Contribution of Phe-7 to Tat-Dependent Export of β-Lactamase in Xanthomonas campestris

Abstract

Strains of Xanthomonas campestris pv. campestris isolated in Taiwan are commonly resistant to ampicillin owing to the constitutive expression of a chromosomally encoded β-lactamase that is secreted into the periplasm. In this study, we found that levels of β-lactamase vary among X. campestris pv. campestris strains, a difference that can be attributed to amino acid substitutions at least at positions 7 and 206, with the former having the major impact. Bioinformatic and PCR analyses indicated that X. campestris pv. campestris possesses tatABC genes and that the signal peptide of X. campestris pv. campestris pre-Bla contains the typical twin-arginine motif (N-R-R-Q-F-L at amino acid residues 3 to 8 in strain X. campestris pv. campestris strain 11), suggesting that Bla is secreted via the Tat pathway. To assess the importance of Phe⁷ in the efficient export of X. campestris pv. campestris Bla, we prepared mutant constructs containing amino acid substitutions and monitored their expression by measuring enzyme activity and detecting Bla protein by Western blotting. The results indicate that replacement of Phe⁷ with Leu severely inhibited Bla export whereas replacement with Pro almost abolished it. Although a change to Arg caused moderate inhibition of export, replacement with Tyr had no effect. These results suggest that for efficient export of Bla by X. campestris pv. campestris, the aromatic-aromatic interactions and stability of protein structure around the twin-arginine motif are important, since only proteins that can attain a folded state in the cytoplasm are competent for export via the Tat pathway.

INTRODUCTION

Production of β-lactamase is the intrinsic mechanism underlying resistance to β-lactam antibiotics in many Gram-negative bacteria. Control of β-lactamase gene (bla) expression is known to involve several regulatory genes, namely, ampR, ampD, ampG, and nagZ; mutations that affect the expression of any one of these regulatory genes can affect the susceptibility of a bacterium to β-lactams (1, 13). In addition, a mutation in the promoter sequence or within a structural gene can also cause changes in resistance patterns (14, 24, 26). β-Lactamase amino acid sequences, including those of the signal sequence, vary widely in Gram-negative bacteria (6). It is known that β-lactamases are secreted into the periplasm after cleavage of the signal sequence. Two pathways have been reported for this translocation, i.e., the general secretory (Sec) pathway characteristic of Escherichia coli enzymes (22) and the twin-arginine translocation (Tat) pathway described for one of the Stenotrophomonas maltophilia β-lactamases (22). Unlike the case of the Sec-dependent pathway, only proteins that can attain a folded state in the cytoplasm are competent for export via the Tat pathway (10, 11).

Xanthomonas campestris pv. campestris is a Gram-negative phytopathogenic bacterium that causes black rot in crucifers (38). In our previous study, we found that strains of X. campestris pv. campestris isolated in Taiwan are commonly resistant to ampicillin (36). To understand the basis of this resistance, we previously cloned and sequenced the responsible bla gene, bla_XCC-1, as well as the upstream ampR1 gene from X. campestris pv. campestris strain 11 (36, 37). These studies showed that (i) bla_XCC-1 is a chromosomal gene encoding a 30-kDa periplasmic enzyme; (ii) the deduced amino acid sequence of bla_XCC-1 reveals identity (51.5%) and conserved domains with S. maltophilia L2 and other Ambler class A/Bush group 2 β-lactamases; (iii) the regulatory gene ampR1, encoding a trans-acting protein (AmpR1) that belongs to the LysR family of bacterial regulators, is located immediately upstream of bla_XCC-1 and transcribed divergently; (iv) AmpR1 is required for the constitutive expression of bla_XCC-1; and (v) homologous β-lactamase and ampR genes are widespread in xanthomonads, as revealed by Southern hybridization using sequence-specific probes. In addition, all ampicillin-resistant Xanthomonas strains tested expressed β-lactamase constitutively, although the levels of β-lactamase activity varied from one strain to another. However, little is known about the translocation of β-lactamase in Xanthomonas.

