|Home | About | Journals | Submit | Contact Us | Français|
Free fatty acids are important antimicrobial substances regulating the homeostasis of colonizing bacteria on epithelial surfaces. Here, we show that meningococci express a functional farAB efflux pump, which is indispensable for fatty acid resistance. However, other than in Neisseria gonorrhoeae, the transcriptional regulator FarR is not involved in regulation of this operon in Neisseria meningitidis. We tested the susceptibility of 23 meningococcal isolates against saturated and unsaturated long-chain fatty acids, proving that meningococci are generally highly resistant, with the exception of serogroup Y strains belonging to sequence type 23. Using genetically determined lipopolysaccharide (LPS)-truncated mutant strains, we show that addition of the LPS core oligosaccharide and hexa-acylation of its membrane anchor lipid A are imperative for fatty acid resistance of meningococci. The sensitivity of the serogroup Y strains is due to naturally occurring mutations within the lpxL1 gene, which is responsible for addition of the sixth acyl chain on the LPS membrane anchor lipid A. Therefore, fatty acid resistance in meningococci is provided by both the active efflux pump FarAB and by the natural permeability barrier of the Gram-negative outer membrane. The transcriptional regulator FarR is not implicated in fatty acid resistance in meningococci, possibly giving rise to a constitutively active FarAB efflux pump system and thus revealing diverse mechanisms of niche adaptation in the two closely related Neisseria species.
Although humans are physiologically colonized by a vast number of microorganisms, invasion of this microflora is efficiently prevented by dermal and mucosal epithelia which form the barriers of the human body. Homeostasis of the colonizing microflora at these barriers is maintained by numerous mechanisms that include the secretion of antimicrobial compounds by the epithelia. Among these compounds are free fatty acids, which act as nonselective antimicrobials against a broad range of microorganisms. Sebaceous glands, producing a mixture of lipids called sebum, are present in the skin, the nasal epithelium, and the oral epithelium of most adults (62, 71). The produced sebum consists mainly of triglycerides which undergo hydrolysis by both host and bacterial lipases to produce free fatty acids (9). In human saliva, lipid content is about 4 to 20 mg/100 ml and includes both saturated (mainly palmitic and stearic acid) and unsaturated (mainly oleic, linoleic, palmitoleic, arachidonic, and docosahexaenoic acid) fatty acids (59, 65). Similarly, fatty acids form a major constituent of vaginal secretions (45). In most cases bacterial susceptibility to free fatty acids depends on organization of the cell wall, with Gram-negative bacteria being more resistant than Gram-positive bacteria (72). Enterobacteriaceae are shielded by their lipopolysaccharide (LPS) from the antimicrobial activity of medium- and long-chain fatty acids (58). Neisseria meningitidis and Neisseria gonorrhoeae are closely related Gram-negative species which—in spite of their similar LPS structures—display strongly differing levels of resistance to antimicrobial fatty acids. N. meningitidis is an exclusive human commensal that is generally highly resistant to free fatty acids and that colonizes the nasopharynx of up to 10% of the healthy population. In rare instances the bacteria can cause devastating invasive infection, resulting in sepsis and meningitis, mostly in young infants and toddlers. N. gonorrhoeae is a human pathogen responsible for genital tract infection (gonorrhea) as well as rare cases of disseminated disease and is susceptible to fatty acids (16, 35, 39, 54). Therefore, gonococci rely on two efflux pump systems for survival on the urogenital mucosa: first, the MtrCDE system exporting hydrophobic agents like bile salts, antibacterial peptides (protegrin-1 and LL-37), and steroidal hormones (16-18, 41, 46) and, second, the FarAB-MtrE system responsible for the export of long-chain fatty acids (34, 37). As the expression of such elaborate efflux systems is energy draining, they are kept under strict transcriptional control (43). Indeed, both systems have been shown to be repressed by the transcriptional regulator MtrR, the mtrCDE system directly (18, 19), and the farAB-mtrE-system indirectly via MtrR-dependent repression of the repressor farR (33, 34). The fatty acid resistance regulator FarR, in turn, binds directly and specifically to the farAB promoter region in both Neisseria species yet assumes a different function in meningococci (33, 55).
