Survival of Campylobacter jejuni 11168H in Acanthamoebae castellanii Provides Mechanistic Insight into Host Pathogen Interactions

Abstract

Campylobacter jejuni is the leading cause of bacterial foodborne gastroenteritis worldwide but is rarely transferred between human hosts. Although a recognized microaerophile, the majority of C. jejuni are incapable of growing in an aerobic environment. The persistence and transmission of this pathogen outside its warm-blooded avian and mammalian hosts is poorly understood. Acanthamoebae species are predatory protists and form an important ecological niche with several bacterial species. Here, we investigate the interaction of C. jejuni 11168H and Acanthamoebae castellanii at the single-cell level. We observe that a subpopulation of C. jejuni cells can resist killing by A. castellanii, and non-digested bacteria are exocytosed into the environment where they can persist. In addition, we observe that A. castellanii can harbor C. jejuni 11168H even upon encystment. Transcriptome analyses of C. jejuni interactions revealed similar survival mechanisms when infecting both A. castellanii and warm-blooded hosts. In particular, nitrosative stress defense mechanisms and flagellum function are important as confirmed by mutational analyses of C. jejuni 11168H. This study describes a new host–pathogen interaction for C. jejuni and confirms that amoebae are transient hosts for the persistence, adaptability, and potential transmission of C. jejuni.

1. Introduction

Campylobacter jejuni is the most prevalent gastrointestinal bacterial pathogen worldwide but is rarely transferred between human hosts, suggesting that its high frequency in infection must relate to survival outside its warm-blooded hosts [1]. It is still ambiguous how this bacterium, an obligate microaerophile, remains omnipresent in the environment, given its inability to grow under atmospheric conditions [1,2]. Although not all ‘environments’ outside of the host may necessarily be aerobic, C. jejuni has been shown to form biofilms in atmospheric conditions as a mode of persistence [3,4] and has also been recognized to transform into a viable but non-culturable (VBNC) state [5,6]. This allows the bacterium to withstand many stresses and remain viable with low metabolic activity for months [7]. Additionally, several studies have investigated C. jejuni’s environmental survival outside of the host with free-living protists [8,9,10,11,12]; however, the mechanistic understanding of this important niche remains poorly understood.

Free-living bacteria populations face constant challenges, particularly relating to predation by protists such as Acanthamoebae species [13,14,15]. Some bacteria have developed mechanisms to defend against killing within amoebic cells [14]. Acanthamoebae spp. have also been shown to have an important role in harboring and dissemination of pathogenic bacteria [13,14,16]. These interactions have been linked to several disease outbreaks of contaminated water [16] and food sources, including livestock [17]. Acanthamoebae spp. are characterized by a biphasic life cycle; a ‘trophozoite’ metabolically active form and a ‘cyst’ state, which is a non-dividing dormant form [18]. Encystment occurs from the trophozoite form, with changes to the cell morphology including cell rounding and thicker cell wall [19]. The thickened cell wall enhances resistance to disinfection treatments and environmental stresses, which has been shown to be a reservoir for several intracellular bacteria pathogens [8,13,16,17].

Acanthamoebae spp. are widely used in research as non-mammalian surrogate organisms to study host interactions [20], including C. jejuni [11,12,21]. Studies have primarily focused on C. jejuni–amoebal interaction population measurements, which enumerate surviving bacteria by colony-forming units, but there is little information on C. jejuni–amoebae interactions at the single-cell level. Some studies have reported on the ability of C. jejuni to be phagocytosed by amoebae and thrive within this niche, where they multiply and “burst” out, leading to the release of C. jejuni into the environment [11]. This observation is seen in many other bacteria pathogens [14]. However, in our previous study we reported uptake of C. jejuni by Acanthamoebae spp. but not proliferation [12]. Instead, we observed that Acanthamoebae spp. are a transient host for C. jejuni, and survival of this bacterium within amoebae enhances its subsequent invasion and survivability in human epithelial cells and amoebae cells [12].

Here, we scrutinize the interaction of C. jejuni and amoebae using single-cell imaging complemented with intra-amoebal C. jejuni transcriptome profiling. We investigate the interaction of C. jejuni strain 11168H with A. castellanii, a keratitis strain from the T4 genotype, on the single-cell level by live-cell imaging using confocal microscopy. We show that sub populations of C. jejuni cells can (i) resist killing by A. castellanii; (ii) are released from A. castellanii through exocytosis; and (iii) contrary to previous reports [22], C. jejuni can maintain this niche upon amoebal encystment. C. jejuni genome is small (∼1.6 Mb), relative to other enteric pathogens [23,24]; therefore, we hypothesized that C. jejuni would use similar strategies to survive internalization by amoebae as it would with its warm-blooded hosts. Using RNA-Seq data and mutagenesis we show that C. jejuni 11168H indeed utilizes similar strategies to survive and persist within amoebae as it does with its warm-blooded hosts. Our data reinforce the observation that A. castellanii are an important transient host for C. jejuni 11168H and may enhance environmental persistence of this prevalent pathogen.