To understand the residues that play important roles at the level of β-lactamase expression, we cloned bla-ampR systems from various X. campestris pv. campestris isolates and expressed them in the same genetic background. The results showed that the amino acid at position 7 of Bla within the signal sequence could significantly affect the level of active enzyme, reflecting not only the efficiency of protein transport to the periplasm but possibly the stability of the peptide as well.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. Luria-Bertani (LB) medium and L agar (20) were used as general-purpose media for cultivation of X. campestris pv. campestris (28°C) and E. coli (37°C). Media were supplemented with the antibiotic ampicillin (50 μg/ml), kanamycin (50 μg/ml), gentamicin (15 μg/ml), or tetracycline (15 μg/ml), as appropriate.

Table 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Characteristics	Source or reference
E. coli DH5α	endA1 hsdR17 (rk− mk+) supE44 thi-1 recA gyrA relA1 ϕ80dlacZΔM15 Δ(lacZYA-argF)	Bethesda Research Laboratories
X. campestris pv. campestris strains
11	Wild-type strain isolated in Taiwan; Apr	39
17	Wild-type strain isolated in Taiwan; Apr	39
85	Wild-type strain isolated in Taiwan; Apr	39
11Aps	Strain 11 blaXCC-1 mutant constructed by inserting a Tcr cartridge into the SmaI site of the blaXCC-1 gene; Aps Tcr	36
Plasmids
pGEM-T Easy	Used for cloning of PCR fragment; Apr	Promega
pGEM11	pGEM-T Easy carrying complete strain 11 ampR and bla genes; Apr	This study
pGEM17	pGEM-T Easy carrying complete strain 17 ampR and bla genes; Apr	This study
pGEM85	pGEM-T Easy carrying complete strain 85 ampR and bla genes; Apr	This study
pGEM11-F7L	pGEM11 with F7→L7 mutation in bla gene; Apr	This study
pGEM11-F7P	pGEM11 with F7→P7 mutation in bla gene; Apr	This study
pGEM11-F7R	pGEM11 with F7→R7 mutation in bla gene; Apr	This study
pGEM11-F7Y	pGEM11 with F7→Y7 mutation in bla gene; Apr	This study
pXEG	pXV64 derivative carrying ColE1 ori; Gmr	This study
pBR11	pXEG carrying complete strain 11 ampR and bla genes; Gmr	This study
pBR17	pXEG carrying complete strain 17 ampR and bla genes; Gmr	This study
pBR85	pXEG carrying complete strain 85 ampR and bla genes; Gmr	This study
pBR11-F7L	pBR11 with F7→L7 mutation in bla gene; Gmr	This study
pBR11-F7P	pBR11 with F7→P7 mutation in bla gene; Gmr	This study
pBR11-F7R	pBR11 with F7→R7 mutation in bla gene; Gmr	This study
pBR11-F7Y	pBR11 with F7→Y7 mutation in bla gene; Gmr	This study
pET-bla	Created by amplifying the blaXCC-1 gene and cloning it as an NdeI-HindIII fragment into pET-30b	This study

DNA methods.

The primers used in PCRs are listed in Table 2. Preparation of plasmid and chromosomal DNA, restriction enzyme digestion, and transformation of E. coli were carried out by standard procedures (28). Plasmids were delivered into X. campestris by electroporation (34). The TatP 1.0 server (http://www.cbs.dtu.dk/services/TatP/) was used for analysis of signal peptides.

Table 2.