Here, we show that the FarAB efflux pump system is functional in N. meningitidis, providing resistance against saturated long-chain fatty acids. Contrary to the findings in N. gonorrhoeae, this efflux pump is not negatively controlled by the transcriptional regulator FarR (34, 55). Additionally, intact LPS is important for a functional FarAB system as the presence of the inner core oligosaccharide as well as hexa-acylation of lipid A was required to maintain protection against palmitic acid.
MC58, an N. meningitidis serogroup B clinical isolate (B:15:P1.7,16), was kindly provided by E. R. Moxon (38). The H44/76 (B:15:P1.7,16) strain used in this study was provided by T. Tønjum (21). The lpxA-deficient H44/76 strain was a kind gift from J. Tommassen (61). Most clinical isolates of N. meningitidis and N. gonorrhoeae were allocated by the National Reference Center for Meningococci (NRZM, Würzburg, Germany); additional clinical isolates of N. gonorrhoeae were provided by the clinical diagnostics department of the Institute of Hygiene and Microbiology (IHM, Würzburg, Germany) (Table (Table1).1). Strains were stored at −80°C, plated on GC agar (BD Difco, Germany) supplemented with PolyVitex (bioMérieux, France), unless indicated otherwise, and grown at 37°C in 5% CO2.
For construction of the ΔlptA mutant strains (Table (Table2),2), parental MC58 ΔsiaD (14), MC58 ΔsiaD ΔlgtA (51), and MC58 ΔsiaD ΔkdtA strains were transformed with plasmid pCS46 harboring a spectinomycin resistance cassette within the open reading frame (ORF) of the lptA gene NMB1638. For construction of plasmid pCS46, a 2-kb fragment of the NMB1638 ORF was amplified from genomic DNA of strain MC58 with oligonucleotides OK178 (5′-TCAATAATCCGGATTCCAAATG) and OK179 (5′-TAACCATCGGCGCCATATTG) and cloned into pCRII blunt (Invitrogen, Germany). The fragment was digested with EcoRI and cloned into pTL1 (67). The resulting plasmid was linearized with inverted PCR using primers OK229 (5′-TGGTCCCTAGGTACGACAATACGGTTTTGTATG) and OK230 (5′-AAAAACCTAGGGTAATCCTGATATTGCAACATTG) and digested with AvrII (restriction sites are underlined in sequences). A nonpolar promoterless spectinomycin resistance cassette was amplified from plasmid pHP45 (49) with OK6 (5′-CCTAGGTTATTTGCCGACTACCTTGGT) and OK7 (5′-CCTAGGATGCGCTCACGCAACTGGT) and inserted via the AvrII sites to obtain pCS46. Inactivation of NMB1638 was confirmed by PCR, DNA sequencing, and increased sensitivity to polymyxin B. For construction of the kdtA deletion mutants, parental MC58 ΔfarR strain (55) was transformed with plasmid pCS32 harboring a spectinomycin resistance cassette within the ORF of the kdtA gene NMB0014, and parental MC58 ΔsiaD strain (31) was transformed with plasmid pCS33 containing a kanamycin resistance cassette within the NMB0014 ORF. For construction of plasmids pCS32 and pCS33, a 2-kb fragment of the NMB0014 ORF was amplified by PCR from genomic DNA of strain MC58 with primers OK110 (5′-TTGGGGCTGTCTTACGACG) and OK111 (5′-AACTCGGATACGGATTGTATC) and cloned into pCRII blunt. The fragment was digested with EcoRI and cloned into pTL1. The resulting plasmid was linearized with inverted PCR using primers OK116 (5′-TGGGCCTAGGGTATGCGTATTACCTGTGCG) and OK117 (5′-CAGCCCTAGGCCGCGATATTGTTGCCACG), and the aforementioned spectinomycin cassette was inserted via the AvrII sites to obtain pCS32. For pCS33, a kanamycin resistance cassette was amplified from plasmid pUC4K (GenBank accession number X06404) with primers OK227 (5′-GGGGAATTCCTAGGCCACGTTGTGTCTCAAAATC) and OK228 (5′-GCTGAATTCCTAGGCGTGAAGAAGGTGTTGC) and inserted instead of the spectinomycin cassette. Inactivation of NMB0014 was confirmed by PCR and DNA sequencing. LPS immunotype was