Skip to main content

Development and validation of a CRISPR interference system for gene regulation in Campylobacter jejuni

Abstract

Background

Campylobacter spp. are the leading cause of bacterial food-borne illness in humans worldwide, with Campylobacter jejuni responsible for 80% of these infections. There is an urgent need to understand fundamental C. jejuni biology for the development of new strategies to prevent and treat infections. The range of molecular tools available to regulate gene expression in C. jejuni is limited, which in turn constrains our ability to interrogate the function of essential and conditionally essential genes. We have addressed this by developing and utilising a CRISPR-based interference system known as CRISPRi in C. jejuni to control gene expression. To achieve this, a catalytically inactive (“dead”) cas9 and sgRNA backbone from the Streptococcus pyogenes CRISPRi system was combined with C. jejuni-derived promoters of predetermined expression activities to develop a CRISPRi-based repression tool in C. jejuni strains M1Cam and 81–176.

Results

The CRISPRi tool was validated through successful repression of the arylsulphatase-encoding gene astA using a range of sgRNA target sequences spanning the astA gene. The tool was also applied to target astA in an M1Cam CRISPR-Cas9 deletion strain, which showed that the presence of an endogenous CRISPR-Cas9 system did not affect the activity of the CRISPRi-based repression tool. The tool was further validated against the hippicurase-encoding gene hipO. Following this, the flagella genes flgR, flaA, flaB and both flaA and flaB were targeted for CRISPRi-based repression, which resulted in varying levels of motility reduction and flagella phenotypes as determined by phenotypical assays and transmission electron microscopy (TEM).

Conclusions

This is the first report of a CRISPRi-based tool in C. jejuni, which will provide a valuable resource to the Campylobacter community.

Peer Review reports

Background

Campylobacter spp. infection is the leading cause of bacterial food-borne illness in humans worldwide[1], with Campylobacter jejuni responsible for ~ 80% of infections[2, 3]. Campylobacteriosis presents as acute gastroenteritis which, although typically self-limiting, can warrant antibiotic treatment. Moreover, accumulating evidence associates campylobacteriosis with chronic sequelae including Guillain-Barré syndrome[4, 5] and inflammatory bowel disease[6, 7]. Consequently, there is an urgent requirement to understand fundamental C. jejuni biology and transmission routes for the development of new strategies to prevent and treat infections.

The implementation of transposon (Tn) mutagenesis in C. jejuni has provided key insights into genes essential for growth and survival[8]. It has also identified several essential or conditionally essential genes of unknown function. There are limited molecular tools available to regulate gene expression in C. jejuni, which in turn hinders investigation into the function of essential and conditionally essential genes. As deletion or inactivation of these genes will likely result in a lethal phenotype, the means to regulate the expression of target genes would provide a useful step forward in gene function analysis in C. jejuni. We aimed to address this by establishing and validating a CRISPR interference (CRISPRi)-based repression tool in C. jejuni to control gene expression.

CRISPRi utilises a catalytically inactive ("dead”) Cas9 protein (dCas9) along with an sgRNA to provide repression of target genes. Whilst conventional CRISPR-Cas9 technology performs targeted double-stranded cleavage of DNA, CRISPRi uses a sgRNA-dCas9 complex which binds to the target sequence and prevents transcription via steric hindrance of the RNA polymerase protein complex[9]. CRISPRi has been successfully developed in a wide range of Gram-negative and Gram-positive bacteria[10], including Escherichia coli[9, 11,12,13], Bacillus subtilis[14, 15]), Staphylococcus aureus[16,17,18], Mycobacteria spp[19,20,21,22] and Vibrio cholerae[23].

A CRISPRi construct constitutively expressing both dCas9 and sgRNA was designed, informed by an activity analysis of candidate C. jejuni promoters, and integrated into the chromosome of the C. jejuni strains M1Cam[24, 25] and 81–176[26]. The S. pyogenes-derived CRISPRi tool was validated against the target gene astA, which encodes an arylsulphatase enzyme that releases sulphates from aryl compounds. Importantly, its activity can be detected in C. jejuni through growth on the chromogenic substrate 5‐bromo‐4‐chloro‐3‐indolyl sulphate (XS)[27], and can be quantified using an nitrophenol assay, which measures the cleavage of nitrophenol from potassium 4-nitrophenyl sulfate by AstA in the presence of tyramine[27, 28]. These phenotypic readouts provided a convenient visual assessment of CRISPRi system functioning (XS agar) alongside a quantitative assessment of AstA activity as an indicator of astA transcription (nitrophenol assay) when assessing C. jejuni CRISPRi strains.

CRISPRi-based repression was developed in two C. jejuni strains, 81–176 and M1Cam. It is important to note that whilst M1Cam contains an endogenous Type II-C CRISPR-Cas9 system, 81–176 does not[29]. Therefore, the use of these two strains not only permitted demonstration of cross-strain functionality of the system, but also enabled the opportunity to explore whether the presence of a native CRISPR-Cas9 could impact functionality of exogenous CRISPRi machinery. This was achieved through the generation of a CRISPR-Cas9 locus deletion in M1Cam.

A key component of the validation experiments was to consider the suitable expression levels of both dCas9 and sgRNAs. It has been shown that excessive expression of S. pyogenes dCas9 in other bacterial species can be toxic[30]. However, sufficient expression of dCas9, as well as sgRNAs, must be achieved in order to create a system that delivers adequate levels of repression[9, 30]. Therefore, an analysis of candidate C. jejuni promoters for use in a CRISPRi system was performed using a LacZ reporter system. This enabled the selection of appropriate promoters to express dcas9 and sgRNA within the CRISPRi constructs. Following this, target regions in the astA gene were identified and used to design a set of sgRNAs. The sgRNAs and dcas9 were used to construct a series of homologous recombination plasmids, designed to deliver the CRISPRi machinery into a pseudogenic region (CjM1_0055-0056/57) in the chromosome of 81–176 and M1Cam. Following this, the astA phenotype of the resulting CRISPRi strains was qualitatively and quantitively evaluated using chromogenic agar plating and nitrophenol assays respectively. Finally, the impact of the endogenous CRISPR-Cas9 system in M1Cam was investigated, through deletion of cas2, cas1, cas9 and the CRISPR array. A series of CRISPRi strains targeting astA were generated in the deletion strain to allow comparison of AstA activity in the presence and absence of the endogenous CRISPR-Cas9 system.

Following astA targeting in M1Cam and 81–176, the CRISPRi-based repression tool was further validated through targeting of the gene hipO, which encodes a hippicurase enzyme. Hippicurase is an enzyme which converts hippuric acid into benzoic acid and glycine[31] and is present in both M1Cam and 81–176. Detection of its activity can be achieved through a hippicurase assay[32]. During the assay, bacterial cells are lysed to release the hippicurase enzyme. Following this, HipO activity is assessed through the addition of ninhydrin, which reacts with glycine and results in the formation of a purple product, which is measured using a spectrophotometer at 570 nm. A series of CRISPRi strains in M1Cam and 81–176 were generated to target hipO at a range of locations from 5’ to 3’, and the hippicurase assay was used to assess HipO activity in each.

Next, genes from the flagella biosynthetic pathway were selected for CRISPRi-based repression. Flagella biosynthesis is a tightly regulated process, whereby following formation of the flagellar T3SS core, the FlgSR response-regulator system is activated through FlgS auto-phosphorylation and subsequent FlgR phosphorylation. FlgR then triggers the expression of σ54 Class II genes, although the mechanism by which it performs this is unclear [33, 34]. Expression of σ54 Class II genes results in the production of the minor flagella filament protein FlaB, and the remaining hook basal body components including the rod and the hook. Following this, the regulator protein FlgM is secreted, lifting its repression of σ28, which then mediates expression of Class III genes, including the major flagella filament gene flaA, resulting in the assembly of the complete flagellum[35].

This pathway was targeted for several reasons. Firstly, motility is a key determinant of virulence in C. jejuni[36, 37], and therefore the flagella biosynthesis pathway is a heavily researched area within the field. Despite this, gaps still exist in the understanding of the regulation of flagella biosynthesis[38, 39]. If CRISPRi-based repression of genes involved in this pathway could be achieved, this would offer a useful tool to explore regulation of flagellar assembly in greater depth. Secondly, from a practical level, C. jejuni motility can be measured using soft-agar stab motility assays, which therefore provided a phenotypic screen to assess the levels of CRISPRi-based repression of genes within the flagella biosynthesis pathway. Finally, although the CRISPRi-based repression tool has been verified through targeting of hipO and astA, applying it to genes within the flagella biosynthesis pathway provided the opportunity to validate the tool with a more complex regulatory system which has still not been fully resolved.

Within the flagella biosynthesis pathway, the genes flgR, flaA and flaB were selected for CRISPRi-based repression. The rationale for choosing these genes was as follows: Firstly, flgR was chosen as its deletion has previously been shown to abolish motility[40], meaning that complete repression of flgR was predicted to result in the same phenotype. In turn, incomplete repression was expected to result in a reduction in motility. Secondly, flaA and flaB were selected as their products form the external flagellar filament[41, 42], and their expression lies further down the flagella biosynthetic pathway in comparison to flgR[33], meaning that the pathway could be targeted at multiple points.

As well as being targeted for CRISPRi-based repression individually, flaA and flaB were also targeted simultaneously using a single sgRNA target location, as the genes share 92.5% sequence homology in both M1Cam and 81–176. The motility of all CRISPRi transformants was measured using soft-agar motility plating to assess the impact of CRISPRi-based repression on the flagella biosynthesis pathway. Following this, selected CRISPRi transformants were further analysed using RT-qPCR to gain insights into the expression levels of target genes, and transmission electron microscopy (TEM) was used to observe the phenotype of selected populations of bacteria.

In summary, we have developed and validated a novel CRISPRi tool for targeted transcriptional repression in C. jejuni.

Results

Activity screening of candidate C. jejuni promoters using a LacZ reporter system

To identify promoters for appropriate expression levels of S. pyogenes dcas9 and sgRNA, a series of candidate C. jejuni promoters were assessed for their expression activities using a LacZ reporter assay (Fig. 1A, i-iv). The promoters chosen for this analysis included those driving the following genes; porA (major outer membrane protein), cat (chloramphenicol acetyltransferase resistance cassette), cas9 (CRISPR-associated protein 9) and metK (methionine adenosyltransferase). These promoters are hereafter referred to as pPorA, pCat, pCas9 and pMetK. The promoters pPorA and pCat were selected due to our previous experience[25], with pPorA being considered a “high-strength” promoter and pCat a “low-strength” promoter. pCas9 was chosen in order to gain an insight into the endogenous levels of expression of C. jejuni cas9. Finally, pMetK was chosen due to its previous use to express cas9 in complementation strains[43].