Primers used in this study

Primer	Sequencea	Positionsb
bla	5′-TCACGCGCAGCAGGTGCACC-3′	+922 to +941
ampR	5′-TCAATCGGCGCGCTGCTGTT-3′	+857 to +876
bla-E	5′-CAGAATTCGATGGCGTGTGTTGCAGTC-3′	+961 to +987
ampR-H	5′-GTTTAAGCTTTTTGCGTTGCAGGTGGGGAA-3′	+918 to +947
bla1E	5′-GGAATTCGATGGCGTGTGTTGCAGT-3′	+962 to +986
bla2P	5′-GGTGCCCAGCTGTTGCTCCAATG-3′	+128 to +150
bla1ER	5′-GAACCGACGGCAGTTGCTGTCCATG-3′	+6 to +30
bla2PR	5′-CATGGACAGCAACTGCCGTCGGTTC-3′	+6 to +30
blaF1-FP	5′-GAACCGACGGCAGCCTCTGTCCATGACC-3′	+6 to +33
blaF2-FP	5′-GGTCATGGACAGAGGCTGCCGTCGGTTC-3′	+6 to +33
blaF1-FR	5′-GAACCGACGGCAGAGACTGTCCATGACC-3′	+6 to +33
blaF2-FR	5′-GGTCATGGACAGTCTCTGCCGTCGGTTC-3′	+6 to +33
blaF1-FY	5′-GAACCGACGGCAGTACCTGTCCATGACC-3′	+6 to +33
blaF2-FY	5′-GGTCATGGACAGGTACTGCCGTCGGTTC-3′	+6 to +33
bla-NdeI	5′-AGGCAGCCATATGTTGAACCGACGGCA-3′	−10 to +17
bla-HindIII	5′-CCGAAGCTTACCCACGAGGTCGCGTA-3′	+900 to +925
tatABC-F	5′-AAGCTTGATCATGGGCAGTTTCAG-3′	−10 to +14
tatABC-R	5′-TCTAGACTAGCTTGCCTTCGGCG-3′	+1549 to +1571

^aRestriction enzyme sites (six bases) and mutation sites (three bases) are underlined.

^bRelative to the translational start site of the specific gene or operon named in the primer.

Construction of plasmids.

Plasmid pXEG, derived from 1.6-kb plasmid pXV64 of Xanthomonas pv. vesicatoria (35) by the cloning of a 1.0-kb PstI fragment containing the Gm^r gene from plasmid pX1918GT (30) together with a 1.2-kb PstI fragment containing multiple cloning sites and the ColE1 ori from pBluescript SK⁺ (31), was used as an E. coli-Xanthomonas shuttle vector. This plasmid was capable of autonomous replication in E. coli and Xanthomonas at copy numbers of approximately 500 and 60, respectively.

The 2.0-kb bla-ampR regions, containing both genes and the intergenic region, from X. campestris pv. campestris strains were obtained by PCR amplification. The primer pair bla-E and ampR-H, containing an EcoRI site and a HindIII site, respectively (Table 2), was used for amplification of the X. campestris pv. campestris strain 11 region. The resultant amplicon was cloned into pGEM-T Easy, generating pGEM11. The fragments amplified from X. campestris pv. campestris strains 17 and 85 using primers bla and ampR (Table 2) were also cloned into pGEM-T Easy, giving rise to pGEM17 and pGEM85, respectively. The 2.0-kb EcoRI-HindIII fragment from pGEM11 and the 1.9-kb EcoRI-SpeI fragments from pGEM17 and pGEM85 were then subcloned into the compatible sites of shuttle vector pXEG, forming expression plasmids pBR11, pBR17, and pBR85, respectively.

Site-directed mutation of Bla.

Plasmids pBR11-F7L, pBR11-F7P, pBR11-F7R, and pBR11-F7Y, which express mutant forms of Bla with F→L, F→P, F→R, and F→Y substitutions at position 7, respectively, were derived from pBR11 using a PCR-based strategy. The steps involved were as follows. (i) A 0.98-kb EcoRI-PvuII fragment, designated fragment PCR-1, encompassing bp 6 to 986 downstream of the bla_XCC-1 translational start site was amplified from pBR11 with primers bla1E and bla1ER. (ii) A 0.28-kb fragment, designated PCR-2, straddling nucleotide (nt) 150 downstream of the ampR1 translational start site to nt 30 downstream of the bla_XCC-1 translational start site was PCR amplified from pBR11 using primers bla2P and bla2PR. (iii) After PCR-1 was annealed to PCR-2, the hybrid was extended with TaKaRa EX Taq (TaKaRa Bio Inc.); this was followed by PCR amplification of the extension product using primers bla1E and bla2P. (iv) The resultant 1.2-kb PCR fragment was cloned into the vector pGEM-T Easy, forming pGEM-F7L. The same procedures were followed for the construction of pGEM-F7P, pGEM-F7R, and pGEM-F7Y, except that primer pairs blaF1-FP/blaF2-FP, blaF1-FR/blaF2-FR, and blaF1-FY/blaF2-FY, respectively, were used in place of bla1ER and bla2PR (Table 2). (v) Nucleotide sequencing was performed to verify that the desired substitutions at codon 7 had indeed occurred. (vi) The 1.1-kb EcoRI-PvuII fragment of pBR11 was replaced with the corresponding region from pGEM-F7L, pGEM-F7P, pGEM-F7R, and pGEM-F7Y, generating pBR11-F7L, pBR11-F7P, pBR11-F7R, and pBR11-F7Y, respectively.