analyzed by enzyme-linked immunosorbent assay (ELISA), or the absence of the LPS oligosaccharide chain was demonstrated by Tricine gel analysis. For construction of the lpxL1 deletion strain, the lpxL1 gene (NMB1418) was amplified with 500 bp flanking both sides using the oligonucleotides lpxL1.3 (5′-GCGCGCGAATTCGATCATTTTTACGTCGCCTC) and lpxL1.4 (5′-GCGCGCGAATTCGCCATTTTCTACGCTTTGC). The PCR product was digested with EcoRI and cloned into pTL1 to obtain pTL1-lpxL1. An inverse PCR with the oligonucleotides lpxL1.5 (5′-CTGCCGCCTAGGACCTGCATCCCCGAAGGATT) and lpxL1.6 (5′-AAGACA CCTAGGGCGAACATCCGGAACAATAT) generated a linearized product which was digested with AvrII and ligated with the spectinomycin cassette as described above. The resulting plasmid was used for transformation of the parental MC58 ΔsiaD strain (31). Inactivation of NMB1418 was confirmed by PCR and direct sequencing. For complementation of the 10-bp deletion within the lpxL1 gene in strain α 24, the parental strain was transformed with the above-mentioned plasmid pTL1-lpxL1, and transformants were picked and screened for complementation by PCR using the oligonucleotides lpxL1.7 (5′-CCGTGTTGAAACA) and lpxL1.8 (5′-AGGGTAGAAATGC). Of the candidate complemented strains, the lpxL1 gene was amplified and sequenced with the oligonucleotides lpxL1.3 and lpxL1.4. For construction of the farAB deletion strains, a 1,700-bp fragment of the farAB operon (NMB0318/0319) was amplified using the oligonucleotides 306 (5′-GCGCGCGAATTCGACGATAATGATGTGCTGGC) and 307 (5′-GCGCGCGAATTCCAATGACGCCGACGGTAAAA). The PCR product was digested with EcoRI and cloned into pTL1, resulting in the plasmid pTL1-farAB. An inverse PCR with the oligonucleotides 302 (5′-GTCGCACCTAGGGTTAATGGCATCCTATCCGC) and 303 (5′-GGACGGCCTAGGGCTGCACCACTTTAATCCAG) generated a linearized product which was digested with AvrII and ligated with the spectinomycin cassette as described above. The resulting plasmid pTL1-farAB-SmR was used for transformation of the parental MC58 and MC58 ΔsiaD ΔkdtA strains. Inactivation of farAB was confirmed by PCR and direct sequencing. All mutant strains were tested for expression of Opa, Opc, and pili by Western blot analysis.
The fatty acid resistance was estimated by a modified efficiency-of-plating (EOP) analysis as described earlier (55). Briefly, 500 bacteria were plated for each strain in triplicate onto GC agar plates with or without supplementation of 150 μg/ml (585 μM) palmitic acid (Sigma, Germany) or 2.8 μg/ml (10 μM) linoleic acid (Sigma, Germany) and incubated overnight at 37°C with 5% CO2. CFU of at least three independent experiments were counted, and EOP values were calculated by dividing the number of CFU on the fatty acid plates by the number of CFU on the control plates. For presentation, resistance was expressed in terms of percentages by multiplying EOP values by 100. A two-tailed unpaired Student's t test was used to calculate statistical significance (P values).
The ability of palmitic acid to penetrate the cell wall and accumulate at the cytoplasmic membrane was measured by a fluorescence assay (27). Bacteria grown overnight on GC agar plates were resuspended in phosphate-buffered saline (PBS) and adjusted to an optical density at 600 nm (OD600) of 0.3 in 1 ml. Fluorescently labeled palmitic acid (BODIPY FL C16; 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-hexadecanoic acid) was obtained from Invitrogen. Bacterial suspension (100 μl) was mixed with 0.5 μl of BODIPY FL C16 (1 μg/μl) and incubated for 5 min on ice, washed, and fixed for 60 min in 1% formaldehyde. Fluorescence of 25,000 bacteria per experiment was measured in a flow cytometer at 530 nm (FACSCalibur; BD Biosciences).
