Bacterial Strains, Plasmids and Media

Bacterial strains and plasmids used in this study are listed in Table 1. Bacteria were routinely grown at 37°C in Luria-Bertani (LB) broth with aeration or on Congo red agar [26]. Antibiotics used were as follows: 50 µg ampicillin (Amp) ml⁻¹; 25 µg chloramphenicol (Cml) ml⁻¹; 50 µg kanamycin (Kan) ml⁻¹.

Table 1

Bacterial strains and plasmids.

Strain or plasmid	Relevant characteristics	LPS*	Source/reference
E. coli K-12
DH5α	endA hsdR supE44 thi-1 recA1 gyrA relA Δ(lacZYA-argF) U169 [80dlacΔ(lacZ) M15)	R	Gibco-BRL
S. flexneri 2a
2457T	wild type strain	S	[26]
ETRM22	2457T icsP- mutant; KanR	S	This study
ETRM230	2457T rmlD::kanR mutant; KanR	R	This study
ETRM233	2457T rmlD- mutant; KanR	R	This study
ETRM240	2457T icsP−/rmlD- mutant; KanR	R	This study
ETRM143	ETRM22 (pIcsP)	S	This study
ETRM117	ETRM22 (pIcsPHA)	S	This study
ETRM118	ETRM22 (pBAD30)	S	This study
ETRM243	ETRM240 (pIcsPHA)	R	This study
ETRM245	ETRM240 (pBAD30)	R	This study
RMA4376	ETRM243 (pRMA727)	S	This study
RMA4377	ETRM243 (pACYC184)	S	This study
Plasmids
pCACTUS	Suicide vector; CmlR; 30°C		[27]
pCACTUS-icsP::kanR	pCACTUS with icsP::kanR gene		This study
pSL1180	Cloning vector; AmpR		[46]
pSL1180-icsP::kanR	pSL1180 with icsP::kanR gene		This study
pACYC184	Cloning vector; CmlR, TetR		[47]
pKTUWE	pACYC184 with kanR gene; KanR		[36]
pKD4	Vector containing FRT-flanked kanR gene		[32]
pKD46	Red lambda plasmid; AmpR; 30°C		[32]
pCP20	FLP recombinase; AmpR, CmlR; 30°C		[32]
pGEMT-Easy	Cloning vector; AmpR		Promega
pGEMT-Easy::icsP HA	pGEMT-Easy with icsPHA gene; AmpR		This study
pGEMT-Easy::kanR	pGEMT-Easy with kanR gene; AmpR		This study
pBAD30	Arabinose-inducible pBAD promoter vector, AmpR		[39]
pIcsP	pBAD30 with icsP gene; AmpR		This study
pIcsPHA	pBAD30 with icsPHA gene; AmpR		This study
pRMA718	pUC1318 containing S. flexneri rfb region; AmpR		[26]
pRMA727	pACYC184 with rmlD gene; CmlR		[26]

Open in a separate window

^*S, smooth LPS; R, rough LPS.

DNA Methods

E. coli K-12 DH5α was used for all cloning experiments. DNA manipulation, PCR, transformation and electroporation was performed as previously described [27], [28]. Anti-HA monoclonal antibody (#H3663) was purchased from Sigma. Rabbit anti-IcsP and anti-IcsA antibodies were prepared as described previously [26], [29]. The antibodies were produced under the National Health and Medical Research Council (NHMRC) Australian Code of Practice for the Care and Use of Animals for Scientific Purposes and were approved by the University of Adelaide Animal Ethics Committee.

Insertion of HA Epitope into IcsP

The sequence encoding the HA epitope (YPYDVPDYA) was inserted into the putatively non-essential IcsP OM loop 5 (based on sequence alignments with E. coli OmpT) using SOE PCR [30], [31]. In the first part of this two-step PCR technique, upstream and downstream amplicons were amplified from S. flexneri 2457T genomic DNA using HA encoding primers (ET18/ET19) and icsP ^- specific primers (ET3/ET10) (Table 2). The two amplicons from this primary PCR were then mixed and used as a DNA template for the second round of PCR with primers ET3/ET10. In this second reaction, the HA encoding regions of the primary PCR amplicons overlap and prime one another to give the final icsP PCR product with an inserted HA epitope. The icsP^HA fragment was then cloned into pGEMT-Easy and primers ET22/ET25 (Table 2) were used to amplify an icsP ^HA product with KpnI and HindIII restriction enzyme sites from this construct. The resultant KpnI-HindIII fragment was digested and sub-cloned into likewise digested pBAD30 to give pBAD30::icsP ^HA, also referred to as pIcsP^HA in text (Table 1). Primers ET22/ET25 were also used to amplify the icsP gene from 2457T genomic DNA, and cloned into pBAD30 to give pBAD30::icsP, referred to as pIcsP in text (Table 1). DNA sequencing was used to confirm that no mutation had been introduced by PCR into the sequence, and the presence of the in-frame HA epitope tag sequence.