DNA sequencing.

Both DNA strands were sequenced by the dideoxy chain termination method (29).

β-Lactamase assays.

Cells of X. campestris pv. campestris from overnight cultures were diluted into fresh LB broth (initial optical density at 550 nm [OD₅₅₀], 0.35), grown to an OD₅₅₀ of about 0.8, and harvested by centrifugation. After sonication, cell extracts were centrifuged and supernatants (cell extracts) were saved and used for Western blotting or as enzyme sources for enzymatic activity assays. The activity of β-lactamase was measured by monitoring the hydrolysis of CENTA, a chromogenic cephalosporin (Merck KGaA, Darmstadt, Germany), at 28°C as previously described (16). The reaction mixture contained cell extract (enzyme source) in 10 mM Tris-HCl (pH 8.0)–1 mM MgCl₂–0.1 mM dithiothreitol–0.1 mM CENTA. Hydrolysis of CENTA was monitored by measuring the increase in OD₄₁₀. One unit of enzyme activity was defined as the amount required to hydrolyze 1 nmol of CENTA/min. Protein content was determined by using Bio-Rad Protein Kit II with crystallized bovine serum albumin as the standard and following the instructions supplied by the manufacturer.

Determination of MICs.

MICs of different β-lactam antibiotics for Xanthomonas strains were determined by the microdilution method recommended by the Clinical and Laboratory Standards Institute, using β-lactams purchased from Sigma (St. Louis, MO).

Cellular fractionation.

Cells of X. campestris pv. campestris were cultured in LB broth containing antibiotics as appropriate. Cells were harvested by centrifugation at 4,000 × g for 20 min, and pellets were resuspended in buffer containing 50 mM Tris (pH 8.0), 18% sucrose, and 1 mM CaCl₂-EDTA. Lysozyme (0.5 mg/ml) was then added, and samples were left on ice for 30 min. After centrifugation at 8,000 × g, supernatants contained periplasmic content, whereas the pellet contained spheroplasts. The resulting spheroplasts were then suspended in 10 mM Tris (pH 8.0) and disrupted by sonication. After centrifugation at 8,000 × g for 10 min, proteins in the supernatant were collected as the cytoplasmic fraction.

Overexpression of recombinant Bla_his.

The bla_XCC-1 gene was amplified and cloned as an NdeI-HindIII fragment into pET-30b (Novagen, Inc., Madison, WI), forming pET-bla. The resulting plasmid, containing a C-terminal His tag, was introduced into E. coli strain BL21, and expression of recombinant protein by transformed E. coli was induced by adding isopropyl-β-d-thiogalactopyranoside (IPTG) to a final concentration of 0.1 mM. The recombinant protein was purified on a His-binding column and checked for purity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Western immunoblotting.

Proteins obtained from identical numbers of cells were separated by SDS-PAGE on 14% gels. Gels were either stained with Coomassie blue dye or transferred to nitrocellulose membranes for Western blotting. Membranes were subsequently incubated with antibodies prepared by immunizing a rabbit with recombinant Bla_his.

Nucleotide sequence accession numbers.

The bla-ampR sequences of X. campestris pv. campestris strains 11, 17, and 85 have been deposited in GenBank under accession numbers JN935203, JN935204, and JN935202, respectively.

RESULTS AND DISCUSSION

Levels of β-lactamase vary among X. campestris pv. campestris strains.

It has previously been shown that strains of Xanthomonas isolated in Taiwan are commonly resistant to ampicillin at a level of 50 μg/ml (36). In this study, to measure the β-lactamase levels of different strains of X. campestris pv. campestris, we assayed cell extracts of three Taiwan isolates, X. campestris pv. campestris strains 11, 17, and 85, for enzymatic activity. Results showed that the β-lactamase activities, measured as hydrolysis of the synthetic substrate CENTA, of X. campestris pv. campestris strains 11 and 17 were similar (220 and 225 nmol/min/mg protein, respectively) and approximately 43% lower than that of strain 85 (390 nmol/min/mg protein).