For sequencing of the lpxL1 gene (NMB1418), a PCR using the oligonucleotides lpxL1.1 (5′-AATCCTTCGGGGATGCAGGT) and lpxL1.2 (5′-TTTCGACTCGAAACGCCTGA) was performed to amplify the gene from cell lysates of N. meningitidis. PCR products were purified and subjected to nucleotide sequencing.
Sequences were analyzed using BioEdit, version 126.96.36.199 (20), and multiple alignments were facilitated with ClustalW (64). Schematic models and chemical structures were drawn with ACD/ChemSketch Freeware Product, version 12.01 (ACD/Labs) and Adobe Illustrator CS, version 11.0.0 (Adobe Systems Inc.).
We previously reported a high intrinsic fatty acid resistance of meningococci to palmitic acid (55). To test whether this high degree of resistance is spread throughout the whole species, we chose to examine a variety of meningococcal strains including invasive and carrier isolates of both hypervirulent and nonvirulent clonal complexes in addition to several reference strains. Therefore, 23 strains (Table (Table1)1) covering the serogroups A, B, C, 29E, W-135, Y, Z, and the capsule null locus (cnl) as well as various sequence types (ST) were tested for their resistance against the long-chain saturated palmitic acid (hexadecanoid acid, C16:0) and the unsaturated linoleic acid (9,12-octadecadienoic acid, C18:2Δ9Δ12). The fatty acid resistance was estimated by an EOP analysis, as described previously (55). Consistent with our earlier findings, nearly all meningococcal isolates displayed a high intrinsic resistance between 80 and 100% against both palmitic and linoleic acid (Fig. (Fig.1).1). Interestingly, we could identify three strains of serogroup Y belonging to ST-23 with unusual sensitivity to palmitic acid. The percentage of resistance for these three strains ranged from 0 to 16% (average, 10% ± 9%) in contrast to 76 to 107% (mean, 92% ± 9%) for all other strains (P < 0.01). Despite this sensitivity, these strains were highly resistant against linoleic acid, showing that resistance against these two compounds is not correlated. The susceptibility of these strains was not due to expression of the serogroup Y capsule as another serogroup Y strain (ST-166) showed full resistance against palmitic acid.
It has been shown before that the antimicrobial activity of long-chain fatty acids is mainly restricted to Gram-positive bacteria (13, 26). This difference may result from the composition of the outer membrane of Gram-negative bacteria, which is effectively active against hydrophobic substances (15, 58). Indeed, our results indicate that the high intrinsic resistance of meningococci is neither due to serogroup nor correlated with virulence. Likewise, expression of capsular polysaccharides was shown to be insignificant since the wild-type serogroup B strain MC58 (B:15:P1.7,16) was as resistant as the constitutively unencapsulated isogenic MC58 ΔsiaD strain against both palmitic and linoleic acid (see Fig. Fig.3).3). It has been shown for Escherichia coli and Salmonella enterica serovar Typhimurium that LPS shields these Gram-negative bacteria from the antimicrobial activity of fatty acids (58). However, the LPS of the Neisseria species is significantly different from enterobacteriaceal LPS. It is anchored in the outer bacterial membrane by lipid A, and two heptoses (Hep I and Hep II) branch from KDO I (2-keto-3-deoxyoctulosonic acid) carrying the short oligosaccharide side chains (47, 70). A structural model of the LPS of the N. meningitidis strain MC58 (serogroup B, ST-32, immunotype L3) integrating recent research and nuclear magnetic resonance (NMR) analysis (28, 47, 50, 66, 75) is depicted in Fig. Fig.2.2. As meningococci are—in contrast to other Gram-negative bacteria—viable without LPS, we tested a ΔlpxA mutant, which is not able to produce any LPS (61). This strain was not able to grow on plates supplemented with palmitic acid (Fig. (Fig.3A),3A), indicating a complete loss of the intrinsic resistance. The growth on linoleic acid was significantly reduced but not completely abolished (Fig. (Fig.3B).3B). Next, we performed stepwise truncations of the oligosaccharide moiety of the LPS. The sites of these truncations are indicated with the respective gene names in the scheme (Fig. (Fig.2).2). Neither deletion of the phosphoethanolamine (PEA) residue of lipid A in a ΔlptA mutant strain nor truncation of the α-chain behind the galactose residue in a ΔlgtA mutant strain led to fatty acid susceptibility. However, deletion of both of these genes entailed a 50% reduction in resistance against palmitic acid. Finally, deletion of the whole carbohydrate structure branching from lipid A in a ΔkdtA mutant strain eliminated the palmitic acid resistance to a minimum and, in combination with deletion of lptA, reversed it completely, whereas the resistance against linoleic acid was not influenced. In conclusion, meningococcal resistance to palmitic acid is dependent on its LPS as a ΔlpxA mutant strain was highly susceptible. More precisely, it was sufficient to delete the whole carbohydrate moiety of the lipooligosaccharide (ΔkdtA) to render meningococci sensitive to palmitic acid.
