Implications of the expression of Enterococcus faecalis citrate fermentation genes during infection

Citrate is an ubiquitous compound in nature. However, citrate fermentation is present only in a few pathogenic or nonpathogenic microorganisms. The citrate fermentation pathway includes a citrate transporter, a citrate lyase complex, an oxaloacetate decarboxylase and a regulatory system. Enterococcus faecalis is commonly present in the gastro-intestinal microbiota of warm-blooded animals and insect guts. These bacteria can also cause infection and disease in immunocompromised individuals. In the present study, we performed whole genome analysis in Enterococcus strains finding that the complete citrate pathway is present in all of the E. faecalis strains isolated from such diverse habitats as animals, hospitals, water, milk, plants, insects, cheese, etc. These results indicate the importance of this metabolic preservation for persistence and growth of E. faecalis in different niches. We also analyzed the role of citrate metabolism in the E. faecalis pathogenicity. We found that an E. faecalis citrate fermentation-deficient strain was less pathogenic for Galleria mellonella larvae than the wild type. Furthermore, strains with deletions in the oxaloacetate decarboxylase subunits or in the α-acetolactate synthase resulted also less virulent than the wild type strain. We also observed that citrate promoters are induced in blood, urine and also in the hemolymph of G. mellonella. In addition, we showed that citrate fermentation allows E. faecalis to grow better in blood, urine and G. mellonella. The results presented here clearly indicate that citrate fermentation plays an important role in E. faecalis opportunistic pathogenic behavior.


Introduction
Since all living organisms contain a certain intracellular level of citrate, this organic acid is commonly found in nature. In addition, citrate is extensively used as preservative in food and beverages. Different bacteria are able to utilize citrate via the citrate lyase pathway. This metabolism consists primarily of a transport system, which incorporates citrate into the cell, a citrate a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 more difficult to treat [23]. For this reason, new insights into the field of enterococci infection prevention are needed, and continuous knowledge generation regarding virulence-associated factors is mandatory. Well characterized virulence factors include the cytolysin CylL, the aggregation substance Agg, the metalloendopeptidase GelE, the extracellular surface protein Esp, the cell surface protein EfaA, the serine protease SprE, the adhesion to collagen protein Ace, fibrinogen Ebp and collagen Acm [26]. Experiments with animal infection models are often useful for identification of differentially expressed genes that could act as virulence traits or fitness factors [27]. Despite this, infection mechanisms, especially the transcriptional modulation occurring in living hosts, are still poorly understood.
In this study, we analyzed the role of citrate fermentation in the pathogenicity of E. faecalis using the insect G. mellonella as infection model. Citrate metabolism deficient-E. faecalis JH2-2 strains resulted less virulent than the wild type, suggesting that this process could be important for this bacterium opportunistic behavior. We also found that an active citrate metabolism allows E. faecalis to grow better in blood and urine, where citrate is present. Furthermore, an α-acetolactate synthase-deficient strain was also less virulent than the wild type suggesting that the increase of internal and external pH provided by pyruvate degradation could promote E. faecalis survival during infection.

Bioinformatic analysis
Enterococcus whole genome sequences were downloaded from the NCBI Refseq database (December 2017). Duplicated genomes were detected and removed prior to analysis. Proteins belonging to E. faecalis TX4000 citrate metabolism were used as queries in local tblastn searches. Cut-off values for searches in Enterococcus genomes were set to 70% of sequence coverage with >70% or 57% amino acid identity for CitE and CitF, respectively. Cut-off values for searches carried out in E. faecalis genomes were 70% of sequence coverage with >85% amino acid identity.