Construction of S. flexneri icsP mutant

The S. flexneri 2457T icsP^- mutant strain was constructed using allelic exchange mutagenesis [27] to inactivate the icsP gene by insertion of a kanamycin resistance gene (kan^R). Initially, the icsP gene was PCR amplified with primers ET3/ET4 containing BamHI and SacI restriction sites from 2457T genomic DNA. The resultant PCR fragment was digested with BamHI and SacI and sub-cloned into likewise digested pSL1180 (Table 1). Further digestion with ClaI allowed insertion of the AccI-AccI digested kan^R gene from pKTUWE (Table 1) to give pSL1180-icsP::kan^R (Table 1). Following re-digestion with BamHI and SacI, the icsP::kan^R fragment was cloned into pCACTUS (Table 1) and transformed into S. flexneri 2457T via electroporation. Allelic exchange mutagenesis was performed as previously described [27]. The icsP::kan^R mutation in the virulence plasmid was confirmed by PCR with primers ET3/ET4 (Table 2) to give the 2457T icsP^- mutant ETRM22 (Table 1).

Construction of S. flexneri icsP⁻/rmlD^- Double Mutant

The S. flexneri 2457T icsP⁻/rmlD^- mutant strain was constructed using a modification of the λ Red recombinase system to initially delete the rmlD gene [32]. Primers ET28/ET29 containing NheI restriction enzyme sites (Table 2) were used to PCR amplify the kan^R gene from pKD4 (Table 1). The amplified product was ligated into pGEMT-Easy and pGEMT-Easy::kan^R was digested with NheI. The NheI-kan^R-NheI fragment was then subcloned into likewise digested pRMA718 [26] to give pRMA718-rmlD::kan^R (Table 1). This plasmid was then digested with BamHI and the rmlD::Kan^R fragment was cloned into the BamHI site of pCACTUS. The pCACTUS-rmlD::Kan^R construct was then electroporated into S. flexneri 2457T and allelic exchange mutagenesis was induced to give the 2457T rmlD::kan^R mutant ETRM230 (Table 1). ETRM230 was transformed with pCP20 at 30°C to flip out the FRT flanked kan^R gene and give the 2457T rmlD^- mutant ETRM233 (Table 1). The rmlD mutation was confirmed by LPS analysis. ETRM233 was further electroporated with pCACTUS-icsP::kan^R (Table 1) and another round of allelic exchange mutagenesis was performed to give the final 2457T rmlD^−/icsP^- double mutant ETRM240 (Table 1).

Analysis of IcsP/IcsP^HA Protein Production

For detection of native IcsP, strains were grown at 37°C in LB broth with aeration for 16 h, subcultured 1/20 into fresh broth and grown for another 3 h to an OD₆₀₀ of ~1. Strains harbouring pIcsP or pIcsP^HA were grown in LB broth containing 0.2% (w/v) glucose for 16 h with aeration, subcultured 1/20 into fresh broth and grown for 1.5 h to an OD₆₀₀ reading of ~0.4. Cultures were then pelleted by centrifugation (2219×g, 10 min, Sigma 3K15 centrifuge), washed 3 times in LB, and unless otherwise stated, induced with 0.03% (w/v) arabinose for 1 h to an OD₆₀₀ of ~1. Cells (5×10⁸) were then harvested by centrifugation and resuspended in 2X sample buffer [33]. Protein samples were solubilised at 100°C for 5 min, separated on SDS 15% polyacrylamide gels, and stained with Coomassie R-250, or subjected to Western immunoblotting on nitrocellulose membrane (Medos) with either polyclonal rabbit anti-IcsP antiserum (at 1/250 dilution) or monoclonal mouse anti-HA (at 1/500 dilution). Detection was performed with goat anti-rabbit (or anti-mouse) horseradish-peroxidase-conjugated antibodies (KPL) and chemiluminescence reagent (Sigma). Benchmark prestained molecular weight markers (Invitrogen) were used as molecular size markers.

