by BamA (Doyle & Bernstein, 2022). While we did not measure protein folding of MspA directly, our flow cytometry experiments showed that MspA M183 point mutants and deletion mutants are detectable on the cell surface of M. smegmatis on a similar or only slightly reduced level as the parent MspA protein suggesting that folding and membrane assembly of MspA is not affected after deletion or mutation of the conserved methionine 183. Thus, we propose that the loss of MspA protein with mutations at position 183 during heat extraction is caused by a strongly reduced thermal stability. This hypothesis is supported by our observation that methionine 183 and tryptophan 40 form a sulfur‐π electron interaction, which are known to stabilize proteins (Daeffler et al., 2012; Panja et al., ²⁰²⁰; Ringer et al., ²⁰⁰⁷). Structural analysis of all proteins in the protein database in 2019 revealed a strong enrichment of methionine close to aromatic amino acids indicating sulfur‐π electron interactions in almost 70% of all proteins with a high resolution structure, but their abundance and importance in stabilizing proteins has not been recognized until recently (Weber & Warren et al., 2019). These structures revealed peaks of potential sulfur‐π interactions at distances between 3.5 and 4 Å for small angles (<10°) and of ~5.5 Å for large angles (70°–90°) (Robert et al., 2006). The distance between methionine 183 and tryptophan 40 in the MspA channel is 5.4 Å with an angle of approximately 70° (Figure S5). Surprisingly, the importance of methionines in close proximity to aromatic amino acids for protein stability has been demonstrated only for a few disease‐relevant proteins (Gacia et al., 2006; Moriki et al., ¹⁹⁹⁷) or model proteins (Chaturvedi & Mahalakshmi, 2013; Valley et al., ²⁰¹²). To our knowledge, our study provides the first direct evidence of a methionine‐aromatic interaction by demonstrating the gain of MspA protein stability after reversing the positions of methionine 183 and tryptophan 40, while both single mutants, MspA M183S and W40S, are not resistant to heat denaturation. These experiments suggest an important stabilizing role of the sulfur‐π electron interaction between methionine 183 and tryptophan 40 in MspA. This interpretation is also consistent with the biophysical analysis of a methionine triple mutant of the β‐barrel protein OmpX of E. coli, which showed a strongly decreased β‐barrel stability, while protein folding was not affected (Chaturvedi & Mahalakshmi, 2013). A close inspection of the OmpX structure indeed revealed that the distance of methionines 18 and 21 to phenylalanine 90 (4.5–6.1 Å) is consistent with the sulfur‐π electron interaction in MspA (Figure S6) and in other proteins (Vetriani et al., 1998). The identification of a potential methionine‐aromatic interaction in OmpX indicates that sulfur‐π electron interactions may also play an important stabilizing role in other β‐barrel outer membrane proteins. This hypothesis warrants further investigation.