2. Materials and Methods

2.1. Strains and Cultures

Campylobacter jejuni strain 11168H, a motile derivative of the original sequence strain NCTC 11168 [25], were stored using Protect bacterial preservers (Technical Service Consultants, England, UK) at −80 °C. Bacteria were streaked on Columbia blood agar (CBA) plates containing Columbia agar base (Oxoid) supplemented with 7% (v/v) horse blood (TCS Microbiology, UK) and grown at 37 °C in a microaerobic chamber (Don Whitley Scientific, England, UK), containing 85% N₂, 10% CO₂, and 5% O₂ for 48 h. Bacteria were grown on CBA plates for a further 16 h at 37 °C prior to use.

Acanthamoeba castellanii (T4 genotype) Culture Collection of Algae and Protozoa (CCAP) 1501/10 (Scottish Marine Institute) were grown to confluence at 25 °C in 75-cm² tissue culture flasks containing peptone yeast and glucose (PYG) media (proteose peptone 20 g, glucose 18 g, yeast extract 2 g, sodium citrate dihydrate 1 g, MgSO₄ × 7H₂O 0.98 g, Na₂HPO₄ × 7H₂O 0.355 g, KH₂PO₄ 0.34 g in distilled water to make 1000 mL, pH was adjusted to 6.5). Amoebae were harvested by scraping the cells into suspension, and viability was determined by staining with trypan blue and counting by a hemocytometer using inverted light microscopy.

2.2. C. jejuni Mutant Transformation

C. jejuni 11168H mutant constructs were obtained from the Campylobacter Resource Facility (http://crf.lshtm.ac.uk/wren_mutants.htm, accessed on 1 August 2022) and strain 11168H was naturally transformed using the previously described biphasic method [26] with some minor changes. Briefly, C. jejuni was cultured on Mueller Hinton (MH) agar for 16 h, cells were resuspended in MH broth to an OD_600nm = 0.5, 0.5 mL of this suspension was added to a 15 mL falcon that contained 1 mL of MH agar, C. jejuni donor DNA 0.1–1 µg was added to the suspension and incubated for 24 h at 37 °C in microaerobic conditions, and the suspension was plated on MH agar with respective antibiotic for up to 5 days. C. jejuni green fluorescent producing bacteria (GFP) strain 11168H_GFP was constructed as previously described [12,27]. C. jejuni constructs 11168HΔcgb, 1168HΔdsbA, and 1168HΔdsbB were a gift from Dr. Aidan. J. Taylor (University of Sheffield). All strains used in the study can be found in Supplementary File S1 (Table S2).

2.3. C. jejuni Invasion and Survival Assay

Invasion and survival assays were performed as previously described [12,28] with minor alterations. Briefly, C. jejuni 11168H and its mutants were incubated with a monolayer of approximately 10⁶ A. castellanii at a multiplicity of infection (M.O.I) of 200:1 for 3 h at 25 °C in 2 mL PYG media (the M.O.I. of 200:1 was chosen to allow maximal internalization of bacteria without impacting amoebae predation through density-dependent inhibition). The monolayer was washed 3× with phosphate buffer saline solution (PBS) and incubated for 1 h in 2 mL of PYG media containing 100 μg/mL of gentamicin. C. jejuni cells were harvested by scraping the amoebae into suspension and centrifuged for 10 min at 350× g to pellet the bacteria and amoebae. Supernatant was discarded and the pellet was suspended in 2 mL of distilled water containing 0.1% (v/v) Triton X-100 for 10 min at room temperature with vigorous pipetting (every 2 min) to lyse the amoebae and release bacteria cells. The suspension was then centrifuged for a further 10 min at 4000× g, the resultant pellet was resuspended in 1 mL PBS and enumerated for colony forming units on CBA plates for up to 72 h at 37 °C microaerobically.