To enable comparative assessment of activity, each promoter was cloned with a lacZ gene and transformed into a pseudogenic chromosomal region in 81–176 and M1Cam using the homologous recombination vector pRC1 (Fig. 1A). A no promoter control was included to detect any promotor read-through that could contribute to lacZ expression in the pseudogenic site. Together, this resulted in a panel of 5 transformants per C. jejuni strain which were then assessed for LacZ activity using a β-galactosidase assay protocol previously developed in C. jejuni to monitor plasmid-based LacZ expression[44].

Fig. 1
figure 1

Activity screening of candidate C. jejuni promoters in M1Cam and 81–176 using a LacZ reporter system. Four C. jejuni promoters, pPorA (A,i), pCatA (A,ii), pCas9 (A,iii) and pMetK (A,iv) were each inserted into a lacZ Kanr reporter construct (v). The reporter constructs were transformed into M1Cam and 81–176 and screened using a β-galactosidase assay (B and C, respectively). In both strains, pPorA demonstrated the greatest activity, followed by pCat and pCas9 which displayed similar activities, and pMetK, which had the lowest activity. **** = p < 0.0001, *** = p < 0.001. N = 3 biological replicates, Statistics performed using an unpaired t-test, error bars = SD

As expected, the results of the β-galactosidase assay (Fig. 1B &C) showed that in both strains, pPorA had the greatest activity, with a significantly higher activity than any of the other promoters, in both M1Cam and 81–176. pCat and pCas9 both showed a similar level of activity in both strains, demonstrating “moderate-strength” activities within the range of the promoter panel tested. Finally, pMetK displayed the lowest activity amongst the promoter panel, with a significantly weaker level of LacZ activity in comparison to pPorA, pCat and pCas9 in both M1Cam and 81–176. Interestingly, a small amount of LacZ activity was detected in the “no promoter” 81–176 strain, indicating that there may be some readthrough from promoters upstream of this region. Overall, the results of the assay provided quantitative assessment of activity of four C. jejuni promoters in strains 81–176 and M1Cam.

Assembly of CRISPRi components into constructs for transformation into C. jejuni

From the data gathered on promoter activities using the β-galactosidase assay, the promoters pCat and pPorA were chosen to express dcas9 and sgRNA, respectively. pCat was chosen to express dcas9 as firstly, it showed a similar level of expression in comparison to the endogenous C. jejuni cas9 promoter, and secondly, it demonstrated a moderate level of expression in comparison to pPorA. This was preferable to avoid dCas9 overexpression whilst also allowing sufficient dCas9 production for repression to occur. pCat was chosen over pCas9 due to its extensive use in our previous studies. Meanwhile, pPorA was chosen to express the sgRNA component to saturate the system with target sgRNA and thereby prevent their levels being a limiting factor in repression.

Next, the astA gene sequence was screened using a sgRNA target generator script[45] to identify a series of 20 bp sgRNA target sequences directly upstream of a 5’-NGG-3’ PAM site in the template strand of the gene, meaning the sgRNAs would target the non-template strand (Fig. 2A). The target generator script contained a series of parameters that were set to minimise the likelihood of off-target effects, which discarded any sgRNAs that were likely to disrupt the transcription of any genes other than astA. Following this, the output target sequences were passed through a sgRNA hairpin checker script[46] to detect hairpin formation between the target sequence and the backbone of the sgRNA. Target sequences where hairpin formation was detected were discarded. Finally, the GC content of the remaining target sequences were calculated and guides with a GC content below 30% and above 60% were discarded. This was because target sequences were integrated into CRISPRi plasmids as primers through inverse PCR, and therefore required an appropriate GC content to enable stable binding of primer and template. This culminated in a list of 14 sgRNA sequences shared by M1Cam and 81–176, targeting a series of regions from 5’ to 3’ along astA, each named astA_N, where N was the location of the first base of the PAM site adjacent to the target region. Although it has been suggested that targeting the 5’ end of genes of interest is the most effective strategy in CRISPRi sgRNA design[9], sgRNAs spanning the length of the gene were selected, as the system had not been trialled in C. jejuni before. However, 2 sgRNAs from the list of 14 were omitted; astA_1369 as it was only 1 bp downstream from astA_1368, and astA_1825 following multiple unsuccessful attempts to construct a plasmid containing this target sequence. This resulted in 12 sgRNA target sequences to be taken forward to generate CRISPRi constructs (Fig. 2B &C).

Fig. 2
figure 2

Design of sgRNA target sequences for integration into an sgRNA scaffold to combine with dcas9 to form a CRISPRi homologous recombination vector. An sgRNA target generator script was used to create a list of sgRNA targets using the astA gene sequence (A). Targets were screened for off-target effects using the genome sequences of 81–176 and M1Cam. Target sequences were then screened to detect any hairpin formation with the sgRNA scaffold backbone and selected for a GC content between 30–60%. The final list of 12 sgRNA target sequences are shown in (B), with their relative locations along the astA gene shown in (C). To express sgRNAs, a scaffold construct with a pPorA promoter and an sgRNA backbone was synthesized (D) and inserted into pRC1 in the opposite orientation to dcas9 driven by pCat (E)

The dcas9 and sgRNA scaffold components were assembled into the homologous recombination vector pRC1, facing in opposite orientations to prevent promoter read-through, to produce a CRISPRi construct constitutively expressing dcas9 and an sgRNA (Fig. 2D &E). Subsequently, only the target region of the sgRNA scaffold was modified to produce each different CRISPRi construct, resulting in 12 plasmids – pRC40-44 and pRC73-80 (Table S2).

Deletion of CRISPR-Cas9 locus in M1Cam

In preparation for the transformation of CRISPRi constructs into M1Cam, a deletion mutant of the CRISPR-Cas9 locus was generated to control for any competition between the endogenous CRISPR-Cas9 and the exogenous CRISPRi machinery. The CRISPR-Cas9 locus in M1Cam is a Type-II system and features cas2, cas1 and cas9 genes and a small CRISPR array[29]. A deletion vector, pRC_CRISPR (Table S2) containing ~ 500 bp left and right homologous flanking sequences to regions up- and down-stream to the 4737 bp CRISPR region and a chloramphenicol acetyltransferase resistance cassette (cat) were used to delete the region from the M1Cam genome. Constructs were confirmed and verified by whole genome sequence (WGS) analysis. This produced strain M1Cam ∆CRISPR_array/cas2/cas1/cas9::cat, hereafter referred to as M1Cam∆CRISPR. WGS showed successful deletion of the region but highlighted a single CTT deletion in an intergenic region located at 1,279,352 bp. The intergenic region was between CjM1Cam_1315, encoding a putative plasmid stabilisation system protein and CjM1Cam_1315, a tRNA-Ser, which is encoded on the reverse strand, meaning that the CTT deletion was downstream of each coding region. SNPs and INDELs can appear during growth and passage of C. jejuni[25, 47, 48], therefore M1Cam∆CRISPR was taken forward for the generation of CRISPRi transformants.

Assessment of CRISPRi-based repression of astA using a nitrophenol assay

Constructs pRC40-44 and pRC73-80 (Table S2) were transformed into a pseudogenic chromosomal region in 81–176, M1Cam and M1Cam∆CRISPR (CjM1_0055-56/0057) to produce 12 CRISPRi transformants targeting astA per strain. Each transformant was named after the sgRNA within the CRISPRi plasmid it had been transformed with (e.g. astA_47) (Table S1). Plating of the transformants onto MH agar supplemented with 100 µg/ml XS enabled qualitative assessment of AstA activity in each strain, with a reduced level of blue colouration of the bacterial growth to varying degrees seen in all 12 CRISPRi transformants across all 3 strains (Fig S1).

To quantify this observation, nitrophenol assays were performed for each set of transformants. For M1Cam, M1Cam∆CRISPR and 81–176, the nitrophenol assay results demonstrated a significant decrease in AstA activity for all sgRNAs compared to their respective parental wild-type (WT) (Fig. 3 A, B & C). Although the efficacy of astA repression appeared to differ between sgRNA target location, this did not prove to be statistically significant. Interestingly, the overall levels of AstA activity within M1Cam were greater than 81–176. This could be seen visually during culture on chromogenic XS agar (Fig S1) and upon quantification with nitrophenol assays, with a mean nitrophenol release of 247.6 µM, 284.6 µM and 97.3 µM for M1Cam, M1Cam∆CRISPR and 81–176, respectively (Fig. 3A, B & C).

Fig. 3
figure 3

Quantification and comparison of AstA activity in 81–176, M1Cam and M1Cam∆CRISPR CRISPRi transformants using a nitrophenol assay. A significant decrease in AstA activity was demonstrated for each sgRNA in all CRISPRi transformants (A, B and C) in comparison to each respective WT (**** = P < 0.0001, ANOVA, N = 3, error bars = SD). Mean percentage decrease calculations for each CRISPRi transformant with respect to the WT in each strain showed no significant difference in CRISPRi efficacy between M1Cam and M1Cam∆CRISPR (D). All sgRNA target locations showed > 78% decrease in detectable AstA activity across all strains. No significant difference in sgRNA efficacy was detected between strains except for astA_85 and astA_1091 between M1Cam and 81–176 (* = p < 0.05, 2-way ANOVA, error bars = SD)

Next, to assess if there was any potential competition between the endogenous CRISPR machinery in M1Cam and the exogenous CRISPRi components that affected CRISPRi repression efficiency, AstA activity levels between M1Cam and M1Cam∆CRISPR strains were compared. The efficacy of each sgRNA for each strain was compared by calculating the mean percentage decrease in AstA activity with respect to each parental WT strain. The percentage decrease for each sgRNA was then compared between strains (Fig. 3D). There was no statistically significant difference between M1Cam and M1∆CRISPR AstA activity decreases for each sgRNA. Implying that there was not competition between the endogenous CRISPR-Cas9 and the exogenous CRISPRi machinery.

Comparisons between the M1Cam, M1Cam∆CRISPR strains and 81–176 strains were performed to assess the efficacy of each sgRNA location between strains. This showed that there was no significant difference in sgRNA efficacy except for sgRNA astA_85 and sgRNA_1091 between M1Cam and 81–176, whereby 81–176 had a greater decrease in AstA activity than M1Cam with these sgRNAs. Additionally, the analysis showed that all sgRNAs provided at least 78% decrease in activity across all strains (Fig. 3D), highlighting the repression efficacy of the CRISPRi-based repression system.

Further validation of CRISPRi-based repression targeting the hippicurase gene hipO

Following successful CRISPRi-based repression of astA, the gene hipO was targeted to further validate the tool. A total of 6 sgRNA target locations from 5’ to 3’ along hipO (Fig. 4A) were identified using the sgRNA target generator script and subsequent quality control described previously. The sgRNA target sequences were integrated into CRISPRi plasmids to produce pRC97-pRC102 (Table S2) which were then used to transform M1Cam and 81–176. Following this, transformants were assessed for HipO activity using a hippicurase assay.