3.2. How does the methionine tryptophan interaction contribute to the stability of MspA?

To maintain their functions, proteins of bacteria growing under extreme environmental conditions evolved to optimize stabilizing interactions (Pavlenok et al., 2012). For example, inter‐subunit salt bridges in proteins from thermophile bacteria have been found to play an important role in protein stability (Vieille & Zeikus, 2001). However, extremely stable proteins are characterized by many other stabilizing interactions including hydrophobic interactions, aromatic interactions, hydrogen bonds and ion pairs (Voulhoux et al., 2003). Interestingly, sulfur‐π electron interactions involving methionine are not mentioned. Molecular dynamics calculations for two protein complexes with critical methionine‐aromatic interactions revealed mean intermolecular interaction energies between 1.4 and 1.9 kcal/mol for a methionine‐tyrosine bond and between 1.9 and 2.8 kcal/mol for a methionine‐tryptophan bond (Vetriani et al., 1998). These values are consistent with an 1–2 kcal/mol range experimentally determined for sulfur‐aromatic interactions in model systems (Bodner et al., 1980). These energies are similar to that of a single salt bridge ranging from 1 to 3 kcal/mol (Struyve et al., 1991). However, ionic interactions are strongly dependent on their local environment (Bosshard et al., 2004), and are not compatible with hydrophobic cores of proteins and are, therefore, often found on the protein surfaces. By contrast, sulfur‐aromatic interactions not only occur over longer distances (5–6 Å vs. <4 Å for salt bridges), but also reduce the desolvation energy penalty in hydrophobic environments (Vetriani et al., 1998). In this regard, it is interesting that W40 and M183 are located in the hydrophobic core of an immunoglobulin‐like β‐sandwich domain (Lewis et al., 2018) of MspA. While immunoglobulins are stabilized by disulfide bonds which do not exist in MspA, our study strongly indicates that the W40‐M183 interaction plays a similar stabilizing role in MspA. W40 and M183 are located in the middle of this domain surrounded exclusively by apolar residues and connect two opposing β‐strands in the MspA monomer (Figure 7). Only two polar residues, R38 and H42, are at the edges of this domain. This hydrophobic core of the MspA monomer highlights the advantages of a sulfur‐π electron interaction for protein stabilization. This conclusion is supported by our observation that no monomers were detected in Western blots of the M183D mutant, which places a charge inside the hydrophobic core (Figure 2), in contrast to His‐tagged MspA (Figure 1), which retains the critical methionine 183 or the MspA M183A mutant (Figure 1b, c), which does not affect the stability of MspA. We previously observed that the MspA monomer can be produced in the cytosol of E. coli as a water‐soluble, folded protein in the absence of detergents and/or lipid membranes (Faller et al., 2004; Heinz, Engelhardt, & Niederweis, ²⁰⁰³) highlighting the monomer's stability. Denaturation of a destabilized monomer during heat extraction from M. smegmatis and subsequent precipitation would explain the absence of MspA proteins with detrimental mutations at positions M183 and W40. Such a mechanism and the location of M183 and W40 inside the hydrophobic core of the MspA monomer, argue against the direct involvement of these amino acids in oligomerisation and membrane insertion of MspA, since the membrane‐compatible β‐barrel of MspA only forms after oligomerization of monomers (Faller et al., 2004). This is exactly what we observed in the surface accessibility experiments. Even the destabilized MspA M183D and Δ183‐184 mutants, which are absent in heat extracts, are detectable on the cell surface of M. smegmatis (Figure 6), demonstrating that they are capable to oligomerize and form a β‐barrel, which then inserts into the outer membrane.

FIGURE 7.

Amino acids in the interaction sphere of methionine 183 and tryptophan 40 in MspA. A portion of a single MspA monomer (PDB#: 1UUN) is shown. The C‐terminal β‐strand 12 is shown as a red ribbon. Surface‐accessible residues face the viewer in panel “a.” Panel “b” was rotated to the left and enlarged. C‐terminal M183 (red) interacts with W40 (purple) in β‐strand 3. With the exception of R38 (orange), predominantly hydrophobic residues are found around M183–L26 (blue), H42 (light blue), F70 (dark green), V151 (light green), A154 (purple), V158 (light green), and L160 (blue). The image was prepared with Chimera 12.

The observation that the M183A mutant is present in extracts of M. smegmatis at a similar level as MspA M2 (Figure 2) is an apparent contradiction to the importance of the M183‐W40 interaction for the stability of MspA. However, it appears that alanine plays a similar stabilizing role as methionine yet by a different mechanism. Alanine is a small amino acid, which may induce a further compaction of the hydrophobic core and substitute for the loss of the sulfur‐pi electron interaction with multiple hydrophobic interaction. This hypothesis is based on the finding that compact structures are an important property of proteins from thermophilic bacteria compared to their mesophilic homologs Sawle & Ghosh, ²⁰¹¹. A high resolution structure of the M183A mutant would be necessary to confirm this hypothesis.

It should be noted that the importance of the W40‐M183 interaction for the stability of the MspA monomer does not explain the absence of C‐terminally His‐tagged MspA in heat extracts from M. smegmatis. The fact that we did not observe a major difference in the heat stability of recombinant C‐terminally tagged MspA variants produced in E. coli compared to MspA protein purified from M. smegmatis (Phale et al., 1998), indicates that the C‐terminal His‐tag is not compatible with membrane integration of MspA in M. smegmatis.