2.4. Live-Cell Imaging

Green fluorescent protein (GFP) producing C. jejuni 11168H_GFP strain was used for confocal live-cell imaging. Time lapse experiments were performed as described above with minor changes. A. castellanii cells were incubated in PYG media supplemented with dextran-conjugated Texas Red (10,000 MW; Thermofischer) at a final concentration of 100 μg/mL overnight. A. castellanii cells were washed three times and approximately 10⁶ of adherent amoebae were infected with C. jejuni 11168H_GFP (M.O.I of 200) in 35 mm μ-Dish devices (IBIDI) prior to imaging. Texas Red-conjugated dextran was monitored in the red region (excitation and emission wavelengths are 595/615 nm) and GFP was monitored in the green region (excitation and emission wavelengths are 488/510 nm).

Confocal laser scanning microscopy images were obtained using an inverted Zeiss LSM 880 confocal microscope (Zeiss, Berlin, Germany). Images were taken at 5 s intervals for time lapse experiments.

2.5. RNA-Seq and Real-Time RT-qPCR

RNA-Seq was used to identify differentially expressed genes between intra-amoebal C. jejuni strain 11168H and the control (which was incubated in the absence of A. castellanii) after a total of 4 h. Experiments were performed as described above; briefly, A. castellanii 10⁶ were infected co-incubated with C. jejuni 11168H (M.O.I of 200) for 3 h and washed three times to remove extracellular bacteria. Gentamycin was added to a final concentration of 100 μg/mL; this culture was incubated for a further 1 h before washing three times and RNA was extracted using Triazole (Sigma Aldrich, England, UK) following manufacturer’s protocol. Ribosomal RNA was depleted using Ribominus (Invitrogen, Scotland, UK), and libraries were prepared using TruSeq^® Stranded mRNA (Illumina, Carlifonia, US). Raw reads were obtained from an Illumina MiSeq paired-end sequencing platform (Illumina). The paired-end reads were trimmed and filtered using Sickle v1.200 [29]; Bowtie2 [30] was used to map the reads against the reference sequence, C. jejuni strains 11168H assembly GCA_900117385.1. Cufflinks suite [31] was used to convert annotations from GFF to GTF format, and Bedtools [32] was used to generate transcript counts per samples. Statistical analysis was performed in R using the combined data generated from the bioinformatics as well as metadata associated with the study (multifactorial design). Adjusted p-value significance cut-off of 0.001, and log fold change cut-off of >1.5, was used for multiple comparison.

Expression of genes of interest were quantified by real-time RT-qPCR and normalized against gyrA. A 1 μg of total RNA of each sample was reverse-transcribed to cDNA using RT² first strand kit (Qiagen) according to manufacturer’s protocol. Quantification of gene expression was achieved by real-time RT-qPCR using Sybr^TM Green real-time PCR master mix using primers generated using PrimerQuest (IDT) (Supplementary File S1 (Table S3)). Real-time RT-PCR was performed in 96-well plates using an ABI PRISM 7300 Real-time PCR System (Applied Biosystems), and the relative gene expression for the different genes was calculated from the crossing threshold (Ct) value according to the manufacturer’s protocol (2^−ΔΔCt) after normalization using the gyrA endogenous control (68).

2.6. Statistical Analysis

Evaluation of C. jejuni interaction with A. castellanii was based on at least 500 amoebae cells per experiment. All experiments presented are at least three independent biological replicates. RNA-Seq analysis were performed in R using the combined data generated from the bioinformatics. Differentially expressed genes were considered significant when the p-value of three independent biological experiments was below 0.001. All other data were analyzed using Prism statistical software (Version 9, GraphPad Software, San Diego, CA, USA), and statistical significance was considered when the p-value was <0.05. All values are presented as standard deviation of at least three independent experiments.

3. Results

3.1. Undigested C. jejuni Are Released Back into the Environment

As a predator, free-living amoebae inhabits aquatic environments, a niche often occupied by C. jejuni [9,19,20]. It is plausible that these two organisms have frequent encounters and have co-evolved. Using green fluorescent protein (GFP) producing C. jejuni strain 11168H (1168H_GFP) and indigestible-conjugated dextran to track the phago–endosomal pathway time-lapse imaging (0–3.5 h post infection (h.p.i)), we studied the interaction of C. jejuni and A. castellanii at the single-cell level. We observed internalization, trafficking of bacteria into digestive vacuoles, and the persistence of undigested C. jejuni (Figure 1).

