Overexpression of Enterococcus faecalis elr operon protects from phagocytosis

Reagents

All reagents were obtained from Sigma-Aldrich (St. Louis, MO), unless otherwise stated.

Bacterial strains and plasmids

Bacterial strains and plasmids used in this work are listed in Table 1. E. faecalis strains were grown in M17 medium supplemented with 0.5 % glucose (GM17) at 37 °C
without aeration. Escherichia coli strains were grown aerobically in Luria-Bertani medium at 37 °C. Plasmid constructions
were first established in E. coli TG1 strain and then transferred into E. faecalis by electrotransformation using a Bio-Rad Gene Pulser Electroporator (Bio-Rad Laboratories).
E. faecalis strains expressing green-fluorescent protein (GFP) were obtained by electroporation
with pMV158-GFP plasmid 24]. Recombinant bacteria were selected by the addition of antibiotics as follows: for
E. faecalis chloramphenicol 4 ?g/ml, tetracycline 4 ?g/ml and erythromycin (Ery) 30 ?g/ml; for
E. coli, chloramphenicol 10 ?g/ml and ampicillin 100 ?g/ml. DNA manipulations were performed
as previously described 25].

Table 1. Strains and plasmids used in this work

Cell line and culture conditions

The RAW 264.7 mouse macrophage cell line (ATCC®?TIB-71) was maintained in DMEM supplemented
with 10 % heat-inactivated fetal bovine serum (FBS) and 2 mM l-glutamine 26]. For phagocytosis assays, cells were seeded at 0.5 × 10⁶/well into 12-well tissue culture plates (TPP, Domique Dutscher, Brumath, France)
and incubated overnight at 37 °C under 6 % CO2. For microscopy experiments, cells
were cultured in tissue culture plates containing poly-L-lysine pretreated coverslips
for microscopy or on Lab-tek chamber slides (Nunc, Domique Dutscher). Comparative
analysis of phagocytosis using either heat-inactivated serum or serum-free media (Macrophage-SFM,
GIBCO, Invitrogen) did not show differences (data not shown). Thus, for practical
reasons we decided to use heat-inactivated serum in all experiments of this work.

Generation of anti-ElrA rat polyclonal antibodies

Recombinant ElrA was purified to produce polyclonal rat anti-ElrA antibodies by Proteogenix
(Oberhausbergen, France). Briefly, a DNA fragment encoding elrA was PCR-amplified from E. faecalis chromosomal DNA using OEF275 and OEF276 primers (Table 2). The PCR product was digested with BamHI/PciI and cloned into purified BamHI/NcoI-digested pET2817 vector backbone, resulting in plasmid VE14047 which was transformed
into E. coli BL21(?DE3). For ElrA production and purification, the resulting recombinant strain
was cultured at 25 °C and induced with 1 mM of IPTG (isopropyl ?-D-1-thiogalactopyranoside)
for 5 hrs. Recombinant 6xHis::ElrA protein was purified under denaturing conditions
on Ni-NTA columns using the QIAexpress kit (Qiagen, Courtaboeuf, France).

Table 2. Primers used in this study

Construction of mutant and over-expressing strains of E. faecalis

The E. faecalis elrA gene is part of a five-gene operon elrA (OG1RF_12055), elrB (OG1RF_12054), elrC (OG1RF_12053), elrD (OG1RF_12052), and elrE (OG1RF_12051) (Fig. 1), encoding putative surface proteins of unknown function. To circumvent the lack
of ElrA production in vitro, we constructed a genetically modified E. faecalis strain harboring the constitutive promoter PaphA3 (hereafter named P⁺), instead of the native promoter, PelrA, upstream of the whole elr operon (i.e., elrA-E, Fig. 1). This genetically modified strain (called P⁺–elrA–E), was constructed by a double cross-over event using the pGhost9 plasmid 27]. Briefly, two overlapping fragments were PCR-amplified from E. faecalis OG1RF chromosomal DNA with primers OEF343/OEF344 and OEF345/OEF346 (Table 2). The two PCR products were then fused by PCR using the external primers OEF344/OEF346,
and the resulting product was cloned into purified XhoI-BamHI-digested pACYC177 vector, resulting in plasmid pVE14142. The PaphA3 promoter was PCR-amplified with primers Vlac1 and Vlac2 from pTCV-lac(PaphA3) plasmid 16]. An EcoRI-BamHI fragment, containing the PaphA3 promoter, was then cloned into EcoRI-BglII-digested pVE14142 vector to obtain plasmid pVE14145. Then, a 2.3 kb XhoI-EaeI fragment from pVE14145 plasmid (containing the promoter and the targeted region)
was cloned into pGhost9 vector to generate the final vector pVE14146. This plasmid
was established in E. faecalis OG1RF strain and a markerless insertion of PaphA3 upstream of the elrA-E operon was performed as previously described 1]. Correct integration of PaphA3 into the chromosomal locus was confirmed by sequencing. All the following mutant
constructs were performed using P⁺?elrA-E strain as a recipient in order to have the same genetic background (Table 1). For the construction of a strain expressing only elrA under the control of PaphA3 promoter, a fused DNA fragment using primers OEF13/OEF595 and OEF596/OEF49 amplified
from OG1RF strain DNA was cloned into pGEM-T easy vector (Promega) to generate pVE14179.
A 4.5 kb PstI fragment was then cloned into PstI-digested pGhost9 to obtain plasmid pVE14450 and established in P⁺–elrA-E strain to obtain the P⁺–elrA-?elrB-E strain. For the construction of a strain expressing elr operon lacking elrA, a 6.6 kb Bst/ApeI fragment from pVE14009 was cloned into Bst/ApeI-digested pVE14146 vector, resulting in pVE14457. This plasmid was established in
P⁺–elrA–E strain to obtain the P⁺-?elrA strain. To inactivate the whole elr operon (i.e., elrA–E), we first generated an in-frame deletion of the whole operon by PCR. For this, we
used OEF15/OEF18 and OEF598/OEF49 primers described for the first PCR. The two PCR
products were fused by PCR using external primers OEF49/OEF15, and the resulting product
was cloned into pGEM-T, resulting in plasmid pVE14178. A BstAPI/AatII 890bp DNA fragment from pVE14178 was then cloned into pVE14145 to generate pVE14455.
The final plasmid was generated by cloning a 4.5 kb XhoI DNA fragment from pVE14455 vector into XhoI-digested pGhost9 to obtain pVE14456. This plasmid was established in P⁺–elrA-E strain and the resulting strain was named P⁺-?elrA-E. All expected modifications or deletions were confirmed by sequencing.