3.3. The rim domain of MspA influences channel properties

We showed that MspA M2 W40M/M183W forms functional channels in lipid bilayer experiments with an average conductance of 2.9 nS (Figure 5d). The conductance distribution of this mutant is narrower than that of MspA M2 indicating more a homogeneous channel population (Figure 5b, d). Our previous studies showed that the conductance of MspA depends on the chemical properties of the amino acids in the pore constriction and/or inside the central channel (Butler et al., 2008). Since methionine 183 and tryptophan 40 are located in the rim domain of MspA, which is partially embedded in the outer membrane of M. smegmatis (Mahfoud et al., 2006), the bilayer experiments indicate that the rim domain of MspA also influences channel conductance. It is conceivable that conformational changes of the rim domain alter the position of amino acids that line the central channel thereby affecting the ion flow through the pore.

4. CONCLUSIONS

We conclude that the role of the conserved C‐terminus of MspA is different from that in outer membrane proteins from Gram‐negative bacteria, which contains the “β signal” (de Cock et al., ¹⁹⁹⁷; Kim et al., ²⁰¹²; Robert et al., ²⁰⁰⁶) and plays an important role in protein folding and membrane integration via the Bam machinery (Doyle & Bernstein, 2019; Doyle & Bernstein, ²⁰²²). This observation, together with the fact that no homologs of the Bam proteins exist in mycobacteria, suggest that the outer membrane protein folding and membrane integration is mediated by a mechanism distinct from that in Gram‐negative bacteria. Taken together, our results demonstrate the importance of the highly conserved C‐terminus of MspA for the structural stability of its monomer. Our findings can be applied to the entire MspA family of the proteins and will help in the rational protein design of MspA and its homologs. Clearly, the tryptophan 40 and methionine 183 residues should not be altered in the many uses of MspA in nanotechnology, for example, in nanopore sequencing of DNA, RNA, and proteins (Brinkerhoff et al., 2021; Manrao et al., ²⁰¹²; Yan et al., ²⁰²¹; Zhang et al., ²⁰²⁰) in which the extreme stability of MspA is an important advantage compared to other pore proteins.

5. EXPERIMENTAL PROCEDURES

5.1. Chemicals and enzymes

Chemicals were of the highest purity available from Sigma Aldrich, Merck, Invitrogen, or Fisher Scientific unless otherwise noted. The lipid 1,2‐diphytanoyl‐sn‐glycero‐3‐phosphocholine (DPhPC) was purchased from Avanti Polar Lipids (Alabaster, AL). The detergent n‐octyl‐polyoxyethylene (OPOE) was purchased from Santa Cruz Biotechnology (Dallas, TX). Middlebrook 7H9 broth and Middlebrook 7H10 agar were from Difco (Franklin Lakes, NJ). Restriction enzymes and other molecular biology reagents were purchased from New England Biolabs (Ipswich, MA). The oligonucleotides were obtained from Integrated DNA Technologies (Coralville, IA).

5.2. Bacterial strains and growth conditions

Escherichia coli DH5α was used for cloning experiments and was routinely grown in Luria‐Bertani (LB) broth or agar plates at 37°C supplemented with 100 μg/ml of hygromycin. Mycobacterium smegmatis ML712, which lacks the four porin genes mspA, mspB, mspC, and mspD (Pavlenok et al., 2022) was used for expression of all mspA mutants in this study. The M. smegmatis ML712 strain was grown at 37 C in 7H9 liquid medium (BD Biosciences) supplemented with 0.2% glycerol and 0.05% Tween 80 or on 7H10 agar (BD Biosciences) supplemented with 0.2% glycerol. Hygromycin was used in concentrations of 50 μg/ml in both 7H9 and 7H10. Detailed description of the strains is in Table S1.

5.3. Plasmid construction

All plasmids used in this study were constructed using standard molecular biology techniques. The plasmid pML844 (Butler et al., 2008) which carries mspA m2 gene (D90N/D91N/D93N/D118R/D134R/E139K) was used as a template for cloning. Additionally, pML844 digested with SphI and HindIII (5.5 kb fragment) was used as a vector for the resulting plasmids. Table S2 describes plasmids used in this study. To delete C‐terminal amino acids, standard PCR with iProof High Fidelity GC master mix (Biorad) was used. The forward COLE1 primer binds 446 nucleotides upstream of 5′‐end of mspA m2 and spans the region encoding SphI restriction site. A reverse primer has a complimentary binding region upstream of a targeted triplet followed by Opal stop codon and HindIII restriction site. After the PCR, the amplified fragment was digested with SphI and HindII and ligated with the vector resulting in pML4149 (MspA M2 Δ184), pML4150 (MspA M2 Δ183‐184), and pML4147 (MspA M2 Δ182‐184). To introduce mutations in mspA m2 we used cite‐directed mutagenesis based on combined chain reaction (CCR) (Bi & Stambrook, 1997). The template was amplified with COLE1 forward and pMS‐SEQ1 reverse primers with addition of mutagenesis primer (Table S3). The amplified fragments were treated with SphI and HindIII and ligated with appropriately digested vector resulting in pML4151 (M2 M183A), pML4152 (M2 M183D), pML4153 (M2 W40S), pML4154 (M2 W181S), pML4167 (M2 W40L/M183A), pML4168 (M2 W40R/M183D), pML4169 (M2 W40M/M183W). All plasmids were confirmed by restriction digestion and Sanger sequencing prior to transformation into M. smegmatis ML712 for protein analysis.