Sucrose Gradient Density Fractionation

Fractionation of the cell whole membrane (WM) into cytoplasmic membrane (CM) and OM fractions was performed by sucrose gradient centrifugation according to the method of Osborn and Munson [34]. In brief, 200 ml cultures were grown and induced with arabinose as described above, harvested by centrifugation (9,800×g, 15 min, 4°C, JA14 rotor, Beckman centrifuge J2-21M), washed in 50 mM Tris-HCl (pH 8.0) and resuspended in 5 ml 10 mM HEPES in 1 mM MgCl₂. The bacterial suspension was then passed through a pre-cooled French Pressure cell (SLM Aminco) once and re-centrifuged to remove cell debri. WM pellets were collected by ultracentrifugation (115,000×g, 1 h, 4°C, 80 Ti rotor, Beckman Coulter Optima L-100 XP ultracentrifuge), solubilised in 0.8 ml 25% (w/w) sucrose in 5 mM EDTA and applied to a 10 ml sucrose gradient of 30–50% (w/w) sucrose in 5 mM EDTA. Centrifugation to equilibrium was performed with a Beckman SW40Ti swing out rotor (217,000×g, 20 h, 4°C, Beckman Coulter Optima L-100 XP ultracentrifuge) and 0.5 ml fractions collected through the pierced bottom of the tube. 10 µl samples of each fraction were resuspended in 2X sample buffer [33] and IcsP protein detected as described above.

Detection of Cell Associated and Soluble IcsA

Whole cell and supernatant bacterial protein extracts were prepared as described previously [35]. IcsA protein was detected from 10 µl whole cell protein extracts and 20 µl supernatant protein extracts. Western immunoblotting was performed as described above, but with a polyclonal rabbit IcsA antibody and a goat anti-rabbit HRP conjugate.

LPS Analysis

LPS samples and gels were prepared as described previously [36], [37].

LPS Depletion-regeneration Assay

Depletion and regeneration of LPS was performed as previously described [29] with the exception that 0.03% (w/v) arabinose induction was included in the final hour of tunicamycin/polymyxin B nonapeptide (PMBN) treatment.

Formaldehyde Fixation of Bacteria for Immunofluorescence (IF) Microscopy

Bacteria were grown and induced as described above 1×10⁸ cells of induced bacteria were then harvested by centrifugation, washed once in PBS and resuspended in 100 µl 3.7% (w/v) formaldehyde (Sigma) in PBS for 20 min at room temperature (RT). Fixed bacteria were then pelleted, washed three times in PBS and resuspended in a final volume of 100 µl PBS.

Quantum Dot (QD) IF Staining and Epi-fluorescence Microscopy

Sterile glass coverslips were placed into wells of a 24-well tray and coated with 10% (v/v) poly-L-lysine solution (Sigma) in PBS for 1 h at RT. Coating solution was aspirated and 5 µL of formaldehyde fixed bacteria were spotted onto coverslips. The tray was then centrifuged to assist adherence of bacteria (Heraeus Labofuge 400R Centrifuge, 2,000×g, 5 min, 20°C). Bacteria were blocked for 1 h at RT with 10% (v/v) foetal calf serum (FCS) diluted in PBS. For labelling, bacteria were incubated for 2 h at RT with mouse anti-HA antibody (Sigma) and rabbit anti-IcsA antiserum diluted 150 and 1100 respectively in PBS containing 10% (v/v) FCS. Bacteria were then washed 3 times with PBS, and then incubated for 1 h at RT with either QD 525 donkey anti-mouse antibody (Invitrogen) or QD 625 donkey anti-rabbit antibody (Invitrogen) diluted 150 and 1100, respectively, in PBS containing 10% (v/v) FCS. After a final 3 more washes with PBS, coverslips were mounted on glass microscope slides with Mowiol 4–88 (Calbiochem). All microscopy images were captured using an Olympus IX-7- Microscope, with a phase contrast 100X oil immersion objective and a 1.5X enlarger, which was controlled by MetaMorph (Version 7.7.1.0, Molecular Devices). All IcsP^HA (525 nm) and IcsA (625 nm) channel images were acquired with 1 sec and 0.1 sec exposures respectively using an X-Cite 120Q lamp set at high intensity as the excitation source. The excitation filter used was FF01-435/40-25 (Semrock) and the emission filters were FF01-525/15-25 and FF01-625/15-25 (Semrock). Semrock FF510-Di01-25×36 dichroics were used.