Fig. 4
figure 4

Further validation of CRISPRi-based repression tool through targeting of hipO. Six target sgRNA locations were identified in the hipO gene in M1Cam and 81–176, distributed from the 5’ to 3’ end of the gene (A). Assessment of HipO activity in M1Cam and 81–176 CRISPRi strains targeting hipO in a range of locations (B). All target locations in 81–176 CRISPRi strains showed significantly decreased HipO activity (**** = p < 0.0001, *** = p < 0.001, one way ANOVA). All target locations except HipO_883 showed significantly decreased HipO activity in M1Cam CRISPRi strains (††† = p < 0.001, †† = p < 0.01, † = p < 0.05, one way ANOVA), with a trend of decreased activity from 5’ to 3’, although no significant difference was found between sgRNA target locations. 81–176 – N = 3 biological replicates, M1Cam – N = 5 biological replicates, error bars = SD. Analysis of relative percentage decrease in HipO activity for each strain (C) revealed that the sgRNAs HipO_676, HipO_883 and HipO_976 showed a significantly greater decrease in HipO activity in 81–176 than in M1Cam. Statistics calculated using two-way ANOVA, *** = p < 0.001, ** = p < 0.01 and * = p < 0.05, error bars = SD

The results of the hippicurase assay showed a significant decrease in HipO activity in 81–176 for all sgRNA locations (Fig. 4B). A significant decrease in HipO activity was detected in M1Cam for all sgRNA locations except HipO_883. Furthermore, the level of HipO activity appeared to decrease with sgRNA locations further from the 5’ region, although the pairwise comparisons were not statistically significant. When relative percentage decreases in HipO activity were calculated, this revealed that sgRNAs HipO_676, HipO_883 and HipO_976 showed a greater decrease in HipO activity in 81–176 in comparison to M1Cam (Fig. 4C). This highlighted that there may be inter-strain differences in the efficacy of targeting certain gene locations with CRISPRi-based repression. Overall, these results further validated the functioning of the CRISPRi-based repression tool by demonstrating repression of hipO in M1Cam and 81–176.

Targeting genes within the flagella biosynthesis pathway with the CRISPRi-based repression tool

The flgR, flaA and flaB gene sequences from M1Cam and 81–176 were screened for sgRNA target sequences using the methods, and selection criteria, described above. This identified 5 sgRNA target sequences in flgR (Fig. 5A), 2 in flaA, and 2 in flaB (Fig. 6A). The low number of sgRNA target sequences found for flaA and flaB was attributed to the fact that flaA and flaB share 92.5% nucleotide identity, meaning that shared sgRNA target sequences were discarded to limit off-target effects. This raised the question as to whether sgRNA target locations shared by flaA and flaB could be used to repress both flaA and flaB simultaneously. To address this, flaA and flaB gene sequences were screened for sgRNA target sequences using whole genome sequences omitting flaB and flaA, respectively. This identified 8 sgRNA target sequences present in both flaA and flaB in M1Cam and 81–176 (Fig. 7A).

Fig. 5
figure 5

Assessing the impact of CRISPRi-based repression of flgR on M1Cam and 81–176 motility. Five sgRNA target locations (A) were identified in flgR in M1Cam and 81–176. Motility plating of CRISPRi transformants (B) showed a significant decrease in motility (* = p < 0.05, ** = p < 0.01 and *** = p < 0.001, One-way T-test, N = 3 biological replicates, error bars = SD) for all sgRNA locations except M1Cam flgR_559 and 81–176 flgR_647, which showed no significant difference in motility

Fig. 6
figure 6

Assessing the impact of CRISPRi-based repression of flaA and flaB independently in M1Cam and 81–176 motility. Two sgRNA target locations unique to each of flaA and flaB were identified (A). Motility analysis of CRISPRi transformants targeting flaA (B) showed a significant motility reduction in M1Cam flaA_761 and flaA_1662, but no reduction in 81–176 flaA_761 and flaA_1662. Motility plating of CRISPRi transformants targeting flaB (C) showed a small but significant reduction in motility in 81–176 flaB_791 and flaB_1592, and a significant decrease in motility in M1Cam flaB_1592 but no significant decrease in M1Cam flaB_1662 (* = p < 0.05, ** = p < 0.01, One-way T-test, N = 3 biological replicates, error bars = SD)

Fig. 7
figure 7

Testing whether flaA and flaB can be simultaneously targeted by a single sgRNA in M1Cam and 81–176. Eight sgRNA target locations were identified in M1Cam and 81–176 flaA and flaB, whereby the red arrow represents shared sequence identity between flaA and flaB (A). Motility plating of CRISPRi transformants targeting flaA and flaB showed a significant motility decrease in M1Cam flaAB_428, flaAB_441, flaAB_527, flaAB_851 and flaAB_103, and 81–176 flaAB_428 and flaAB_1034 (B) (* = p < 0.05, ** = p < 0.01 and *** = p < 0.001, One-way T-test, N = 3 biological replicates, error bars = SD)

Following this, each of the 17 sgRNA target sequences were used to replace AstA_523 in the CRISPRi plasmid pRC44, to form plasmids pRC56, pRC57, pRC59, pRC61, pRC62, pRC103, pRC104, and pRC106-pRC116 (Table S2). These plasmids were then used to transform WT M1Cam and WT 81–176, resulting in 34 successful transformants. Phenotypic assessment of the motility of each transformant was then carried out using soft-agar motility plating.

In M1Cam and 81–176 CRISPRi transformants targeting flgR, there was a significant reduction in motility for all sgRNA target locations (Fig. 5B) except flgR_559 in M1Cam and flgR_647 in 81–176. The pattern of sgRNA target location efficacy differed between strains, with 81–176 flgR_559 displaying the greatest decrease in motility, whilst M1Cam flgR_559 showed no significant reduction in motility. Instead, the sgRNA target location demonstrating the greatest decrease in motility in M1Cam was flgR_913. The sgRNA flgR_647 showed a significant decrease in motility in M1Cam but no significant decrease in 81–176.

Next, M1Cam and 81–176 CRISPRi transformants targeting flaA were assessed for motility. Interestingly, no significant decrease in motility was shown for sgRNA target locations flaA_761 and flaA_1662 in 81–176 (Fig. 6B), whilst M1Cam demonstrated a significant decrease in motility for both locations, with flaA_761 showing the greatest reduction (Fig. 6B).

Targeting of flaB using sgRNA target locations flaB_761 and flaB_1592 revealed a small but significant reduction in motility in 81–176 (Fig. 6C). In M1Cam, targeting of flaB_761 resulted in no reduction in motility, but targeting of flaB_1592 resulted in a significant decrease in motility (Fig. 6C).

Following these results, M1Cam and 81–176 CRISPRi transformants targeting sgRNA locations present in both flaA and flaB were assessed for motility (Fig. 7B), to investigate whether both genes could be targeted for repression simultaneously using a single sgRNA. As with the results in targeting flgR, flaA and flaB, sgRNAs targeting flaA and flaB simultaneously displayed differing efficacies between strains M1Cam and 81–176. M1Cam showed a significant decrease in motility for sgRNA target locations flaAB_428, flaAB_441, flaAB_527, flaAB_851 and flaAB_1034, with the latter exhibiting the greatest reduction in motility (Fig. 7B). Meanwhile, 81–176 demonstrated a significant decrease in motility for sgRNA target locations flaAB_428 and flaAB_1034, with the former showing the largest reduction in motility (Fig. 7B).

Abolition of motility was not achieved in any of the M1Cam and 81–176 CRISPRi transformants targeting flgR, flaA, flaB, or flaA and flaB simultaneously. Overall, the data showed that the CRISPRi-based repression tool was able to repress the target gene(s) with a resultant decrease in motility, however it was unable to deliver full repression of the target gene(s).

To investigate transcript levels of genes targeted for CRISPRi-based repression quantitative reverse transcription PCR (RT-qPCR) was performed using a selection of CRISPRi transformants that demonstrated the overall greatest decrease in motility in each strain. These included the M1Cam CRISPRi transformants targeting sgRNA sequences flgR_913, flaA_761, flaB_1592 and flaAB_1034, and the 81–176 transformants targeting flgR_559 and flaAB_428.

RT-qPCR analysis of selected CRISPRi transformants targeting flgR, flaA, flaB, and flaA/flaB

RNA was extracted from M1Cam flgR_913, flaA_761, flaB_1592 and flaAB_1034, and 81–176 flgR_559 and flaAB_428. cDNA was generated from the RNA and used for RT-qPCR analysis. For each target gene, primers (Table S3) were designed to amplify two qPCR products, one in a region upstream of the sgRNA target location and one in a region downstream of it. These were notated as the gene name followed by “U” or “D”, whereby “U” was the product upstream the sgRNA target location and “D” was the product downstream of it (e.g., flgR_U and flgR_D). This was to gain an insight into how the binding of the sgRNA:dCas9 complex affected the transcription upstream and downstream of the target site. For the RT-qPCR analysis of CRISPRi transformants targeting flgR, primers were included for flaA and flaB PCR products to provide an insight of the downstream effects of altered flgR expression. The housekeeping gene gyrA, was used as a control to normalise experimental cycle thresholds against[8].

The RT-qPCR results from M1Cam flgR_913 showed a significant decrease in flgR_U, flgR_D and flaB products in comparison to M1Cam (Fig. 8A). flgR_D demonstrated the greatest fold change of 0.03, equating to ~ 33.3-fold decrease, in comparison to flgR_U which displayed a fold change of 0.59, equating to ~ 1.6-fold decrease. This indicated greater repression of flgR downstream of the sgRNA:dCas9 binding site at flgR_913 compared to the region upstream of the site. No significant reduction in flaA was detected compared to M1Cam, suggesting flaA transcription was not significantly reduced following the detected repression of flgR. However, flaB showed a significant fold change of 0.11, equating to ~ 8.8-fold decrease.

Fig. 8
figure 8

RT-qPCR analysis of selected M1Cam and 81–176 CRISPRi transformants targeting flgR, flaA, flaB and flaA/flaB. RT-qPCR was performed on M1Cam flgR_913 (A), flaA_761 (B), flaB_1592 (C), flaAB_1034 (D), 81–176 flgR_559 (E) and flaAB_428 (F) to gain an insight into levels of repression elicited in each gene. RT-qPCR products lying upstream (U) and downstream (D) of the sgRNA target locations were designed for each gene, denoted as gene_U and gene_D respectively. The genes flaA and flaB were included in RT-qPCR analyses for M1Cam flgR_913 and 81–176 flgR_559 as downstream indicators of the consequences of flgR repression. * = p < 0.05, ** = p < 0.01 and **** = p < 0.0001. Statistics calculated using One-way T-tests of ∆∆ct values, N = 3 biological repeats, error bars = 95% CI

RT-qPCR analysis of M1Cam flaA_761 showed no significant fold change in flaA_U or in flaA_D compared to M1Cam (Fig. 8B). Meanwhile, M1Cam flaB_1592 demonstrated a significant negative fold change for flaB_D of 0.04, equating to ~ 25.6-fold decrease (Fig. 8C). However, there was no significant fold change for flaB_U. As the sgRNA:dCas9 binding site at flaB_1592 lies 139 bp from the 3’ end of the gene (flaB is 1731 bp long), these results suggest that the beginning of the flaB transcript was transcribed but potentially stopped short at flaB_1592.

