The Meningococcal Cysteine Transport System Plays a Crucial Role in Neisseria meningitidis Survival in Human Brain Microvascular Endothelial Cells

Neisseria meningitidis colonizes at a nasopharynx of human as a unique host and has many strains that are auxotrophs for amino acids for their growth. To cause invasive meningococcal diseases (IMD) such as sepsis and meningitis, N. meningitidis passes through epithelial and endothelial barriers and infiltrates into blood and cerebrospinal fluid as well as epithelial and endothelial cells. However, meningococcal nutrients, including cysteine, become less abundant when it more deeply infiltrates the human body even during inflammation, such that N. meningitidis has to acquire nutrients in order to survive/persist, disseminate, and proliferate in humans. This was the first study to examine the relationship between meningococcal cysteine acquisition and the pathogenesis of meningococcal infections. The results of the present study provide insights into the mechanisms by which pathogens with auxotrophs acquire nutrients in hosts and may also contribute to the development of treatments and prevention strategies for IMD.

binding protein (cab), respectively ( Fig. 1) and represent the cysteine transport system (CTS). In the present study, we constructed an insertion mutation in the cbp gene and analyzed its biological function in meningococcal nutrient virulence.
Infectious abilities in human brain microvascular endothelial cells (HBMEC) were initially examined. The growth rate of the cbp mutant in assay medium (AM) (see Materials and Methods) was approximately one-third that of the wild-type strain in a 4-h incubation with shaking (data not shown). The adhesion ability of the cbp mutant HT1959 strain was reduced to approximately one-third that of the wild-type strain ( Fig. 2A), which might reflect the growth rate difference in AM. On the other hand, the number of internalized bacteria decreased to approximately 1/100 that of wild-type strain NIID84 Δcps (Opa Ϫ , Opc ϩ/Ϫ ) ( Fig. 2B; see also Fig. S1 in the supplemental material). It was important that the same phenotype was observed in the other meningococcal strains, NIID512 (Opa ϩ , Opc Ϫ ) and NIID521 (Opa ϩ , Opc ϩ/Ϫ ) ( Fig. 2A and B; see also Fig. S1). In order to confirm the relationship between insertional mutation and infection defects, HT 2077, a cbp mutant complemented with the wild-type cbp ϩ gene expressed from the tac promoter in trans at the ggt locus ( Fig. 1) was constructed and examined in HBMEC. It is important that the insertion of genes at the ggt locus did not affect meningococcal infection of human endothelial and epithelial cells (32,33,38,42,43). While the Cbp protein was expressed in HT2077 (Fig. 2D), the infectious defect was not restored in HT2077 ( Fig. 2B and C). In contrast, the infectious defect in HBMEC was restored by complementation with three wild-type cts genes (cbp ϩ , ctp ϩ , and cab ϩ ) in trans at the ggt locus in the HT2080 N. meningitidis strain ( Fig. 1 and 2B and C). It is important that the same results were again obtained in the NIID512 and NID521 genetic backgrounds (Fig. 2). The transcription of cts genes in these mutants examined  Takahashi et al. by reverse transcriptase PCR (RT-PCR) was consistent with the infection recovery in Fig. 2; cbp mRNA was detected in NIID84 Δcps, HT2080, and HT2077, in which only the cbp gene was complemented, but ctp and cab mRNA were detected only in NIID84 Δcps and HT2080, indicating that the cbp, ctp, and cab genes were transcriptionally comprised of one operon at the cts gene allele (Fig. S2). Taken together, these results suggested that all three cts genes, but not cbp alone, were required for the meningococcal infection of HBMEC. Moreover, considering the expression profiles of Opa and Opc (Fig. S1), which also affect meningococcal infection of cultured human cells (44)(45)(46), these proteins could not be related to the results observed in this study.
Therefore, the subsequent experiments were performed using the NIID84 Δcps strain only.
Cysteine uptake activity via the CTS correlated with the meningococcal infection of HBMEC. While all three proteins (Cbp, Ctp, and Cab) were clearly required for the efficient internalization of HBMEC, the relationship between the pathogen's infectious abilities and cysteine transport via the CTS remained unknown. Therefore, we investigated cysteine uptake activity. In our previous studies, L-glutamate uptake activity via the GltT-GltM transport system was examined in buffer A containing 20 mM NaCl for 20 s (32) because L-glutamate uptake is drived by a Na ϩ gradient (31). However, under the same conditions, [ 35 S]cysteine was imported very inefficiently into N. meningitidis (see Fig. 3A compared to Fig. S3A). Moreover, the addition of NaCl did not promote uptake (data not shown). Thus, in the present study, synthetic medium (SM) devoid of cysteine (see Materials and Methods) was used instead of buffer A containing NaCl. Since increases in NaCl concentrations in SM did not stimulate meningococcal cysteine uptake (Fig. S3C), meningococcal cysteine uptake was examined in SM containing 60 mM NaCl in the present study. Under these conditions, [ 35 S]cysteine was imported less efficiently in cbp mutant HT1959 than in wild-type strain NIID84 Δcps (Fig. 3A). The low efficiency of cysteine uptake was not restored in HT2077 (a cbp mutant complemented with a cbp ϩ gene only) but recovered in the HT2080 N. meningitidis strain (a cbp mutant complemented with all three cts ϩ genes) (Fig. 3A). The correlation observed with infectious abilities in HBMEC ( Fig. 2B and C) suggested that the efficient influx of extracellular cysteine via the CTS into N. meningitidis contributed to the efficient meningococcal internalization into HBMEC (see Discussion).