As none of the LPS modifications conferring susceptibility to fatty acids is likely to occur accidentally in wild-type meningococci, they cannot explain the unusual susceptibility of the serogroup Y, ST-23 strains. It has recently been published that a high percentage of meningococcal isolates carry mutations in the lpxL1 gene leading to penta-acylated lipid A (12). As shown in Fig. Fig.2,2, normal meningococcal lipid A is anchored in the outer membrane via six acyl chains. The lpxL1 gene is responsible for addition of the last lauroyl chain to the 2′ end of the glucosamine disaccharide (68). To check the 23 meningococcal strains tested during the fatty acid susceptibility assays (Table (Table1),1), we sequenced the lpxL1 gene of each strain. Indeed, each of the three serogroup Y, ST-23 strains with increased susceptibility to palmitic acid harbored one of the seven described inactivating mutations (12) (Fig. (Fig.4A).4A). Whereas strain DE 6853 contained a type V mutation, a deletion of an adenosine in a stretch of seven adenosines, strains α 24 and DE 7671 had a deletion of 10 nucleotides and thus a type VI mutation (12). Unlike the findings by Fransen and colleagues, the latter two deletion mutants had one point mutation changing the adenosine at position 707 to a guanosine, and the 10-bp deletion was shifted one nucleotide downstream. In contrast, no mutations were found within the lpxL1 gene of the 20 palmitic acid-resistant isolates. To account for the possibility that these mutations and thus the penta-acylated lipid A may be responsible for the unusual sensitivity to palmitic acid, we generated an lpxL1 deletion mutant in the unencapsulated MC58 ΔsiaD strain and performed susceptibility assays. Indeed, the absence of the sixth acyl chain totally abolished the resistance against the C16 fatty acid (Fig. (Fig.4B).4B). In contrast, no effect on susceptibility to linoleic acid could be observed. In addition, we complemented the type VI mutation in strain α 24 by reintegration of the wild-type lpxL1 copy into the native chromosomal locus. The resulting strain showed fully restored resistance against palmitic acid (Fig. (Fig.4B),4B), proving that an intact lpxL1 gene and thus hexa-acylated lipid A molecules are important for the resistance of meningococci to palmitic acid. The susceptibility to fatty acids in Staphylococcus aureus is correlated with an increased binding of palmitic acid to the cytoplasmic membrane (29). To test whether accumulation of palmitic acid on the bacterial cell wall or incorporation into the cytoplasmic membrane is responsible for the increased palmitic acid susceptibility due to penta-acylated lipid A, we performed a fluorescence assay (27) using BODIPY-labeled palmitic acid. However, analysis by flow cytometry resulted in no different fluorescence acquisition of the MC58 ΔsiaD parental or the MC58 ΔsiaD ΔlpxL1 strain (Fig. (Fig.4C).4C). Consequently, hexa-acylation of lipid A is necessary for the intrinsic resistance of meningococci against palmitic acid irrespective of C16 uptake.