Insertion of a HA Epitope into IcsP

IcsP is 60% identical in primary amino acid sequence to the E. coli protease OmpT and computer structure modelling predicts that both proteins possess similar β-barrel structures (Fig. 2A). Based on sequence alignments with E. coli OmpT (Fig. 2B), a HA epitope tag (YPYDVPDYA) was hence inserted into the OM loop 5 of IcsP (Fig. 2B) using SOE PCR (as described in the Methods). An area within this loop region with some sequence diversity between IcsP and OmpT was selected as we reasoned that this sequence variability might allow the protein to accommodate the epitope insertion with little disturbance to the overall structure. The OM loop 5 region was also selected to increase the chance of surface detection by antibodies. OM loops 2 and 4 were not selected for HA tag insertion to avoid the proposed catalytic residues present in OmpT [38] which also exist in IcsP (Fig. 2C). The IcsP^HA (as well as IcsP) coding regions were placed in front of the pBAD promoter in pBAD30 [39] to allow expression control with arabinose. Expression of IcsP/IcsP^HA was confirmed by Western immunoblotting with anti-IcsP or anti-HA antibodies (Fig. 3A & B, lanes WM and Fig. 1A & C).

Open in a separate window

Figure 2

Putative structure of IcsP and location of HA epitope insertion in IcsP.

(A) IcsP was modelled using the SWISS-MODEL Protein Modelling Server (http://swissmodel.expasy.org//SWISS-MODEL.html) (left) and compared to the structure of OmpT (PDB 1I78) (right); (B) Schematic diagram of IcsP showing the location of the OM loops 1 - 5; (C) Amino acid sequence alignment of IcsP (AF001633) and OmpT (P09169) showing 60% identity (green shaded regions). The black-boxed amino acids in the sequence of OmpT refer to active site residues found in OM loops 2 and 4 38], and the purple-boxed amino acids refer putative LPS binding sites 38]. The location of the HA epitope (YPY DVP DYA) insertion into the putatively non-active OM loop 5 (L5) is indicated by the red arrow in (A) – (C).

Open in a separate window

Figure 3

Analysis of IcsP/IcsP^HA subcellular localisation by sucrose density gradient centrifugation of WM.

S. flexneri icsP^- strains harbouring either pIcsP^HA and pIcsP were grown in LB for 1.5 h to an OD₆₀₀ reading of ~0.4, washed 3 times, and induced with arabinose for 1 h (OD₆₀₀ ~1). Whole membranes (WM) were prepared by French press lysis, and subjected to sucrose density gradient centrifugation 48] as described in the Methods. Fractions (0.5 ml; numbered 1-15) were collected and samples of each were electrophoresed on SDS 15% polyacrylamide gels for staining with Coomassie Blue and Western immunoblotting with anti-HA or anti-IcsP antibodies; (A) The results obtained for icsP^- [pIcsP^HA]; (B) The results obtained for icsP^- [pIcsP]. The migration positions of the Benchmark Prestained Marker (M) Standards (Invitrogen) are indicated on the left in kDa. The major OM proteins OmpF+OmpC and OmpA are indicated by the two arrowheads in lane 2. Lane C corresponds to the control lanes containing WC samples of icsP^- [pIcsP^HA] induced with 0.2% (w/v) arabinose for protein overexpression (positive with both anti-HA and anti-IcsP by Western immunoblotting) as indicated by the arrows on the right. The sucrose fractions in lanes 2 – 4 contain most of the OM proteins, and the sucrose fractions in lanes 8 – 14 contain most of the inner (cytoplasmic) membrane proteins.

IcsP^HA Activity on IcsA

To determine whether insertion of a HA epitope into the OM of IcsP affected IcsP’s protease activity on IcsA, pellet and supernatant protein preparations of 2457T, icsP^- and arabinose-induced icsP^- strains expressing pIcsP^HA, pIcsP and pBAD30 were subjected to Western immunoblotting with an anti-IcsA antibody. The full length 120 kDa IcsA protein was detected in cell pellet samples of all strains as expected (Fig. 1B, lanes 1, 3, 5, 7 and 9), while the 95 kDa cleaved form of IcsA was only detected in the supernatant sample of 2457T, icsP^- [pIcsP^HA] and icsP^- [pIcsP] (Fig. 1B, lanes 2, 6 & 8). These results suggest that the insertion of the HA tag into the OM loop 5 of IcsP does not affect its ability to cleave IcsA, and hence the IcsP^HA protein is functional.

