Evaluation of Chlamydia pecorum major outer membrane protein vaccine a management tool in wild koala (Phascolarctos cinereus) populations

Abstract

Chlamydiosis is a significant disease affecting Eastern Australian koalas (Phascolarctos cinereus), contributing to the decline of some koala populations, necessitating investigations into appropriate management strategies to address chlamydiosis in wild koala populations. The aim of this study was to investigate the effect of a Chlamydia pecorum recombinant Major Outer Membrane Protein (rMOMP) vaccine as a potential strategy for managing chlamydiosis at a population level. This study comprised a blinded, randomised placebo-controlled trial, encompassing different koala populations where chlamydiosis is having differing effects. Wild koalas were recruited into a vaccination or a placebo treatment group and followed for 12 months, with recapture and resampling at 2, 6 and 12 months post vaccination. Vaccination stimulated a significant plasma anti-MOMP IgG response and greater IL-17 and TNFα mRNA fold change from rMOMP stimulated leukocytes, however, did not boost pre-existing immune responses, from natural infection, in koalas. The observed immunological stimulation did not translate to any effect on chlamydiosis or chlamydial shedding in our study populations. These findings highlight the necessity of improving our understanding of what constitutes a protective immune response in koalas to guide the development of a more effective vaccine. This study evaluated the estimated effect of vaccination necessary to achieve management outcomes predicted by modelling studies. It is possible that vaccination has a more modest effect and could benefit koala populations with a lower disease prevalence or be useful in conjunction with additional management strategies.

Introduction

Koalas (Phascolarctos cinereus) have been listed as an endangered species within the Australian states of New South Wales (NSW), Queensland and the Australian Capital Territory, as these populations have declined and are forecast to further decline by approximately 26–50% over the next three generations¹. A significant factor contributing to this decline is chlamydiosis, the disease caused by Chlamydia pecorum². Chlamydiosis results in ocular, urinary³ and reproductive tract pathology^4–7 reducing fertility in both female and male koalas^7,8. Chlamydiosis is present in most surveyed wild populations on mainland Australia^9–11 affecting 28–73% of individuals^9,12. Although antibiotic therapy can resolve chlamydial shedding and some aspects of ocular and urinary tract disease, its use for disease management at a population level is limited by the logistics of prolonged and frequent therapeutic administration^13,14, the effect of these antibiotics on the microbiome¹⁵ and the likelihood of reinfection or recrudescence following cessation of therapy^11,16. Vaccination against C. pecorum offers potential as a tool to achieve the two primary goals of chlamydiosis management: minimising transmission and mitigating the severity of disease^17,18.

To ascertain the efficacy of vaccination to address chlamydiosis in wild koala populations, it is important to build stronger evidence regarding its effect in diseased koalas. While modelling suggests that vaccination could potentially reverse population declines of koala populations impacted by chlamydiosis^19, this assumes a vaccine efficacy of 35–75%. Evidence suggests that vaccination can reduce the median age at which a koala develops signs of disease by 3 years (5 years to 8 years) and decrease chlamydial related deaths by 64.7% (8.5–3%)²⁰. A C. pecorum recombinant Major Outer Membrane Protein (rMOMP) vaccine can induce humoral^17,21,22 and Th1 and Th17 associated cell mediated immune responses in koalas^18,21. However, there is limited evidence comparing vaccine elicited immune stimulation to immune responses following natural infection^17,23,24 and the effect of vaccine elicited immune stimulation on reducing transmission or severity of disease remains unclear. There is evidence suggesting vaccination can improve ocular chlamydiosis^25,26 however these studies did not compare to a placebo group and thus, disease improvement may have occurred regardless²⁷. Where vaccination trials have included free ranging C. pecorum positive, subclinical koalas, vaccination has been associated with reduced chlamydial shedding for up to six months^17,24 but sustained benefits beyond this timeframe are likely necessary for effective management in wild populations. The effect of vaccination on reproductive disease also remains unexplored. Considering reproductive disease is a significant factor driving population decline^8,28 exploring this aspect is necessary if vaccination is intended to be employed to address chlamydiosis in wild populations.

Chlamydiosis has a substantial impact on some wild koala populations²⁹ yet the ability of an rMOMP vaccination to mitigate transmission or chlamydiosis remains unexplored in this context. Vaccination has been employed as a management strategy in other wild animal populations facing substantial disease impacts, with varying success^30–32. For example, vaccination reduced disease severity and mortality in wild rabbit (Oryctolagus cuniculus) populations heavily impacted by myxomatosis³². However, where vaccination has not boosted natural immune responses, for instance, in wild dogs (Canis familiaris) exposed to canine Distemper virus, canine Adenovirus and canine Parvovirus, vaccination had limited beneficial effects³¹. While there is evidence suggestion a triple dose vaccination can enhance pre-existing natural immune responses greater than natural infection as compared to unvaccinated koalas²²there is limited evidence indicating a single dose vaccine can achieve a similar outcome^17,24. To effectively employ vaccination as a strategy against chlamydiosis in high-disease prevalence populations or in koalas with pre-existing natural immune responses, it is crucial to investigate whether a single dose vaccination enhances and prolongs natural immune responses compared to a placebo-control group.