To identify the compartment containing intracellular C. jejuni, we followed the time-course of the phago–endosomal pathway by incubating amoebae for 16 h with indigestible fluorophore conjugated to a complex branched glucan, dextran-Texas Red, and tracked localization. In line with a previous study [8], we observed aggregation of C. jejuni 11168H at a certain position on A. castellanii prior to entering the amoebae. This aggregation phenotype was termed “backpack” formation [33] (Figure 1a); we also observed formation of small bacteria-filled vacuoles (Figure 1b), which ultimately formed larger compartments that colocalized with dextran; this indicated that intracellular C. jejuni are trafficked into “digestive-vacuoles” (Figure 1c,d).

Whilst most of the C. jejuni were lysed within these compartments, a fraction of the bacteria survived. This was based on retained integrity of the classical helical-shape morphology of C. jejuni. Next, we sought to observe the fate of these intact C. jejuni cells within these compartments (Figure 2).

A. castellanii were incubated with C. jejuni 11168H_GFP for 3.5 h and monitored over a period of 80 s. Live-cell imaging revealed that these undigested bacteria are exocytosed back into the environment (Figure 2 and Supplementary Movie S1). Similar to other phagocytes [34,35], A. castellanii accumulated cellular waste, such as digested bacteria, and the dextran are excreted out of the cell, thus also releasing undigested bacteria back into the environment. We did not observe this phenotype using a non-pathogenic E. coli strain DH5α tagged with the fluorophore mCherry (Supplementary File, Figure S2).

3.2. C. jejuni Maintains a Niche within Amoebic Cyst

Contrary to previous reports [11,21,28], we did not observe intracellular multiplication of C. jejuni 11168H and lysis of amoebae, but we did observe C. jejuni 11168H_GFP within the cyst form of A. castellanii (Figure 3).

C. jejuni 11168H_GFP was incubated with A. castellanii for 72 h at 25 °C in PYG media before imaging with confocal microscopy. Out of 500 amoebic cysts that were enumerated in each biological replicate, an average of ∼30% contained GFP-producing bacteria (Figure 3a,b). Previously, it was reported that intracellular survival of C. jejuni undergoes drastic physiological changes, and recovery required pre-culturing under oxygen-limiting conditions (anaerobic) [36]; however, we were unable to recover viable colony forming units (cfu) from this culture. This suggested that C. jejuni 11168H may undergo metabolic reprogramming within A. castellanii cysts.

3.3. Intra-Amoebal C. jejuni Transcriptome

To identify the key regulatory changes triggered by survival of C. jejuni within A. castellanii, RNA-Seq was used to determine differentially expressed genes in strain 11168h intra-amoebae relative to the control at 4 h post infection. There was a large difference in gene expression between intra-amoebae C. jejuni and the control; a total of 94 genes were differentially transcribed (>1.5-fold; p-value < 0.001): 72 were up-regulated whilst 22 were down-regulated in the intra-amoebae C. jejuni (

Table 1. Significantly differentially expressed genes in intra-amoebae C. jejuni compared to the control in strain 11168H after 4 h post infection.