To determine the mechanistic basis of variations in β-lactamase levels in different X. campestris pv. campestris strains, we cloned and sequenced the PCR fragments containing the complete ampR and bla genes plus the ampR-bla intergenic region from strains 11, 17, and 85. Sequence analysis revealed that (i) all of the Bla and AmpR proteins encoded were 296 and 291 amino acid (aa) residues, respectively, instead of 295 and 289 aa as previously determined for strain 11 (36); (ii) all ampR and ampR-bla intergenic regions (100 bp) were identical in nucleotide sequence; and (iii) two amino acid substitutions occurred in Bla: one at aa 7 within the signal peptide, where strains 11 and 17 had a Phe residue but strain 85 had a Leu residue; and the other at aa 206, where strain 11 had a His residue but strains 17 and 85 had a Gln residue. These results indicate that differences in Bla levels among strains may have resulted from differences in the Bla signal sequence, coding region, or both, since all three ampR genes and the promoters of ampR and bla, both of which are located in the ampR-bla intergenic region, had identical sequences.

X. campestris pv. campestris Bla appears to be secreted via the Tat-dependent pathway.

During our recent study of phage export, we isolated an xpsD mutant derivative of strain 17 that was deficient in type IV secretion, a Sec-dependent pathway. This mutant, designated strain 17-xpsD::Gm, was found to be deficient in the secretion of several extracellular enzymes that depend on the Sec pathway but to retain a wild-type level of ampicillin resistance (50 μg/ml) (15). This result indicated that secretion of X. campestris pv. campestris Bla is not Sec dependent. A bioinformatic analysis of β-lactamase signal sequences suggested that Tat-dependent translocation of β-lactamases is more likely to occur in plant bacteria such as Xanthomonas, Stenotrophomonas, or Mesorhizobium than in clinical bacteria (22). Consistent with this, our preliminary results showed that the MIC of ampicillin was only 8 μg/ml for the tatC mutant of strain 11 (17). Taken together, these findings indicate that secretion of Bla is Tat dependent in X. campestris pv. campestris.

Bacterial Tat-dependent signal peptides are similar to Sec-type signal peptides in that both contain a positively charged amino terminus (N region), a hydrophobic core (H region), and a carboxy-terminal domain (C region) containing the signal peptidase cleavage site. Several features of the bacterial Tat-dependent signal peptides are notable. (i) A twin-arginine motif (Z-R-R-x-ϕ-ϕ) is located in the N region of the Tat signal sequence, where Z is any polar residue, ϕ is a hydrophobic residue, and x is any residue (3, 21). (ii) The H region of Tat signal peptides typically is slightly longer, containing 13 to 20 uncharged residues, and less hydrophobic than the H region of Sec substrates. (iii) Twin-arginine signal peptides often have one or more positively charged residues, called Sec avoidance signals, in the C-terminal region just upstream of the signal peptidase cleavage site (4). Replacement of either arginine residue in the twin-arginine motif can have effects ranging from a modest reduction in the export rate to a complete export block, depending on the individual signal peptide and the specific translocation system involved (19). Moreover, differences in amino acids within the signal sequence also cause variations in the level of resistance to β-lactams (25). Thus, in addition to primary structure and expression, the rate of translocation into the periplasm can also modulate the level of resistance conferred by a particular β-lactamase gene. A sequence analysis of the complete ampR-bla regions from X. campestris pv. campestris strains 11, 17, and 85 using the TatP program revealed the presence of a 25-aa signal peptide in the Bla protein from each strain. At aa 3 to 8 within the signal peptide, both X. campestris pv. campestris strains 11 and 17 had N-R-R-Q-F-L and strain 85 had N-R-R-Q-L-L, which matched the twin-arginine motif (3, 21). Compared to other bacteria, in which the N region of Tat signals is typically longer than that of Sec signals (8, 21, 32), X. campestris pv. campestris Bla has an extremely short N region, with the twin arginine residues being located 3 residues from the N terminus. Two residues upstream of the end of the C region (A-W-A) was a positively charged arginine residue, an arrangement similar to that of a Sec avoidance signal.