Ezrin accumulation was observed beneath the cbp N. meningitidis mutant. We previously demonstrated that the transient uptake of L-glutamate upon meningococcal adhesion to HMBEC triggered meningococcal internalization into HBMEC and the concomitant accumulation of ezrin beneath meningococci (33), which is a marker for the meningococcal stimulation of host cell signaling for internalization (47). In order to establish whether the same strategy is applicable to cysteine uptake, we analyzed changes in the host cell cytoskeleton upon meningococcal infection using indirect N. meningitidis mutants to HBMEC and effects of complementation of the cbp ϩ or cbp ϩ ctp ϩ cab ϩ genes in the cbp mutant on bacterial infection. The numbers of bacteria were measured as CFU. Levels of internalized bacteria were determined as numbers of bacteria recovered after the gentamicin treatment. Each value represents the mean Ϯ standard error of mean (CFU per 10 4 HBMEC) of results from at least four experiments. Open, filled, light gray, and dark gray bars in panels A, B, and C indicate the bacterial number of or percent internalized N. meningitidis cts ϩ strains (NIID84 Δcps, NIID512 Δcps, and NIID521 Δcps), cbp::cat (HT1959, HT2056, and HT2063), and Δcps mutants in which the cbp ϩ or cts ϩ (cbp ϩ ctp ϩ cab ϩ ) genes were ectopically complemented (HT2077, HT2078, and HT2079 or HT2080, HT2081, and HT2082), respectively (see Table 1). *, P Ͻ 0.001; #, P Ͻ 0.02 (significantly different from the cts ϩ strain or the Δcps mutant complemented with cts ϩ genes). (D) Western blotting for Cbp proteins. Bacterial extracts equivalent to an OD 600 of 0.025 were analyzed by Western blotting. Lanes 1 and 2 represent molecular markers corresponding to 33  immunofluorescence to monitor the localization of ezrin (Fig. 4). While the accumulation of ezrin was minimal in HBMEC not infected with N. meningitidis (Fig. 4, left panels), it was clearly detected beneath wild-type N. meningitidis strain NIID84 Δcps (Fig. 4, middle panels). It is important that ezrin also accumulated beneath the cbp HT1959 mutant as efficiently as beneath the wild-type strain (Fig. 4, right panels), while the ]cysteine and a final NaCl concentration of 60 mM in SM. Transport was examined for 5 min due to its lower efficiency than L-glutamate uptake (32). Open, filled, light gray, and dark gray bars indicate the cysteine uptake rate (nmol/min/mg protein) of NIID84 Δcps (cts ϩ ), HT1959 (NIID84 Δcps cbp), HT2077 (NIID84 Δcps cbp ggt::cbp ϩ ), and HT2080 (NIID84 Δcps cbp ggt::cts ϩ ), respectively. #, P Ͻ 0.02 (significantly different from the cts ϩ strain); *, P Ͻ 0.005 (significantly different from the Δcps cts ϩ strain). (B) Time course of L-cysteine uptake. The assay was performed in the presence of 100 M L-[ 35 S]cysteine and a final NaCl concentration of 60 mM in SM. The open circles with a solid line and filled circles with a dashed line indicate cysteine uptake by NIID84 Δcps (cts ϩ ) and HT1959 (NIID84 Δcps cbp), respectively. (C) Concentration dependence of L-cysteine import by strains NIID84 Δcps (cts ϩ ) and HT1959 (NIID84 Δcps cbp). The assay was performed in SM containing 60 mM NaCl for 5 min. (D) Lineweaver-Burk plot of data shown in panel C relative to NIID84 Δcps (cts ϩ ). A formula deduced from the data is also shown in the same panel. The K m value of meningococcal [ 35 S]cysteine uptake in SM was calculated to be approximately 114 M. ezrin accumulation was not observed in the ⌬gltT ⌬gltM N. meningitidis strain (32,33). This result indicated that the cbp mutant was not defective for stimulation of the ezrin accumulation but was defective in another mechanism(s).
The intracellular survival rate of the cbp mutant was reduced to the level seen with the ⌬gshB mutant. Since defective L-glutamate uptake reduced glutathione synthesis and concomitantly decreased intracellular survival (31)(32)(33)(34), it is possible that the invasion defect in the cbp mutant can be explained by bacterial survival/persistence in HBMEC after internalization. In order to examine this possibility, the meningococcal survival rate in HBMEC was assessed in a 4-h incubation under our experimental conditions (Fig. 5A). The intracellular bacterial number of wild-type N. meningitidis strain NIID84 Δcps gradually decreased during the course of the incubation, and after a 4-h incubation, approximately 60% of the bacteria remained in HBMEC (Fig. 5B). In contrast, the intracellular bacterial number of the cbp N. meningitidis mutant markedly decreased during the 4-h incubation (Fig. 5B). The rate of the decrease (6% of the zero time number) in the cbp mutant was markedly less than that (60% of the zero time number) of the wild-type strain. These results suggest that the cbp mutation decreased intracellular survival rates in HBMEC, which may have contributed to the apparent decrease in the internalized bacterial number in HBMEC (Fig. 2B).
Glutathione concentrations did not markedly decrease in the cbp N. meningitidis mutant. Previous studies reported that L-glutamate uptake is important for the synthesis of glutathione, a ubiquitous antioxidant in all living cells (48), and leads to N. meningitidis resistance to the neutrophil oxidative burst (34) and increased intracellular survival in human cells (31,33). Since cysteine is also one of the components of glutathione, we constructed the gshB (glutathione synthetase) HT2120 mutant and examined the relationship between glutathione synthesis and the apparent ability of the cbp N. meningitidis mutant to internalize into HBMEC (Fig. 2B). While meningococcal adhesion to HBMEC was slightly affected by the cbp and gshB mutations, a marked decrease was not observed (Fig. S4A). On the other hand, the internalization of the cbp HT1959 mutant into HBMEC was approximately 100-fold less efficient than that of wild-type strain NIID84 Δcps, while that of the gshB HT2120 mutant was approximately 3-fold less efficient than that of wild-type strain NIID84 Δcps ( Fig. S4B and C). These results suggest that glutathione synthesis was less likely to be related to the apparent meningococcal internalization (intracellular survival) into HBMEC.
