The plasmid-encoded Ipf and Klf fimbriae display different expression and varying roles in the virulence of Salmonella enterica serovar Infantis in mouse vs. avian hosts

Abstract

Salmonella enterica serovar Infantis is one of the prevalent Salmonella serovars worldwide. Different emergent clones of S. Infantis were shown to acquire the pESI virulence-resistance megaplasmid affecting its ecology and pathogenicity. Here, we studied two previously uncharacterized pESI-encoded chaperone-usher fimbriae, named Ipf and Klf. While Ipf homologs are rare and were found only in S. enterica subspecies diarizonae and subspecies VII, Klf is related to the known K88-Fae fimbria and klf clusters were identified in seven S. enterica subspecies I serovars, harboring interchanging alleles of the fimbria major subunit, KlfG. Regulation studies showed that the klf genes expression is negatively and positively controlled by the pESI-encoded regulators KlfL and KlfB, respectively, and are activated by the ancestral leucine-responsive regulator (Lrp). ipf genes are negatively regulated by Fur and activated by OmpR. Furthermore, induced expression of both klf and ipf clusters occurs under microaerobic conditions and at 41°C compared to 37°C, in-vitro. Consistent with these results, we demonstrate higher expression of ipf and klf in chicks compared to mice, characterized by physiological temperature of 41.2°C and 37°C, respectively. Interestingly, while Klf was dispensable for S. Infantis colonization in the mouse, Ipf was required for maximal colonization in the murine ileum. In contrast to these phenotypes in mice, both Klf and Ipf contributed to a restrained infection in chicks, where the absence of these fimbriae has led to moderately higher bacterial burden in the avian host. Taken together, these data suggest that physiological differences between host species, such as the body temperature, can confer differences in fimbriome expression, affecting Salmonella colonization and other host-pathogen interplays.

Fig 1. Genetic organization and phylogenetic distribution of the ipf and klf clusters.

The pESI-encoded ipf and klf gene clusters were compared using the Easyfig tool with homologous clusters found in the nr database. Nucleotide comparison between the related clusters is illustrated by the shades of grey that indicate the degree of sequence homology (in %). Gaps in the grey areas point to the lack of sequence similarity, and protein functions (regulator, usher, chaperone and fimbrial subunit) are color-coded. (A) The ipf cluster encodes four ORFs: IpfA, a fimbrial protein; IpfB, the fimbrial chaperone; IpfC, the fimbrial usher; and IpfD, a putative fimbrial adhesin. Homologs of the ipf cluster were found in Salmonella enterica subsp. diarizonae strain 11–01854 (GenBank: CP011292.1) and in Salmonella enterica subsp. VII integrative and conjugative element ICESe3 region, strain SARC16 (sequence ID: FN298495.1). (B) The klf cluster in S. Infantis 119944 contains 12 ORFs encoding the major fimbrial subunit (KlfG), usher (KlfD), chaperone (KlfE) and four minor subunits (KlfC, KlfF, KlfH and KlfI). KlfJ, KlfK and KlfA have unknown function and KlfL and KlfB are unique to pESI and function as regulators (see Fig 8). Similarity to the fae cluster in enterotoxigenic E. coli (GenBank: CP002730) is shown at the upper line of the alignment. Homologous clusters among S. enterica serovars were found in S. Anatum str. USDA-ARS-USMARC-1735 (GenBank: CP007584.2); S. Bareilly str. CFSAN000189 (GenBank: CP006053); S. Bredeney str. CFSAN001080 (GenBank: CP007533); S. Schwarzengrund str. CVM19633 (GenBank: CP001127.1); S. Montevideo str. USDA-ARS-USMARC-1921 (GenBank: CP007540.1); S. Typhimurium strain FORC_015 (GenBank: CP011365); and S. Cubana str. CFSAN002050 (GenBank: CP006055.1). The G+C content is illustrated by the blue-red histogram, at the top of the panel, while G+C > 50% is shown in red and G+C < 50% is shown in blue.

Fig 2. Klf homologs harbor different alleles of the fimbrial major subunit KlfG.

