Knockout mutagenesis of the kpsE gene of Campylobacter jejuni 81116 and its involvement in bacterium–host interactions


  • Editor: George Mendz

Correspondence: Benjamin N. Fry, Biotechnology and Environmental Biology, RMIT University, PO Box 71, Bundoora VIC 3083 Australia.
Tel.: 61 3 9925 7100; fax: 61 3 9925 7110; e-mail:


Campylobacter jejuni is a common cause of bacterial enteritis. The surface capsular polysaccharides are important for this bacterium to survive in the environment, but little is known about their involvement in bacterium–host interactions. This study showed that the C. jejuni capsular polysaccharides play an important role in adherence to and invasion of human embryonic epithelial cells. However, no significant role of capsular polysaccharides was shown in colonization of the chicken gut.


The main cause of Campylobacter-associated human enteritis is Campylobacter jejuni, which is a species commonly found as a commensal in the gastrointestinal tract of poultry (Wassenaar et al., 1997; Wassenaar & Blaser, 1999). It has been suggested that the virulence factors of Campylobacter may be distinct from factors required for colonization, including living as a commensal in the chicken intestinal tract (Fry et al., 2000). The capsular polysaccharides (CPSs) of C. jejuni appear to contribute to virulence (Bacon et al., 2001; Guerry et al., 2002). However, as the CPSs of C. jejuni may undergo phase variation (Szymanski et al., 2003), which can result in the expression of different CPSs, it is possible that the bacterium–host interactions will vary within strains. Thus, the importance of the CPS for the bacterium's survival in the environment makes this polysaccharide an appropriate molecule for studying host–bacterium relationships.

Materials and methods

Bacteria and culture cells

Descriptions of the bacteria and plasmids used in this study are provided in Table 1. Campylobacter jejuni 81116 was grown at 42°C in microaerobic conditions on Skirrow agar plates supplemented with 10% horse blood. Escherichia coli DH5α (Hanahan, 1983) was used as a host for pBluescript SKII as well as all recombinant plasmids. It was also used as a negative control for adhesion and invasion assays, and was grown at 37°C in Luria–Bertani (LB) medium, with or without ampicillin (100 μg mL−1 or kanamycin (50 μg mL−1. Salmonella enteritis serovar Typimurium (STM I) was used as a positive control for adhesion and invasion assays and was grown in LB broth. Human embryonic epithelial (INT-407) cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM).

Table 1.   Bacteria and plasmids used in this study
Strains or plasmidsRelevant propertiesReferences/source
C. jejuni strains
 81116Penner 6 serotype(Palmer et al., 1983)
 81116 kpsE81116 transformed with pBkEkmRThis study
E. coli strain
 DH5αF/φ80 lacZΔM15(Hanahan, 1983)
Salmonella serovar Typhimurium strain (STM1)Attenuated vaccine strain (aroA)(Alderton et al., 1991)
 pBlCloning vector pBluescript SKII(Short et al., 1998)
 pMW2Vector with C. jejuni kanamycin resistance cassetteM. Wösten
 pBkE-1A 614-bp PCR product of the kpsE gene from C. jejuni 81116 in pBlThis study
 pBkE-2Carries a 614-bp fragment and a 334-bp fragment of the C. jejuni 81116 kpsE geneThis study
 pBkEkmRpBkE-2 containing a KanR cassette inserted between the 614-kb and 334-bp fragmentsThis study