Bacterial strains and cultures
Bacterial strains and plasmids used in this study are listed in Table 1. E. coli strain DH5α was used as an intermediate host for cloning, and E. coli EC101 was used as host for pGhost9 constructs. E. coli strains were grown at 37˚C under aerobic conditions in Luria-Bertani medium (LB), or on LB agar plates. Ampicillin (100 μg/ml), or erythromycin (150 μg/ml, 250 μg/ml) were included in the medium to select cells harboring ampicillin-or erythromycin-resistant plasmids.
For urine or blood growth experiments, E. faecalis strains were grown overnight in LB supplemented with glucose 0.5% (LBG), at 37˚C. Cultures were subsequently diluted 100 x in 10 ml pre-warmed LBG, and further incubated at 37˚C. When an OD 600 value of 0.1 was reached, cultures from each strain were centrifuged (10000 x g for 2 min) and resuspended in sterile human urine [4] or defibrinated mouse blood [3] and incubated at 37˚C. Samples were collected at different times to determine CFU/ml or prepared for observation by fluorescence microscopy.

Construction of E. faecalis JH2-2 mutant strains
The E. faecalis JH2-2-Citstrain was constructed by interrupting the oadD gene by a single recombination event using the thermosensitive vector pGhost9 [28]. A fragment of 417 bp comprising the 3´end of oadH and the 5´end of oadD was amplified by PCR using chromosomal DNA of E. faecalis JH2-2 as template. Forward primer EfoadH (5´-GGGCTGTCAGA AGAAGCTTAGCTAGTTG-3´) introduced a HindIII site, while reverse primer EfoadD_Up (5´-ACATGAATTCCTGTTACCGTACCTG-3´) introduced an EcoRI site. After digestion, the PCR product was ligated into the corresponding sites of the pGhost9 vector. The resulting plasmid, named pmCit (Table 1), was introduced into E. coli EC101, isolated, and then electroporated into the E. faecalis JH2-2 strain. Citrate-deficient strain JH2-2-Cit -(Table1) was constructed as described in [18], and insertion verified by PCR.
pBV153 plasmid for complementation was constructed as follows. The promoter region PcitM was PCR amplified with primers EcoPr (5´-GTAGATGAATTCCAAAAAAATAATG CA-3´) and PrNde (5´-GATCAACCATATGTCTTCTTTCCTAAT-3´) using pBM02 plasmid as template (Table 1) [30]. The PcitM PCR product was cloned in pGemT-easy (Promega), this resulting plasmid was subsequently digested with EcoRI, and the released fragment was ligated in the EcoRI site of pBM01 plasmid (Table 1) [30]. Desired PcitM orientation was determined by restriction analysis and sequencing in the University of Maine, DNA sequencing Facility, US DNA Sequencing.

Construction of a fluorescent reporter plasmid
The plasmid bearing the promoter-gfp and -cherry transcriptional fusion is derived from the pTLGR plasmid [29]. The promoter region of the citrate divergent operons citHO and oadHDB-citCDEFX-oadA-citMG, was amplified by PCR with primers BamHprom-Up (5´-AGGGGATCC ATTACTAAAGATGTAAAC-3´) and BamHprom-Lo (5´-TTAGGATCCTAAATATTCTTTC CC-3´) which introduced a BamHI restriction site, using chromosomal DNA of E. faecalis JH2-2 as template. The fragment was digested with BamHI and ligated in the same site of pTLGR plasmid. Fragment orientation was determined by PCR, and confirmed by sequencing. After isolation, the pTLGR-Pcit plasmid was electroporated into the E. faecalis JH2-2 strain.

Infection and survival experiments
E. faecalis strains were grown overnight in LB medium without shaking at 37˚C. Then, the bacterial cultures were diluted in 50 ml of LB medium to a final OD of 0.1 and grown at 37˚C without shaking, until exponential phase was reached (4,5h). The cultures were centrifuged, resuspended in PBS 1X + 20% glycerol and frozen at -80˚C. Inoculums prepared in this way were plated to determine CFU/ml. Larvae were inoculated by direct injection into the hemocoel using a Hamilton syringe 705 equipped with a repeating dispenser. Each larvae group of 16 individuals was inoculated with a fixed CFU/larva ratio ranging from 9 x 10 6 to 6 x 10 7 . After injection, larvae were incubated at 30˚C. Survival of the individuals was monitored every two to four hours, by direct observation and gently touching non-motile larvae to evidence movement response or confirm death. Survival of the larvae group was evaluated until 72 h post-infection.