The results for M1Cam flaAB_1034 showed significant fold changes for flaA_U, flaA_D, flaB_U and flaB_D (Fig. 8D). Interestingly, both flaA_U and flaB_U displayed ~ 2.6 and ~ 2.0-fold increase respectively. Meanwhile flaA_U and flaB_U demonstrated 0.25 and 0.48-fold changes, equating to ~ 4.0 and ~ 2.1-fold decreases respectively. As expected, this indicated that the region upstream of the sgRNA:dCas9 binding site at flaAB_1034 in flaA and flaB was transcribed to a greater level in comparison to M1Cam, whilst the region downstream was transcribed to a lower level.

In 81–176 flgR_559, the RT-qPCR analysis showed no significant fold change in flgR_U, but demonstrated a significant negative fold change of 0.32 for flgR_D, equating to ~ 3.1-fold decrease (Fig. 8E). Like the results for M1Cam flgR_913, no significant fold change was exhibited in flaA, but a significant negative fold change of 0.39 was displayed in flaB, equating to ~ 2.6-fold decrease.

Finally, RT-qPCR analysis of 81–176 flaAB_428 showed significant negative fold changes of 0.25 and 0.08 in flaA_D and flaB_U, equating to ~ 4.0 and ~ 12-fold decreases respectively (Fig. 8F). No significant fold changes were demonstrated in flaA_U and flaB_D. This indicated that for flaA, targeting of flaAB_428 resulted in significant repression downstream of the sgRNA target location, but not upstream of it. Conversely, for flaB, targeting of flaAB_428 resulted in significant repression upstream of the sgRNA target location, but not downstream of it.

Overall, the RT-qPCR analysis showed that the CRISPRi-based repression tool elicited varying levels of repression of target genes in CRISPRi strains M1Cam flgR_913, flaB_1592, flaAB_1034, 81–176 flgR_559 and flaAB_428. No repression of flaA was detected in M1Cam flaA_761.

It is worth noting that the motility assays and RT-qPCR reported a mean value which was representative of the whole population of bacteria used. Therefore, this did not account for the possibility of population heterogeneity which has been reported in C. jejuni[47],whereby complete repression may have occurred in some bacteria but not others. To investigate the phenotype of each population and assess the possibility of sub-populations, TEM was used to visualise populations of bacteria from each CRISPRi transformant population, to identify whether any population-level heterogeneity was observed, for example bacteria with flagella and bacteria without flagella.

TEM analysis of selected CRISPRi transformants targeting flgR, flaA, flaB, and flaA/flaB

TEM analysis of M1Cam flgR_913, flaA_761, flaB_1592 and flaAB_1034, and 81–176 flgR_559 and flaAB_428 was performed to provide a gross visual overview of phenotypic presentation of flagella within each population. This was to ascertain whether there was a consistent phenotype within each population, or whether there were patterns of phenotypic heterogeneity – for example, aflagellate bacteria and fully flagellated bacteria. This was partly motivated by previous studies that have identified that WT C. jejuni can display within-population phenotypic heterogeneity[47, 49, 50].

TEM of in vitro grown WT M1Cam bacteria demonstrated a phenotype with bipolar flagella, which served as a baseline for comparison of M1Cam CRISPRi transformants (Fig. 9A). In contrast, the M1Cam flgR_913 populations showed a variety of phenotypes (Fig. 9B). Predominantly, an aflagellate phenotype was observed, with occasional individual bacteria showing short to mid-length single or bipolar flagella, and some with full-length bipolar flagella. The M1Cam flaA_761 populations contained bacteria with full-length flagella (Fig. 9C). Similarly, M1Cam flaB_1592 bacteria appeared extremely similar to M1Cam bacteria, with full-length bipolar flagella (Fig. 9D). Finally, the M1Cam flaAB_1034 bacteria grossly exhibited short bipolar flagella, often with one slightly longer flagellum (Fig. 9E).

Fig. 9
figure 9

TEM visualisation of M1Cam CRISPRi transformants targeting flgR, flaA, flaB and flaA/flaB. TEM analysis showed that CRISPRi transformants M1Cam flgR_913 (B), and flaAB_1034 (E) showed different phenotypes in comparison to M1Cam (A). M1Cam flgR_913 showed an aflagellate phenotype with some individuals exhibiting bipolar flagella (B). M1Cam flaA_761 (C) an M1Cam flaB_1592 (D) showed no difference in phenotype compared to M1Cam, and M1Cam flaAB_1034 showed short bipolar or single flagella (E). TEM images displayed were selected as representative examples of images taken from 3 biological replicates. Scale bar = 500 nm

TEM of in vitro grown WT 81–176 bacteria showed bipolar flagella against which 81–176 CRISPRi transformants were compared (Fig. 10A). The 81–176 flgR_559 bacteria showed a similar distribution of phenotypes to M1Cam flgR_913, with most of the bacteria visualised displaying an aflagellate phenotype, with infrequent observation of a few bacteria with short to mid-length single or bipolar flagella, and some bacteria with full-length bipolar flagella (Fig. 10B). The 81–176 flaAB_428 bacteria showed short to mid-length bipolar flagella (Fig. 10C).

Fig. 10
figure 10

TEM visualisation of 81–176 CRISPRi transformants targeting flgR and flaA/flab. TEM analysis showed that CRISPRi transformants 81–176 flgR_559 (B) and 81–176 flaAB_428 (C) displayed different phenotypes in comparison to 81–176 (A). 81–176 flgR_559 grossly demonstrated an aflagellate phenotype with some bacteria exhibiting bipolar flagella (B). 81-176_flaAB428 demonstrated short to mid-length flagella (C). TEM images displayed were selected as representative examples of images taken from 3 biological replicates. Scale bar = 500 nm

Discussion

Taken together, the results demonstrate that the S. pyogenes-derived CRISPRi repression tool developed here with selected sgRNA target sequences could elicit effective repression of the biosynthetic genes astA and hipO in C. jejuni strains M1Cam and 81–176. The results also showed that the CRISPRi-based repression tool could be applied to genes within the complex flagella biosynthetic pathway, resulting in varying levels of motility reduction and flagella phenotypes.

When developing the CRISPRi-based repression tool, it was necessary to carefully consider the levels at which dCas9 and the sgRNAs were expressed. Previously it has been shown that overexpression of dcas9 in combination with sgRNAs containing one of a series of specific sequences in the first 5 nucleotides proximal to the PAM (known as a “bad seed” sequence) can be toxic in bacterial cells[30]. Conversely, insufficient levels of dCas9 and sgRNAs result in low repression efficacy[51]. Balancing these requirements is essential to form a functioning CRISPRi tool. Promoter screening using a LacZ reporter assay provided a valuable insight into promoter activity which informed the decision to use pCat and pPorA for dCas9 and sgRNA expression respectively. Whilst the combination of pCat and pPorA in expressing dcas9 and sgRNA respectively resulted in successful astA and hipO repression with selected sgRNA target sequences, the trial of other promoters in the CRISPRi-based repression tool could allow a range of repression levels. For instance, pCat could be swapped with a less active promoter such as pMetK in driving dCas9 and therefore elicit a lower level of repression. Indeed, promoters outside of the four used in this study could also be assessed for their activity and be used to drive dcas9 expression to achieve a range of repression levels. Finally, the addition of an inducible promoter driving dcas9 expression would allow tuning of repression through alteration of cellular dCas9 levels, which has been used previously in pooled CRISPRi screens[10, 52], and in functional analysis of genes of interest[53].

The location of sgRNA targets across the length of the astA or hipO genes from 5’ to 3’ did not appear to have a significant impact on the efficacy of the system. Aside from dcas9 and sgRNA expression levels, the location of regions targeted by sgRNAs has been shown to be highly important in CRISPRi design[9, 54]. Concentration of targets at the 5’ end of the gene (within the first 5% of the open reading frame) have typically been favoured due to the formation of a transcriptional road block by the CRISPRi complex before most of the gene has been transcribed[9, 45]. To determine if this held true in C. jejuni, a range of sgRNAs targeting from 5’ to 3’ in astA and hipO were evaluated. The results of the CRISPRi repression of astA showed that there was no significant difference between each of the sgRNAs used along the length of the gene in both M1Cam and 81–176. When targeting hipO there was also no significant difference between sgRNA location in M1Cam and 81–176.

Although sgRNA target location did not have a significant effect on the repression of astA and hipO, some sgRNA target locations demonstrated significantly different repression efficacies between strains. For example, comparison of relative decreases in AstA activity between M1Cam and 81–176 showed that sgRNAs astA_85 and astA_1091 demonstrated a significantly greater efficacy in 81–176 than M1Cam. An equivalent comparison between HipO activity decreases showed that sgRNAs HipO_676, HipO_883 and HipO_976 also showed a significantly greater efficacy in 81–176 than M1Cam. This highlights the importance of validating results using multiple strains, if practical to do so.

Another key element in this validation work was the comparison of the application of CRISPRi-based repression in M1Cam and the CRISPR-Cas9 region deletion mutant M1Cam∆CRISPR. Generation of the mutant enabled comparison of CRISPRi function in both strains, in order to detect any competition between the endogenous CRISPR-Cas9 system in M1Cam and the S. pyogenes-derived CRISPRi components. Competition between these elements could potentially affect CRISPRi repression efficacy, and therefore had to be considered. For instance, one theoretical example of competition could be the interaction of S. pyogenes sgRNA with the endogenous C. jejuni Cas9 protein. This could potentially deplete the amount of S. pyogenes sgRNA available for complexing with S. pyogenes dCas9 and thereby affect repression efficacy. Moreover, should interaction between S. pyogenes sgRNA and C. jejuni Cas9 occur, this could result in genomic DNA cleavage at sgRNA target sites, although this would require the presence of appropriate C. jejuni PAM recognition sites. Another theoretical example of competition could be between the S. pyogenes sgRNA and endogenous C. jejuni tracrRNA:crRNA complex. If the C. jejuni tracrRNA:crRNA complex was able to interact with the S. pyogenes dCas9 protein, this would lead to reduced dCas9 interaction with the S. pyogenes sgRNA and potentially lower CRISPRi repression efficacy.