As yet, our results indicate that meningococci display a high intrinsic resistance to long-chain fatty acids that relies on the carbohydrate composition of the LPS and independently also on the acylation of lipid A. In gonococci, fatty acid resistance is conferred by the farAB-encoded efflux pump system (34, 37). This efflux pump is composed of the membrane fusion protein FarA, the cytoplasmic membrane transporter protein FarB, and the outer membrane channel MtrE (Fig. 5A and B) (7, 17, 18, 34). As depicted in Fig. Fig.5A,5A, the transcriptional repressor FarR negatively controls transcription of this operon (33), supported by the heterodimeric integration host factor (IHF), which also binds to the farAB promoter region, stabilizes FarR-DNA binding, and is probably responsible for bending of the target DNA (32). Deletion of FarR led to expression of the efflux pump system and thus to export of the antibacterial compounds (Fig. (Fig.5B).5B). In meningococci, a homologue of the FarAB system has been described, but the transcriptional repressor FarR could not be implicated in fatty acid resistance (55). In order to test whether the high intrinsic resistance of N. meningitidis prevents experimental quantification of genetic mutations, we constructed a farR-deficient strain in a fatty acid-sensitive ΔkdtA strain. However, this fatty acid-sensitive strain showed no enhanced resistance upon deletion of the repressor farR (Fig. (Fig.5C),5C), indicating that FarR is not involved in fatty acid resistance in meningococci. To exclude the possibility that the concentration of palmitic acid used, 150 μg/ml, is too low to trigger a FarR-dependent effect, we determined the MIC of palmitic acid for strain MC58 and the corresponding farR deletion strain. In accordance with our previous results, no differences were observed; the MIC for MC58 and MC58 ΔfarR was reached at 1,500 μg/ml, supporting our findings that FarR cannot be implicated in meningococcal fatty acid resistance.
We previously demonstrated that N. meningitidis FarR specifically binds to the farAB promoter region (55). Nonetheless, FarR is irrelevant for fatty acid resistance of meningococci. To further analyze whether this implies the insignificance of the FarAB system for fatty acid resistance in meningococci, we generated farAB-deficient strains in the wild-type MC58 as well as in the palmitic acid-sensitive MC58 ΔsiaD ΔkdtA strain. Interestingly, inactivation of this operon completely abolished the palmitic acid resistance of the wild-type strain, whereas the survival rate with linoleic acid was unchanged (Fig. (Fig.5C).5C). Additionally, expression analysis showed no differential farAB transcription comparing the wild type with the farR-deficient strain (data not shown). This indicates that the palmitic acid resistance of meningococci is maintained by active efflux of this compound and additionally by an intact LPS molecule yet is not negatively controlled by FarR. Alternatively, these two mechanisms could act in cooperation to provide fatty acid resistance, which seems unlikely as each single mutation already led to complete palmitic acid sensitivity.
Long-chain fatty acids are well known for their antimicrobial activity and have been utilized in food microbiology for many years as additives for food preservation (25). However, the reason for their antimicrobial action had not been pinpointed precisely; fatty acids were shown to block electron transport, inhibit oxygen and amino acid uptake, uncouple the oxidative phosphorylation, and inhibit bacterial fatty acid synthesis (39, 58, 74). Interestingly, the antimicrobial activity of long-chain fatty acids is mainly restricted to Gram-positive bacteria (13, 26). This difference may result from the composition of the outer membrane of Gram-negative bacteria, which is effectively active against hydrophobic substances (15, 58). The two closely related Gram-negative species N. meningitidis and N. gonorrhoeae are both exclusively found in humans but show very different adaptations to their host. N. meningitidis is a commensal, colonizing the nasopharynx of approximately 10% of the population, and is only rarely associated with invasive infection. In contrast, N. gonorrhoeae is a human pathogen which causes asymptomatic and symptomatic infections mainly of the genital tract. Their outer membranes—like the outer membranes of all other Gram-negative bacteria—are dominated by LPS. The LPS structure of these Neisseria species is basically consistent with the model in Fig. Fig.2,2, except for the outer carbohydrate residues, which are highly diverse in meningococci, leading to 12 different immunotypes (56); in contrast, gonococci possess an additional glucosyl transferase encoded by lgtD, which adds an N-acetylgalactosamine at the terminal galactose residue of the α-chain (22). Here, we demonstrate a generally high intrinsic resistance of multiple clinical isolates of N. meningitidis to the long-chain fatty acids palmitic and linoleic acid. The resistance to palmitic acid was shown to be dependent on the LPS as a ΔlpxA mutant strain was highly susceptible. Further analysis revealed that it was sufficient to