Creation of a KpsE mutant

The CPS transporter gene (kpsE), which is flanked by kpsT and kpsD (Karlyshev & Wren, 2001), was used to create a CPS mutant in C. jejuni 81116. A suicide vector was constructed using the neighbor-joining method described by Li et al. (2003). First, a 614-bp DNA fragment, containing 400 bp of the kpsE gene and 214 bp of flanking sequence from the neighboring kpsT gene, was amplified with primer kpE-1 (TGATGAATTCGGAGCTGTTGGAGATCCT, forward), which is complementary to the flanking kpsT gene sequence, and primer kpE-2 (GGTTGGATCCGTATGAACCTTAACTCTTGC, reverse), which is complementary to the kpsE gene sequence. The primers kpE-1 and kpE-2 contained an EcoRI and a BamHI restriction site, respectively, at the 5′ end. The PCR product was cloned into the multiple cloning site of pBluescript SKII, and restriction enzyme analysis was used to confirm that the recombinant plasmid pBkE1 had been obtained. Second, primers kpE-3 (TTCAGGATCCAACTATAGAAGCTACT) and kpE-4 (TAGCTCTAGATTGATCCTCTGCTGAA), which include a BamHI and an XbaI restriction site, respectively, were used to amplify a 348-bp fragment, which contained 251 bp of the 3′ end of the kpsE gene and 97 bp of the 5′ end of the upstream kpsD gene. The amplified fragment was inserted into pBkE-1 between the BamHI and XbaI sites of the multiple cloning site. The resulting intermediate plasmid was named pBkE-2. The kanamycin cassette was isolated from vector pMW2 by digesting it with BamHI, and a Geneclean DNA purification kit was used to purify the 1.4-kb fragment from the agarose gel. The plasmid pBkE-2 was cut at the unique BamH1 site to insert the kanamycin cassette, to obtain the suicide plasmid called pBkEkmR, which was subsequently used to naturally transform C. jejuni 81116 as described previously (Wassenaar et al., 1993). The plasmid construct was also used as a probe in Southern blotting experiments after labeling with digoxigenin-dUTP as described by the manufacturer (Roche Molecular Biochemicals). Samples for electron microscopy examination were prepared as described by Karlyshev et al. (2001).

Adherence and invasion assays

Approximately 2.0 × 105 INT-407 cells in DMEM were seeded into 24-well plates. After 2 days, the semiconfluent monolayer cells were prewashed twice with DMEM. Bacteria (2 × 108) were added to the wells, and incubated at 37°C (5% CO2). After 3 h, the unattached bacteria were washed three times with agitation in phosphate-buffered saline (PBS), and the cells were disrupted with 2.5% Triton X-114 in 200 μL of DMEM by incubating for 15 min. The binding of bacteria to epithelial cells was quantitated by counting viable C. jejuni cells from serial dilutions of lysates. Invasion assays were performed in a similar way. However, after the bacteria had been allowed to adhere to the monolayer, wells were washed three times and then incubated for 3 h in DMEM containing 250 μg gentamicin mL−1 to kill the extracellular bacteria. The cells were subsequently washed three times with PBS and lysed with Triton X-114. The suspensions were diluted, and viable bacteria were calculated as described above. Adhered bacteria were determined as a percentage of inoculated bacteria, and invasion was quantified as internalized bacteria relative to adhered bacteria. All assays were conducted in duplicate and repeated independently three times. The significance of differences in adhesion and invasion abilities between samples was determined using Student's t-test. A P-value<0.05 was defined as significant.

In vivo colonization assay

Groups of seven chickens were divided randomly and equally into a test group and a control group, maintained in separate isolators, and provided with food and water ad libitum. The chickens were inoculated by feeding 0.1 mL of a bacterial suspension, which contained 106 CFU of C. jejuni in PBS. Before inoculation, a cloacal swab was taken from chickens to confirm that C. jejuni was absent prior to challenge. At day 7, a cloacal sample was taken to confirm that the chickens had become shedders. Ten days after challenge, the chickens were sacrificed by decapitation. The cecum and colon were aseptically removed, and C. jejuni cells were quantified by plating out serial dilutions onto Skirrow agar plates containing 250 μg mL−1cephaperazon to reduce the growth of any other bacteria present in the chicken gut (Wassenaar et al., 1993). The levels of colonization were given as CFU per gram of cecal or colon contents per individual chicken, and the detection limit for colonization was 102 CFU g−1of cecal or colon contents (Ahmed et al., 2002). The difference between the two groups was analyzed using Student's t-test, and the level of colonization and challenge dose were calculated as log10 number of C. jejuni organisms (Stern et al., 1990).