Despite these potential sources of competition, the nitrophenol assay results showed no detectable difference in the CRISPRi repression efficacy between M1Cam and M1Cam∆CRISPR. No significant differences between relative percentage decrease in AstA activity were detected between M1Cam and M1∆CRISPR for each sgRNA location. This suggested that any potential competition between the systems did not detectably interfere with the repression capability of the CRISPRi tool.

The reason why no difference was found could be because S. pyogenes and C. jejuni CRISPR-Cas9, although both being Type-II CRISPR systems, have some fundamental differences. For example, X-ray crystallography analysis has shown that C. jejuni sgRNA forms a unique triple helix structure when in complex with C. jejuni Cas9, which has not been observed in S. pyogenes sgRNA:Cas9 interactions[55]. This means that the C. jejuni tracrRNA:crRNA complex may not be able to successfully interact with S. pyogenes Cas9. Secondly, the C. jejuni and S. pyogenes Cas9 proteins each have a different PAM recognition motif (5’-NGG-3’ and 5’-NNNVRYM-3’ respectively)[55], indicating that if either the S. pyogenes dCas9 or the C. jejuni Cas9 could harness tracrRNA:crRNA or sgRNAs respectively, it is unlikely that the complexes would be able to bind to target DNA effectively.

Following validation of the CRISPRi-based repression tool against astA and hipO, the tool was used to target the genes flgR, flaA, flaB, and both flaA and flaB (simultaneously) within the highly complex flagella biosynthesis pathway. Soft-agar motility plating provided a convenient means to assess the phenotypic impact of targeting each gene(s). Following this, a selection of CRISPRi transformants were examined in-depth using RT-qPCR to gain an insight into the expression of genes targeted for repression. Furthermore, TEM analysis was used to provide an overview of the gross phenotypic presentation of each selected CRISPRi transformant.

Taken together, these results provided the opportunity to observe broad patterns in the application of CRISPRi-based repression. Firstly, at the population level, no complete abolition of motility was observed during targeting of each gene, although significant reductions in motility were achieved during CRISPRi-based repression of all genes except for flaA in 81–176. Secondly, the results reinforced the observations made during CRISPRi-based repression of hipO that there are between-strain differences in sgRNA target location efficacy. This pattern was particularly profound in this set of experiments, whereby the most effective sgRNA target location differed between M1Cam and 81–176 for all genes that showed a reduction in motility (flgR, flaA, flaB). Finally, TEM analysis of selected CRISPRi transformant populations indicated the possibility of within-population variance in the efficacy of CRISPRi-based repression, whereby phenotypes appeared to vary between bacteria in the population.

In addition to broad observations, insights into the outcomes of targeting each gene using CRISPRi-based repression were made. Firstly, targeting of flgR demonstrated that the greatest reduction in motility was achieved using sgRNA locations flgR_913 and flgR_559 in M1Cam and 81–176 respectively. Previous studies have indicated that deletion of flgR results in an aflagellate, non-motile phenotype[28, 33, 40], therefore the reduction rather than complete abolition of motility in CRISPRi targeting of flgR in M1Cam and 81–176 indicated that full repression, within the total population of bacteria, was not being achieved. To interrogate this further, RT-qPCR was performed using RNA extracted from M1Cam flgR_913 and 81–176 flgR_559, and TEM analysis was carried out on populations of each.

The RT-qPCR results for M1Cam flgR_913 showed that flgR expression was significantly decreased, with a significant reduction in qPCR products upstream and downstream of the sgRNA:dCas9 binding site at flgR_913. The results also showed a significant decrease in flaB expression. As a flagella Class II gene, flaB expression is driven by σ54, whose transcriptional activity is regulated by the FlgSR system and FlhF[34, 39]. This indicates that the decrease in flgR expression could have resulted in a decrease in flagella Class II gene expression (including flaB) through this pathway. Interestingly, flaA expression was not significantly decreased. As a flagella Class III gene, flaA expression is driven by σ28, which is itself a flagella Class II gene[34]. If flagella Class II genes had decreased expression, then a decrease in flaA expression may have been expected as a result. Phenotypic analysis of M1Cam flgR_913 using TEM provided further context for the motility and RT-qPCR results. TEM imaging broadly showed that the M1Cam flgR_913 population had an aflagellate phenotype, with occasional individual bacteria displaying mid- to full-length single or bipolar flagella. During soft-agar motility plating, these sub-populations may have grown and contributed to the motility measured on the plates. Furthermore, as the RT-qPCR results were an overall (i.e., averaged) representation of expression of flgR, flaA, and flaB in the entire population, it is possible that these sub-populations may have had little to no reduction in expression of these genes, which could have contributed to the expression measured.

The RT-qPCR results for 81–176 flgR_559 showed a similar pattern to M1Cam flgR_913. However, there was no significant decrease in expression upstream of the sgRNA:dCas9 binding site at flgR_559, only downstream. This indicated that flgR expression was significantly decreased following the sgRNA target sequence location. Expression of flaB was significantly decreased, whilst flaA expression was not. TEM analysis of 81–176 flgR_559 showed that like M1Cam flgR_913, the predominant phenotype was aflagellate, with fewer mid- to full-length single or bipolar flagella. The presence of these sub-populations may have contributed to the motility and RT-qPCR results recorded for 81–176 flgR_559 as described above for M1Cam. This indicated that the CRISPRi-based repression of flgR in both strains resulted in a similar outcome – formation of aflagellate and partial to fully flagellated sub-populations. This population heterogeneity was an important observation, as it indicated that CRISPRi transformant populations should not be assumed to always exhibit fully uniform phenotypes. This phenomenon has also been described in CRISPRi studies in E. coli [56].

Next, targeting of flaA for CRISPRi-based repression showed that only M1Cam CRISPRi transformants demonstrated a significant reduction in motility. M1Cam flaA_761 demonstrated the greatest decrease in motility, but no abolition of motility (as assessed using motility agar). Typically, deletion of flaA results in a non-motile phenotype with a short flagella comprising FlaB[57]. This indicated that full repression of flaA expression was not occurring. RT-qPCR using RNA extracted from M1Cam flaA_761 showed no significant decrease in flaA expression, and TEM analysis of M1Cam flaA_761 populations showed full length bipolar flagella. Overall, this indicated that despite the observed decrease in motility, significant repression of flaA was not detected by RT-qPCR and the TEM analysis showed no reduction in flagella length. Whilst the flagella seemed to be produced, it is possible that the FlaA proteins were truncated, which in turn could have impacted flagellum structure and movement. Neither 81–176 flaA_761 nor flaA_1662 showed a motility decrease in comparison to WT 81–176. This highlighted that on top of inter-strain differences in optimal sgRNA target sequence locations, there are inter-strain differences as to whether a gene can be repressed at all.

Interestingly, 81–176 and M1Cam CRISPRi transformants targeting flaB displayed differing motility results. 81–176 flaB_761 and 81–176 flaB_1592 demonstrated a small reduction in motility, whilst M1Cam flaB_761 showed no reduction and M1Cam flaB_1592 exhibited a large decrease in motility. Reduction but no abolition of motility has been reported following deletion of flaB, with full length flagella comprising FlaA[57]. RT-qPCR of M1Cam flaB_1592 showed a significant decrease in flaB expression downstream of the sgRNA target location site, but no decrease upstream of the site. The location at flaB_1592 was fairly close to the 3’ end of the gene, as flaB is 1731 bp, indicating that the RT-qPCR could suggest that the full transcript may not have been produced. The TEM analysis showed that the M1Cam flaB_1592 population had full-length bipolar flagella. This indicated that motility can still be impacted despite the presence of full-length bipolar flagella. Like M1Cam flaA_761, despite the presence of full flagella, it is possible that the FlaB protein could be truncated, thereby impacting the flagellum structure and movement.

The shared sequence identity of flaA and flaB allowed the opportunity to trial sgRNA to target sequences present in both genes, to determine if a single sgRNA could be used to target two genes for repression at the same time. Targeting of both flaA and flaB in M1Cam revealed that sgRNA target location flaAB_1034 showed the greatest decrease in motility, however no abolition of motility was achieved. RT-qPCR showed that flaA and flaB expression was significantly decreased downstream of the sgRNA:dCas9 binding site. This implied that both flaA and flaB were targeted at population level, although heterogeneity in flaA and flaB targeting at a single-cell level, cannot be ruled out. Interestingly, flaA and flaB expression was significantly increased upstream of the sgRNA:dCas9 binding site. TEM analysis of M1Cam flaAB_1034 populations showed that the prevailing phenotype was the presence of short bipolar flagella. This indicated incomplete formation of the flagella, which may have been due to reduced levels of FlaA and FlaB. Alternatively, incomplete flagella may have been due to truncation of FlaA and FlaB, resulting in incorrect filament structure. The shorter flagella were likely a factor in the observed decrease in motility.

Simultaneous targeting of flaA and flaB in 81–176 revealed that sgRNA target sequence flaAB_428 showed the greatest decrease in motility. RT-qPCR of 81–176 flaAB_428 showed a significant decrease in flaA expression downstream of the sgRNA:dCas9 binding site, but not upstream, whilst flaB expression was decreased upstream of the binding site but not downstream. This indicated that flaB may have been incompletely transcribed – for instance, transcription could have been resumed following the sgRNA:dCas9 binding site, although a within-gene promoter was not identified in the flaB sequence[58]. Nevertheless, these findings highlighted that despite targeting identical sgRNA target sequences within flaA and flaB, the outcome resulted in different patterns of repression. It is also worth considering that although two genes were targeted, the sgRNA expression level was not increased, meaning that the same amount of sgRNAs used previously in these experiments to target one gene were shared across two genes. This in turn may have resulted in incomplete, or lower, repression for each gene.

TEM analysis of 81–176 flaAB_428 bacteria demonstrated a gross phenotype of short- to mid-length bipolar flagella, which may have been a result of reduced FlaA and FlaB levels, or truncated FlaA and FlaB proteins, leading to incomplete flagella formation. This in turn likely contributed to the decrease in motility observed. Taken together, the targeting of flaA and flaB in M1Cam and 81–176 using a single sgRNA target sequence showed that it is possible to target two genes with shared sequence identity for repression simultaneously.

Overall, we have demonstrated that an S. pyogenes-derived CRISPRi-based repression tool can be integrated onto the chromosome of two C. jejuni strains and be used to repress the expression of the target genes astA and hipO. Whilst the presence of a CRISPR-Cas9 system in M1Cam could have been a factor in CRISPRi repression efficacy, deletion of it did not elicit any significant detectable difference in CRISPRi repression. We have also demonstrated that genes within the flagella biosynthetic pathway (flgR, flaA, flaB and flaA/flaB) can be targeted using the CRISPRi-based repression tool. This is the first report of a CRISPRi-based repression tool in C. jejuni and provides a basis for future application and development as a highly versatile molecular tool for the regulation of gene expression. We believe that it will be a valuable addition to the Campylobacter molecular-toolbox to better understand this important pathogen, with the aim of identifying means to reduce colonisation of food-producing animals and to treat or prevent infection in humans and other animals.