Xanthomonas oryzae pv. oryzae has been experimentally shown to possess a twin-arginine translocation system (7). The genomic sequences of three X. campestris pv. campestris strains have been determined (9, 23), whereas that of strain 17 was partially completed (http://xcc.life.nthu.edu.tw) while this report was being prepared; a tat operon (tatABC) was found to be present in each of these genomes. Using a strain 17-specific primer pair (Table 2) complementary to the flanking regions of the tat operon, we were able to amplify a 1.6-kb fragment from the strain 11, 17 and 85 genomes, confirming the presence of a tat operon in these strains (17). These findings, taken together with the presence of a conserved bacterial Tat-dependent signal peptide, suggest that X. campestris pv. campestris Bla proteins are processed and translocated to the periplasm via the Tat pathway.

Expression of β-lactamase by bla-ampR constructs from different X. campestris pv. campestris strains.

To investigate the effects of sequence difference on the levels of Bla production in the same genetic background, we cloned the entire ampR-bla sequences of the three strains as described in Materials and Methods and expressed them in strain 11Ap^s, an ampicillin-sensitive mutant (36). E. coli transformants carrying the resulting plasmids (pGEM11, pGEM17, and pGEM85) showed no β-lactamase activity, a result similar to our previous finding that the promoters of bla_XCC-1 and ampR1are inactive in E. coli (37). The levels of enzyme expressed in strains 11Ap^s(pBR11), 11Ap^s(pBR17), and 11Ap^s(pBR85) were determined by assaying β-lactamase activity in cell extracts prepared from these strains. The β-lactamase activities of strains 11Ap^s(pBR11) and 11Ap^s(pBR17) were 2,045 and 2,475 nmol CENTA/min/mg protein, respectively; these levels were 10 and 12 times that expressed by strain 11, apparently because of higher copy numbers of the cloned bla genes (Table 3). The approximately 21% difference between the levels expressed by strains 11Ap^s(pBR17) and 11Ap^s(pBR11) suggests that the Gln residue at position 206 in Bla from strain 17 results in better enzyme performance than the His found in the corresponding position of Bla from strain 11. Notably, the β-lactamase activity expressed by strain 11Ap^s(pBR85) was only 48.5% of that expressed by strain 11Ap^s(pBR11) (Table 3). Thus, whereas a His²⁰⁶→Gln substitution (pBR11 versus pBR17) caused a minor effect, the Phe⁷→Leu substitution (pBR17 versus pBR85) appeared to have a major impact on the β-lactamase expression level of X. campestris pv. campestris, indicating that the properties of the amino acid at residue 7 are functionally important. The fact that strain 11Ap^s(pBR85) exhibited the lowest β-lactamase activity of the three constructs was surprising, since the enzyme level of strain 85 was higher than that of strains 11 and 17 (see above). Thus, it is possible that another factor(s) is involved in the regulation of bla expression in X. campestris pv. campestris.

Table 3.

Effects of substitutions at aa 7 and 206 of X. campestris pv. campestris Bla on levels of β-lactamase activity^a

Strain	β-Lactamase
	Amino acid sequence	Mean enzyme activityb ± SEM	Relative activity (%)
11	Phe7-His206c	220 ± 60	10.8
11Aps	—d	20 ± 10	1.0
11Aps(pBR11)	Phe7-His206c	2,045 ± 430	100.0
11Aps(pBR17)	Phe7-Gln206c	2,475 ± 580	121.1
11Aps(pBR85)	Leu7-Gln206	990 ± 230	48.5
11Aps(pBR11-F7L)	Leu7-His206	500 ± 110	24.5
11Aps(pBR11-F7P)	Pro7-His206	110 ± 40	5.4
11Aps(pBR11-F7R)	Arg7-His206	890 ± 170	43.6
11Aps(pBR11-F7Y)	Tyr7-His206	2,185 ± 540	106.9

^aβ-Lactamase activity was measured in strain 11Ap^s, an ampicillin-sensitive mutant derived from strain 11, carrying the wild-type bla gene from strain 11, 17, or 85 and in strain 11 Bla with a substitution at aa 7. All of these bla genes were maintained in plasmid form. Strains 11 and 11Ap^s were used as controls.