Upregulated
Gene ID	Gene Name	Description of Product	Log2 Fold Change	Adjusted p-Value
Cj1586	cgb	Single domain hemoglobin	4.663339	1.88 × 10−26
Cj0974		Uncharacterized protein	4.417067	6.57× 10−5
Cj1340c	Maf1	Motility accessory protein	4.349186	6.28 × 10−34
Cj0380c		Uncharacterized protein	4.241269	1.77 × 10−4
Cj0972		Uncharacterized protein	4.14400	1.08 × 10−5
Cj0758	grpE	Response to hyperosmotic and heat shock	3.963416	1.38 × 10−20
Cj0971		Uncharacterized protein	3.459789	2.23 × 10−4
Cj0241c	herA	Bacteriohemerythrin	3.326818	1.75 × 10−2
Cj0877c		Uncharacterized protein	3.258874	3.88 × 10−3
Cj0200c		Periplasmic protein	3.249386	6.69 × 10−12
Cj1650	fliS	Flagellar export chaperone [37]	3.201313	3.33 × 10−19
Cj0836	ogt	Methyltransferase	2.961825	2.21 × 10−13
Cj1242	ciaC	Campylobacter Invasion antigen C	2.946553	5.82 × 10−38
Cj0201c	imB	Integral membrane protein	2.902751	6.00 × 10−3
Cj0921c	peb1A	Amino acid transporter	2.902712	2.54 × 10−24
Cj0865	dsbB	Protein-disulfide oxidoreductase	2.727325	9.52 × 10−11
Cj0030		Uncharacterized protein	2.711549	7.47 × 10−25
Cj1465	flgN	Flagellar hook protein [38]	2.704280	5.90 × 10−12
Cj0251c		Highly acidic protein	2.676403	1.56 × 10−5
Cj0829c		Uncharacterized protein (Putative CoA binding domain containing protein)	2.664411	1.76 × 10−5
Cj1503c	putA	Putative proline dehydrogenase/delta−1-pyrroline-5-carboxylate dehydrogenase	2.646897	9.64 × 10−54
Cj1537c	acsA	Acetyl-coenzyme A synthetase	2.643273	9.49 × 10−16
Cj0243c		Uncharacterized protein	2.508166	5.98 × 10−6
Cj1450	ciaI	Putative ATP/GTP-binding protein	2.461822	3.79 × 10−17
Cj0919c		Aspartate/glutamate/glutamine transport system permease protein	2.401703	7.08 × 10−10
Cj0459c		Uncharacterized protein	2.291369	5.87 × 10−13
Cj1075	fliW	Flagellar assembly factor	2.260313	3.55 × 10−8
Cj0018c	dba	Disulfide bond formation protein	2.260119	1.52 × 10−6
Cj1668c		Periplasmic protein	2.168892	2.44 × 10−10
Cj0967		Periplasmic protein	2.103994	1.08 × 10−4
Cj1682c	gltA	Citrate synthase	2.097224	1.27 × 10−20
Cj0920c	hisM	Putative ABC-type amino-acid transporter permease protein	2.09015	9.49 × 10−16
Cj0864		Putative periplasmic protein-thioredoxin-like protein	2.063197	1.12 × 10−3
Cj0728		Putative periplasmic protein	2.052897	3.32 × 10−4
Cj0917c	cstA	Carbon starvation protein A	2.017821	6.06 × 10−21
Cj0979c		Putative secreted nuclease	1.958595	5.65 × 10−8
Cj0156c		Ribosomal RNA small subunit methyltransferase E	1.919746	1.64 × 10−5
Cj0528c	flgB	Flagellar basal body rod protein	1.882962	1.65 × 10−23
Cj1107	clpS	ATP-dependent Clp protease adapter protein	1.869786	1.03 × 10−6
Cj0429c		Uncharacterized protein	1.863965	2.77 × 10−12
Cj1464	flgM	Flagellar biosynthesis protein	1.844621	7.16 × 10−11
Cj0580c	hemN	Heme chaperone	1.842605	8.88 × 10−3
Cj0922c	PEB1C	Aspartate/glutamate/glutamine transport system ATP-binding protein	1.836333	8.68 × 10−21
Cj0949c		peptidyl-arginine deiminase; involved in Arginine and proline metabolism	1.829841	6.17 × 10−7
Cj1495c		Uncharacterized protein	1.805563	5.34 × 10−24
Cj0959c		Membrane protein insertion efficiency factor	1.783311	6.27 × 10−3
Cj0872	dsbA	Thiol:disulfide interchange protein	1.769298	2.13 × 10−4
Cj0898		Histidine triad (HIT) family protein	1.748533	1.21 × 10−7
Cj1563c		Putative transcriptional regulator	1.743043	1.47 × 10−3
Cj1462	flgI	Flagellar P-ring protein	1.741476	2.47 × 10−11
Cj1681c	cysQ	CysQ protein-amino acid metabolism	1.732868	7.20 × 10−4
Cj0573		GatB/YqeY family protein	1.705464	5.93 × 10−5
Cj0175c	cfbpA	Iron-uptake ABC transporter substrate-binding protein	1.694355	2.13 × 10−3
Cj0420		Periplasmic protein	1.688196	4.43 × 10−8
Cj0509c	clpB	Stress-induced multi-chaperone system	1.674056	1.11 × 10−10
Cj1463	flgI	Flagellar biosynthesis protein	1.658851	2.68 × 10−6
Cj0540		Exporting protein	1.647181	1.42 × 10−7
Cj1025c		Uncharacterized protein	1.631584	2.18 × 10−7
Cj1380	dsbC	Periplasm bi-functional thiol oxidoreductase	1.631086	8.26 × 10−16
Cj1383c		Uncharacterized protein	1.628912	1.77 × 10−4
Cj1382c	fldA	Flavodoxin I	1.599526	8.35 × 10−13
Cj0428		Uncharacterized protein	1.589477	1.56 × 10−22
Cj0076c	lctP	L-lactate transporter permease	1.57782	1.97 × 10−9
Cj0398	gatC	Glutamyl-tRNA (Gln) aminotransferase subunit C	1.564965	1.01 × 10−4
Cj0465c	ctb	Group 3 truncated hemoglobin	1.547660	1.76 × 10−5
Cj1338c	flaB	Flagellin B	1.543761	3.16 × 10−11
Cj1199		Putative iron/ascorbate-dependent oxidoreductase	1.535012	1.60 × 10−10
Cj1189c	cetB	Bipartate energy taxis response protein	1.529662	1.76 × 10−5
Cj1094c	yajC	Preprotein translocase accessory complex subunit	1.523906	3.14 × 10−5
Cj0547	flaG	Flagellar protein	1.52304	4.91 × 10−4
Cj1670c	cgpA	N-acetylgalactosamine (GalNAc)-containing glycoprotein	1.518031	2.82 × 10−3
Cj1106		Thioredoxin-Single domain hemoglobin	1.500635	4.56 × 10−9
Downregulated
Cj0620		Uncharacterized protein	−4.25725	2.28 × 10−3
Cj0637c	mrsA	Catalyzes the reversible oxidation-reduction of methionine sulfoxide to methionine	−3.94117	4.67 × 10−5
Cj1276c		Integral membrane protein	−3.78944	8.12 × 10−3
Cj1282	mrdB	Rod shape-determining protein	−3.76646	8.31 × 10−3
Cj1493c		Integral membrane protein	−3.41854	5.51 × 10−3
Cj0648		Membrane protein	−3.30497	7.05 × 10−3
Cj1080c	hemD	Uroporphyrinogen-III synthase	−3.29741	6.57 × 10−3
Cj1132c		Uncharacterized protein	−2.68281	6.86 × 10−6
Cj1448c	kpsM	Capsule polysaccharide export system inner membrane	−2.53961	1.35 × 10−6
Cj1423c	hddC	D-glycero-alpha-D-manno-heptose 1-phosphate guanylyl transferase	−2.17572	7.22 × 10−4
Cj0267c		Integral membrane protein	−2.08538	2.86 × 10−6
Cj0773c		Methionine transport system permease protein	−2.0194	9.28 × 10−11
Cj0644		TatD-related deoxyribonuclease	−1.9908	9.34 × 10−5
Cj1567c	nuoM	NADH dehydrogenase I chain M	−1.9148	5.20 × 10−5
Cj0944c		Periplasmic protein	−1.78725	2.43 × 10−3
Cj0935c		Putative sodium:amino-acid symporter family protein	−1.77122	2.23 × 10−4
Cj1264c	hydD	Putative hydrogenase maturation protease	−1.74266	1.73 × 10−6
Cj0357c	plsY	Acyl phosphate: glycerol−3-phosphate acyltransferase	−1.69479	3.97 × 10−3
Cj0411		ATP/GTP binding protein	−1.67294	1.76 × 10−5
Cj0717		ArsC family protein	−1.64102	5.36 × 10−5
Cj1388		Endoribonuclease L-PSP	−1.60661	2.64 × 10−3
Cj1205c	radA	DNA repair and recombination	−1.51410	9.99 × 10−8