5.4. Extraction of MspA proteins from outer membrane of M. smegmatis

Mycobacterium smegmatis ML712 harboring plasmids for expression of mspA mutants was grown as described in previous section. When cultures reached late log‐phase the cells were harvested at 5000g for 15 min at 4°C. Cells were washed once in phosphate‐buffered saline (PBS) and were subjected to heat‐selective extraction procedure which yields predominantly MspA protein (Heinz & Niederweis, 2000). The cells were resuspended in 150 mM NaCl, 25 mM Na₂HPO₄/NaH₂PO₄, 0.5% (v/v) n‐octylpolyoxyethylene (OPOE), pH 7.5 buffer. The ratio of cell mass to the buffer volume was 1:3. The mixture was then incubated at 100°C for 30 min. Extracted soluble MspA proteins with minor contaminants were separated from insoluble material by centrifugation at 16,000g for 30 min at 4°C. Protein concentration of the lysate was measured by BCA kit according to manufacturer's protocol.

5.5. Analysis of protein levels of MspA mutants in detergent extracts of M. smegmatis

MspA protein samples after detergent extraction were loaded and separated on a denaturing 10% polyacrylamide gel followed by Coomassie stain. In parallel, separated proteins were transferred onto PVDF membrane and stained with rabbit antiserum followed by HRP‐conjugated mouse anti‐rabbit antibodies for the MspA detection. Unless otherwise specified, equal amount of protein samples were loaded onto each lane, specifically, 2 μg of protein for SDS‐PAGE, and 50 ng of protein for Western blots. Resulting polyacrylamide gels were scanned with Epson Perfection V700 PHOTO scanner (Epson), and blots were obtained by ChemiDoc XRS+ system (BioRad).

5.6. Lipid bilayer measurements

Lipid bilayer experiments were performed in a custom made lipid bilayer apparatus as previously described (Heinz & Niederweis, 2000). Briefly, a Teflon cuvette with 10 ml volume is separated into to compartments (cis‐ and trans‐) by a wall with an aperture of approximately 1 mm in diameter. Ag/AgCl electrodes were bathed in a 1 M KCl, 10 mM HEPES, pH 7.4 electrolyte solution. The cuvette was primed on both side of the aperture with 2% diphytanoylphosphatidylcholine (DPhPC; Avanti Polar Lipids) in chloroform. Lipid membranes were painted across aperture from a solution of 1% of DPhPC in n‐decane. The samples were added to both sides of the cuvette. Baseline and detergent‐containing buffers were examined to exclude contamination and detergent interference. Single channel conductances for more than 100 pores per protein sample were recorded. The current traces were recorded at −10 mV potential. The currents were recorded using a Keithley 428 Amplifier at a sampling rate of 33 kHz, and digitized by a computer equipped with Keithley Metrabyte STA 1800 U interface. The data were recorded with Test Point 4.0 software (Keithley). The raw data were analyzed using IGOR Pro 5.03 (WaveMetrics) using a macro provided by Dr. Harald Engelhardt. The data were further analyzed in SigmaPlot 11.0 (Systat Software) to generate graphs shown here.