Materials and methods

Bacterial strains and growth conditions

All C. jejuni strains were cultured on Mueller–Hinton (MH) agar (Oxoid) supplemented with appropriate antibiotics and substrates. All C. jejuni strains were grown at 42 °C under microaerophilic conditions (5% O2, 5% CO2, 5% H2 and 85% N2) in a M95 Variable Atmosphere Workstation (Don Whitley, Shipley, United Kingdom). For long term storage, bacterial strains were scraped into Microbank (ProLab Diagnostics) cryovials using 1 μl inoculation loops and stored at -80 °C. C. jejuni strains from freezer stocks were resuscitated by streaking onto MH agar using 1 μl inoculation loops, followed by growth for 48 h. Bacterial growth was then re-streaked onto MH agar plates and harvested for use after 16 h, by scraping into MH broth using an L-shaped spreader. Bacterial strains used in this study are detailed in Table S1.

Plasmids and primers

Plasmids generated in this study are detailed in Table S2. Primers used in this study are detailed in Table S3.

Molecular biology

All PCRs to generate lacZ reporter constructs used Q5 High Fidelity Polymerase (New England Biolabs). A plasmid (pSV009) that we have previously reported for homologous recombination[25] (Table S2) was used as a template vector in the design of all lacZ reporter constructs. pSV009 features two 400 bp homologous flanking arms to the M1Cam pseudogenic region CJM1_0055– 0057 which has > 98% sequence identity in 81–176. pSV009 was modified to remove a pCat promoter and add a MluI site, yielding pRC1 (Table S2).

Primers ES02, RC20, RC118 and RC129 were used with primer RC18 (Table S3) to amplify and fuse the promoters pPorA, pCat, and pMetK to the lacZ gene amplified from E. coli K12 genomic DNA. Primers RC21 and RC18 (Table S3) were used to amplify the lacZ gene without a promoter. PCR products were digested with XhoI and BamHI (New England Biolabs) and ligated with pRC1 to produce pRC2, pRC3, pRC27 and pRC28 (Table S2).

Selection of target sequences and sgRNA design

The DNA sequences for each target gene in M1Cam and 81–176 (Genbank accession numbers: CP012149.1 and CP022551.1, respectively) were downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/nucleotide/). The Python script sgRNA_design_main.py[45] was used to generate a list of sgRNA target sequences for each gene. The lists were then cross-referenced, and only sgRNA target sequences that appeared in both M1Cam and 81–176 lists were taken forward. These sequences were then entered into the Python script gRNAInteractionScreen_ThymeFeb2016 2.py to omit sequences containing hairpin formation with the sgRNA backbone[46]. Primers were designed for integration of sgRNA target sequences into sgRNA plasmids (Table S2). DNA construct and sgRNA target illustrations were prepared using SnapGene Viewer (GSL Biotech; available at snapgene.com).

Construction of CRISPRi plasmids

An sgRNA expression construct was designed in silico, to include a pPorA sequence, sgRNA target sequence astA_47, a sgRNA scaffold from S. pyogenes and a terminator sequence and ordered as a GeneArt Strings DNA Fragment (Thermofisher), which was digested with BamHI and PstI and ligated with pRC1 to produce pRC35. This plasmid was used as a template for inverse PCR to create all other sgRNA plasmids (Table S2). The dcas9 sequence was amplified from the plasmid pdCas9 (Addgene plasmid # 46,569) using primers RC133 and RC135 (Table S3) to fuse the pCat sequence to dcas9. The product was digested with XhoI and AgeI and ligated into each respective sgRNA plasmid to produce all CRISPRi plasmids (Table S3).

Transformation of C. jejuni strains

Natural transformation of C. jejuni was performed using an adapted biphasic method[59, 60] with EcoRI-methylated DNA[61]). Resuscitated bacteria were scraped into 1 ml MH broth using an L-shaped spreader, diluted to an OD600nm of 0.5 and 500 μl was added on top of 2 ml of MH agar with 5% (v/v) defibrinated horse blood (Oxoid) ​​in a 30 ml universal tube (Greiner Bio-One). Cells were then incubated at 42 °C under microaerophilic conditions for 3 h. 300 ng methylated DNA was added to the bacteria in 20 μl dH2O, which were then incubated at 42 °C under microaerophilic conditions for 3-5 h, followed by selection plating on MH-Blood plates supplemented with appropriate antibiotics, and growth in microaerophilic conditions for 3–5 days.

Construction of M1Cam CRISPR array mutant M1Cam∆CRISPR

The CRISPR locus targeted for deletion spanned a 4965 bp region containing the CRISPR array, cas2, cas1, and cas9 (M1Cam genome location 1,431,507–1,436,472 GenBank: CP012149.1). The CRISPR locus deletion construct was designed in silico by adding ~ 500 bp sequence flanking either side of the CRISPR locus to a chloramphenicol acetyltransferase cat resistance cassette[25] to yield a homologous recombination construct. NdeI and MfeI restriction sites were added at the 5’ and 3’ end and the construct ordered as a GeneArt (Thermofisher) plasmid. The plasmid was digested with NdeI and MfeI and the construct was gel-extracted before ligation into gel-extracted pRC1 digested with NdeI and MfeI. Ligations were transformed into chemically competent E. coli DH5α cells (New England Biolabs) followed by selection-plating on LB-Cm-Amp agar plates. This yielded pRC_CRISPR_Del (Table S2). M1Cam CRISPR locus deletion strains were PCR verified using primers RC182 and RC183 (Table S3). Sanger sequencing of homologous recombination sites using primers RC182 and RC183 was performed to check for scar-less recombination (Table S3).

RNA extraction and RT-qPCR

C. jejuni strains were grown in liquid culture for 16 h or until the OD600nm reached 0.4. 2 ml RNAprotect Bacteria Reagent (Qiagen) was added to 1 ml of each culture in a 15 ml Falcon tube (Greiner), vortexed and incubated for 5 min at room temperature. Tubes were centrifuged for 10 min at 5000 × g, and the supernatant decanted. RNA extraction was performed on pellets using a Qiagen RNeasy mini kit according to the manufacturer’s protocol.

Reverse transcription was carried out using LunaScript RT SuperMix Kit (New England Biolabs). 1 μg of RNA was added to a 20 μl reaction containing a final concentration of 1 × LunaScript RT SuperMix. In parallel, 1 μg of RNA was added to a 20 μl reaction containing a final concentration of 1 × No RT control mix. Reactions were incubated in a thermocycler at 25 °C for 2 min, 55 °C for 10 min and 95 °C for 1 min.

qPCR was performed using 3 biological replicates in triplicate using Luna Universal qPCR Master Mix (New England Biolabs). The gene gyrA was used as a housekeeping gene[8]. 1 μl cDNA was added to a 20 μl reaction containing a final concentration of 1 × Luna Universal qPCR Master Mix, and 0.25 μM of each primer. No-RT control samples were included to check for gDNA contamination, along with negative controls using water. The qPCR was performed using a RotorGene Q (Qiagen) using the SYBR setting. Thermocycling conditions included an initial denaturation at 95 °C for 1 min, denaturation at 95 °C for 15 s, extension at 60 °C (45 cycles), and melt curve from 60–95 °C. Following every qPCR run, melt curves were analysed to verify the presence of a single curve for each reaction. Cycle threshold (CT) values were recorded and ∆∆CT values calculated using the following equations:

WT ∆CT = WT Ct (gene of interest) – WT Ct (housekeeping gene).

CRISPRi ∆CT = CRISPRi Ct (gene of interest) – CRISPRi Ct (housekeeping gene).

∆∆CT = (WT ∆CT) – (CRISPRi ∆CT).

Analysis of ∆∆CT values was carried out using previously described methods[62]. ∆∆CT values from each biological replicate were used to calculate a mean ∆∆CT and 95% confidence interval (CI) for each gene. A One-way T-test was performed to determine significance of each ∆∆CT value. Fold changes were then calculated for graphical representation, whereby fold change = 2−∆∆CT.

β-galactosidase assays

The β-galactosidase assay protocol was adapted from previously described methods[44, 63]. Resuscitated C. jejuni strains were harvested into 1 ml MH broth using an L-shaped spreader and diluted to an OD600nm of 0.2, which was used to prepare a starter culture of OD600nm of 0.005 in MH broth. Cultures were grown in microaerophilic conditions at 200 RPM at 42 °C for 16 h or until the OD600nm reached 0.4. 40 μl culture was removed and added to 60 μl permeablisation solution (100 mM Na2HPO4, 20 mM KCl, 2 mM MgSO4, 0.8 mg/ml hexadecyltrimethylammonium bromide, 0.4 mg/ml sodium deoxycholate, 5.4 μl/ml beta-mercaptoethanol) and incubated at 30 °C for 30 min. 600 μl substrate solution (60 mM Na2HPO4, 40 mM NaH2PO4, 1 mg/ml o-nitrophenyl-β-D-Galactoside (ONPG), 2.7 μl/ml β-mercaptoethanol) was added and the time taken for the mixture to develop a pale-yellow colour was recorded. 700 μl stop solution (1 M Na2CO3) was added and the mixture centrifuged at 20,000 × g for 1 min. 1 ml reaction mixture was carefully removed and OD420nm measured. Miller Units were calculated using the following calculation:

Miller Units = 1000 × OD420nm/(OD600nm of culture samples)x(volume taken[0.02 ml])x(reaction time(min))).

At least 3 independent experiments were carried out to determine β-galactosidase activity for each strain.

Nitrophenol assay

To quantify AstA activity, freezer stock strains were streaked onto MH agar plates supplemented with the appropriate antibiotics and grown under microaerophilic conditions for 48 h. Each strain was re-streaked onto 3 plates and then grown under microaerophilic conditions for a further 18 h. Arylsulphatase activity was measured using previously described methods[28, 64]. Briefly, each plate was scraped into 1 ml PBS using an L-shaped spreader, and then diluted with PBS (pH 7.2) to an OD600nm of 1.0. Each sample was divided into 2-paired 1 ml aliquots and centrifuged at maximum speed for 5 min. For each pair, one aliquot was washed with 1 ml buffer AB1 (0.1 M Tris·Cl, pH 7.2 (Sigma)) and centrifuged at 10,000 × g for 5 min. The pellet was washed and centrifuged at 10,000 × g for 5 min, followed by resuspension in 1 ml AB1. The same process was carried out with the other aliquot of the pair using buffer AB2 (2 mM tyramine (Sigma) in 0.1 M Tris·Cl, pH 7.2). Following this, 200 µl of each paired aliquot, per sample, was added to 200 µl of freshly prepared buffer AB3 (20 mM nitrophenyl sulphate (Sigma) in 0.1 M Tris·Cl, pH 7.2) and incubated for 1 h at 37 °C. The reactions were stopped by addition of 0.2 M NaOH (Sigma) and the OD410nm for each sample was measured, the samples in AB1 were used as blanks for paired samples in AB2.