In order to further clarify whether glutathione participates in intracellular survival, the content of glutathione was measured (Fig. 6). In N. meningitidis strains grown in gonococcal (GC) medium base (GCB), the intracellular glutathione concentration in the The HBMEC monolayer was infected with wild-type cts ϩ (middle) and cbp::cat (right) N. meningitidis strains. A noninfected HBMEC monolayer is also shown in the left panels. Bacteria and HBMEC were observed by phase-contrast microscopy (upper panels). Ezrin was immunostained with anti-ezrin monoclonal antibodies (MAb) and Alexa Fluor 488-conjugated rabbit anti-mouse IgG (green channel) (lower panels).
⌬gshB mutant was approximately 50 mol/mg protein, representing the background level of glutathione during the 4-h incubation (Fig. 6). Under the same conditions, wild-type N. meningitidis strain NIID84 Δcps had an intracellular glutathione concentration of approximately 800 mol/mg protein, while that in the cbp mutant was approximately 400 mol/mg protein (Fig. 6). When the strains were grown in AM, intracellular glutathione concentrations slightly decreased in the wild-type N. meningitidis strain but remained unchanged in the cbp mutant. Similar results were observed for N. meningitidis strains infecting HBMEC (Fig. 6). Moreover, the reduction in the glutathione amount was statistically recovered in a cbp mutant complemented with HT2080 cts genes but not in a cbp mutant complemented with a HT2077 cbp gene (Fig. S5A). These results suggest that cysteine uptake via the CTS accounts for some, but not all, of the glutathione synthesis.
Resistance to reactive oxygen species (ROS) was not largely defective in the cbp N. meningitidis mutant. L-Glutamate imported via GltT-GltM transporters was previously shown to be utilized in the synthesis of glutathione for resistance to H 2 O 2 (34), a membrane-permeable ROS that damages proteins and DNA (49). Since glutathione concentrations were reduced to approximately 50% of the level in the wild-type strain (Fig. 6), decreases in the glutathione concentration in the cbp mutant may have reduced the intracellular survival rate in HBMEC. In order to examine this possibility, sensitivity to H 2 O 2 and paraquat, which also generates toxic oxygen species during respiration within the cytoplasm after penetrating the cells (50,51), was investigated in vitro (Fig. 7). The levels of sensitivity to 0.25 and 0.5 mM H 2 O 2 did not significantly differ among the wild-type, cbp mutant, and ⌬gshB N. meningitidis strains (Fig. 7A). However, at 1 mM H 2 O 2 , the cbp mutant was approximately 5-fold more sensitive than the wild-type N. meningitidis strain under conditions in which the ⌬gshB mutant was approximately 13-fold more sensitive than the wild-type N. meningitidis strain (Fig. 7A). Under these conditions, the cbp mutant was approximately 3-fold less sensitive to 1 mM H 2 O 2 than the ⌬gshB mutant (P Ͻ 0.001). These results suggest that a cysteine transport deficiency resulted in N. meningitidis being slightly more sensitive to 1 mM H 2 O 2 . The slight reduction of resistance to 1 mM H 2 O 2 was statistically recovered in a cbp mutant complemented with HT2080 cts genes but not in a cbp mutant complemented with a HT2077 cbp gene (Fig. S5B). Regarding paraquat, no significant differences were observed with 2 mM paraquat among the wild-type, cbp mutant, and ⌬gshB N. meningitidis strains (Fig. 7B). At 10 mM paraquat, no significant differences were noted between the wild-type and cbp mutant N. meningitidis strains, while the ⌬gshB mutant was approximately 7-fold more sensitive than the wild-type and cbp mutant N. meningitidis strains (Fig. 7B). However, at 50 mM paraquat, the cbp mutant was approximately 3-fold more sensitive than the wild-type strain whereas the ⌬gshB mutant was approximately 240-fold more sensitive than the wild-type strain (Fig. 7B). The reduction of resistance to 50 mM paraquat was statistically recovered in a cbp mutant complemented with HT2080 cts genes but not in a cbp mutant complemented with a HT2077 cbp gene (Fig. S5C). Collectively, these results with respect to sensitivities to H 2 O 2 and paraquat indicated that decreases in glutathione concentrations due to CTS defects did not markedly affect meningococcal resistance to ROS.
Cysteine uptake via the CTS was essential for meningococcal growth in the presence of less than 300 M cysteine. We investigated meningococcal growth in liquid media. In the complete medium GCB, the cbp mutant grew as efficiently as the CTS Role in N. meningitidis Survival in HBMEC ® wild-type N. meningitidis strain (Fig. 8A), indicating that the growth of the cbp mutant was not defective in liquid medium. In previous experiments using meningococcal defined medium, Neisseria chemically defined medium (NCDM), developed by Catlin (30), was commonly used for N. meningitidis (31,34). However, under our experimental conditions, most of the N. meningitidis strains did not grow in NCDM or grew to an optical density at 600 nm (OD 600 ) of ϳ0.4 from 0.1 at the starting point even though N. meningitidis had been precultured on GC agar (data not shown) (38). In order to  overcome this issue, we used our original SM, which was basically the same as NCDM but was supplemented with a one-fourth volume of originally reconstituted MCDB131 devoid of NaCl and cysteine (see Materials and Methods and Table S2 in the supplemental material). Thus, SM did not contain NaCl or cysteine, and these reagents were supplemented if indicated. In SM supplemented with 60 mM NaCl, wild-type N. meningitidis strain NIID84 Δcps did not grow when cysteine was not supplemented (Fig. 8B). On the other hand, supplementation of cysteine completely suppressed the growth defect (Fig. 8B), ensuring that meningococcal growth in SM was completely dependent on cysteine. It is important that the supplementation of even 50 M cysteine was sufficient for meningococcal growth in SM (Fig. 8B). Regarding the growth of the cbp mutant, it also did not grow in SM devoid of cysteine (Fig. 8C). The supplementation of increasing concentrations of cysteine gradually suppressed the growth defect in the cbp mutant, while at least 300 M cysteine was required in order to completely suppress the growth defect (Fig. 8C). These results suggest that cysteine uptake via the CTS is crucial for meningococcal growth under conditions of restricted concentrations of cysteine. Furthermore, these results are consistent with the K m value of meningococcal [ 35 S]cysteine uptake in SM being approximately 114 M (Fig. 3D). Since the concentration of the intracellular cysteine pool is very limited (approximately 100 M) (52), these results suggest that the observed marked reduction in bacterial internalization into HBMEC was mostly due to the meningococcal starvation of cysteine in HBMEC (see Discussion).