Determination of in vivo bacterial loads
For bacterial count in hemolymph, 45 larvae were inoculated with a total of 9 x 10 6 CFU/larva for each strain. Fifteen larvae were separated from the group at 0, 24 and 48 h post-inoculation, three sub-groups for each time were formed hemolymph extracted, pooled together and suspended in cold Insect Physiological Saline (IPS) buffer (150 mM sodium chloride, 5 mM potassium chloride, 10 mM Tris HCl pH 6.9, 10 mM EDTA and 30 mM sodium citrate) [33] with Triton X-100 0.03% v/v. Pools were vortexed, incubated at room temperature for 15 seconds and then plated in LBG for CFU counting. The assay was carried out in duplicate, with three technical replicates for each time point.

Hemocyte collection and fluorescence assays
At different times, hemolymph was extracted and pooled from larvae inoculated with E. faecalis strains JH2-2/pTLGR or JH2-2/pTLGR-Pcit carrying the fluorescent plasmid. Pools were diluted 1/100 in cold IPS, centrifuged at 700 x g, washed twice and finally resuspended in PBS 1X containing glucose 5 mM, MgCl 2 1 mM, and CaCl 2 0.5 mM [33]. Samples were fixed with formaldehyde and slides mounted in VectaShield (Vector Laboratories, BIOARS S.A., Argentina). Images were acquired in a Nikon E600 microscope, using a 60×1.4 WD Plan-ApoVC objective with a Nikon DXM1200 digital camera using ACT-1 software. Fiji software [34] was used to pseudocolor images.

Data analysis
Data analysis was done using R software. Survival curves were constructed according to the Kaplan-Meier method, using the LogRank and Holm-Sidak tests [35] for multiple comparisons.
To determine differences between bacterial count in blood, urine and hemolymph, a oneway analysis of variance was performed for each time point and the Tukey´s test was selected for multiple comparisons. In our model the fixed effect are the strains and random effect is the time. For the hemolymph experiment the experimental unit is the pool of 5 larvae; for the experiment with blood and urine the experimental unit is each tube where bacteria was grown. CFU/ml data was converted to Log(CFU/ml) and then analysis was performed, Normality of the Log(CFU/ml) data was confirmed with Shapiro-Wilks test. P value was set at 0.05 in all cases.

Diversity of citrate fermentation pathways in the Enterococcus genus
Citrate fermentation-associated genes were searched in silico in the Enterococcus genomes available at the NCBI RefSeq database. A total of 1478 genomes comprising different enterococci species were selected for further analysis. Using the citrate lyase citE and citF genes from E. faecalis TX4000 strain, we found that more than half of the genomes (758) encoded these genes. Consequently, we focused the search on the E. faecalis species. From 501 available genomes, we found that citrate metabolism is present in almost all of them, since 500 genomes encoded the citrate lyase genes. Next, all the genes of the cit cluster in the E. faecalis TX4000 strain were found by BLAST searches in the 500 E. faecalis cit + genomes analyzed indicating that citrate metabolism is highly conserved among E. faecalis strains.
Gene context analysis of Enterococcus genus representative strains revealed that Type I citrate metabolism cluster is present only in E. faecium [8,36] These results suggest that this metabolism could be playing an important role in these microorganisms ability to grow, persist or colonize different niches. In particular, the genes and arrangement described for the E. faecalis JH2-2 strain [18] (Fig 1B) are conserved in all of the strains, despite their different origins (hospital, water, food or plant). In all of the E. faecalis analyzed genomes, the transcriptional activator CitO, responsible for the induction of the cluster in the presence of citrate, the citrate transporter CitH, the citrate lyase complex (CitD, CitE, CitF) and its accessory proteins (CitC, CitG and CitX), the four subunits of the membrane OAD (OadH, OadA, OadD and OadB) and the soluble OAD (CitM) were found encoded.