Vaccination has been used to successfully create a protective barrier around non-infected wildlife populations and therefore offers potential to create a barrier between C. pecorum positive and negative populations. For example, vaccination of wild foxes (Vulpes vulpes) against rabies created immune belts that prevented the spread of rabies disease from affected areas into adjacent non-affected areas^33,34. An estimated vaccine coverage of 60% prevented encroachment of infection into neighbouring vulnerable populations³⁴. It is unknown what level of vaccine coverage would be required to create a protective immune barrier around C. pecorum negative koala populations. In South-west Sydney, a koala population free of C. pecorum infection lies in close proximity to populations where infection is present^35,36. Exploration of vaccination as a strategy to reduce transmission events in the interface zone between these positive and negative populations, and therefore slow or prevent disease incursion into valuable disease-free koala populations, is warranted.

The objective of this study was to evaluate the efficacy of a single dose rMOMP vaccine in koala populations with differing prevalences of chlamydial infection, with the view to assess this strategy as a disease management tool at a population level. This study was designed as a blinded, randomised placebo-controlled trial, encompassing three distinct populations: one critically impacted by chlamydiosis; another where chlamydiosis is present at a lower prevalence, with limited reproductive disease, which geographically neighbours a third population where chlamydial infection is absent. This study examined the effect of vaccination on both systemic humoral and cell-mediated immunity and assessed if vaccination enhanced natural immune responses, and thereby provide benefits to infected koalas. The effect of vaccination on chlamydiosis and chlamydial shedding was evaluated, with the intent to predict its ability to reduce chlamydial disease and transmission. Furthermore, data on infection incidence was included as a preliminary investigation into the likely efficacy of vaccination in preventing disease incursions into C. pecorum naïve koala populations.

Materials and methods

Ethics

All methods were performed in accordance with the relevant guidelines and regulations. All handling and sampling of koalas was conducted under the University of Sydney Animal Ethics Approval Number 2019/1547 and 2023/2253 and NPWS Scientific Licence SL102331. No animals within this study were euthanised. This study is reported in accordance with ARRIVE guidelines.

Vaccination

The vaccine antigens comprised three whole recombinant Major Outer Membrane Proteins representing C. pecorum ompA genotypes A, F and G³⁷. The circulating strains within the study populations were broadly matched to the genotypes used within this vaccine^10,38. These antigens were combined with a three-component adjuvant, containing polyribosinic: polyribocytidic acid (poly I: C), polyphosphazene polyelectrolytes, and a synthetic host defence peptide (VIDO-Intervac, University of Saskatchewan, Canada). This vaccination has been tested previously in koalas^{17,21,22,25,26,39}. Vaccination and placebo injections (Hartmann’s solution) were administered subcutaneously (0.5 ml) into the interscapular region at first capture.

Experimental design and statistics

Three different populations were recruited into this study, with different attributes (Fig. 1; Table 1), to address the study aims. The first, Gunnedah on the Liverpool Plains of NSW, is critically impacted by chlamydiosis and the increase in C. pecorum prevalence over the past 14 years has coincided with a dramatic decline in the number of breeding females, and subsequent increase in the average age of the population^10,12,29. In the second population, the Southern Highlands, C. pecorum infection and chlamydiosis is present at lower levels than Gunnedah. Chlamydiosis does not critically impact this population and there is minimal evidence of reproductive disease (Table 1). This population is divided into two separate study sites, “Appin” (Interface) and “Mittagong”. The third population, South-west Sydney, is considered free of chlamydiosis and C. pecorum infection (Table 1), is genetically distinct and has a lower genetic diversity compared to other populations within the Sydney region, consistent with a recent bottle neck⁴⁰. All analyses were conducted within the R Statistical Environment⁴¹.

Fig. 1.

Map of koala populations included in this study. Map was generated with QGIS Geographic Information System (Version 3.34) https://www.qgis.org⁴³.

Table 1.

Summary of pre-vaccination number affected/total number of C. pecorum positive cases, chlamydiosis (urinary tract, ocular disease, paraovarian cysts), and number of individuals included in the vaccinated and placebo groups for the populations South-west sydney, Southern highlands and Gunnedah.

	South-west Sydney	Southern Highlands,  Appin site, (Interface)	Southern Highlands,  Mittagong site	Gunnedah
C. pecorum positive koalas	0% (0/10)	11% (2/18)	33% (9/27)	87% (41/47)
Evidence of urinary disease	0% (0/10)	11% (2/18)	25% (7/27)	32% (15/47)
Evidence of ocular disease	0% (0/10)	27% (5/18)	37% (10/27)	47% (22/47)
Evidence of paraovarian cysts (females only)	0% (0/5)	12.5% (1/8)	0% (0/14)	60% (18/30)
Vaccinated koalas	100% (10/10)	100% (18/18)	55% (15/27)	60% (28/47)
Placebo koalas	0% (0/10)	0% (0/10)	45% (12/27)	40% (19/47)
Median tooth wear class, pre-molar 442#	2 lines	2 lines	2 lines	Flat

# Tooth Wear Class refers to the degree of wear on the occlusal surface of upper pre-molar 4 in the dental arcade, with the following sequential wear categories possible: 1 line; 2 lines; circle; flat.