Preparation of protein extracts, SDS gel electrophoresis, and immunoblot analysis

Total protein extraction from bacteria, SDS-PAGE, and Western blot immunodetection
were carried out using standard methods (24) with some modifications. Strains were
grown at 37 °C overnight and then diluted 100-fold and grown under the same conditions
to an OD₆₀₀~1. Protein crude extract was obtained by trichloroacetic acid (TCA) precipitation
by mixing 800 ?l of bacterial culture with 200 ?l of ice cold TCA solution (100 %
w/v). The protein pellet was then obtained by centrifugation and recovered directly
into SDS sample buffer. Anti-ElrA antibody was used at a dilution of 1:500 for Western
blot immunodetection.

RNA isolation and Northern blotting

Total RNA was extracted as previously described 28]. Northern blots were performed on 40 ?g of total RNA separated on a 0.9 % denaturing
agarose gel as previously described 29]. Specific oligonucleotides OEF9 and OEF212 were used to detect elrA transcripts. Oligonucleotides were labelled with [?-32P]-ATP and T4 polynucleotide
kinase (NEB Biolabs) according to the recommendations of the manufacturer (NEB Biolabs).
Analysis was performed from RNA extracted from two independent experiments.

Phagocytosis assay with RAW macrophages

Fluorescent E. faecalis were grown on GM17 plates containing erythromycin (GM17-Ery), with a single colony
subsequently being selected and grown overnight in GM17-Ery broth. A 100 ?l aliquot
was then transferred into 10 ml of fresh GM17-Ery and incubated until cultures reached
an OD₆₀₀ ~1. Bacteria were then pelleted by centrifugation, washed three times with PBS, and
adjusted to a concentration of 1 × 10⁹ CFU/ml in supplemented DMEM. The number of bacteria present in each suspension was
confirmed by plating onto solid GM17-Ery.

For phagocytosis experiments, adherent RAW cells were infected with fluorescent E. faecalis at a multiplicity of infection (MOI) of 100:1 (bacterium/cell ratio). After 30 min
of interaction, cells were washed twice with PBS, recovered with cell dissociation
buffer (GIBCO, Invitrogen), washed again, and finally fixed in 3 % paraformaldehyde
(PFA) solution. Fluorescence of RAW cells due to infecting bacteria was detected by
a flow cytometer in the FL-1 channel. The phagocytosis index (PI) was calculated using
the percent of fluorescent macrophages after E. faecalis wild-type (WT) strain infection and applying the following formula: PI = (percent
of fluorescent macrophages after infection X 100 /percent of fluorescent macrophages
after WT infection) 30], 31]. Results are expressed as the mean ± SEM from three independent experiments usually
performed in duplicate or triplicate.

Bacterial adhesion assay

To separate adhesion from subsequent steps of phagocytosis, cells were pretreated
30 min with 1 ?g/ml of cytochalasin D (CytD), an actin polymerization inhibitor, as
described 17]. A CytD (1000X) stock solution in DMSO was prepared according to manufacturer’s recommendations
and stored at ?20 °C. DMEM supplemented medium (see above) was used to dilute stock
solution. RAW cells were seeded at 1 × 10⁶/well into 6-well tissue culture plates (TPP, Dominique Dutscher) and incubated O/N
at 37 °C under 6 % CO2. Macrophages pre-treated with CytD were first washed twice
with fresh medium and then infected at a MOI of 100:1, similar to phagocytosis analysis
above; CytD-untreated and uninfected macrophages were used as negative controls. Fluorescence
in RAW cells due to infecting bacteria was detected by flow cytometry. Adhesion Index
(AI) = (percent of GFP⁺ macrophages pre-treated with CytD, after infection by the E. faecalis mutant strain X 100/percent of GFP⁺ macrophages pre-treated with CytD after WT infection).