^bSpecific activity expressed in nanomoles of CENTA hydrolyzed per minute per milligram of protein.

^cWild-type sequence.

^d—, the bla gene was disrupted.

Phe⁷ of X. campestris pv. campestris Bla is important for enzyme export.

To assess the importance of aa 7 in X. campestris pv. campestris Bla in controlling the level of β-lactamase activity, we constructed four mutant Bla proteins by site-directed mutagenesis of pBR11 to change Phe⁷ to Leu⁷, Pro⁷, Arg⁷, or Tyr⁷ while retaining His²⁰⁶. A functional analysis of mutant Bla proteins, summarized in Table 3, revealed that (i) a conservative Phe⁷→Tyr substitution did not affect β-lactamase levels, (ii) a Phe⁷→Arg substitution resulted in a 56% reduction in β-lactamase activity, (iii) a Phe⁷→Leu substitution reduced enzymatic activity by 75%; and (iv) a Phe⁷→Pro substitution almost abolished β-lactamase activity. Notably, the combination of Leu⁷ and Gln²⁰⁶ in pBR85 retained higher activity than the combination of Leu⁷ and His²⁰⁶ in pBR11-F7L (48.5% versus 24.5%), supporting the interpretation that Gln may function better than His at position 206. To observe the subcellular distribution of X. campestris pv. campestris Bla proteins, cells harboring different X. campestris pv. campestris bla-ampR constructs were fractionated and the fractions were separately resolved by SDS-PAGE and analyzed by Western blotting. Results of the Western blot analyses are shown in Fig. 1 (A, cell extract; B, periplasm; C, cytosol) and summarized as follows. (i) As shown in Fig. 1A, two bands were observed in samples from strain 11Ap^s(pBR11) carrying the wild-type bla gene (lane 3) and in strains 11Ap^s(pBR11-F7L) (lane 4), 11Ap^s(pBR11-F7R) (lane 6), and 11Ap^s(pBR11-F7Y) (lane 7) producing mutant bla gene products. The two bands together represent the total bla gene products synthesized by the bacteria. The fast-moving bands were apparently mature Bla proteins, as they exhibited the same mobility as mature Bla from strain 11 (Fig. 1B, lane 1) and their molecular mass (∼29 kDa) was similar to that calculated from the predicted amino acid sequences. The slow-moving bands with slightly higher molecular masses appeared to be the pre-Bla proteins. (ii) No bla gene product was detected in bla mutant strain 11Ap^s, as expected (lane 2). (iii) The amount of Bla in strain 11Ap^s(pBR11-F7L) (Fig. 1B, lane 4) was about twice that in strain 11 (Fig. 1B, lane 1), consistent with the observed ∼2-fold increase in β-lactamase activity compared to that of strain 11 (Table 3). Compared to strains 11Ap^s(pBR11), 11Ap^s(pBR11-F7R), and 11Ap^s(pBR11-F7Y), very low levels of total bla products were observed in strain 11Ap^s(pBR11-F7L) (Fig. 1A, lanes 3, 4, 6, and 7). However, more Bla was detected in the periplasm of strain 11Ap^s(pBR11-F7L) than in that of strain 11Ap^s(pBR11-F7R) (Fig. 1B, lanes 4 and 6). (iv) Strains 11Ap^s(pBR11) and 11Ap^s(pBR11-F7Y) produced much larger amounts of the bla gene products than did strain 11 (Fig. 1A, lanes 1, 3, and 7), consistent with their 10-fold higher levels of β-lactamase activity than that of strain 11 in enzyme assays (Table 3). A large amount of Bla was also produced by strain 11Ap^s(pBR11-F7R), although it was slightly less than those produced by strains 11Ap^s(pBR11) and 11Ap^s(pBR11-F7Y) (Fig. 1A, lane 6). (v) In strain 11, only mature Bla was detected; the major portion was found in the periplasm, and only trace amounts were detected in the cytoplasm (Fig. 1A, B, and C, lanes 1). In contrast, in strains 11Ap^s(pBR11), 11Ap^s(pBR11-F7R), and 11Ap^s(pBR11-F7Y), the amounts of bla products found in the cytoplasm were larger than those found in the periplasm (Fig. 1C, lanes 3, 6, and 7); in strains 11Ap^s(pBR11-F7R) and 11Ap^s(pBR11-F7Y), most of them were retained as pre-Bla in the cytosol (Fig. 1C, lanes 6 and 7). It was also noted that about equal amounts of Bla and pre-Bla were retained in the cytosol of strain 11Ap^s(pBR11) (Fig. 1C, lane 3). The retention of mature proteins in the cytosol indicated that signal peptide cleavage occurred while pre-Bla was still inside the inner membrane. This also suggested that export had been a limiting factor after cleavage. The proteins retained in the cytosol of strain 11Ap^s(pBR11-F7R) were mostly pre-Bla, suggesting that the processing step instead of export had been the bottleneck resulting in the accumulation of these preproteins in the cytosol. These observations suggest that the bla gene product of strain 11 at wild-type levels can be appropriately processed and translocated but overexpressed bla products produced by multiple copies of the cloned genes overwhelm the capacity of the system. Since only proteins that can attain a folded state in the cytoplasm are competent for export via the Tat pathway (10, 11), the overexpressed bla products may not be folded efficiently for export. In E. coli, rapid induction of TorA-green fluorescent protein often results in the accumulation of misfolded, inactive protein, which binds to the membrane and is slowly degraded (2). Degradation of unprocessed Tat substrates in X. campestris pv. campestris may not be as efficient as in E. coli, since large amounts of pre-Bla were found to be retained in the cytoplasm of X. campestris pv. campestris. (vi) In strain 11Ap^s(pBR11-F7P), which exhibited an extremely low level of β-lactamase activity in an enzyme assay (Table 3), pre-Bla and Bla proteins were almost undetectable (Fig. 1A, B, and C, lanes 5). There are two possible reasons for this phenomenon, namely, that the mutant pre-Bla was somehow more efficiently degraded and that the synthesis of this Phe⁷→Pro mutant bla product was greatly reduced. Whether the latter is the case remains to be tested, but it is reasonable to predict that a Phe⁷→Pro substitution, which not only eliminates the phenyl ring of Phe⁷ but also causes destabilization of the protein structure, can cause a severe denaturation of the proteins, since proline can destabilize α-helices and β-strands by introducing kinks and turns into these structures (27). The severely denatured proteins, which become excellent substrates for degradation, were then subject to rapid degradation (12).