G. mellonella as a model to study the role of citrate fermentation in virulence
The connection between citrate metabolism and aroma compound production pathways has been extensively studied as well as the regulatory mechanisms involved [8,[18][19][20][21][22]37]. Although many E. faecalis strains are known opportunistic pathogens, the relationship between citrate utilization route and pathogenicity has not been previously analyzed.
Thus, G. mellonella was selected as a suitable infection model to investigate this link. Use of this moth´s larvae as an alternative model has proven useful in simple infection experiments, giving fast and reliable results, which at the same time correlate with those obtained with more traditional models [27,[38][39][40]. Thus, to determine if genes responsible for citrate fermentation are expressed in E. faecalis-infected larvae, a fluorescent probe was used. The divergent promoter region of the cit cluster was cloned in the pTLGR fluorescent reporter plasmid [29], generating the pTLGR-Pcit construct (Table 1). In pTLGR-Pcit, PcitCL (the promoter of the catabolic operon) controls the expression of GFP whereas PcitH (the promoter of the citrate transporter and regulator) controls the expression of the Cherry protein (Fig 2A).
E. faecalis JH2-2 strain was electroporated with pTLGR or pTLGR-Pcit plasmids. Next, E. faecalis strains harboring pTLGR or pTLGR-Pcit plasmids were grown in LB and LB with 0.5% citrate (LBC); to assess promoter induction samples were withdrawn after 16 hs of growth at 37˚C and observed by fluorescence microscopy. As shown in Fig 2B, in LB a few cells showed a faint GFP and Cherry fluorescence indicating a low basal activity of the promoters [18]; on the contrary, as expected in LBC E. faecalis cells showed strong fluorescent signals, demonstrating induction of both promoters. No fluorescence was detected in the empty pTLGR plasmid (not shown).
Next, G. mellonella larvae groups were injected with 1 x 10 7 CFU/larva of E. faecalis JH2-2 pTLGR-Pcit or pTLGR. Hemolymph of sample individuals was extracted at different time points after inoculation and fluorescent bacteria were detected. After 24 hours of inoculation (Fig 2C), evidence of active cit promoters was found, both free in the hemolymph and associated to hemocytes. A strong fluorescent signal is observed around the hemocytes, indicating that several bacteria are in contact with them, at 48 h. Also during the assay, melanization and deterioration of larvae health conditions was observed for individuals inoculated with E. faecalis JH2-2 pTLGR-Pcit or pTLGR strains (Fig 2C). Production of melanin by G. mellonella larvae is part of the insect innate immune response; melanin is synthesized after infection and it is often found around encapsulated microorganisms [41]. In our case, distinctive dark spots were observed after 24 h of infection while at 48h melanization extended to the rest of the larva as a consequence of infection progression. No fluorescence was detected in the hemolymph of larvae inoculated with E. faecalis JH2-2 harboring empty pTLGR plasmid (not shown).