The first aims of this study were to: (1) investigate whether vaccination stimulated humoral and cell-mediated immune responses in wild koalas, compared to a placebo-control cohort and (2) if this vaccine boosts natural immune responses in infected koalas, compared to a placebo-control cohort. The selected immune responses, IFNγ, IL-17, TNFα and IL-10, were chosen due to evidence linking these cytokines to variations in chlamydiosis expression in koalas^44–46, as well as their known stimulation following rMOMP vaccination^17,25. To address these aims, the koalas from Gunnedah and Southern Highlands (Mittagong site) populations were recruited into vaccination and placebo-control groups. A sample size of 10 koalas per treatment group was determined by a power analysis⁴⁷. This is predicted to achieve a power of 80%⁴⁸, assuming a 60% difference between vaccinated and unvaccinated groups, consistent with previous data²⁴ and at the upper range (35% – 75%) of a desired effect of vaccination¹⁹. There was a significant difference between populations on immune responses, where Gunnedah koalas had greater concentrations at time zero, compared to Mittagong. Therefore, the cohorts were examined separately. The effects of treatment (vaccination or placebo), time following vaccination and their interaction, on anti-MOMP plasma IgG concentrations and antigen-stimulated fold change of four different cytokines were tested with separate Linear Mixed Effects Models (LMMs), using the ‘lme4’ package⁴⁹. Individual koala was treated as the random factor. Data were log-transformed to meet the assumptions of the LMMs. To investigate if vaccination further boosted natural immune responses, the effect of treatment, time following vaccination, their interaction, and population on anti-MOMP plasma IgG values and antigen-stimulated fold change of four different cytokines were tested with separate LMMs. Individual koala was treated as the random factor. Again, data were log-transformed to meet the assumptions of the LMMs. This analysis included all vaccinated and placebo-control koalas from Gunnedah and Southern Highlands (Mittagong site) populations who were C. pecorum positive at the beginning of the study.

The second aim of the study was to assess if vaccination affected ocular, urinary and reproductive tract chlamydiosis or chlamydial shedding, thus conferring benefits to populations impacted by chlamydiosis. To analyse the effect of vaccination on signs of chlamydiosis (ocular and urinary), the effect of treatment (vaccination or placebo), time following vaccination, their interaction, and population on ocular and urinary tract disease was tested in koalas from Gunnedah and Southern Highlands (Mittagong site) populations with Cumulative Link Mixed Models (CLMM), using the ‘clmm2’ function constructed with the ‘ordinal’ package⁵⁰. Individual koala was included as the random factor. CLMM allows the analysis of ordinal response variables while using random effects⁵¹. To evaluate the effect of vaccination on paraovarian cysts, only the Gunnedah population was examined as only two koalas outside of this population had paraovarian cysts detected throughout the study. The effect of treatment (vaccination or placebo), time following vaccination, and their interaction on paraovarian cyst score was tested with a CLMM with individual koala included as the random factor.

To test the effect of vaccination on chlamydial shedding, the effect of treatment (vaccination or placebo), time following vaccination, their interaction, and population on C. pecorum, delta cycle threshold (ΔCT) values were tested with a LMM, with individual koala as the random factor. ΔCT values were used to normalise C. pecorum CT against swab yield (koala beta-actin DNA) from the same swab sample and generated as ΔCT = CT of C. pecorum - CT of beta-actin. Chlamydial shedding is expressed as the inverse of C. pecorum ΔCT values because, as chlamydial shedding decreases, C. pecorum ΔCT values increase.

The third aim of the study was to evaluate if vaccination against chlamydiosis could reduce the likelihood of transmission of infection into a neighbouring C. pecorum negative population. To examine this aim, changes to chlamydial shedding and the detection of positive cases were analysed in vaccinated koalas in the Appin cohort, Southern Highlands population. Appin is a geographic region where there is potential for encroachment of chlamydiosis into a neighbouring C. pecorum negative population, South-west Sydney. To test the effect of vaccination on chlamydial shedding in this population, the effect of time following vaccination on C. pecorum ΔCT values was tested with a LMM, with individual koala as the random factor. No placebo group was recruited in the South-west Sydney population or Appin cohort as the trial was enacted as a management strategy to prevent and aimed to achieve the highest level of vaccine coverage possible within the geographical area.

Koala capture and recruitment

Koalas were captured using the ‘noose and flag’ technique⁵². Alfaxalone (Jurox), was used to sedate the koalas at a dose rate of 1.8 mg/kg by intramuscular injection and then oxygen and isoflurane was administered to effect through a fitted mask to anaesthetise the koala. One veterinarian (SJS) assessed and sampled all koalas. At initial assessment, koalas were classified as either diseased or non-diseased, based on external signs of chlamydiosis, presence of perineal staining “wet bottom,” or ocular disease⁵³. These criteria were chosen as they can improve with antibiotic treatment¹⁴ and because infection status of individuals was unknown at initial capture. Following this classification, koalas within the Gunnedah population and Mittagong site were assigned randomly to one of two treatment groups, vaccination, or placebo. The veterinarian who examined the koalas and administered the treatment was blinded to treatment group allocation. VHF collars (transmitter model M3420, ATS Australia) were fitted to the koalas so they could be radio-tracked and visually monitored from a distance at 6–8 week intervals throughout the study, and recaptured, clinically evaluated and sampled at 2, 6 and 12 months following vaccination.