Fluorescence and electron microscopy

Raw macrophages were seeded in 12-well cell culture plates on a glass slide and infected
with GFP-labeled E. faecalis wild-type (WT) or P⁺–elrA-E strains at a MOI of 1:100, with uninfected macrophages serving as negative control.
After 30 min of interaction, macrophages were washed twice with PBS, fixated and immunolabeled
with Streptococcus group D antiserum (BD Diagnostics, Le Pont de Claix, France) as previously described
4]. Fluorescence was examined using a Carl Zeiss microscope (Axiovert 200 M, in the
ApoTome mode) at MIMA2 platform (INRA, Jouy en Josas). Images were processed with
Axiovision version 4.6 (Carl Zeiss).

Imaging of bacterial-cells interaction was performed using a Hitachi S-4500 scanning
electron microscope (SEM) at the MIMA2 imaging platform. Macrophages were seeded in
12-well cell culture plates and infected with either E. faecalis wild-type (WT) or P⁺–elrA-E strains at a MOI of 1:100, with uninfected macrophages serving as negative control.
After 30 min of interaction, macrophages were washed twice with PBS, recovered with
cell dissociation buffer (GIBCO, Invitrogen), washed again, and suspended in a fixative
solution and treated as previously described 32].

Preparation of bacterial samples for transmission and scanning electron microscopy
was performed as previously described 32], 33]. Thin-sections and negative-stains were observed with a Zeiss EM902 electron microscope
operated at 80 kV (MIMA2 – UR 1196 Génomique et Physiologie de la Lactation, INRA,
plateau de Microscopie Electronique, 78352 Jouy-en-Josas, France). Microphotographies
were acquired using MegaView III CCD camera and analyzed with the ITEM software (Eloise
SARL, Roissy CDG, France).

Microbial adhesion to solvents

Microbial adhesion to solvents (MATS) analysis was carried out as described previously
by Bellon-Fontaine and collaborators 21]. In brief, a single colony of each of the E. faecalis strains studied was subcultured four times in BHI and harvested at stationary phase.
Bacterial cells were centrifuged at 5000 ×g for 8 min and washed twice in 0.15 M NaCl
and re-suspended to a final OD₄₀₀ ~0.8. Bacterial suspensions (2.4 ml) were vortexed for 1 min with 0.4 ml of highest
purity grade chloroform (Sigma-Aldrich), hexadecane (Sigma-Aldrich), ethyl acetate
(Merck), or decane (Merck). The emulsion was left to stand for 20 min to allow complete
phase separation, and the OD₄₀₀ of 1 ml from the aqueous phase was measured. Affinity of the cells for each solvent
(% affinity) = ((OD_f-OD_i)/ OD_i)x100 where OD_i is the initial optical density of the bacterial suspension before mixing with the
solvent, and OD_f the final absorbance after mixing and phase separation. Analysis was performed twice
in triplicate.

Mouse peritonitis model

The mouse experiments were approved by the Institutional Animal Use and Care Committee
at the Università Cattolica del Sacro Cuore, Rome, Italy (permit number Z21, 1 November
2010), and authorized by the Italian Ministry of Health, according to the Legislative
Decree 116/92, which implemented the European Directive 86/609/EEC on laboratory animal
protection in Italy. Animal welfare was routinely checked by veterinarians of the
Service for Animal Welfare.

Virulence of strains OG1RF, ?elrA, and P⁺–elrA-E was tested as described previously 4]. The inoculum size was confirmed by determining the number of CFU on brain heart
infusion agar. Each inoculum was 10-fold diluted in 25 % sterile rat fecal extract
prepared from a single batch as previously described 34]. Groups of 10 ICR outbred mice (Harlan Italy Srl, San Pietro al Natisone, Italy)
were challenged intraperitoneally with 1 ml of each bacterial inoculum, housed five
per cage, and fed ad libitum. A control group of mice was injected with 25 % sterile rat fecal extract only. Survival
was monitored every 3 to 6 h. In another set of experiments, groups of mice were killed
24 h postinfection, and livers and spleens were removed, weighed, homogenized, and
serially diluted in saline solution for colony counts.

Statistical analysis

Statistics were performed using GraphPad Prism (Version 4.00 for Windows, GraphPad
Software, San Diego California, USA). One-way analysis of variance (ANOVA) was followed
by Dunnett’s multiple-comparison test when comparing multiple groups for one factor.
For animal experiments, survival estimates were constructed by the Kaplan-Meier method
and compared by log rank analysis, and comparisons with P values of 0.05 were considered to be significant.