DISCUSSION
Investigation of the role of metabolism in virulence was recently recognized as a priority equivalent to studying classical virulence factors (24,26,53). While cysteine levels in eukaryotes need to be tightly regulated by controlling the degradation of this amino acid (54,55) and transport of it from the extracellular milieu into the cell and vice versa (56), bacterial cysteine uptake has not yet been elucidated in detail (57)(58)(59)(60). N. meningitidis cysteine transport and its role in the pathogenesis of meningococcal infections have yet to be clarified. To the best of our knowledge, while all of the experiments in this study were done only by the use of in vitro models and with a limited number of meningococcal strains, this is the first study to have suggested a relationship between meningococcal cysteine uptake and the pathogenesis of meningococcal infections.
Although the N. meningitidis strains used in the present study were cysteine auxotrophs (Fig. 8), a genome-scale metabolic model suggested that N. meningitidis is capable of the de novo synthesis of all amino acids, including cysteine (40,61). However, experimental data revealed contrasting results; some N. meningitidis strains, namely, MC58 (62) and HB-1 (40), grew in minimal defined media, while strain ␣522 failed to grow (62). Previous studies also demonstrated that cysteine auxotrophs differed among N. meningitidis strains (30) and that many N. meningitidis strains had cysteine auxotroph characteristics with respect to their growth in the presence of low cysteine concentrations (38)(39)(40). While SM used in the present study contained approximately 3 mM inorganic sulfur molecules, which were mainly SO 4 2Ϫ (see Table S2 in the supplemental material), NIID84 Δcps did not grow in SM if cysteine was supplemented (Fig. 8B). Although further experiments are required to obtain morecomprehensive information, the following may be considered: N. meningitidis potentially has a set of genes for de novo cysteine synthesis in nature, and no mutation was found in the open reading frame (ORF) of any CTS genes in the genomes of 14 N. meningitidis strains recorded in NCBI database (data not shown). However, their products might not be always expressed in all N. meningitidis strains because of some negative on the transcription or translation, or enzymes other than CTS enzymes might affect cysteine assimilation but not acquisition for N. meningitidis. In fact, the levels of cysteine acquisition ability were different between meningococcal strains and seemed not to be related to the sequence type (see Fig. S6 in the supplemental material) or to the invasive/noninvasive phenotype (63). Regardless of the nature of the cysteine auxotroph in N. meningitidis, the meningococcal ability to uptake cysteine from the surrounding environment would be advantageous for N. meningitidis for the following reasons: N. meningitidis preferably uses cysteine (cystine) as a sulfur source (41), and de novo cysteine synthesis from inorganic sulfur molecules is more laborious and energetically less efficient than cysteine uptake from the surrounding environment because cysteine synthesis requires multiple enzymatic steps (41) with at least one ATP molecule (61). Furthermore, nonessential sulfur is not stored by N. meningitidis (36). Therefore, N. meningitidis may have to continuously and quickly acquire cysteine for its survival, persistence, and proliferation in humans. Additional experiments are required for more complete understanding of the meningococcal cysteine acquisition.
Three pathogenic bacteria with auxotroph characteristics for cysteine have evolved their own strategies to acquire cysteine in host cells; intracellular Legionella pneumophila acquires nutrients that include cysteine by promoting host proteasomal degradation (25) and Anaplasma phagocytophilum obtains cysteine by promoting host autophagy (64). Francisella tularensis uses glutathione as a cysteine source for intracellular multiplication by ␥-glutamyl aminopeptidase (GGT) (52), which is the same strategy used for meningococcal proliferation in CSF (38). However, the results obtained so far have indicated that the intracellular persistence of N. meningitidis was not supported by meningococcal GGT because N. meningitidis strain HT2022, in which the ggt gene was disrupted by the insertion of all three wild-type cts genes (Fig. 1), was internalized as efficiently as the wild-type strain ( Fig. 2B and C). Regarding pathogenic bacteria that cause meningitis, such as Haemophilus influenzae and Streptococcus pneumoniae, while limited information is currently available on the role of cysteine in the pathogenesis of meningitis, the cysteine synthesis gene in S. pneumoniae was shown to be transcriptionally upregulated in the presence of low cysteine concentrations (65). While meningococcal glutamate uptake was also regulated by Na concentration (31), the meningococcal cysteine uptake activity was not regulated by Na (Fig. S3C), leading to the speculation that the activity might be not altered by environmental conditions.
Compared to the large reduction in the intracellular survival rate (Fig. 5B), [ 35 S]Cys uptake in the cbp mutant decreased to only approximately 40% of the level seen in the wild-type strain (Fig. 3A). The discrepancy might have been due to the experimental conditions used for analysis of [ 35 S]Cys uptake in this study; the [ 35 S]Cys obtained in our laboratory tended to bind nonspecifically to meningococci as well as to any chemical filter (e.g., polyvinylidene difluoride [PVDF] and glass) even though they were washed with excess amounts of phosphate-buffered saline (PBS) containing 2% saponin (data not shown). Since no other cysteine uptake system was found in the meningococcal genome database uploaded (data not shown), the CTS would seem to be unique with respect to meningococcal cysteine acquisition, and it was also shown that the [ 35 S]Cys uptake in the cbp mutant could not be observed even for 5 min of incubation with a high concentration of [ 35 S]Cys in buffer A instead of SM (Fig. S3A). Thus, it would also seem to be possible that the [ 35 S]Cys uptake examined in SM in this study was overestimated against the true function of cysteine uptake via the CTS in N. meningitidis and that methods used to address the problems regarding [ 35 S]Cys uptake could be further improved.