Plasma anti-major outer membrane protein (MOMP) IgG ELISA

All concentrations of reagents, blocking and incubation conditions were determined by optimisation experiments using chequerboard titrations. In the optimised assay, 96 well flat bottom plates (Greiner Bio-One #650101) were coated with 2 µg of rMOMP serovar G (produced by methods previously described)²² in carbonate-bicarbonate buffer (Sigma-Aldrich #C3041) at 100 µl/well at 4 °C overnight. Wells were emptied and then blocked with 300 µl of 5% skim milk powder in PBS with Tween (0.05%) (PBST) at 37 °C for 1 hour. Wells were emptied and koala plasma was applied to the wells at a concentration of 1:400 at 100 µl/well and incubated at 37 °C for 1 hour. Wells were then emptied and washed five times with PBST with a microplate washer (Biorad model 1575). An in-house sheep anti-koala-IgG antibody¹⁸ 100 µl was added to each well at a dilution of 1:8000 and incubated at 37 °C for 1 hour. Following incubation, wells were washed as described above. Horseradish peroxidase conjugated Rabbit anti-sheep IgG 100 µl (Abcam #ab6747) was added to each well (1:20000 dilution) and incubated at 37 °C for 1 hour. Following incubation, wells were washed as described above, using PBS only. Finally, 100 µl of 3,3’, 5, 5’ Tetramethylbenzidine (Sigma-Aldrich #T5525) was added to each well and allowed to develop at room temperature, in the dark, for 20 min before 100 µl of 1M H₂SO₄ was added to each well. Optical density (OD) was read at 450 nm. Each plate contained a blank well (no antigen or plasma), the same negative control from a captive, C. pecorum negative koala and a standard curve based on doubling dilutions of a known strong positive sample, selected during optimisation, from a C. pecorum positive koala. The inter-assay coefficient of variation was < 15% and the intra-assay coefficient of variation was accepted if it was < 10%. Each sample was run in a well coated with antigen and one without antigen, to allow quantification of non-specific background. All samples were run in duplicate, and the mean was calculated from these replicates. All OD values were blank adjusted; the OD value of the no-antigen well was subtracted from the OD value of the antigen and plasma well for each sample. The highest concentration standard was given a nominal value of 32. The other values were calculated relative to that standard, based on dilution. The OD values of the samples were compared to the standard curve using a 4-parameter logistic curve function.

Leukocyte stimulations

Two millilitres RPMI medium (Sigma-Aldrich #R7388), was incubated for 30 min at 37 °C. The same volume of heparanised blood from each koala was centrifuged at 3000 g for five minutes and the leukocyte fraction (buffy coat) was aspirated and suspended in the pre-incubated media. In duplicate, 5 µg of rMOMP genotype G in 25 µl of foetal calf serum (FCS) or, in the case of negative controls FCS 25 µl, was added to 220 µl of the buffy coat suspension. Suspensions were then incubated for 12 h at 37 °C, with air permeable cap, and then 750 µl of RNA later (Sigma-Aldrich #R0901) was added to each sample and the sample stored at room temperature for 24 h, then frozen at −20 °C until processing.

RNA extraction and cDNA synthesis

RNA was extracted from the leukocyte suspensions using the RiboPure – Blood kit (Invitrogen #AM1928) according to the manufacturer’s protocol. Extracts were treated with RNAse-free DNAse, (Thermofisher #EN0521) and cDNA synthesis was then performed using the RevertAid First Strand cDNA Synthesis Kit (K1622) according to the manufacturer’s protocol.

Cytokine analysis

A qPCR assay was used to estimate expression of the following genes: GAPDH, IFNγ, IL-17A, Il-10 and TNFα^45,54. Reactions were made to a final volume of 20 µl consisting of 10 µl of SsoAdvanced^tm Universal SYBR Green Supermix (BioRad), 0.5 µM of each primer with 6 µl of H₂O (GAPDH) and 0.3 µM of primer with 6.8 µl of H₂O (IFNγ and IL-17A IL-10 and TNFα) and 2 µl of DNA. Cycling conditions for the different genes were applied as previous^45,54. All samples were run in duplicate with negative controls at each step. IFNγ, IL-10 and TNFα and IL-17A expression for each koala for each time point were normalised to GAPDH by the 2^−ΔΔCT method, where ΔΔCT = (CT of target − CT of GAPDH) − (CT of target − CT of GAPDH) and presented as a fold change relative to the unstimulated sample.