Citrate metabolism deficiency impairs E. faecalis virulence in G. mellonella
Taking into consideration the above data, contribution of the cit cluster to E. faecalis pathogenesis in this insect model was evaluated. Larvae group survival was monitored up to 72 h postinfection and resulting data were analyzed through Kaplan-Meier curves [35,42].
Next, citrate fermenting-deficient E. faecalis JH2-2 strains were used to establish cit genes contribution to this bacterium infection capacity. A JH2-2-Citstrain, impaired at citrate use (final CFU/ml in LBC (1.47 ± 0.01) x10 8 vs (1.89 ± 0.04) x10 8 for the wildtype, S2 Fig), was constructed with an interruption within the cit gene cluster (Table 1). We carried out larvae inoculations with a total of 4 x 10 7 CFU/larva of this strain. Despite the low survival percentage observed for the wild type (WT) strain, almost no larvae mortality was obtained over the course of a 72 h experiment for the JH2-2-Citstrain (Fig 3A). Even with a higher inoculum concentration (6 x 10 7 CFU/larva) this mutant strain resulted barely lethal to G. mellonella (S1 Fig). A representative group of larvae inoculated with the JH2-2-Citstrain can be observed in Fig 3F. Remarkably, these low lethality levels are similar to those observed with the nonpathogenic innocuous L. lactis IL1403 strain. These findings suggest that larvae mortality induced by E. faecalis could be dependent on the presence of an active citrate metabolism.
In order to extend our knowledge about the role of citrate metabolism in E. faecalis infection of G. mellonella, mutants in the subunits of the membrane decarboxylase OAD were used. The JH2-2-OadAstrain cannot metabolize citrate beyond oxaloacetate; on the contrary the JH2-2-OadBstrain is capable of slowly degrading oxaloacetate to pyruvate by the action of the cytoplasmic OadAHD complex [21]. Thus, survival of G. mellonella inoculated with these strains was monitored to analyze involvement of the mOAD complex in E. faecalis virulence. Since lethality is influenced by inoculum concentration, we injected larvae with several CFU values to observe variations in strain pathogenicity. Virulence of the JH2-2-OadBstrain in G. mellonella larvae was tested in a CFU range varying from 9 x 10 6 to 3 x 10 7 CFU/larva. This strain resulted less virulent than the WT when inoculated at CFU of 9 x 10 6 , 1 x 10 7 and 3 x 10 7 To corroborate the role of citrate metabolism in E. faecalis JH2-2 virulence, the mOAD complex mutation was complemented by expressing the OadA subunit in trans using the pOadA plasmid (Materials and methods and Table 1). JH2-2-OadAcells harboring pOadA recovered their ability to metabolize citrate resulting in an increase in final CFU/ml counts for cells grown in LBC ((2.1 ± 0.09) x10 8 CFU/ml vs (1.57 ± 0.02) x10 8 CFU/ml for the mutant, S2  Fig). Once the cit + phenotype was confirmed, the complemented strain (JH2-2-OadA -/ pOadA) was used to inoculate G. mellonella larvae, again a range of CFUs were assayed. As shown in Fig 3D, JH2-2-OadA -/pOadA cells were as virulent as the WT (P>0.001) for the lowest concentrations (9 x 10 6 and 1 x 10 7 CFU/larva), with values of 3 x 10 7 CFU/larva the complemented strain was slightly more virulent than the WT (P<0.001) (S1 Fig). A group of larvae inoculated with 1 x 10 7 CFU/larva of the complemented strain is shown in Fig 3F where body melanization to the same extent to that of the WT strain can be observed.
Repizo et al [21] showed that the JH2-2-OadAstrain cannot metabolize citrate and, consequently, does not show growth improvement in the presence of this compound. On the other hand, in LB growth medium supplemented with citrate, JH2-2-OadBis able to reach final OD 600 levels similar to those of the parental strain, with a delay in the beginning of the second growth phase, i.e., when citrate pathway is induced [18]. Thus, despite the ability of JH2-2-OadBto grow in citrate-containing media (LBC), it seems that under infection conditions, the delay observed in its growth is strongly detrimental for the cells and they cannot cope with the immune system of G. mellonella as well as the WT. This leads to higher larvae survival rates. Nonetheless, citrate metabolism deficiency makes strains less virulent than the WT, confirming the observation made for the JH2-2-Citstrain.
Given the connection between aroma compound production from pyruvate and citrate metabolism (Fig 1C), an α-acetolactate synthase deficient strain (JH2-2-AlsS -) was also used in survival experiments and compared with the other strains tested. The JH2-2-AlsS - [37] strain is unable to further convert pyruvate into α-acetolactate. As a consequence, this strain exhibits growth deficiency in the presence of pyruvate at pH 5.5, and is unable to grow at pH 4.5. This indicated that the pathway involved in aroma compound generation is an important mechanism which allows growth in acidic media [37]. Inoculums of JH2-2-AlsSwere prepared and used to inject G. mellonella larvae. After examination of larvae health status during 72h, KM curves were plotted (Fig 3E). When 1 x 10 7 CFU/larva were injected, JH2-2-AlsSwas less virulent than the WT (P<0.001). With a 3 x 10 7 inoculum, the AlsS deficient strain was as virulent as the WT (P = 0.37). suggesting that the "pyruvate to acetoin" pathway could be affecting E. faecalis virulence.