The cbp mutant grew normally (similarly to the wild-type strain) in GCB (Fig. 8A), eliminating the possibility that a mutation in meningococcal CTS itself affected essential N. meningitidis growth. It should be noted that the cbp mutant grew better than the wild-type strain in GCB at the stationary phase (Fig. 8A), which might be related to the negative effect of cysteine for bacterial growth (discussed below). On the other hand, under cysteine-limited conditions, while the supplementation of at least 50 M cysteine was sufficient to suppress the growth defect in cysteine-free SM in the wild-type strain (Fig. 8B), supplementation of more than 300 M cysteine was required for the complete suppression of growth of the cbp mutant (Fig. 8C). These results suggest that meningococcal cysteine uptake mediated via the CTS plays an essential role in N. meningitidis growth under conditions of low cysteine concentrations. On the basis of the cysteine concentrations seen in humans, meningococcal cysteine uptake via the CTS is advantageous for N. meningitidis (Fig. 9); the concentration of cysteine in the human nasopharynx remains unclear, but the nasopharynx is considered to be one of the nutritionally richest niches in humans because cysteine is supplied from the digestion of food. Alternatively, cysteine may be supplied from nasopharyngeal commensal bacteria. On the other hand, when N. meningitidis enters and passes through human epithelial and endothelial cells by endocytosis, the intracellular cysteine concentration is very low at approximately 100 M (52). When the organism enters the bloodstream, the concentration of cysteine in human blood plasma is only approximately 30 to 60 M (65, 66). During inflammation with sepsis, the cysteine concentration is significantly decreased (67,68). N. meningitidis attaches again to endothelial cells comprising the BBB, passes through cells or loosens tight junctions (47), and enters the CSF. The concentration of cysteine in the CSF is very limited at less than 1 M (38). Thus, cysteine concentrations are very low in humans (69) because mammals synthesize cysteine from methionine and serine, and cysteine may also be supplied from glutathione (70), which is an abundant and ubiquitous molecule in all organs and cell types  (Fig. 8A), N. meningitidis may have two cysteine transport systems: a CTS that functions under low-cysteine conditions (higher affinity for cysteine) and an unidentified cysteine transporter that may function only under cysteine-rich conditions (lower affinity for cysteine). While this study showed only the intracellular functions of cysteine uptake via the CTS in N. meningitidis (C), this physiological function may be extensively applied to all human milieus for meningococcal infections in humans on the basis of the following concept. During the course of meningococcal invasion, until the point at which N. meningitidis reaches the human cells, N. meningitidis encounters at least 3 types of environments in humans: (A) the environment outside the human (e.g., nasopharynx), (B) the environment inside humans but not the intracellular milieu (e.g., blood and CSF), and (C) the intracellular milieu (in epithelial and endothelial cells). Cysteine is not limited in the environment represented in panel A because sufficient cysteine is supplied from the digestion of food. However, cysteine availability is very limited inside human environments (B and C) due to nutrient immunity with respect to cysteine. Cysteine concentrations in extracellular environments in humans (B) in blood (ϳ50 M) and CSF (ϳ1 M) are lower than in the intracellular milieu (A). The limitation of cysteine availability is also found in the intracellular environment, in which the concentration of cysteine is estimated to be approximately 100 M. Thus, N. meningitidis has to "dive" into humans to cause IMD, in which context the milieu cysteine availability is always limited by the human nutrient immunity.
CTS Role in N. meningitidis Survival in HBMEC ® (48). The limitation of the availability of "free" cysteine in humans may be regarded as a type of nutrient immunity. Therefore, the meningococcal cysteine acquisition via the CTS may play an important role as a "nutrient virulence factor" for N. meningitidis infections in humans with nutrient immunity for cysteine (Fig. 9).
In the present study, we used unencapsulated N. meningitidis strains for efficient internalization by eliminating the multiple effects of the capsule on meningococcal infections in humans. As discussed in our previous study (33), the meningococcal capsule may affect the intracellular survival ability (Fig. 5). However, since the majority of internalized bacteria are considered to remain unencapsulated until a phase variation occurs (8,71), the intracellular survival of the unencapsulated state may also be important in the N. meningitidis infectious cycle. Furthermore, since the capsule expression does not hinder the import of amino acid into N. meningitidis (34), cysteine uptake in N. meningitidis is physiologically advantageous for the infection of humans regardless of the capsule expression state.
While cysteine is an important amino acid because it contains sulfur, which plays a vital role in the catalytic sites of many enzymes and participates in ion and redox metabolism (72), excessive amounts of cysteine are also harmful for many bacteria (73). For example, Escherichia coli, which is a prototroph for cysteine, did not synthesize excess amounts of (74) or store (75) cysteine by feedback pathways. This may also be the case for N. meningitidis because the supplementation of a higher concentration of cysteine appeared to inhibit meningococcal growth in SM at the late logarithmic phase (Fig. 8B) and the addition of 1 mM cysteine also appeared to inhibit meningococcal growth in SM (data not shown). Since cysteine auxotrophic organisms have to control intracellular cysteine concentrations at its uptake only, it seems reasonable to hypothesize that slow cysteine uptake via the CTS in N. meningitidis is among its advantageous biological functions.
The exploitation of host nutrients by pathogens is a major factor in host-pathogen interactions and is one of the fundamental aspects of infectious diseases; however, our knowledge of microbial metabolism in hosts is limited. Further studies on the mechanisms underlying microbial (including meningococcal) "nutritional virulence" and metabolism in vivo will provide a more comprehensive understanding of the central aspect of infectious diseases and result in the development of treatments and prevention strategies, such as the development of inhibitors for folate metabolism in pathogenic bacteria (76) or for iron acquisition using antibodies against N. meningitidis transferrin-binding proteins (77).