A nitrophenol standard curve was prepared by performing serial 1:2 dilutions using 200 µM p-Nitrophenol (Sigma) in buffer AB1 (0.1 M Tris·Cl, pH 7.2 (Sigma)) and obtaining OD410nm values for each concentration. The standard curve was then used to convert OD410nm values obtained from the nitrophenol assay into µM nitrophenol released. At least 3 independent experiments were carried out to determine arylsulphatase activity for each strain.

Hippicurase assay

To quantify HipO activity, freezer stock strains were streaked onto MH agar plates supplemented with the appropriate antibiotics and grown under microaerophilic conditions for 48 h. Each strain was re-streaked onto 3 plates and then grown under microaerophilic conditions for a further 18 h. Hippicurase activity was measured using a previously described method[32]. Each plate was scraped into 1 ml PBS using an L-shaped spreader, and then diluted with PBS to an OD600nm of 1.0 and pelleted by microcentrifugation for 5 min at 10,000 × g at room temperature. Pellets were washed with 1 ml HIP buffer (2% w/v sodium Hippurate in 100 mM Tris–HCl, pH 8.0) and pelleted as previously described. Pellets were resuspended in 1 ml HIP buffer and incubated, along with a 1 ml cell-free control aliquot of HIP buffer, for 30 min at 37 °C. Suspensions were pelleted and 900 µl supernatant added to 500 µl ninhydrin reagent (3.5% ninhydrin in 1:1 acetone:butanol mixture) and incubated at 95 °C for 20 min. Samples were centrifuged for 1 min at 10,000 × g at room temperature to remove debris, and OD570nm recorded using the no-cell HIP buffer-ninhydrin reagent mixture as a control. If necessary, samples were diluted in 60% ethanol before OD570nm measurement. At least 3 independent experiments were carried out to determine hippicurase activity for each strain.

Motility assays

Freezer stock strains were streaked onto MH agar plates supplemented with the appropriate antibiotics and grown under microaerophilic conditions for 48 h. Each strain was re-streaked onto 3 plates and then grown under microaerophilic conditions for a further 18 h. Motility was measured using previously described methods[25, 65] as follows.

The bacterial growth from each plate was scraped into 1 ml MH broth using an L-shaped spreader, and then diluted with MH broth to an OD600nm of 0.4. Three x 1 μl aliquots of each suspension were stabbed, using a 10 μl pipette tip, into MH broth with 0.4% select agar (Sigma) to provide 3 stabs for each technical replicate, followed by 16 h growth under microaerophilic conditions. The diameter of bacterial growth was measured using a ruler. At least 3 separate biological replicates were performed for each strain tested.

Transmission electron microscopy

C. jejuni strains were grown in liquid culture in MH broth until an OD600nm of 0.4 was reached. 600 μl of each culture was gently pipetted into a 1.5 ml Eppendorf and 600 μl of 8% paraformaldehyde (PFA) (v/v) in PBS (pH 7.2) was added, followed by incubation at room temperature for 45 min. Following this, suspensions were washed twice with PBS, followed by centrifugation for 5 min at 4000 × g, and resuspended in 100 μl PBS. 20 μl suspension was adsorbed onto glow discharged 400 mesh copper grids with a carbon film attached for 30 s. The grids were rinsed twice in dH2O and stained for 30 s with 1% aqueous uranyl acetate (UA). Grids were then viewed using a Tecnai G2 (FEI) transmission electron microscope operated at 200 keV. Around 500–1000 bacterial cells were surveyed across each TEM grid.

Data and statistical analysis

Statistical analysis and graphical presentation were performed using GraphPad Prism (GraphPad Software, San Diego, California USA, www.graphpad.com).

Availability of data and materials

All data generated or analysed during this study are included in this published article [and its supplementary information files]. The datasets used and/or analysed, including TEM images, during the current study are available from the corresponding author on reasonable request.

References

  1. Kirk MD, Pires SM, Black RE, Caipo M, Crump JA, Devleesschauwer B, et al. World Health Organization Estimates of the Global and Regional Disease Burden of 22 Foodborne Bacterial, Protozoal, and Viral Diseases, 2010: A Data Synthesis.von Seidlein L, editor. PLoS Med. 2015;12(12):e1001921.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Mason J, Iturriza-Gomara M, O’Brien SJ, Ngwira BM, Dove W, Maiden MCJ, et al. Campylobacter infection in children in malawi Is common and is frequently associated with enteric virus co-infections. Hold GL, editor. PLoS One. 2013;8(3)e59663.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. de Vries SPW, Linn A, Macleod K, Maccallum A, Hardy SP, Douce G, et al. Analysis of Campylobacter jejuni infection in the gnotobiotic piglet and genome-wide identification of bacterial factors required for infection. Sci Rep. 2017;7:44283.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Kuwabara S, Yuki N. Axonal Guillain-Barré syndrome: Concepts and controversies. Lancet Neurol. 2013;12:1180–8. Elsevier.

    Article  PubMed  Google Scholar 

  5. Scallan Walter EJ, Crim SM, Bruce BB, Griffin PM. Incidence of Campylobacter-associated Guillain-Barré syndrome estimated from health insurance data. Foodborne Pathog Dis. 2020;17(1):23–8.

    Article  PubMed  Google Scholar 

  6. Riddle MS, Gutierrez RL, Verdu EF, Porter CK. The chronic gastrointestinal consequences associated with campylobacter. Curr Gastroenterol Rep. 2012;14(5):395–405.

    Article  PubMed  Google Scholar 

  7. Axelrad JE, Olén O, Askling J, Lebwohl B, Khalili H, Sachs MC, et al. Gastrointestinal Infection Increases Odds of Inflammatory Bowel Disease in a Nationwide Case-Control Study. Clin Gastroenterol Hepatol. 2019;17(7):1311-1322.e7.

    Article  PubMed  Google Scholar 

  8. de Vries SPW, Gupta S, Baig A, Wright E, Wedley A, Jensen AN, et al. Genome-wide fitness analyses of the foodborne pathogen Campylobacter jejuni in in vitro and in vivo models. Sci Rep. 2017;7(1):1251.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Qi LS, Larson MH, Gilbert LA, Doudna JA, Weissman JS, Arkin AP, et al. Repurposing CRISPR as an RNA-guided platform for sequence-specific control of gene expression. Cell. 2013;152(5):1173–83.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Todor H, Silvis MR, Osadnik H, Gross CA. Bacterial CRISPR screens for gene function. Curr Opin Microbiol. 2021;59:102–9. Elsevier Ltd.

    Article  CAS  PubMed  Google Scholar 

  11. Larson MH, Gilbert LA, Wang X, Lim WA, Weissman JS, Qi LS. CRISPR interference (CRISPRi) for sequence-specific control of gene expression. Nat Protoc. 2013;8(11):2180–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Rousset F, Cui L, Siouve E, Becavin C, Depardieu F, Bikard D. Genome-wide CRISPR-dCas9 screens in E. coli identify essential genes and phage host factors 2. bioRxiv. 2018;14(11).

    Google Scholar 

  13. Vigouroux A, Cordier B, Aristov A, Alvarez L, Özbaykal G, Chaze T, et al. Class-A penicillin binding proteins do not contribute to cell shape but repair cellwall defects. Elife. 2020;1:9.

    Google Scholar 

  14. Peters JM, Colavin A, Shi H, Czarny TL, Larson MH, Wong S, et al. A comprehensive, CRISPR-based functional analysis of essential genes in bacteria. Cell. 2016;165(6):1493–506.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Hawkins JS, Silvis MR, Koo BM, Peters JM, Osadnik H, Jost M, et al. Mismatch-CRISPRi Reveals the Co-varying Expression-Fitness Relationships of Essential Genes in Escherichia coli and Bacillus subtilis. Cell Syst. 2020;11(5):523-535.e9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Zhao C, Shu X, Sun B. Construction of a gene knockdown system based on catalytically inactive (“dead”) Cas9 (dCas9) in Staphylococcus aureus. Appl Environ Microbiol. 2017;83(12):e00291-17.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Stamsås GA, Myrbråten IS, Straume D, Salehian Z, Veening JW, Håvarstein LS, et al. CozEa and CozEb play overlapping and essential roles in controlling cell division in Staphylococcus aureus. Mol Microbiol. 2018;109(5):615–32.

    Article  PubMed  Google Scholar 

  18. Sato’o Y, Hisatsune J, Yu L, Sakuma T, Yamamoto T, Sugai M. Tailor-made gene silencing of Staphylococcus aureus clinical isolates by CRISPR interference. PLoS One. 2018;13(1):e0185987.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Rock JM, Hopkins FF, Chavez A, Diallo M, Chase MR, Gerrick ER, et al. Programmable transcriptional repression in mycobacteria using an orthogonal CRISPR interference platform. Nat Microbiol. 2017;2(6):16274.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Choudhary E, Thakur P, Pareek M, Agarwal N. Gene silencing by CRISPR interference in mycobacteria. Nat Commun. 2015;6(1):1–11.

    Article  Google Scholar 

  21. Hecht N, Becher M, Korman M, Vishkautzan M, Gur E. Inter- and intramolecular regulation of protein depupylation in Mycobacterium smegmatis. FEBS J. 2020;26:febs.15245.

    Google Scholar 

  22. Feng S, Liu Y, Liang W, El-Sayed Ahmed MAEG, Zhao Z, Shen C, et al. Involvement of Transcription Elongation Factor GreA in Mycobacterium Viability, Antibiotic Susceptibility, and Intracellular Fitness. Front Microbiol. 2020;23:11.

    Google Scholar 

  23. Caro F, Place NM, Mekalanos JJ. Analysis of lipoprotein transport depletion in Vibrio cholerae using CRISPRi. Proc Natl Acad Sci U S A. 2019;116(34):17013–22.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Friis C, Wassenaar TM, Javed MA, Snipen L, Lagesen K, Hallin PF, et al. Genomic characterization of Campylobacter jejuni strain M1. PLoS One. 2010;5(8).

    Article  PubMed  PubMed Central  Google Scholar 

  25. de Vries SPW, Gupta S, Baig A, L’Heureux J, Pont E, Wolanska DP, et al. Motility defects in Campylobacter jejuni defined gene deletion mutants caused by second-site mutations. Microbiol (United Kingdom). 2015;161(12):2316–27.