Citrate metabolism enhances E. faecalis growth in insect hemolymph, blood and urine
In previous sections we showed that citrate metabolism plays an important role during E. faecalis infection of G. mellonella. Consequently, an analysis of cit cluster induction in other animal fluids associated to common diseases caused by E. faecalis was performed. To this end, E. faecalis JH2-2-pTLGR or pTLGR-Pcit was grown in mouse defibrinated blood, and activity of Pcit promoters was analyzed at different time points. GFP and Cherry fluorescent signals were detected as early as 2 hours after exposition to blood (Fig 4A), indicating that both promoters were quickly induced. No fluorescence was detected at time 0 for pTLGR-Pcit or with the empty pTLGR plasmid during a 24 h assay. When E. faecalis JH2-2 pTLGR-Pcit was incubated in human urine, induction of Pcit promoters was observed after 2 h with increasing signal during the time assayed (Fig 4B). These results suggest that citrate metabolism in E. faecalis is probably being induced during blood and urine infection and could have a role in infection of higher animals as well as in G. mellonella. To confirm this hypothesis in G. mellonella, bacterial load after infection was determined. Larvae groups were infected with the WT and citrate metabolism-mutant strains. At different time points, hemolymph was extracted from sample individuals, diluted and plated onto an appropriate medium to determine total bacterial CFU/ ml of hemolymph.
Comparative analysis of CFU/ml at times 0, 24 and 48 h post-inoculation, allowed us to confirm that, after 24h of inoculation, the JH2-2 WT strain can effectively reach higher CFU values in vivo than JH2-2-Cit -, JH2-2-OadAand JH2-2-OadB - (Fig 5A, P<0.01 for all the WT vs mutant comparisons). Moreover, at 24 and 48 h no statistically significant differences were found for the total bacterial count in the hemolymph between mutant strains (Fig 5A, P>0.4 for all the comparisons).
These results suggest that citrate metabolism-deficient strains are unable to proliferate as well as the WT in the hemolymph and are probably eliminated by the larvae immune system, thus provoking a less aggressive infection.
Since we showed that citrate genes are induced in blood, we analyzed the growth behavior of the various mutants and the OadAcomplemented strain to determine if their differential capacity for citrate utilization can be correlated with a differential growth. As shown in Fig 5B, no differences in growth were found among strains up to 6 h (P = 1). But, after 24 h of growth, E. faecalis JH2-2 and JH2-2-OadA -/pOadA reached a higher cell density than citrate metabolism mutants (Fig 5B). Statistical analysis indicated that these differences were significant (P<0.05), showing that induction of citrate metabolism could give E. faecalis WT strain an advantage in the overall growth process.