Multiple sequence alignment of the KlfG from the eight klf clusters shown in Fig 1B was performed using CLUSTALW and BOXSHADE tools. Identical amino acids are shown in black and similar amino acids are shown in grey. The conserved signal peptide sequence preceding the Sec cleavage site was predicted by the SignalP 4.1 program and is indicated at the N-terminus of the proteins.

Fig 3. Heterologous expression of Ipf and Klf in surrogate E. coli cells.

Non-fimbriated E. coli strain ORN172 carrying the entire ipf operon (pBAD18::ipf, lanes 1 and 2), the klf operon (pBAD18::klf, lanes 4 and 5), or the empty vector (pBAD18, lane 3) that was used as a negative control was grown in N-minimal medium supplemented with 50 mM L-arabinose (inducing conditions) or 1 M glucose (suppressing conditions). Cultures supernatants that were enriched with surface structures were collected after a shearing treatment, subjected to TCA precipitation and separated on a 12% SDS-PAGE. Arrows show the bands that were isolated from the gel and identified by LC-MS/MS. The proteins that were detected by LC-MS/MS are summarized in the table below. The score value presents the cumulative protein score based on summing the ion scores of the unique peptides identified for that protein. Coverage displays the percentage of the protein sequence covered by the identified peptides. PSMs show the total number of identified peptide sequences (peptide spectrum matches) for the protein, including those redundantly identified. The area displays the average area of the three unique peptides with the largest peak area.

Fig 4. The ipf and klf clusters encode structurally-distinct fimbriae.

(A, B): Fimbriae-less E. coli ORN172 expressing ipfABCD under the tetracycline-inducible promoter (PtetA) were grown in N-minimal medium supplemented with 100 ng/ml anhydrotetracycline (AHT; A) or in the absence of the inducer (B). Cultures harboring ipfABCD were visualized by AFM and TEM. Panels A and B show an AFM height image (I), an enlarged AFM deflection image (II) and a TEM image (III). Highlighted box in (I) indicate the area shown in AFM deflection image (II). Color bar indicates the Z-range. Scale bars, 1 μm (I), 0.25 μm (II), 1 μm (III). (C, D) E. coli ORN172 harboring klfBCDEFGHIJKA under control of an arabinose-inducible promoter (Para) were grown in N-minimal medium supplemented with 0.5 mM arabinose (C), or 1 M glucose (D). Bacterial cells were negatively stained with 0.5% phosphotungstic acid (PTA) and imaged by TEM. Scale bars, 500 nm. Black and red arrows indicate the Klf fimbriae or flagella, respectively.

Fig 5. klf and ipf genes are induced under microaerobiosis.

(A) RNA was extracted from S. Infantis wild-type strain grown to the stationary phase under microaerobic conditions in LB, N-minimal medium pH 7 and N-minimal medium pH 5.8. Quantitative RT-PCR analyses were conducted to determine the fold change in the transcription of ipfC and klfD in cultures grown in LB or N-minimal medium pH 5.8 relative to their transcription in N-minimal medium pH 7.0. (B) qRT-PCR analyses were performed to determine the fold change in the transcription of ipfA, ipfB, ipfC and ipfD in cultures grown in LB to the stationary phase under microaerobic conditions or under aerobic conditions, relative to their transcription at the late logarithmic phase under aerobic conditions. (C) qRT-PCR analyses were performed to determine the fold change in the transcription of klfC, klfD, klfE and klfG in cultures grown in LB to stationary phase under microaerobic or aerobic conditions, relative to their transcription at the late logarithmic phase under aerobic conditions. One way ANOVA with Dunnett's Multiple Comparison Test were performed to determine statistical significance. ns, not significant; *, P<0.05; **, P<0.01; ***, P<0.001.

Fig 6. Klf and Ipf are induced at the avian physiological temperature.