Koala clinical examination

The following procedures were performed at timepoint 0 (recruitment), 2, 6 and 12 months post vaccination. To grade urinary tract disease a wet bottom score (WBS) was allocated, based on a modification of the Griffith (2010)⁵³ scoring system; only scores 0–6 were used because scores 7–9 require prolonged behavioural observations, not amenable to field examinations. Urinary tract disease was then classified as either normal (0), mild (1–2), or severe (3+) based on the WBS. Both eyes were assessed for ocular discharge, conjunctival proliferation, and conjunctival chemosis, each with a score between 0 and 3, and a total cumulative score allocated to each eye⁵³. Ocular disease was then classified as either normal (0), mild (1–5) or severe (6+) depending on the total eye score.

Ultrasound was performed with the koala in dorsal recumbency^55–57. For female koalas without pouch young, the transducer was placed inside the pouch and reproductive structures were visualised by scanning the region cranio-lateral to the bladder. If paraovarian cysts were identified, they were classified as unilateral or bilateral. A paraovarian cyst score was allocated with 0 = no abnormalities detected, 1 = unilateral paraovarian cysts identified and 2 = bilateral paraovarian cysts identified. If paraovarian cysts were not detected, a thorough scan of the caudal abdomen was performed to confirm their absence and normal ovaries were identified. Koalas with pouch young were not examined ultrasonographically.

Specimens for detection of C. pecorum DNA were collected from the conjunctiva bilaterally and the female urogenital sinus or male urethra. For conjunctival sampling, swabs (Copan, Interpath, 160 C) were placed into the conjunctival fornix and rotated five times. In females, swabs were inserted 2–3 cm into the urogenital sinus and rotated five times. In males, the penis was everted, and the swab placed 2 cm into the urethra and rotated five times. All swabs were stored at −20 °C until processing.

Urogenital and ocular swab DNA extraction and qPCR

DNA was extracted from ocular and urogenital swabs using the MagMAX™ CORE Nucleic Acid Purification Kit #A32702. A sterile swab was included as a control for contamination during the extraction process. A multiplex qPCR assay was used to quantify the following genes from each extracted swab sample, P. cinereus beta actin (HEX), 23 S Chlamydia (ROX) and C. pecorum ompB (FAM); primers from Hulse et al., 2018⁵⁸. PCR reactions were made to a final volume of 20 µl consisting of 10 µl of master mix (SensiFAST™ Probe, Bioline, London, UK), 400 µM of each primer, 200 µM of each probe, 4.4 µl of dH₂O and 2 µl of DNA. The cycling conditions included initial denaturation for two minutes at 98 °C, followed by 40 cycles of denaturation for 15 s at 98 °C and a combined annealing and extension step for 30 s at 58°C. Samples were classified as positive if there was amplification of C. pecorum ompB or low amplification of 23 S Chlamydia only, given the multicopy nature of 23 S and the negligible likelihood that ocular or urogenital Chlamydiae are other than C. pecorum⁵⁸. In this study, all positive cases had amplification of both C. pecorum ompB and 23 S Chlamydia. The limit of detection for the C. pecorum ompB assay was determined using a Probit regression analysis, detecting 86 gene target copies per reaction with 95% confidence, between the Ct-values of 34–35.

Results

Analysis of vaccine-induced immune responses in wild Koalas

Vaccination had a significant effect on some immune responses; however, this was only evident in the Mittagong cohort. The vaccinated group in the Mittagong cohort exhibited greater anti-MOMP plasma IgG concentrations and greater IL-17 and TNFα expression in rMOMP-stimulated peripheral leukocytes post vaccination compared to the placebo group (Tables 2 and 4). There was a significant effect of time on plasma anti-MOMP IgG (Table 2) and IL-17 expression (mRNA) fold change from rMOMP stimulated peripheral blood leukocytes (Tables 2 and 3) where both treatment cohorts had an increase in these immune responses over time.

Table 2.

The effect of time, treatment and the interaction of time and treatment in the Mittagong cohort on systemic plasma anti-MOMP IgG concentrations and cytokine expression (mRNA) fold change from rMOMP stimulated peripheral blood leukocytes. Significant effects are in bold.

Dependent variable	Explanatory variables	F value	P value
Plasma anti-MOMP IgG	Time	F3,94= 9.31	P = 0.001
	Treatment	F1,33= 11.66	P = < 0.001
	Time x Treatment	F3,94= 2.96	P = 0.032
IFNγ	Time	F3,50= 0.31	P = 0.816
	Treatment	F1,19= 1.70	P = 0.207
	Time x Treatment	F3,50= 0.15	P = 0.923
IL-17	Time	F3,75= 3.21	P = 0.022
	Treatment	F1,32= 1.15	P = 0.281
	Time x Treatment	F3,75= 3.04	P = 0.032
TNFα	Time	F3,67= 2.56	P = 0.065
	Treatment	F1,29= 0.48	P = 0.499
	Time x Treatment	F3,67= 3.12	P = 0.033
IL-10	Time	F3,6o= 2.06	P = 0.117
	Treatment	F1,24= 0.49	P = 0.490
	Time x Treatment	F3,6o= 1.75	P = 0.166

Table 4.