Localisation of IcsP/IcsP^HA Protein to the OM

Since IcsP^HA was modified compared to IcsP and expressed from the pBAD promoter, the presence of IcsP^HA protein exclusively in the OM was confirmed by using sucrose density gradient centrifugation. The WM of icsP^- [pIcsP^HA] and icsP^- [pIcsP] were fractionated into CM and OM on sucrose gradients, and fractions subjected to SDS 15% polyacrylamide gel electrophoresis, prior to visualisation by Coomassie Blue staining and Western immunoblotting with anti-HA and anti-IcsP. Analysis of the sucrose gradient samples showed that fractions which were enriched with OM proteins (OmpF, OmpC, and OmpA) [40] contained the majority of the 36 kDa IcsP^HA and IcsP proteins (Fig. 3A & B, lanes 2 to 5). These results indicate that IcsP^HA is localised to the OM similar to IcsP, and that the HA insertion did not result in any dramatic disruption of IcsP protein localisation.

Detection of Arabinose Induced IcsP^HA Expression by Western Immunoblotting

Wild-type S. flexneri 2457T showed optimal expression levels of native IcsP at 3 h (Fig. 1A). To investigate the conditions required for comparable IcsP^HA expression, whole cell lysate samples were prepared from icsP^- [pIcsP^HA] and icsP^- [pBAD30] induced with 0%, 0.003%, 0.006%, 0.0125%, 0.025%, 0.05%, 0.1% or 0.2% (w/v) arabinose for 1 h. Western immunoblotting with anti-IcsP showed that a band of ~36 kDa was detected for 2457T and icsP^- [pIcsP^HA] induced with 0.025%, 0.05%, 0.1% and 0.2% (w/v) (w/v) arabinose (Fig. 1C, lanes 1, 3–6), with expression levels comparable to native IcsP for icsP^- [pIcsP^HA] observed between 0.025%–0.05% (w/v) arabinose induction (Fig. 1C, lanes 5 & 6). Induction at 0.03% (w/v) arabinose was hence chosen for all subsequent experiments.

Cell Surface Detection of IcsP^HA Distribution in S. flexneri 2457T and the Effect of Tunicamycin Treatment

Having established an induction protocol that closely approximates the level of expression of IcsP^HA to native IcsP, we next attempted to detect IcsP^HA at these levels on the surface of 2457T icsP^- using indirect QD IF microscopy. Strains icsP^- [pIcsP^HA] and icsP^- [pBAD30] were cultured with 0.03% (w/v) arabinose, fixed, and then probed for HA using a primary anti-HA antibody and a secondary QD 525 conjugated antibody. However, no IcsP^HA was detected on the cell surface, and intensity scans of icsP ^- [pIcsP^HA] were equivalent to icsP^- [pBAD30] (Fig. S1).

Since IcsP^HA could not be detected on the cell surface of icsP^- [pIcsP^HA], we reasoned that the presence of LPS Oag may mask the detection of surface IcsP as this has previously been shown for IcsA [21]. An LPS Oag depletion-regeneration assay [29] was hence carried out on icsP^- [pIcsP^HA] and icsP^- [pBAD30] induced with 0.03% (w/v) arabinose, followed by IF labelling. This assay involves the use of an inhibitor of the WecA enzyme necessary for Oag subunit biosynthesis using tunicamycin, and polymyxin B nonapeptide (PMBN) was used to improve OM penetration. Upon removal of these two chemicals from growing bacteria, LPS Oag is regenerated [29]. Analysis of the resulting LPS by SDS-PAGE and silver staining showed that icsP^- [pIcsP^HA] and icsP^- [pBAD30] samples treated with tunicamycin/PMBN (TP) had depleted LPS Oag (Fig. 4A, lanes 5 & 6), with Oag production restored (R) upon removal of TP (Fig. 4A, lanes 7 & 8). Untreated (U) and PMBN treated (P) samples showed no inhibition of Oag biosynthesis as expected (Fig. 4A, lanes 1–4).