delete kdtA, and thus the whole carbohydrate moiety of the LPS, to render meningococci sensitive to palmitic acid. The resistance against linoleic acid was not influenced by this mutation. Interestingly, this effect, which is dependent on the fatty acid chain length, was also observed for ΔkdtA mutants of E. coli and S. Typhimurium (58). Thus, the carbohydrate moiety of LPS seems to be necessary for resistance against palmitic acid. As mutations in kdtA or lpxA most likely do not occur in nature, they could not account for the susceptibility of three serogroup Y, ST-23 strains tested in this study. These strains were the only meningococcal strains highly susceptible to palmitic acid. It has recently been reported that more than 10% of clinical meningococcal isolates contain natural mutations leading to penta-acylated lipid A (12). The gene responsible for the addition of the last acyl chain, lpxL1, is homologous to the E. coli lpxL (htrB) gene, a lauroyl transferase (2, 5, 68). However, in this case comparisons with E. coli are not easily made because neisserial lipid A is structurally distinct, expressing a symmetric acylation with three fatty acids attached to each of the two glucosamine residues of the lipid A backbone (30, 63). The last two acyl chains added by lpxL1 and lpxL2 are secondary lauric acids, esterified to the hydroxylated primary fatty acids at the 2 and 2′ positions (10, 73). The presence of the last lauric acid added by lpxL1 is necessary for induction of full inflammatory responses in both N. meningitidis and N. gonorrhoeae (10, 12, 40) and seems to be important for intracellular survival of gonococci (48). Here, we demonstrate that the deficiency of lpxL1 expression in the serogroup Y, ST-23 strains or in the lpxL1 deletion mutants is sufficient for reversing the resistance of meningococci to palmitic acid. It has been shown for S. aureus that the susceptibility to fatty acids is correlated with an increased binding of palmitic acid to the cytoplasmic membrane (29). In meningococci, however, there was no difference in the amounts of adsorbed palmitic acid between wild-type and lpxL1 deletion strains. The farAB-encoded efflux pump system enables N. gonorrhoeae to survive in hydrophobic environments (34, 41). The transcriptional regulator FarR, which controls expression of farAB in N. gonorrhoeae, could not be connected with the same function in N. meningitidis (33, 55). We show that the FarAB efflux pump system is also functional and required for meningococcal palmitic acid resistance yet is not negatively controlled by the transcriptional regulator FarR. However, an intact LPS molecule seems to be necessary to maintain the active efflux as truncation of the oligosaccharide moiety or incomplete acylation of lipid A rendered the bacteria susceptible. It is possible that these two mechanisms work in unison, with an intact LPS structure being required for the correct insertion of the efflux pump proteins into the outer membrane, as has been described for the gonococcal MtrCDE system and the E. coli porins (36, 44, 52). However, loss of the FarAB system and truncation of the LPS structure may also independently lead to fatty acid sensitivity. Furthermore, the lpxL1 deletion and, thus, the absence of one out of six acyl chains may change the outer membrane of meningococci since a decrease in the number of acyl chains linked to lipid A may destroy the integrity of the outer membrane, resulting in defects of the permeability barrier maintained by the polyanionic LPS (11, 42, 60). Losing this important barrier function, LPS may not be able to maintain its protection of the efflux pump proteins from exposure, as has been shown for porins (24, 36). Altogether, our results show that the closely related N. meningitidis and N. gonorrhoeae contain the same efflux pump systems responsible for the export of hydrophobic compounds yet evolved different concepts of control. In both species, these systems are dependent on an unperturbed permeability barrier and, thus, an intact LPS for their functionality. While gonococci rely on the MtrRCDE efflux pump system for survival in the urogenital tract (23, 57), meningococci have a functional MtrCDE system but no functional repressor MtrR (1, 46, 53). Instead, natural downregulation of mtrC expression by IHF-dependent insertions of Correia elements prevents a constitutively active MtrCDE efflux pump (6, 53). Furthermore, gonococci need the FarAB system to survive highly hydrophobic environments like the rectum (37, 41), whereas meningococci harbor a functional FarAB efflux pump that is not controlled by FarR but, rather, is constitutively active and thus responsible for the overall high resistance to fatty acids.
We thank N. Trzeciak and S. Hebling for expert technical assistance and T. T. Lâm for support with the flow cytometer. E. R. Moxon, T. Tønjum, J. Tommassen, and P. van der Ley as well as the National Reference Center for Meningococci (NRZM, Würzburg, Germany) kindly provided strains and mutants for this study.
This work was supported by the German Research Foundation within SFB479 (B2 to O.K. and M.F.).
Published ahead of print on 26 March 2010.