Log² Fold change cut-off = >1.5. Adjusted p-value cut-off = <0.001. Genes highlighted in bold were chosen for RT-PCR. Annotations for Table 1 were based on NCBI reference sequence annotations (https://www.ncbi.nlm.nih.gov/refseq/, accessed on 1 August 2022). References indicate where annotations were obtained. For the complete dataset, see Supplementary Table S1.

).

To verify our RNA-Seq results, RNA was extracted independently of the RNA-Seq experiments for real-time RT-PCR. The first genes were cgb which encode a single-domain hemoglobin and ctb which encode a truncated hemoglobin from group III of globins; both have a major role in mediating protection of C. jejuni against nitric oxide and nitrosative stress [39]. dsbA and dsbB both encode periplasmic thiol oxidoreductases that repair disulfide bond in proteins that have been damaged under stress [40,41,42]; cstA, encode carbon starvation protein A; clpB encode an ATP-dependent protease, and grpE encode a heat-shock protein; all of which are involved in C. jejuni stress response and niche adaptation [43,44]. We also determined the expression of peb1A, which encodes an amino-acid binding protein in addition to facilitating cell adhesion [45], and the expression of Cj0971, which encodes an uncharacterized protein. A BLAST (https://blast.ncbi.nlm.nih.gov/) accessed on 1 August 2022) search of Cj0971 protein sequence revealed the product of Cj0971 as a filamentous hemagglutinin; this product mediates adhesion and invasion into host cells in other bacteria [46].