Open in a separate window

Figure 4

Effect of tunicamycin on the LPS of strains expressing IcsP^HA.

(A) Smooth LPS 2457T icsP^- strains harbouring pIcsP^HA or pBAD30 were grown to an OD₆₀₀ reading of ~0.8 in LB, washed 3 times, and treated without TP (U), with PMBN only (P) or with TP treatment for 2 h. Strains were then induced with 0.003% (w/v) arabinose for 1 h, washed 3 times, and grown for an additional 3 h for restoration (R) of LPS Oag; (B) Rough LPS 2457T icsP^-/rmlD^- strains harbouring pIcsP^HA and either pRMA727 or pACYC184 (as indicated) were grown to an OD₆₀₀ reading of ~0.4 in LB, washed 3 times, and induced with arabinose for 1 h. LPS from strains described in (A) and (B) were isolated and detected by silver staining as described in the Methods. The first 15 Oag RUs are indicated on the side of each gel. Each lane contains ~2 x 10⁸ bacterial cells of each strain; (C) Western blots on whole cell lysates obtained from strains in (B) were probed with rabbit anti-IcsP antiserum. The size of the full length IcsP^HA protein (~36 kDa) is indicated. Each lane contains 5 x 10⁷ bacterial cells of each strain.

Following successful depletion of LPS Oag, QD IF microscopy was performed on the above mentioned samples. IcsP^HA could not be detected on untreated or PMBN treated icsP ^- [pIcsP^HA] cells, as expected (Fig. S2A). However, surface IcsP^HA was detected on icsP ^- [pIcsP^HA] cells treated with TP, suggesting that LPS Oag is able to mask antibody accessibility to IcsP^HA (Fig. S2A). Interestingly, IcsP^HA appeared to be distributed asymmetrically over the cell surface of the majority of cells examined and fluorescence intensity line-scans revealed that IcsP^HA localised preferentially to one pole of non-septating cells and the septa of septating cells (Fig. S2A). To determine if the distribution of IcsP^HA on the cell surface segregated to either the new pole or the old pole, additional staining of the bacteria was conducted with anti-IcsA antibodies since IcsA is known to localise to the old cell pole [5]. Again, IcsP^HA could not be detected on untreated or PMBN treated icsP ^- [pIcsP^HA] cells (Fig. 5A & B) but could be detected after TP treatment (Fig. 5C). Peak IcsP^HA detection was consistently observed at the new pole (opposing IcsA at the old pole) and at the septum of septating bacteria as shown by line-scans (Fig. 5C). As expected, no surface IcsP^HA was detected for untreated, PMBN treated, or TP treated samples of icsP^- [pBAD30] when stained for IcsP^HA (Fig. S2B) or double stained for IcsP^HA and IcsA (Fig. S3). Notably, IcsA on icsP^- [pBAD30] was detected at higher amounts laterally and at the septa of bacteria (Fig. S3) as previously seen for ΔicsP strains [13].

Open in a separate window

Figure 5

Effect of LPS Oag-depletion in detection of surface IcsP^HA.

Smooth LPS 2457T icsP^- strains harbouring pIcsP^HA were subcultured in LB broth to an OD₆₀₀ reading of ~0.8, washed 3 times in LB, and then further cultured for 2 h; (A) in the absence of TP, (B) in the presence of PMBN only, or (C) in the presence of TP. Arabinose was included in the final hour of treatment at a concentration of 0.03% (w/v). Samples were then fixed and subjected to QD IF using antibodies to HA epitope and IcsA. Non-septating and septating cells (upper and lower rows respectively) are shown for each treatment group. Representative bacteria are shown. Scan = Single line-scans measuring the fluorescence intensity of IcsP^HA detected along the surface of the bacterium, Bars = 1 µm, Arrows = direction of line-scan, Arrow heads = septa, Control = grown in absence of both tunicamycin and PMBN, P = PMBN, TP = tunicamycin/PMBN, GL = Grey level, Phase = phase contrast image, IcsP^HA = image of fluorescence at 525 nm, IcsA = image of fluorescence at 625 nm, Merge = overlay of IcsP^HA and IcsA images.