Systemic plasma anti-MOMP IgG concentrations and cytokine expression (mRNA) from rMOMP stimulated peripheral blood leukocytes of vaccinated and placebo Koalas within Mittagong cohort. Data is presented as mean (95% confidence intervals). Significant differences are in bold.

Immune variable	Time 0		2 months		12 months
	Vaccinated	Placebo	Vaccinated	Placebo	Vaccinated	Placebo
Plasma anti-MOMP IgG	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.77 (0.29, 1.05)	0.00 (0.00, 0.00)	1.31 (0.88, 1.87)	0.42 (0.06, 0.70)
IFNγ	1.67 (1.17, 2.17)	1.51 (0.91, 2.11)	1.91 (1.52, 2.30)	1.71 (1.23, 2.18)	1.75 (1.14, 2.35)	1.40 (0.67, 2.13)
IL-17	3.57 (2.60, 5.32)	2.81 (1.58, 4.03)	4.70 (4.06, 5.45)	2.49 (1.53, 3.22)	5.13 (4.36, 5.91)	3.38 (2.13, 4.63)
TNFα	2.41 (1.94, 2.87)	3.94 (2.94, 4.93)	4.02 (3.32, 4.73)	3.66 (2.86, 4.45)	4.22 (3.33, 5.11)	3.66 (2.77, 4.50)
IL-10	2.28 (1.74, 2.82)	1.67 (1.31, 2.02)	2.01 (1.88, 2,31)	1.92 (1.58, 2.27)	2.11 (1.43, 2.79)	2.57 (2.21, 2.94)

Table 3.

The effect of time, treatment and the interaction of time and treatment in the Gunnedah cohort on systemic plasma anti-MOMP IgG concentrations and cytokine expression (mRNA) fold change from rMOMP stimulated peripheral blood leukocytes. Significant effects are in bold.

Dependent variable	Explanatory variables	F value	P value
Plasma anti-MOMP IgG	Time	F3,104= 0.39	P = 0.752
	Treatment	F1,39= 0.96	P = 0.334
	Time x Treatment	F3,104= 0.39	P = 0.758
IFNγ	Time	F3,133= 1.50	P = 0.221
	Treatment	F1,45= 1.44	P = 0.236
	Time x Treatment	F3,133= 2.06	P = 0.153
IL-17	Time	F3,109= 3.74	P = 0.010
	Treatment	F1,41= 0.09	P = 0.755
	Time x Treatment	F3,109= 0.18	P = 0.905
TNFα	Time	F3,115= 1.74	P = 0.153
	Treatment	F1,43= 1.09	P = 0.294
	Time x Treatment	F3,115= 1.28	P = 0.282
IL-10	Time	F3,108= 1.37	P = 0.254
	Treatment	F1,44= 0.81	P = 0.370
	Time x Treatment	F3,108= 1.20	P = 0.313

Analysis of the effect of vaccination on immune responses in C. pecorum positive Koalas

Vaccination did not significantly boost pre-existing natural immune responses in C. pecorum positive koalas from the Mittagong and Gunnedah cohorts, when compared to a placebo-control group. C. pecorum positive koalas in both the vaccinated and placebo cohorts had an increase of their plasma anti-MOMP IgG concentrations over time (Table 5).

Table 5.

The effect of treatment, time, the interaction of treatment and time and population in C. pecorum positive koalas, on systemic plasma anti-MOMP IgG concentrations and cytokine expression (mRNA) fold change from rMOMP stimulated peripheral blood leukocytes. Significant effects are in bold.

	IgG	IFNγ	IL-17	TNFα	IL-10
Treatment	F1,56= 0.08 P = 0.886	F1,61= 0.01 P = 0.903	F1,68= 0.02 P = 0.885	F1,49= 1.57 P = 0.215	F1,51= 0.32 P = 0.574
Time	F3,127= 9.24 P = < 0.001	F3,167= 1.71 P = 0.161	F3,168= 1.77 P = 0.156	F3,127= 1.36 P = 0.254	F3,132= 2.244 P = 0.086
Time X Treatment	F3,127= 1.76 P = 0.155	F3,167= 1.04 P = 0.372	F3,168= 0.19 P = 0.897	F3,127= 0.92 P = 0.433	F3,132= 1.02 P = 0.383
Population	F2,46= 1.24 P = 0.301	F2,55= 1.36 P = 0.155	F2,56= 1.25 P = 0.288	F2,45= 0.98 P = 0.409	F2,45= 0.82 P = 0.443

Analysis of the effect of vaccination on chlamydiosis

There was no significant effect of vaccination treatment (F_1,88= 0.06, P = 0.808), time (F_3,225= 1.96, P = 0.114) or the interaction of treatment and time (F_3,225= 1.50, P = 0.214) on urinary tract chlamydiosis. There was a significant effect of population (F_2,155= 2.71, P = 0.050), Gunnedah koalas had a greater severity of urinary tract chlamydiosis relative to the Mittagong cohort.