    Google Scholar 

  26. Korlath JA, Osterholm MT, Judy LA, Forfang JC, Robinson RA. A Point-Source Outbreak of Campylobacteriosis Associated with Consumption of Raw Milk. J Infect Dis. 1985;152(3):592–6.

    Article  CAS  PubMed  Google Scholar 

  27. Yao R, Guerry P. Molecular cloning and site-specific mutagenesis of a gene involved in arylsulfatase production in Campylobacter jejuni. J Bacteriol. 1996;178(11):3335–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Hendrixson DR, DiRita VJ. Transcription of σ54-dependent but not σ 28-dependent flagellar genes in Campylobacter jejuni is associated with formation of the flagellar secretory apparatus. Mol Microbiol. 2003;50(2):687–702.

    Article  CAS  PubMed  Google Scholar 

  29. Pearson BM, Louwen R, Van Baarlen P, Van Vliet AHM. Differential distribution of type II CRISPR-Cas systems in agricultural and nonagricultural Campylobacter coli and Campylobacter jejuni isolates correlates with lack of shared environments. Genome Biol Evol. 2015;7(9):2663–79.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Cui L, Vigouroux A, Rousset F, Varet H, Khanna V, Bikard D. A CRISPRi screen in E. coli reveals sequence-specific toxicity of dCas9. Nat Commun. 2018;9(1):1912.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Hani EK, Chan VL. Expression and characterization of Campylobacter jejuni benzoylglycine amidohydrolase (hippuricase) gene in Escherichia coli. J Bacteriol. 1995;177(9):2396–402.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Park SF. The use of hipO, encoding benzoylglycine amidohydrolase (hippuricase), as a reporter of gene expression in Campylobacter coli. Lett Appl Microbiol. 1999;28(4):285–90.

    Article  CAS  PubMed  Google Scholar 

  33. Wösten MMSM, Wagenaar JA, Van Putten JPM. The FlgS/FlgR Two-component Signal Transduction System Regulates the fla Regulon in Campylobacter jejuni. J Biol Chem. 2004;279(16):16214–22.

    Article  PubMed  Google Scholar 

  34. Joslin SN, Hendrixson DR. Activation of the Campylobacter jejuni FlgSR two-component system is linked to the flagellar export apparatus. J Bacteriol. 2009;191(8):2656–67.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Lertsethtakarn P. Ottemann KM & Hendrixson DR Motility and chemotaxis in Campylobacter and Helicobacter. Annu Rev Microbiol. 2011;65(1):389–410.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Caldwell MB, Guerry P, Lee EC, Burans JP, Walker RI. Reversible expression of flagella in Campylobacter jejuni. Infect Immun. 1985;50(3):941–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Guerry P. Campylobacter flagella: not just for motility. Trends Microbiol. 2007;15:456–61. Elsevier.

    Article  CAS  PubMed  Google Scholar 

  38. Burnham PM, Hendrixson DR. Campylobacter jejuni: collective components promoting a successful enteric lifestyle. Nat Rev Microbiol. 2018;16:551–65. Nature Publishing Group.

    Article  CAS  PubMed  Google Scholar 

  39. Li X, Ren F, Cai G, Huang P, Chai Q, Gundogdu O, et al. Investigating the Role of FlhF Identifies Novel Interactions With Genes Involved in Flagellar Synthesis in Campylobacter jejuni. Front Microbiol. 2020;24(11):460.

    Article  Google Scholar 

  40. Jagannathan A, Constantinidou C, Penn CW. Roles of rpoN, fliA, and figR in expression of flagella in Campylobacter jejuni. J Bacteriol. 2001;183(9):2937–42.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Logan SM, Harris LA, Trust TJ. Isolation and characterization of Campylobacter flagellins. J Bacteriol. 1987;169(11):5072–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Galkin VE, Yu X, Bielnicki J, Heuser J, Ewing CP, Guerry P, et al. Divergence of quaternary structures among bacterial flagellar filaments. Science. 2008;320(5874):382–5.

    Article  CAS  PubMed  Google Scholar 

  43. Dugar G, Leenay RT, Eisenbart SK, Bischler T, Aul BU, Beisel CL, et al. CRISPR RNA-Dependent Binding and Cleavage of Endogenous RNAs by the Campylobacter jejuni Cas9. Mol Cell. 2018;69(5):893-905.e7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Wösten MMSM, Boeve M, Koot MGA, Van Nuenen AC, Van Der Zeijst BAM. Identification of Campylobacter jejuni promoter sequences. J Bacteriol. 1998;180(3):594–9.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Wang T, Guan C, Guo J, Liu B, Wu Y, Xie Z, et al. Pooled CRISPR interference screening enables genome-scale functional genomics study in bacteria with superior performance-net. Nat Commun. 2018;9(1):2475.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Thyme SB, Akhmetova L, Montague TG, Valen E, Schier AF. Internal guide RNA interactions interfere with Cas9-mediated cleavage. Nat Commun. 2016;7(1):11750.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Esson D, Mather AE, Scanlan E, Gupta S, De Vries SPW, Bailey D, et al. Genomic variations leading to alterations in cell morphology of Campylobacter spp. Sci Rep. 2016;6:38303.

  48. Pascoe B, Williams LK, Calland JK, Meric G, Hitchings MD, Dyer M, et al. Domestication of Campylobacter jejuni NCTC 11168. Microb Genomics. 2019;5(7):e000279.

    Article  Google Scholar 

  49. Frirdich E, Biboy J, Huynh S, Parker CT, Vollmer W, Gaynor EC. Morphology heterogeneity within a Campylobacter jejuni helical population: the use of calcofluor white to generate rod-shaped C. jejuni 81–176 clones and the genetic determinants responsible for differences in morphology within 11168 strains. Mol Microbiol. 2017;104(6):948–71.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Sher AA, Jerome JP, Bell JA, Yu J, Kim HY, Barrick JE, et al. Experimental Evolution of Campylobacter jejuni Leads to Loss of Motility, rpoN (σ54) Deletion and Genome Reduction. Front Microbiol. 2020;6(11):2781.

    Google Scholar 

  51. Depardieu F, Bikard D. Gene silencing with CRISPRi in bacteria and optimization of dCas9 expression levels. Methods. 2020;1(172):61–75.

    Article  Google Scholar 

  52. Wang T, Wang M, Zhang Q, Cao S, Li X, Qi Z, et al. Reversible gene expression control in Yersinia pestis by using an optimized CRISPR interference system. Appl Environ Microbiol. 2019;85(12):e00097-19.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Zhang R, Xu W, Shao S, Wang Q. Gene Silencing Through CRISPR Interference in Bacteria: Current Advances and Future Prospects. Front Microbiol. 2021;31:12.

    Google Scholar 

  54. Hawkins JS, Wong S, Peters JM, Almeida R, Qi LS. Targeted transcriptional repression in bacteria using CRISPR interference (CRISPRi). Methods Mol Biol. 2015;1311:349–62.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Yamada M, Watanabe Y, Gootenberg JS, Yamada M, Watanabe Y, Gootenberg JS, et al. Crystal Structure of the Minimal Cas9 from Campylobacter jejuni Reveals the Molecular Diversity in the CRISPR-Cas9 Systems Article Crystal Structure of the Minimal Cas9 from Campylobacter jejuni Reveals the Molecular Diversity in the CRISPR-Cas9 Systems. Mol Cell. 2017;65(6):1109-1121.e3.

    Article  CAS  PubMed  Google Scholar 

  56. Silvis MR, Rajendram M, Shi H, Osadnik H, Gray AN, Cesar S, et al. Morphological and transcriptional responses to CRISPRi knockdown of essential genes in Escherichia coli. MBio. 2021;12(5).

    Article  PubMed  Google Scholar 

  57. Wassenaar TM, Bleumink-Pluym NMC, Van Der Zeijst BAM. Inactivation of Campylobacter jejuni flagellin genes by homologous recombination demonstrates that flaA but not flaB is required for invasion. EMBO J. 1991;10(8):2055–61.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Dugar G, Herbig A, Förstner KU, Heidrich N, Reinhardt R, Nieselt K, et al. High-Resolution Transcriptome Maps Reveal Strain-Specific Regulatory Features of Multiple Campylobacter jejuni Isolates. PLoS Genet. 2013;9(5):e1003495.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Holt JP, Grant AJ, Coward C, Maskell DJ, Quinlan JJ. Identification of Cj1051c as a Major determinant for the restriction barrier of Campylobacter jejuni Strain NCTC11168. Appl Environ Microbiol. 2012;78(22):7841–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Van Vliet AHM, Wooldridge KG, Ketley JM. Iron-responsive gene regulation in a Campylobacter jejuni fur mutant. J Bacteriol. 1998;180(20):5291–8.

    Article  PubMed  PubMed Central  Google Scholar 

  61. Beauchamp JM, Leveque RM, Dawid S, DiRita VJ. Methylation-dependent DNA discrimination in natural transformation of Campylobacter jejuni. Proc Natl Acad Sci U S A. 2017;114(38):E8053-61.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Wilhelm J, Pingoud A. Real-time polymerase chain reaction. Chembiochem. 2003;4(11):1120–8.

    Article  CAS  PubMed  Google Scholar 

  63. Miller JH. Experiments in molecular genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1972.

    Google Scholar 

  64. Davis L, Young K, DiRita V. Genetic manipulation of Campylobacter jejuni. Curr Protoc Microbiol. 2008;10:8A.2.1–8A.2.17.

  65. Cohen EJ, Nakane D, Kabata Y, Hendrixson DR, Nishizaka T, Beeby M. Campylobacter jejuni motility integrates specialized cell shape, flagellar filament, and motor, to coordinate action of its opposed flagella. PLoS Pathog. 2020;16(7).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

The authors thank Dr. Karin Mueller and Dr. Filomena Gallo of the Cambridge Advanced Imaging Centre for their support and assistance with TEM imaging.

Funding

This work was funded by the Wellcome Trust (funder number: 102163/B/13/Z) through a PhD training studentship awarded to RC. ES was funded by BBSRC Institute Strategic Programme Microbes in the Food Chain [BB/R012504/1 and its constituent project BBS/E/F000PR10349 (Theme 2, Microbial Survival in the Food Chain)].

Author information

Authors and Affiliations

Authors

Contributions

RC, ES, RAF, JP and AJG conceptualised the study. RC, ES and AJG designed the methodology. RC carried out the investigation. AJG supervised the study. RC and AJG prepared and wrote the original draft. RC, ES, RAF, JP and AJG reviewed and edited the draft. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Andrew J. Grant.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Costigan, R., Stoakes, E., Floto, R.A. et al. Development and validation of a CRISPR interference system for gene regulation in Campylobacter jejuni. BMC Microbiol 22, 238 (2022). https://doi.org/10.1186/s12866-022-02645-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12866-022-02645-4

Keywords