Inactivation of the kpsE gene and expression of the capsule polysaccharide by kpsE mutants

The plasmid pBkEkmR containing the kpsE gene was created successfully (Fig. 1). This plasmid was then used to introduce a kanamycin resistance gene into the kpsE coding regions of the C. jejuni 81116 genome. The resulting mutant construct was then used to transform C. jejuni 81116 by natural transformation. The presence and orientation of the kanamycin cassette in the kpsE mutant transformants were confirmed by PCR using primers kpE-1 and kpE-4, which flanked the inserted fragment (Figs 1a and 2), to confirm a double crossover event between the donor plasmid and the acceptor genome. Additionally, Southern blots were performed to ensure the presence of the KanR cassette in the correct location (data not shown). Only constructs containing the KanR gene in the same transcriptional orientation as the kpsE gene were used to transform C. jejuni, thus reducing any possible polar effects. Furthermore, polar effects are unlikely, as the kpsE gene is situated in an operon dedicated to capsule biogenesis. Polysaccharide extracts of the kpsE mutant and its parent strain were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) followed by silver and Alcian blue staining to evaluate their CPS profiles. The kpsE mutant showed the same Lipooligosaccharide (LOS) profile as the parent strain, whereas the inactivation of the kpsE gene resulted in the disappearance of the C. jejuni capsule (Fig. 3). Electron microscopy also demonstrated that the kpsE mutants did not express any CPS molecules (data not shown).

Figure 1.

 A schematic representation of the construction of the kpsE mutant construct (a). Locations of the endonuclease recognition sites EcoRI, BamHI, and XbaI were denoted by E, B, and X, respectively. The resulting plasmid pBkEkmR (b).

Figure 2.

 PCR-verification of the insertion of the KanR cassette in the kpsE mutant. The PCR products were obtained with the primers kpE-1 and kpE-4, which flank the kpsE gene. Lanes 1 and 3, are λ DNA digested with PstI, while lanes 2 and 4 are the wild type and the kpsE mutant, respectively.

Figure 3.

 Gel electrophoresis profiles of LOS and CPS of Campylobacter jejuni wild type and the derivative mutants visualised by silver stain (a) and Alcian blue stain (b). The kpsE mutant (lanes 2 and 4) no longer expressed the CPS but showed the original LOS molecule.

Analysis of adhesion and invasion capabilities of the CPS mutant

In order to examine the effect of the kpsE mutation on adhesion and invasion abilities, the kpsE mutant and its parent strain were exposed to the epithelial cell line INT-407. As shown in Table 2, the capacity of the kpsE mutant to adhere to the INT-407 cells was reduced 20-fold, and the invasion capability of the kpsE mutant was two-fold lower than that of the wild type (P<0.02).

Table 2.   Adhesion and invasion abilities of the Campylobacter jejuni parent strain and its derivative mutant*
% of inoculum% of wild type% adhered cells% of wild type
  • *

    STM1 and E. coli DH5α were used as controls.

  • The percentage of adhesion and invasion is given with reference to the wild-type strain 81116, with the wild type taken as 100%.

  • Invasion is given as the percentage of adhered cells that invaded.

  • §

    The numbers in parentheses are the P-values of the Student's t-test analyses.

811164.62 ± 0.561002.4 ± 0.9100
kpsE mutant0.27 ± 0.15.8 (0.01)§1.2 ± 0.550 (0.02)§
STM 14.10 ± 1.1 7.8 ± 1.6 
E. coli3.60 ± 2.0 0.01 ± 0.005 

Colonization of the chicken intestinal tract

Colonization of the chicken intestinal tract was assessed by enumeration of bacteria in the cecum and the colon at day 10 postinoculation. The results from this study showed that the kpsE mutant and the parental strain had different colonization potentials in this animal model. All the chickens were colonized (cecum and colon); however, different levels of colonization were observed (Table 3). The CPS mutant and the C. jejuni wild type were recovered from the cecal and colon contents of all seven chickens at day 10 (Table 3). Although the CPS mutant showed a slight decrease in colonization ability, the counts for this strain were similar to those of the wild type strain at day 10 when the experiment was terminated. This indicated that, in strain 81116, CPS does not play a significant role in colonization of the chicken intestine. No detectable C. jejuni organisms were found in the control group, which was inoculated with PBS.

Table 3.   Ability of the kpsE mutant and the wild type (Campylobacter jejuni 81116) to colonize the cecum and colon of 1-day-old chickens
StrainInoculumChickens colonized*Colonization level
CecumColonCecum contentsColon contents
  • *

    Number of chickens demonstrating bacterial colonization after 10 days per total number of chickens.