There was no significant effect of vaccination treatment (F_1,88= 0.48, P = 0.484) or the interaction of treatment and time (F_3,225= 0.07, P = 0.973) on ocular chlamydiosis. There was a significant effect of population (F_2,155= 2.71, P = 0.048), Gunnedah koalas had a greater severity of ocular chlamydiosis relative to the Mittagong cohort. There was also a significant effect of time (F_3,225= 2.73, P = 0.043) where both vaccinated and placebo koalas had an improvement to their ocular chlamydiosis over time.

In female koalas, there was no significant effect of time (F_3,125= 2.40, P = 0.064) or the interaction of treatment and time (F_3,125= 1.04, P = 0.375), on paraovarian cyst score. There was a significant effect of treatment (F_1,57= 17.54, P = < 0.001) due to an imbalance at recruitment where placebo koalas had unilateral and bilateral cysts detected on ultrasound more frequently than the vaccinated group. This occurred due to chance, as koalas were allocated to treatment group based on external signs expected to respond to treatment. Of the female koalas that did not have paraovarian cysts at the start of the trial, three vaccinated and one placebo koala developed paraovarian cysts over 12 months.

Analysis of the effect of vaccination on chlamydial shedding

There was no significant effect of vaccination treatment (F_1,88= 0.47, P = 0.492), time (F_3,225= 1.08, P = 0.352) or the interaction of time and treatment (F_3,225= 0.04, P = 0.981) on C. pecorum urogenital ∆CT values. There was a significant effect of population (F_2,85= 22.55, P = < 0.001), with koalas within the Gunnedah cohort (Fig. 2) showing greater chlamydial shedding (i.e. lower urogenital ΔCT) compared to koalas within the Mittagong cohort (Fig. 3). New detected urogenital positive cases occurred at the same frequency between the vaccinated and placebo cohort (1:1 at two and six months and 3:3 at 12 months). Chlamydial shedding could not be compared in these koalas due to the small sample size. At the start of the study the prevalence of ocular chlamydial infection was 4%. Over the study period, four new ocular positive cases were detected, all of which were vaccinated koalas. No placebo koalas became positive at the ocular site.

Fig. 2.

Urogenital ΔCT values of vaccinated and placebo koalas within the Gunnedah population at time 0 (pre-vaccination) and 2, 6 and 12 months post vaccination.

Fig. 3.

Urogenital ΔCT values of vaccinated and placebo koalas within the Mittagong cohort, Southern Highlands population, at time 0 (pre-vaccination) and 2, 6 and 12 months post vaccination.

Analysis of the effect of vaccination on reducing transmission risk at an interface between C. pecorum-positive and negative population

Within the Appin site, at the interface between C. pecorum positive and negative populations, where all koalas were vaccinated, the number of C. pecorum positive koalas increased over the study period. Some koalas died due to natural causes (unknown) as they were wild animals. Pre-vaccination, the number of koalas that tested positive within this site was 11% (2/18). This increased to 31% (5/16) whereby an additional three vaccinated koalas became positive at two months. At two months post vaccination, two koalas were removed from the study due to death from natural causes. The same five koalas remained positive at six months post vaccination. The incidence then decreased to 21% (3/14) at 12 months post vaccination, due to the loss of an additional two koalas by natural causes, who had been positive at two and six months. One koala resolved their infection, and one new positive case was detected at 12 months post vaccination. At the Appin site, there was a significant effect of time (F_3,46= 1.75, P = 0.04) on C. pecorum urogenital ∆CT values, with chlamydial shedding increasing over time following vaccination.

Discussion

Despite vaccination stimulating an immune response, this study was unable to provide evidence that a single dose of the vaccine trialled in this study reduced transmission of infection or could mitigate the impact of chlamydiosis in these population scenarios. Vaccination stimulated a significant plasma anti-MOMP IgG response and increased IL-17 and TNFα mRNA responses to rMOMP stimulation of leukocytes, compared to a placebo-control in the Mittagong cohort. This was not evident in the Gunnedah population, which had a high prevalence of pre-existing responses to natural infection. Vaccination did not enhance immune responses in infected koalas within the Mittagong cohort or Gunnedah population. The key finding, that the stimulation of these immune responses following vaccination, did not translate to any observed effect on chlamydiosis or chlamydial shedding in the Mittagong cohort, suggests that the immune response elicited was not appropriate in either magnitude, or type, to yield beneficial outcomes in this population. This suggests that enhancing magnitude or type of immune stimulation warrants further investigation and that there is a need to gather further evidence on what comprises an effective response, to optimise outcomes. As this study was limited to the analysis of selected immune responses, it remains possible that the activation of additional immune pathways may be critical to fully understanding the koala’s immune response to infection.