We also determined the expression of ciaC and ciaI encoding Campylobacter invasion antigens, both of which are essential for invasion and survival within host cells in vitro [47,48]. kpsM encodes a transport permease protein that is important for transportation of the capsular polysaccharide across the inner membrane, and its mutation leads to non-encapsulated C. jejuni [49]; mrsA encodes a methionine sulfoxide reductase, involved in oxidative stress repair of proteins containing methionine [50]; mrdB encodes a rod determining protein (RodA). Mutation to rodA leads to loss of rod shape in bacilli bacteria [51], and Cj1276c encodes an uncharacterized protein.

cgb and ctb had significantly increased transcripts in the RNA-Seq results; real-time RT-PCR indicated a significant (p < 0.05) ∼4.12-fold and a non-statistically significant ∼1.95-fold increase in expression, respectively. The two thiol oxidoreductases (dsbA and dsbB) also showed significant increase in expression in the RT-PCR experiments (∼2.74-fold and ∼2.25-fold, respectively) in line with the RNA-Seq results. cstA, clpB, and grpE were all significantly transcribed in the RNA-Seq results, and RT-PCR showed a significant increased expression of cstA (∼2.74-fold), grpE (∼2.39-fold), and a non-statistically significant increase in expression of clpB (∼1.94-fold). In line with our RNA-Seq results, peb1A (∼2.79-fold), ciaC (∼3.14-fold), ciaI (∼2.10-fold), and Cj0971(∼2.29-fold) were all significantly expressed. kpsM (∼−2.58-fold), mrdB (∼−2.55-fold), and Cj1276c (∼2.44-fold) were all significantly downregulated in our RT-PCR results, whilst mrsA (∼−1.41-fold) had a non-statistically significant reduction; this was reflective of our RNA-Seq results (Figure 4a).

Our RNA-Seq data revealed genes that were previously identified to be important during interactions of C. jejuni with its warm-blooded hosts but also genes that may be required during C. jejuni survival within amoebae host. Therefore, we generated 13 mutants in C. jejuni strain 11168H (Supplementary File S1: Table S2) that showed differential expression from our newly generated RNA-Seq data. We then tested their roles in survival within A. castellanii by enumerating CFU at 4 h post infection (Figure 4b). We chose to test cgb, clpB, cstA, flgB, dsbA, dsbB, dsbC, flaA (although this gene did not make the >1.5-fold cutoff, it was upregulated in our RNA-seq by ∼1.46-fold), flaB, kpsM, mrdB, Cj0420, and Cj0979c mutants.

All mutants tested showed reduced intracellular survivability within A. castellanii relative to the 11168H parent strain. Our analyses indicated that survival within A. castellanii is multifactorial; however, it is likely that some factors may favor survival within amoebae more than C. jejuni warm-blooded hosts. Interestingly, mutation to cgb, flgB (Flagellar basal body rod protein), flaA (major flagellin), and flaB (minor flagellin) showed no survival within A. castellanii. To test whether this reduction was influenced by amoebal uptake, we enumerated bacteria prior to gentamycin treatment (Supplementary File S1; Figure S1b). Our results show that whilst 11168H∆cgb uptake by amoebae was similar to the WT levels, mutation to flgB, flaA, and flaB showed reduced uptake by A. castellanii, revealing the importance of an intact flagellar during amoebae interaction.

4. Discussion

Campylobacter jejuni is the leading cause of foodborne gastroenteritis [1,2]. Understanding its interactions with ubiquitous free-living amoeba in the environment is important and this may shed light on pre-adaptation to survival and virulence mechanisms required when C. jejuni encounters warm-blooded avian or mammalian hosts.

In this study, we investigated the interactions of C. jejuni 11168H with A. castellanii at the single-cell level. We present evidence of previously unreported interactions and propose part of the C. jejuni’s evolution and life-cycle is cohabitation with free-living protist (Figure 5). C. jejuni enters the trophozoites phago–endosomal pathway through phagocytosis; a fraction of the bacteria resists amoebal digestion and they are exocytosed. Previously, we reported that C. jejuni 11168H was more invasive of mammalian cells and more resistant to killing subsequent to transient internalization by Acanthamoebae spp. Here, our observation reveals that these more persistent bacteria are also capable of withstanding amoebal encystation.