(A) qRT-PCR showing the fold change in the transcription of ipfA, ipfB, ipfC and ipfD in cultures grown in LB to the stationary phase under microaerobic conditions at 27°C or 41°C, relative to 37°C. (B) qRT-PCR showing the fold change in the transcription of klfC, klfD, klfE and klfG in cultures grown in LB to the stationary phase under microaerobic conditions at 27°C or 41°C, relative to 37°C. One way ANOVA with Dunnett's Multiple Comparison Test were performed to determine statistical significance. ns, not significant; *, P<0.05; **, P<0.01; ***, P<0.001. (C) Whole cell lysates of S. Infantis strains expressing a 2HA-tagged version of IpfD and KlfC that were grown in LB to the stationary phase under microaerobic conditions at 37°C and 41°C were separated on an SDS-PAGE. Western blotting using anti-HA antibody and anti-RpoD (as a loading control) are shown. The double band shown for KlfC-2HA and IpfD-2HA represents the premature and the signal peptide-cleaved forms of the proteins. IpfD-2HA and KlfC-2HA bands densitometry (normalized to the corresponding RpoD bands) are presented relative to the wt, under the RpoD blot.

Fig 7. OmpR, Lrp and Fur are involved in pESI fimbriae regulation.

(A) Total RNA was extracted from S. Infantis wild-type (wt) strain and ten isogenic null mutants harboring deletion in key regulatory genes. qRT-PCR was applied to define the fold change in klfD transcription between the wild-type background and the mutant strains. All cultures were grown in LB under microaerobic conditions at 37°C. (B) S. Infantis wild-type and ten derivative deletion mutant strains expressing KlfC-2HA were grown in LB under microaerobic conditions at 37°C. Bacterial lysates were separated by SDS-PAGE followed by western blotting using antibodies against hemagglutinin and DnaK as a loading control. (C) The same strains were grown in LB under microaerobic conditions at 41°C. Bacterial lysates were analyzed by western blotting using antibodies against hemagglutinin and RpoD. (D) S. Infantis wild-type strain (wt), its isogenic lrp null strain (lrp), an lrp mutant strain complemented with the lrp gene expressed from pWSK29 (lrp/ pWSK29::lrp), an lrp mutant harboring the empty vector (lrp/ pWSK29), all expressing KlfC-2HA from a low copy-number vector (pACYC184) were grown in LB broth under microaerobic conditions at 41°C. Wild-type strain harboring the empty plasmid pACYC184 (vector) was also included as a negative control. Bacterial lysates were separated by SDS-PAGE followed by western blotting using antibodies against hemagglutinin and RpoD. KlfC-2HA bands densitometry (normalized to the corresponding DnaK bands) are presented relative to the wt, under the DnaK blot. (E) qRT-PCR was applied to determine the fold change in ipfD transcription between the wild-type background and the mutant strains as in panel (A). (F) S. Infantis wild-type and ten isogenic mutant strains expressing IpfD-2HA were grown in LB under microaerobic conditions at 37°C. S. Infantis wild-type strain harboring the pWSK29 (vector) was used as a control for the western blotting. Bacterial lysates were separated by SDS-PAGE followed by western blotting as in panel (B). (G) The same strains were grown in LB under microaerobic conditions at 41°C. Bacterial lysates were analyzed by western blotting using antibodies against hemagglutinin and RpoD. (H) S. Infantis wild-type strain (wt), its isogenic fur mutant strain (fur), a fur mutant strain complemented with the fur gene expressed from pACYC184 (fur/ pACYC184::fur), and a fur mutant harboring the empty vector (fur/ pACYC184), all expressing IpfD-2HA from a low copy-number vector (pWSK29) were grown in LB broth under microaerobic conditions at 37°C. Bacterial lysates were separated by SDS-PAGE followed by western blotting using antibodies against hemagglutinin and RpoD. IpfD-2HA bands densitometry (normalized to the corresponding RpoD bands) are presented relative to the wt below the RpoD blot. (I) S. Infantis wild-type (wt) strain and its isogenic fur null mutant were grown at 37°C to the stationary phase under microaerobic conditions in LB or in LB supplemented with Dip (to a final concentration of 0.2 mM). qRT-PCR was applied to determine the fold change in ipfA transcription of the different cultures, relative to the wt strain grown in LB.

Fig 8. KlfB and KlfL are positive and negative regulators, respectively of the klf operon.