MATERIALS AND METHODS
Bacterial growth conditions. N. meningitidis strains (stored at -80°C) were routinely grown on GC agar plates at 37°C in a 5% CO 2 atmosphere (78). Brain heart infusion (Becton, Dickinson, USA) agar containing 3% defibrinated horse blood (Nihon Biotest, Japan) was used in the selection of kanamycinresistant N. meningitidis strains (63). E. coli was grown on L plates or in L broth liquid cultures at 37°C. When required, antibiotics were added at the following concentrations: kanamycin at 150 g/ml, chloramphenicol at 5 g/ml, and spectinomycin at 75 g/ml for N. meningitidis and kanamycin at 50 g/ml and ampicillin at 50 g/ml for E. coli. The N. meningitidis strains used in the present study are listed in Table 1. GC broth (GCB) contained the following (per liter); proteose peptone, 15 g; NaCl, 5 g; soluble starch, 0.5 g; K 2 HPO 4 , 1 g; KH 2 PO 4 , 4 g; IsoVitaleX enrichment (Difco), 10 ml; 1 M NaHCO 3 , 5 ml; 1 M MgCl 2 , 10 ml. SM in the present study basically consisted of NCDM (30), except that NaCl and cysteine were omitted (details were shown in Table S2 in the supplemental material). NaCl and cysteine were added when required. Additionally, due to another nutrient requirement(s), most meningococcal strains in our laboratory did not grow in NCDM (data not shown). In order to overcome this issue, MCDB131 (Thermo Fisher) lacking cysteine and NaCl, which was specifically manufactured by Cell Science & Technology Inst. Inc. (Japan), was added to a one-fourth volume in a final volume of 10 ml NCDM (details of the contents are shown in Table S2).
Construction of meningococcal mutants. N. meningitidis strains that did not produce a capsule were constructed by transformation with the purified chromosomal DNA of HT1034 (⌬siaB ⌬siaD::kan) (63), and kanamycin-resistant clones were selected ( Table 1). The transformation of N. meningitidis strains was performed as described previously (78).
In order to construct the N. meningitidis cbp insertion mutant, a 0.8-kb DNA fragment from N. meningitidis H44/76 chromosomal DNA containing the cbp gene was amplified with primers cbp-1 and cbp-2 (Table S1) by the use of PrimeSTAR Max DNA polymerase (TaKaRa Bio, Japan) and was cloned into the SmaI site of the pMW119 vector (Nippon Gene, Japan) (4.2 kb) to construct pHT1319 (5 kb). The 5-kb DNA region of pHT1319 was amplified with primers cbp-3 and cbp-4, which separated the cbp structural gene into 400 and 425 bp, respectively, by the use of PrimeSTAR Max DNA polymerase and was ligated with a chloramphenicol resistance gene (cat) to construct pHT1328. A 1.8-kb DNA fragment containing the cbp structural gene insertionally disrupted with the cat gene was amplified with primers cbp-1 and cbp-2 from pHT1328 and transformed into N. meningitidis strains, and chloramphenicol-resistant clones were selected, resulting in cbp insertional mutants (Table 1).
A cbp mutant complemented with the cbp ϩ gene at the ggt allele was constructed as follows. A 0.8-kb cbp gene was amplified with primers pTTQ-1-cbp-5 and M13-47-cbp-6. A 4.3-kb DNA fragment was also amplified from pTTQ18 (79) with primers pTTQ-2(POϩRBS) and M13-47 reverse (Table S1). These two DNA fragments were connected by the use of SLiCE (80). The resultant plasmid was named pHT1381. In order to insert a spectinomycin resistance gene (spc) downstream of the cbp gene, a 5.1-kb DNA fragment was amplified from pHT1381 with primers pTTQ-5 and pTTQ-6 (Table S1). A 1-kb spc gene was also amplified from pHT154 (81) with primers pTTQ-5=(15mer)-M13-RV-Long and pTTQ-6=(15mer)-M13-47-Long. These two DNA fragments were connected by the use of SLiCE, resulting in a plasmid that was named pHT1386. A 3-kb DNA fragment containing the Ptac, cbp, and spc genes was amplified from pHT1386 with primers ggt-5=(15mer)-ptac-13 and ggt-3=(15mer)-ptac-14 and was inserted into a BstXI site (at the middle of the ggt coding region) of pHT624, in which the ggt gene harbored by pHT195 (81) was subcloned into pMW119, resulting in pHT1387. A 4.5-kb DNA fragment containing the ggt::Ptac cbp spc genes was amplified with primers ggt-1 and ggt-2 (82) and transformed into the N. meningitidis cbp::cat mutants, and spectinomycin-resistant (Spc r ) clones were selected, resulting in cbp insertional mutants ectopically complemented with the cbp gene expressed from the tac promoter at the ggt allele (Table 1).
A cbp mutant complemented with all three cts genes (cbp, ctp, and cab) at the ggt allele was constructed as follows: a 3.3-kb DNA fragment containing cts genes was amplified from NIID84 chromosomal DNA with primers cts operon-1 and cts operon-2 and cloned into the SmaI site of pMW119, resulting in pHT1391. In order to insert the spc gene downstream of the cab gene, a 7.6-kb DNA fragment was amplified from pHT1391 with primers M13-47=(15mer)-pMW(MCS)-down-1 and M13-RV=(15mer)-pMW(MCS)-down-2 (Table S1). A 1-kb spc gene was amplified from pHT154 with the universal primers M13-RV and M13-47, and these two DNA fragments were connected by the use of SLiCE, resulting in pHT1393. A 4.3-kb DNA fragment containing cts genes and the spc gene was amplified from pHT1393 with primers ggt-5=(15mer)-pMW119-F and ggt-3=(15mer)-pMW119-R (Table S1) and cloned into a BstXI site of pHT624 by the use of SLiCE, resulting in a plasmid that was named pHT1395. A 6.3-kb DNA fragment containing the ggt::cbp ctp cab spc genes was amplified from pHT1395 with primers ggt-1 and ggt-2 and transformed into the N. meningitidis cbp::cat mutants, and Spc r clones were selected, resulting in cbp insertional mutants ectopically complemented with the cts genes expressed from their own promoter at the ggt allele (Table 1).