Discussion
E. faecalis is a natural member of the gastro-intestinal microbiota of warm-blooded animals and insect guts. The intrinsic ability of this bacterium to resist harsh conditions allows it to persist in hospital environments and to survive host defenses [43]. Nevertheless, many of these microorganisms are associated with food production and some strains also possess probiotic properties [22,44,45]. Accordingly, E. faecalis is often found in diverse types of fermented food products, making them part of the human diet around the world [22,45]. In this context, citrate fermentation is a desired trait of lactic acid bacteria since it contributes to aroma development [2,8,37].
In this study, citrate fermentation operons were detected in 758 out of 1478 genomes analyzed comprising 17 species of the Enterococcus genus. Although citrate-negative strains were found in some of the species remarkably all of the E. faecalis genomes analyzed encoded the thirteen genes necessary for citrate fermentation. Furthermore, this Type II cit cluster was found conserved independently of strain origin, suggesting the importance of pathway preservation in this species.
In this work, we further demonstrate that the presence of cit genes contributes to the pathogenic behavior of E. faecalis in the G. mellonella model. Citrate metabolism-defective strains showed a reduced capacity of infection. When the cit cluster is interrupted (JH2-2-Citstrain), E. faecalis ability to metabolize citrate is abolished and, as a consequence, a remarkably lower mortality of G. mellonella was observed (Fig 3A). On the other hand, in mOAD mutants, citrate metabolism is affected to different extents; the oadA mutant allows the conversion of citrate to oxaloacetate but not to pyruvate (Fig 1C) [21], whereas the oadB mutant conserves a soluble OAD complex (OadADH) which could allow the complete conversion of citrate to pyruvate, but at a lower rate [21]. These metabolic differences between the JH2-2-Citstrain and the mOAD mutants can probably account for the observed differences in the G. mellonella survival rates after inoculation (Fig 3). In addition, an α-acetolactate synthase deficient strain (JH2-2-AlsS -) was also found less virulent than the E. faecalis WT strain.
Examples of metabolic pathways associated with virulence in E. faecalis are not common, Maadani et al. reported that an E. faecalis strain unable to metabolize ethanolamine was less virulent in the C. elegans model [46]. Ethanolamine is found in the intestine and can be used as a source of both carbon and nitrogen. The capacity to use this organic compound has been related with intestinal pathogens, for example, Salmonella species and Listeria monocytogenes [47].
Its ability to survive in hospital environments and to infect immunocompromised patients has made E. faecalis a commonly found cause of bacteremia and urinary tract infections [23][24][25]48]. In fact, many of the sequenced enterococci strains available at the NCBI GenBank database were isolated from blood or urine of hospitalized patients. In this work, activation of the citrate degradative pathway was observed in larvae but also in urine and blood (Fig 4). In this media, concentration of preferred carbon sources such as glucose or fructose may fluctuate between 1 to 6 mM or 0.01 to 0.5 mM, respectively [49]. Under these conditions, only when glucose concentration reaches the higher value its repressive effect could be relevant on the cit cluster reducing approximately 50% the activity of PcitHO promoter and 15% the activity of the catabolic operon PcitCL promoter [20]. Nevertheless, this concentration could allow cometabolism of citrate and glucose. In urine, glucose concentration can reach 1.1 mM [49] a value too low to cause citrate pathway repression. As a consequence, citrate present in both media could be used as a carbon and energy source during E. faecalis infection.
It has been proposed that the ability of E. faecalis to cause infection would not only implicate an organized regulation of several virulence factors and expression of genetic determinants, but also an adaptation of the bacterial cell physiology during the infection process [3]. This suggests that various metabolic pathways could contribute differently to virulence and its ability to persist in diverse environments.
Transcriptomic data of E. faecalis grown in blood and urine has shown that several known virulence factors, such as the fsrB (EF1821), gelE (EF1818), cpsC (EF2493), ace (EF1099) and efaA (EF2076) were modulated in their expression under both growth conditions [3,4]. Also, the data reflected the cells necessity for fast adjustments to withstand nutritional changes [3,4]. In fact, the iron, manganese, sugar and oligo-peptide ABC-transport systems were found up-regulated. Furthermore, in agreement with our results the gene cluster responsible for citrate metabolism (EF3315-27) was up-regulated ( Fig 1C). These results correlate with those obtained in a recent study of E. faecalis infection in subdermal chambers, where citrate metabolism induction was also observed [50]. Transcription of pyruvate metabolism enzymes was also modified during growth in blood and urine: α-acetolactate synthase (alsS, EF1213), αacetolactate decarboxylase (alsD, EF1214), pyruvate dehydrogenase complex (pdH, EF1353-56) and lactate dehydrogenase (ldH, EF0255) were induced ( Fig 1C); whereas phosphotransacetylase (EF0949), acetate kinase (acK, EF1983, pyruvate formate lyase (pflB, EF1613), alcohol dehydrogenase (adhE, EF0900) were repressed (Fig 1C). The up-regulation of citrate and some enzymes of pyruvate metabolism favor the proton consuming reactions and also contribute to increase the Acetyl-CoA pool that can be used as a precursor of the FASII pathway also up-regulated in blood and urine [3,4].
These analyses demonstrated that E. faecalis is capable to adapt its physiology depending on its surroundings and that the induction of virulence factors is probably a piece of a larger physiological adaptation displayed when E. faecalis invades a niche as a pathogen.
In this work, we demonstrated that citrate metabolism gives E. faecalis an advantage to grow in blood and urine. We have also proved that an active citrate metabolism allows the bacterium to proliferate inside living G. mellonella larvae to higher CFU/ml hemolymph counts. This suggests that citrate metabolism is a key feature in E. faecalis, possibly allowing this microorganism to overcome adverse conditions resulting in better growth conditions. Blancato.