Cell Surface Detection of IcsP^HA Distribution in Rough LPS S. flexneri 2457T

To further investigate the effect of Oag on IcsP distribution, a 2457T icsP⁻/rmlD^- double mutant was constructed to independently assess the distribution of IcsP^HA in a rough LPS background. Mutation of the rmlD gene in S. flexneri results in a strain which is unable to synthesise the precursor deoxythymidine diphosphate (dTDP)- rhamnose required for Oag repeat units and, hence, results in a rough LPS phenotype [29], [41]. Analysis of the resulting LPS conferred by icsP⁻/rmlD^- strains expressing pIcsP^HA and pBAD30 showed that rough LPS was observed for both (Fig. 4B, lanes 1 & 2), with a band consistent with the size of the IcsP^HA protein (36 kDa) detected only in the icsP⁻/rmlD^- [pIcsP^HA] sample by Western immunoblotting, as expected (Fig. 4C, lane 1).

Fixed samples of icsP⁻/rmlD^- [pIcsP^HA] and icsP⁻/rmlD^- [pBAD30] cells were then probed for both IcsP^HA and IcsA using the same QD IF staining protocol as previously conducted for smooth strains. Similarly to LPS Oag-depleted icsP ^- [pIcsP^HA], IcsP^HA was detected on the bacterial surface most predominately at the new pole (Fig. 6A) and the septum (Fig. 6B). Again, the majority of septating cells had higher peak IcsP^HA intensity at the septum than new poles of non-septating cells (line-scans Fig. 6A & B). Single staining of these cells for IcsP^HA was also conducted and yielded the same localisation results (Fig. S4). As expected, IcsP^HA was not detected on the icsP⁻/rmlD^- [pBAD30] strain in IF microscopy experiments when either single (Fig. S4) or double stained (Fig. 6B). Again, IcsA on icsP⁻/rmlD^- [pBAD30] was detected at higher amounts laterally and at the septum (Fig. 6A & B).

Open in a separate window

Figure 6

Localisation of IcsP^HA on the surface of rough LPS 2457T icsP^-/rmlD^-.

Rough LPS 2457T icsP^-/rmlD^- strains harbouring pIcsP^HA (left columns) or pBAD30 (right columns) were subcultured in LB broth for 1.5 h to an OD₆₀₀ reading of ~0.4. Cultures were then washed 3 times in LB, induced with 0.03% (w/v) arabinose for 1 h, fixed, and subjected to QD IF using antibodies for HA epitope and IcsA. Non-septating and septating life stages are shown in A and B respectively. Representative bacteria are shown. Scan = Single line-scans measuring the fluorescence intensity of IcsP^HA detected along the surface of the bacterium, Bars = 1 µm, Arrows = direction of line-scan, Arrow heads = septa, GL = Grey level, Phase = phase contrast image, IcsP^HA = image of fluorescence at 525 nm, IcsA = image of fluorescence at 625 nm, Merge = overlay of IcsP^HA and IcsA images.

To again demonstrate the effect of LPS Oag masking of IcsP^HA, smooth LPS structure was restored in the icsP⁻/rmlD^- strain expressing pIcsP^HA by transforming pRMA727 carrying a functional rmlD gene. Analysis of the resulting LPS conferred by icsP⁻/rmlD^-[pIcsP^HA][pRMA727] showed restored smooth LPS phenotype (Fig. 4B, lane 3), and when probed for IcsP^HA by QD IF microscopy, was barely detectable (Fig. 7A). The control icsP⁻/rmlD^- [pIcsP^HA] strain carrying pACYC184 conferred a rough LPS phenotype (Fig. 4B, lane 4) and IcsP^HA was detected by IF microscopy with the same distribution observed previously (Fig. 7B).

Open in a separate window

Figure 7

Complementation of rmlD restores LPS masking effect on IcsP^HA detection.

Rough LPS 2457T icsP^-/rmlD^- strains harbouring pIcsP^HA, and either (A) pRMA727 or (B) pACYC184, were subcultured in LB broth for 1.5 h to an OD₆₀₀ reading of ~0.4. Cultures were then washed 3 times in LB, induced with 0.03% (w/v) arabinose for 1 h, fixed, and subjected to QD IF using antibodies for the HA epitope. Representative bacteria are shown. Scan = Single line-scans measuring the fluorescence intensity of IcsP^HA detected along the surface of the bacterium, Bars = 1 µm, Arrows = direction of line-scan, Arrow heads = septa, GL = Grey level, Phase = phase contrast image, IcsP^HA = image of fluorescence at 525 nm.