A N. meningitidis mutant deleted of the gshB gene was constructed by transformation with a 2-kb DNA fragment containing the ⌬gshB::spc construct amplified from pHT1058 (33), and Spc r clones were selected and named HT2120 ( Table 1).
Production of anti-Cbp protein rabbit antiserum. A 768-bp DNA fragment devoid of the N-terminal putative hydrophobic domain of Cbp (18 amino acids) was amplified with a set of primers  (81). RT-PCR. Bacteria grown on GC agar plates were suspended in 20 ml of GC to an OD 600 of 0.1 and continuously cultured to mid-log phase (OD 600 of ϳ0.6) at 37°C with shaking. The total RNA was isolated from the harvested bacteria using a FastGene RNA Premium kit (Nihon Genetics, Japan). RT-PCR was performed using a SuperScript III One-Step RT-PCR system with Platinum Taq DNA polymerase (Invitrogen) and approximately 0.1 g of total RNA according to the manufacturer's instructions. Reverse transcriptase reactions (50°C for 30 min) were omitted (-RT in Fig. S2 in the supplemental material) to confirm that the amplicon was derived from RNA. The products were visualized by electrophoresis in a 2% agarose gel followed by ethidium bromide staining.
Tissue culture. HBMEC were cultivated as described previously (63). In each experiment, HBMEC were seeded in a culture flask or dish or on a cover glass and were used within 2 days of reaching confluence.
Assessment of host cell-associated and internalized bacteria. HBMEC were cultivated on gelatincoated 96-well tissue culture plates (Iwaki, Japan) at 37°C for 2 days in a 5% CO 2 atmosphere to a concentration of 1 ϫ 10 4 cells/well. Two hours prior to bacterial infection, the culture medium was replaced with assay medium (AM), which was MCDB131 (Invitrogen) supplemented with 10% fetal bovine serum (FBS), 90 g ml Ϫ1 heparin, and 3 mM glutamine. The bacterial suspension was prepared in AM at an OD 600 of 0.05, which corresponded to approximately 5 ϫ 10 6 CFU/100 l. The multiplicity of infection (MOI) was 500, a condition previously used in other studies (45,83), to examine efficient N. meningitidis internalization (33,42). A 100-l portion of the bacterial suspension was inoculated onto host cell monolayers in duplicate for each assay, and the reaction mixtures were incubated at 37°C for 4 h in a 5% CO 2 atmosphere. In order to evaluate bacterial adherence, the monolayers were washed with prewarmed AM four times to remove nonadherent bacteria. Adherent bacteria were released by the addition of phosphate-buffered saline (PBS) containing 2% saponin, and CFU levels were assessed on GC agar plates with appropriate dilutions to count cell-adherent bacteria. In order to evaluate internalized bacteria, AM containing 150 g ml Ϫ1 of gentamicin was added, and cultures were incubated for 1 h in order to kill all extracellular bacteria. Almost all (Ͼ99.999%) of up to 5 ϫ 10 7 extracellular meningococci were killed under these experimental conditions (data not shown). The amounts of internalized bacteria that were not killed by the gentamicin treatment were assessed by the addition of PBS containing 2% saponin and by plating on GC agar after appropriate dilutions were made to count bacterial numbers as CFU levels. Results are expressed as means Ϯ standard deviations (SD), and bacterial numbers were statistically compared using the two-tailed Student's t test.
Western blotting. N. meningitidis, grown on GC agar plates, was suspended in PBS. Bacteria were harvested at an OD 600 of 20 and resuspended in 1 ml 1ϫ SDS buffer and boiled for 10 min. SDS-PAGE and Western blotting were performed as described previously (38).
Bacterial L-cysteine uptake assay. The L-cysteine transport assay was performed as follows. N. meningitidis strains grown overnight on GC agar plates at 37°C in 5% CO 2 were suspended in 2 ml buffer A (50 mM potassium buffer [pH 7.0] and 0.5 mM MgCl 2 ) (31, 32) and harvested by centrifugation at 10,000 ϫ g at 4°C for 2 min. The bacterial pellet was resuspended in 2 ml buffer A and centrifuged again. These procedures were repeated 3 times to wash bacteria. The resultant pellets were suspended in 1 ml buffer A to adjust the bacterial concentration at an OD 600 of 10. Ten microliters of 300 M L-[ 35 S]cysteine (America Radiolabeled Chemicals) (specific activity, 1,075 Ci/mmol) was mixed with 10 l of a 3-foldhigher concentration of SM containing 180 mM NaCl under standard conditions. A 3-fold concentration of L-[ 35 S]cysteine was added to the reaction mixture when substrate dependency for CTS was examined ( Fig. 3C; see also Fig. S3A). The mixture was prewarmed at 37°C for 5 min. Assays were initiated by the addition of 10 l of the bacterial suspension and was incubated at 37°C for 5 min. Further incubations were performed to optimize the incubation time ( Fig. 3B; see also Fig. S3B). The reaction was terminated by the removal of substrates by subsequent maximal centrifugation at 4°C for 1 min. The resultant pellet (bacteria) was washed 3 times with 0.5 ml PBS containing 2% saponin. The resultant pellet was resuspended in 50 l buffer A. Radioactivity was assessed by scintillation counting in an 8-ml vial containing 5 ml Filtron-X (National Diagnostics, USA). The nonspecific binding of L-[ 35 S]cysteine to bacteria was assessed in control bacterial samples that were inactivated at 70°C for 30 min, and the value obtained was subtracted from all experimental values. Protein concentrations were measured using a bicinchoninic acid (BCA) protein assay kit (Thermo), with bovine serum albumin (BSA) used as a standard. L-Cysteine transport values were expressed as means Ϯ SD representing nanomoles of substrate transported per minute per milligram of bacterial protein.