Legionella pneumophila and Vibrio cholerae are known to multiply within Acanthamoebae spp. and even survive long-term within amoebic cysts [52,53]. Additionally, V. cholerae was shown to multiply within A. castellanii cysts, which ultimately leads to the destruction of the amoebic cell [53]. Some studies have reported multiplication of C. jejuni within amoebae [11,54] and destruction of amoebic cells through lysis [8]. We did not observe multiplication of C. jejuni 11168H in our model; however, we observed that long-term co-incubation of C. jejuni 11168H with amoebae led to transformation of amoebal trophozoites to cysts, which contained GFP producing C. jejuni. To our knowledge, this is the first study that shows evidence of C. jejuni within amoebae cysts. The cystic stage of amoeba is known to remain dormant for years and withstand diverse environmental stresses [19,55]. It is plausible that amoebic cysts may offer C. jejuni extra protection from environmental elements, this particular interaction has been reported in L. pneumophila [56]. We speculate that, following favorable conditions such as transmission to a host (avian or mammalian), this “Trojan horse” would transform into trophozoites potentially leading to the release of its bacteria cargo. However, at this point, it is unclear whether encystment is actively induced by C. jejuni. Additionally, we do not exclude that some other C. jejuni strains may behave differently during interactions with Acanthamoebae spp.; indeed, previously we showed interactions of C. jejuni with amoebae was not only strain dependent but also host cell dependent [12].

The transient internalization of C. jejuni in amoebae has the potential to extend our understanding of how this important pathogen interacts with its natural avian and mammalian hosts. Our transcriptome data of intra-amoebal bacteria suggested that C. jejuni utilizes similar factors to interact and survive within amoebae as it does with its natural hosts, adding weight to our previous hypothesis [12]. Similar to the interactions with its natural host [57,58], the nitrosative stress response genes, cgb and ctb, were among the most abundant transcripts in intra-amoebal C. jejuni, indicating nitric oxide (NO^•) response. NO^• is an antimicrobial agent and an important component of the host immune system [59,60]. Disruption of C. jejuni cgb showed significant (p < 0.05) reduction in survival in our model, emphasizing the importance of NO^• defense, but also similarities between A. castellanii and C. jejuni warm-blooded hosts.

Survival and persistence within a host depends upon sensing stress and responding accordingly [61]. Our data also showed upregulation of the stress response in intra-amoebal C. jejuni, particularly gene products associated with DnaK, which are involved in suppressing protein aggregation consequently to environmental stress and are important to C. jejuni niche adaptation [57,62,63,64]. Other upregulated proteostasis genes were of the thiol–disulfide interchange system that repairs protein disulfide bonds, including redox stress [65]. This system is also important during C. jejuni interaction with its hosts [57,62,66,67], and mutation to dsbABC showed reduced survivability in our amoebae model. Motility and taxis are also indispensable determinants of invasion and survival within C. jejuni warm-blooded hosts [23,68,69]; our RNA-Seq data showed that some of the key genes involved in biosynthesis and maintenance of the flagellar were up-regulated. Disruption of flgB and the two heavily O-glycosylated flagellins, flaA and flaB, the former of which was shown to be involved in host cell invasion [70], showed decreased invasion to A. castellanii. Additionally, mutation to flgB was previously shown to be important in the secretion of campylobacter invasion antigens and other modulatory proteins through the putative flagellar-type 3 secretion system (FT3SS) [47,48,71]. This interesting find suggests that internalization by A. castellanii is partly bacterial dependent and emphasizes the importance of a fully intact and functional flagellar in C. jejuni-host interaction [72]. Whether it is motility, per se, and/or the non-motility properties of the C. jejuni flagellar that are important for these interactions, remains to be fully evaluated.

C. jejuni has a plethora of nutrient uptake and transport systems [57,62,73], which are complemented by the central carbon metabolism and electron transport system. We observed differential expression of these genes, indicating that intra-amoebic C. jejuni finely tunes its metabolic requirements to adapt to this niche. Both transient and obligate intracellular pathogens are characterized by their differential expression of their metabolic needs [74,75]. We observed increased up-regulation of metabolic genes in the intra-amoebic C. jejuni relative to the control, indicating that the intracellular environment of A. castellanii is nutrient-restrictive. This was further supported by the up-regulation of the energy taxis signaling response system, cetB [62,76].

We also observed significant differential expression of uncharacterized genes; given that some of these genes were highly upregulated, they might be of functional importance to the transient intracellular lifestyle of C. jejuni, and characterization of their products might be key to advancing our understanding of this enigmatic pathogen. The novel observation that amoebae are capable of harboring C. jejuni 11168H necessitates further investigation on the survival of this pathogen outside the warm-blooded host. This study highlights that amoebae and C. jejuni warm-blooded hosts have similar properties, which makes A. castellanii a useful model to study C. jejuni-host interactions.