5.7. MspA sequence alignment and sequence logo plots

To identify MspA homologs we removed the highly variable signal peptide (first 27 residues) from the MspA porin sequence from Mycolicibacterium smegmatis (WP_011727317.1). Then, BlastP (Altschul et al., 1990) was run on the NCBI Blast server using this query sequence against the refseq_select_prot (https://www.ncbi.nlm.nih.gov/refseq/refseq_select/; updated 2023/11/15) database, with default settings, except that taxonomy was restricted to the Mycobacteriaceae family (taxid:1762) and “max target sequences” was increased to 5000. This returned 261 sequeneces, from which seven sequences covering <80% of the query length were manually deselected. The remaining 254 sequences were accessed, and the alignment was visualized and downloaded, via NCBI's “MSA Viewer.” The alignment was then uploaded to the WebLogo server version 2.8.2 (Crooks et al., 2004) (https://weblogo.berkeley.edu).

5.8. Flow cytometry for surface detection of MspA proteins

The M. smegmatis strains ML16/pMS2 (empty vector), ML16/pML844 (MspA M2), ML16/pML4150 (M2 ΔM183/ΔN184), ML16/pML4151 (M2 M183A), ML16/pML4152 (M2 M183D), and ML16/pML4169 (M2 W40M/M183W) were used for surface detection experiments. Cultures were grown to mid‐log phase and fixed with 4% paraformaldehyde for 1 h at room temperature. The cells were washed three times with PBS containing 0.02% Tyloxapol (PBST). Then, the cells were blocked with rabbit serum for 30 min at room temperature followed by washes with PBST buffer. The primary staining was done using the MspA antiserum (1:100 dilution) for 1 h at room temperature followed by three washes with PBST. Mouse anti‐rabbit Alexa 488‐labeled antibodies were added in the dilution 1:1000 for 1 h at room temperature. Flow cytometry was done on BD Accuri C6 Plus Flow Cytometer (BD Biosciences). The data were gated using unstained cells. Data analysis was done using FCS Express 7 (De Novo Software). The one‐way ANOVA analysis was done using SigmaPlot 11.0 software (Systat Software). Data were considered significant when p ≤ 0.05.

5.9. Analysis of cell size and morphology of M. smegmatis porin mutants

The M. smegmatis strains SMR5 (wt), ML16 and ML712 were grown to mid‐log phase, stained with DMN‐Trehalose, fixed with 4% paraformaldehyde and visualized by microscopy using the BioTek Cytation 5 Cell Imaging Multimode Reader. The quantitative analysis of unstained, fixed cells was done by flow cytometry as described above.

AUTHOR CONTRIBUTIONS

Mikhail Pavlenok: conceptualization (equal), formal analysis (equal), investigation (lead), writing—original draft (lead), writing—review and editing (supporting); Rashmi Ravindran Nair: formal analysis (supporting), investigation (supporting), writing—original draft (supporting); R. Curtis Hendrickson: formal analysis (supporting), investigation (supporting), writing—original draft (supporting); Michael Niederweis: conceptualization (equal), funding acquisition (lead), project administration (lead), supervision (lead), writing—original draft (supporting), writing—review and editing (lead).

FUNDING INFORMATION

This study was supported by the National Institutes of Health (NIH) grant R21 HG010543 and R01 HG012553 to Michael Niederweis. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

CONFLICT OF INTEREST STATEMENT

Michael Niederweis and Mikhail Pavlenok, along with the University of Alabama at Birmingham, hold patents describing the use of MspA for nanopore sequencing of DNA.

Supporting information

FIGURE S1. Logo plot of MspA homologs.

FIGURE S2. Western blot analysis MspA protein levels in whole cell lysates of M. smegmatis.

FIGURE S3. Quantitative analysis of expression levels of MspA mutants in M. smegmatis.

FIGURE S4. Analysis of the cell dispersity of M. smegmatis porin mutants.

FIGURE S5. Geometry of the W40‐M183 interaction in MspA.

FIGURE S6. Location of methionines in the outer membrane β‐barrel protein OmpX.

TABLE S1. Strains used in this study.

TABLE S2. Plasmids used in this study.

TABLE S3. Oligonucleotides used in this study.

TABLE S4. Statistical analysis of the surface accessibility experiments with MspA mutants.

Pavlenok M, Nair RR, Hendrickson RC, Niederweis M. The C‐terminus is essential for the stability of the mycobacterial channel protein MspA . Protein Science. 2024;33(3):e4912. 10.1002/pro.4912

DATA AVAILABILITY STATEMENT

All data are contained within the manuscript. Plasmids and plasmid maps are available upon request from Mikhail Pavlenok or Michael Niederweis.