  • Bacterial counts are presented as the mean CFU per gram of cecum or colon contents ± SD.

C. jejuni 811162 × 1077/77/71.6 × 107± 1.1 × 1061.7 × 106± 1.0 × 105
CPS mutant2 × 1077/77/75.1 × 105± 2.1 × 1041.04 × 104± 2.0 × 103


CPS is an important cell surface component of C. jejuni. To obtain more information concerning the contribution of the kpsE gene to bacterium–host interactions, some biological activities associated with bacterial virulence were analyzed: bacterial attachment, cell invasion, and the ability to colonize the intestinal tract of chickens. For this purpose, C. jejuni 81116, a human clinical isolate, which has been described as a good colonizer of the chicken intestinal tract (Cawthraw et al., 1996; Ahmed et al., 2002), was used. To determine the role of CPS in bacterium–host relationships, a mutant defective in the expression of CPS was created by inactivating one of the capsule transporter genes, kpsE, in the C. jejuni 81116 chromosome. The constructed plasmids used to create a CPS-defective mutant contained more than the minimal length of flanking DNA necessary to allow homologous recombinations as suggested by Wassenaar et al. (1993). Two other kps genes of C. jejuni, kpsM and kpsT, which are located in the same region of the chromosome as the kpsE genes, have been studied previously (Karlyshev et al., 2000; Bacon et al., 2001). Like the kpsE gene, both these genes are involved in the capsule transport mechanism. However, the kpsM and kpsT genes are ABC transporter genes, whereas the kpsE gene is involved in polymer transport through the bacterial cell surface. According to the E. coli group II CPS models, proteins encoded by the kpsE gene are needed to connect export machinery of the inner membrane proteins, KpsM and KpsT, directly to other proteins in the outer membrane, thereby allowing the nascent polymer to cross the periplasmic space (Whitfield & Roberts, 1999; Arrecubieta et al., 2001).

Analysis of the polysaccharides of the kpsE mutant by SDS-PAGE followed by silver staining and Alcian blue staining in addition to electron microscopy studies showed that the mutant did not express a capsule. As the CPS mutant still expressed complete LOS molecules, this confirms previous reports that the capsule of C. jejuni 81116 is expressed independently of the LOS molecule (Aspinall et al., 1995; Muldoon et al., 2002).

In this study, the adhesion and invasion abilities of the kpsE mutant were significantly reduced compared to the parent strain. These results are consistent with a previous study showing that the CPS contributes to the adhesiveness and invasiveness of C. jejuni 81176 (Bacon et al., 2001).

Campylobacter jejuni 81116 is well characterized in terms of its potential to colonize the chicken intestine (Cawthraw et al., 1996; Ahmed et al., 2002). In this study, the role of CPS in colonization of the chicken gut was investigated. The kpsE mutant showed a slightly reduced level of colonization, indicating that other factors, including LOS and glycoproteins, may be involved in the colonization process of C. jejuni 81116.

In conclusion, whereas adherence may be considered as a common phenomenon in C. jejuni, the invasiveness of C. jejuni associated with the infectious process is strain-specific (Konkel et al., 1992). The data that emerged from our study showed that the CPS of C. jejuni 81116 is involved in the adherence process prior to the uptake of bacterial cells by human epithelial cells. Furthermore, CPS molecules are not the only factors involved in the adhesion and invasion ability of C. jejuni 81116. Previously, other molecules, including glycoproteins, have been shown to play an important role as adhesins and colonization factors (Karlyshev et al., 2004). Therefore, this study also suggests that in strain 81116, inactivation of the kpsE gene did not have an effect on the synthesis of C. jejuni glycoproteins. Further studies are required to improve our understanding of the molecular mechanisms involved in the colonization process. Such knowledge may lead to new approaches for reducing or eliminating C. jejuni from chickens, and hence reducing the number of human infections.


The help of Associate Prof. Ann Lawry with the electron microscopy studies is gratefully acknowledged. BMB is a recipient of a scholarship provided by the QUE project, Faculty of Dentistry, University of Indonesia, Indonesia.