Fig 1.

Effect of residue 7 on Bla protein transport in X. campestris. Cells were grown in LB medium to an OD₅₅₀ of 0.8, and bacterial pellets were fractionated into cell extract (A), periplasm (B), and cytoplasm (C) as described in Materials and Methods. Proteins in each fraction were resolved by SDS-PAGE and analyzed by Western blotting using anti-Bla antibodies containing a C-terminal His tag. Lanes: 1, wild-type strain 11; 2, bla mutant strain 11Ap^s; 3, strain 11Ap^s(pBR11) carrying the cloned wild-type bla gene; 4, strain 11Ap^s(pBR11-F7L); 5, strain 11Ap^s(pBR11-F7P); 6, strain 11Ap^s(pBR11-F7R); 7, strain 11Ap^s(pBR11-F7Y). Lanes 4 to 7 contained samples from strains carrying cloned mutant bla genes expressing Leu (L), Pro (P), Arg (R), and Tyr (Y), respectively, at Bla residue 7.

The consensus Phe residue after the invariant Arg residues (called Phe⁺² by some researchers) is the most conserved amino acid of the bacterial twin-arginine motifs, being found in approximately 80% of Tat signal peptides (33). At least three twin-arginine motifs with Phe⁺² have been tested for amino acid substitution, namely, those of SufI and TorA in E. coli and xylanase C (XlnC) in Streptomyces lividans, and various degrees of effect on protein export have been observed (18, 19, 33). It has been known that aromatic-aromatic side chain interactions play an important role in the stabilization of protein structures (5). This notion at least partly explains our observation that replacement of Phe⁷ (equivalent to +2 in the case of E. coli) with nonaromatic amino acids reduced Bla export to various degrees, suggesting that the aromatic-aromatic interactions and protein structure stability around the twin-arginine motif are important for efficient export of Bla by X. campestris pv. campestris.