(A) RNA was extracted from S. Infantis wild-type (wt) strain and three isogenic mutants containing null mutations in klfA, klfB and klfL grown in LB under microaerobic conditions at 37°C. qRT-PCR was conducted to define the fold change in klfD transcription between the wild-type background and the mutant strains. (B) S. Infantis wild-type and the klfA, klfB and klfL mutant strains expressing KlfC-2HA were grown to the stationary phase in LB under microaerobic conditions at 37°C. Wild-type strain harboring the empty plasmid pACYC184 (vector) was included as a negative control. Bacterial lysates were separated by SDS-PAGE followed by western blotting using antibodies against hemagglutinin and DnaK. KlfC-2HA bands densitometry (normalized to the corresponding DnaK bands) are presented relative to the wt, below the DnaK blot. (C) S. Infantis wild-type and the klfB mutant strain expressing KlfC-2HA were grown to the stationary phase in LB under microaerobic conditions at 41°C. Expression of KlfC-2HA was determined as in (B).

Fig 9. The expression profile of ipf and klf in the mouse model.

Female C57BL/6 mice (five animals per group) were infected orally with ∼8×10⁸ CFU of S. Infantis strains harboring luciferase reporter fusion with ipf (ipf::lux) (A, C) and klf (klf::lux) (B, D). Twenty four hours p.i. mice were sacrificed and their intact GI tract was removed and imaged immediately using a photon-counting in-vivo imaging system. (A and B) Bacterial loads in the cecum, colon and ileum are indicated by total CFU counted in the entire organ. Each dot indicates the count in an entire organ in a single animal. (C and D) The detected bioluminescence signal is shown as pseudocolor images, with variations in color representing light intensity at a given location. The color bar indicates relative signal intensity (as photons s⁻¹ cm⁻² sr⁻¹) and the minimal and maximal values measured are indicated in the box below each image. Different organs are indicated as follow: jejunum (J); ileum (I); cecum (CE); and colon (C).

Fig 10. klf and ipf are significantly expressed in the cecum of infected chicks.

Two groups of one day old White Leghorns chicks were infected orally with ∼1×10⁷ CFU of wild-type S. Infantis harboring the ipf::lux (A, C) and klf::lux (B, D) reporter strains. Twenty four hours p.i. chicks were sacrificed and their intact GI tracts as well as their liver were removed and imaged immediately using a photon-counting in-vivo imaging system. (A and B) Bacterial loads in the duodenum, jejunum and ileum, cecum, colon and liver is indicated by a CFU count per organ. The geometrical mean in each organ is shown by a solid horizontal line. (C and D) The color bar indicates relative signal intensity and the minimal and maximal values measured are shown in the box below the color bar. Different organs are designated as follow: duodenum (D) jejunum (J); ileum (I); cecum (CE); colon (C) and liver (L).

Fig 11. Ipf and Klf contribute differently to S. Infantis infection in mouse vs. chicken.

(A-C) C57BL/6 mice were intragastrically infected with ~6×10⁶ CFU of a mixed (1:1) inoculum containing the wild-type S. Infantis (harboring pWSK129; Km^R) and klf (B) or ipf (C) null mutant strain (harboring pWSK29; Amp^R). A mixed inoculum of two S. Infantis wild-type strains carrying pWSK29 or pWSK129 was used as a control (A). Four days p.i., mice were sacrificed and tissues were harvested aseptically, homogenized and plated on selective XLD agar plates for bacterial enumeration. Each dot represents a competitive index (CI) value in one mouse in a single organ (cecum, colon or ileum). The CI was calculated as [mutant/wild-type]_output/[mutant/wild-type]_input. (D and E) Two groups of one day old SPF White Leghorns chicks were infected orally with ~1×10⁷ CFU of 1:1 mixed inoculum containing the wild-type S. Infantis (harboring pWSK129) and ipf or klf deletion mutant strains (harboring pWSK29). Three days p.i. chicks were sacrificed and the indicated organs were homogenized and plated on XLD agar plates supplemented with ampicillin or kanamycin for bacterial numeration. The CI was calculated as above.