Observations of meningococci and ezrin accumulation by immunofluorescence staining. HB-MEC monolayers, grown on a cover glass at 37°C in a 5% CO 2 atmosphere for 2 days, were infected with N. meningitidis in AM at an MOI of 500 for 4 h. The infected HBMEC monolayers were washed four times with 500 l prewarmed AM, fixed with 4% paraformaldehyde-PBS for 15 min, permeabilized with 0.2% Triton X-100 -PBS, and blocked with 2% BSA-PBS for 30 min. The resultant monolayers were incubated with the anti-ezrin monoclonal 3C12 antibody (Santa Cruz, USA) and diluted 100-fold for 45 min and then with the Alexa Fluor 488-conjugated F(ab´) fragment of anti-mouse IgG (Invitrogen) and then diluted 200-fold for 45 min under moist and dark conditions. Glass coverslips were mounted on a glass slide with ProLong Gold antifade reagent (Invitrogen). Infected HBMEC and the attached meningococci were observed using a BX51 microscope with a 100ϫ oil immersion objective.
Monitoring meningococcal survival in HBMEC. A meningococcal intracellular survival assay in HBMEC was performed as follows. Confluent HBMEC monolayers seeded on gelatin-coated 24-well tissue culture plates (Iwaki) were infected with meningococci at an MOI of 5,000 for 2 h in order to obtain the maximal amount of internalized bacteria (32). A high dosage of meningococci did not exert any cytotoxic effect on HBMEC (data not shown). Monolayers were washed with AM four times and then incubated with AM containing 150 g ml Ϫ1 gentamicin for 1 h in order to kill extracellular bacteria. Gentamicincontaining AM was replaced with AM, and the reaction mixtures were incubated further at 37°C in 5% CO 2 to monitor the number of bacteria in HBMEC. This time point was defined as time zero. Intracellular bacteria were measured at time zero and after 1, 2, 3, and 4 h, and bacterial numbers were counted as CFU by plating on GC agar plates after appropriate dilutions. Survival percentages were calculated by the following formula: 100 ϫ (CFU at the indicated hour[s]))/(CFU at time zero). The results were expressed as means Ϯ SD.
Measurement of glutathione concentrations in bacteria only or bacteria that infected HBMEC. N. meningitidis strains grown on GC agar plates at 37°C overnight in 5% CO 2 were suspended in 15 ml AM or GCB in order to adjust the bacterial concentration to an OD 600 of 0.15/ml and were then incubated at 37°C for 4 h in 5% CO 2 , or bacteria were used to infect HBMEC seeded on a 150-mm-diameter dish at an MOI of 500 for 4 h. Bacterial samples were prepared as described previously (33). In brief, harvested bacteria were suspended in 1 ml H 2 O and were centrifuged at 10,000 ϫ g at 4°C for 2 min. This procedure was repeated 3 times. The resultant pellets were suspended in 100 l of 0.5% sulfosalicylic acid, and bacteria were lysed using three freeze/thaw cycles, briefly sonicated, and centrifuged again at 14,000 ϫ g at 4°C for 10 min. The glutathione concentration in the supernatant was measured with a total glutathione quantification kit (Dojindo, Japan). Protein concentrations were measured with a BCA protein assay kit. Glutathione concentrations in bacteria were expressed as means Ϯ SD of nanomoles of glutathione per milligram of bacterial protein.
Assay of sensitivity to H 2 O 2 and paraquat in vitro. N. meningitidis strains grown on GC agar plates at 37°C overnight in 5% CO 2 were suspended in 1.5 ml GCB, and the bacterial suspension was prepared in 1 ml GCB at an OD 600 of 0.01, which corresponded to 1 ϫ 10 7 bacteria/ml. The indicated concentration of H 2 O 2 (0.25, 0.5, or 1 mM) or paraquat (2, 5, or 50 mM), which represents the conditions normally applied to in vitro neisserial analyses (51,(84)(85)(86)(87), was added to 1 ml of the bacterial suspension and incubated at 37°C for 15 min (for H 2 O 2 ) or for 60 min (for paraquat) with gentle shaking. Surviving bacteria were counted as CFU by plating on GC agar plates after appropriate dilutions. The CFU of bacteria not treated with H 2 O 2 or paraquat at time zero was defined as 100%, and the survival rate was calculated using the following formula: 100 ϫ (CFU treated with H 2 O 2 or paraquat)/(CFU not treated with H 2 O 2 or paraquat).
Monitoring of N. meningitidis growth in liquid media. N. meningitidis strains grown on GC agar plates at 37°C overnight in 5% CO 2 were suspended in 10 ml buffer A and centrifuged at 7,000 ϫ g for 2 min. The bacterial pellet was resuspended in 10 ml buffer A and centrifuged again. This procedure was repeated 3 times. The resultant pellets were suspended in 5 ml buffer A. Bacterial suspensions were prepared at an OD 600 of 0.1 in 10 ml GCB or SM supplemented with the indicated concentration of cysteine. Since the cysteine solution was made from cysteine HCl salt, the same amount and concentration of NaOH solution were added for neutralization. Bacteria were cultivated at 37°C with shaking at a rate of 160 rpm/min. Meningococcal growth was monitored as OD 600 by the use of a SmartSpec Plus spectrophotometer (Bio-Rad).
Statistical analyses. Results are expressed as means Ϯ SD. Results of determinations of adhered and internalized bacterial numbers, ratios of internalized/adhered bacteria, glutathione measurements, L-cysteine uptake measurements, and meningococcal survival measurements in HBMEC were compared using the two-tailed Student's t test, and P values of Ͻ0.05 were considered to be significant.