This vaccine is unlikely to mitigate the impact of chlamydiosis in these population scenarios in the present study. We could not demonstrate that this vaccine could prevent or mitigate the advancement of ocular, urinary or reproductive chlamydiosis when compared to a placebo-control. We vaccinated koalas with chlamydial disease because previous research has suggested that vaccination of koalas with ocular chlamydiosis could have therapeutic benefits^25,26. However, the inclusion of a placebo-control in our study indicated that ocular chlamydiosis can improve without treatment, consistent with historical infection trials²⁷. Hence, the therapeutic effect observed in previous research is unlikely to be solely attributable to vaccination. It is important to note, however, that in alternate population scenarios, vaccination was associated with reduced development of chlamydial ocular and urinary disease²⁰. The present study was the first to examine the effect of vaccination on paraovarian cysts in koalas. A limitation of this study was that the majority of the female koalas within the Gunnedah population had developed paraovarian cysts prior to the commencement of the study. Therefore, unsurprisingly we were unable to demonstrate that this vaccine could resolve paraovarian cysts; as the fibrosis^3,4 involved in this pathology is likely to be irreversible^59,60. Based on previous evidence suggesting vaccination had therapeutic advantages, we hypothesised that vaccine induced immune stimulation could prevent the onset of fibrosis following infection. However, within the Mittagong cohort, vaccination did not prevent the development of paraovarian cysts in three vaccinated koalas. This was in comparison to one placebo koala; the small sample size precludes testing of this difference. Due to the field-based nature of this study, reproductive disease was detected solely through ultrasound and therefore it was not possible to evaluate the vaccine’s impact on early-stage reproductive disease. Regardless, these results suggest that alternative strategies are needed to prevent reproductive pathology and mitigate population declines influenced by chlamydiosis.

We were unable to demonstrate that this vaccine is likely to be effective at reducing transmission in wild koala populations where C. pecorum is prevalent or be effective to create a protective barrier for adjacent vulnerable populations. The vaccine in this study did not significantly reduce chlamydial shedding or prevent new infections when compared to a placebo-control cohort in any population scenario. The contrast with findings from previous research^17,24 may be explained by the higher prevalence of C. pecorum in the present study, which could have resulted in more transmission opportunities and may have overwhelmed the vaccine’s protective effect. However, while Gunnedah had a high prevalence of infection, population was accounted for in analysis of the effect of vaccination on chlamydial shedding; there was no effect of vaccination on chlamydial shedding in the Mittagong cohort which had a lower infection prevalence. Although the level of vaccine efficacy required to reduce transmission remains uncertain, modelling¹⁹ and insights from sexually transmitted infections in humans⁶¹ suggest the desired efficacy should be at least 35%−75%. Our study was designed with an assumption of a 60% effect, based off a previous koala vaccine trial²⁴. It is possible that this vaccine may have a smaller effect, which this study did not have the statistical power to detect. Due to the high level of coverage likely needed to prevent incursions of infection into neighbouring vulnerable populations³⁴ this vaccine is unlikely to be effective as a protective barrier to prevent chlamydial incursion. Despite vaccination efforts in the Appin site, chlamydial shedding increased. Thus, alternative management strategies, such as capture, removal and treatment⁶² potentially in conjunction with the installation of physical barriers⁶³ would most likely be needed to protect neighbouring C. pecorum naïve populations.

Research into a strategic vaccination approach could be considered for wild koala populations where there is a high prevalence of infection. Our study aimed to detect a significant effect of vaccination in a cohort of positive and negative koalas due to evidence that vaccination could provide therapeutic benefits^25,26 and boost pre-existing immune responses^17,23 and we aimed to investigate an approach which would be the most logistically feasible. Given the inability of this vaccine to boost pre-existing natural immune responses or reduce chlamydial shedding or disease in infected koalas, vaccinating C. pecorum positive koalas with this vaccine is unlikely to be effective under the studies proposed assumptions. An alternative strategy to implementing vaccination in populations with moderate to high infection rates would be targeting immunologically naïve koalas, specifically koalas prior to reaching breeding age^19,32 and the inclusion of an antibiotic treatment regime to lower infection prevalence. This approach requires further research prior to implementation. Although our study did not demonstrate that vaccination prevents infection, the study design lacked the statistical power to determine if vaccination significantly prevents infection in koalas without pre-existing immune responses compared to a placebo-control group. To investigate this strategy, a large-scale, long-term study is suggested to be conducted in a population with a prevalence of C. pecorum that balances the likelihood of finding naïve animals and likelihood of subsequent exposure.

One hypothesis for the lack of observed beneficial effects from the immune stimulation elicited by this vaccine, is that an inadequate magnitude of response was stimulated. The magnitude of immune stimulation by vaccination in the Mittagong cohort was comparable to the natural responses observed in koalas from the Gunnedah cohort. The magnitude of immune response stimulated by this vaccination is therefore unlikely to be sufficient in reducing transmission or mitigating severe chlamydial pathology as, despite many Gunnedah koalas possessing evidence of immune responses following infection, both transmission rates and the prevalence of advanced chlamydiosis is high. Modifications to the vaccine protocol could potentially enhance the magnitude of immune stimulation. Providing additional vaccine boosters is unlikely to substantially enhance immune responses, as vaccination of infected koalas did not achieve this in our study. An alternative approach is to trial mucosal administration at the urogenital site, as this has the potential to establish robust immunity at the site of primary transmission in koalas. Mucosal immunity, specifically at the urogenital site is important for preventing infection and chlamydial disease in mice⁶⁴ and humans⁶⁵. Mucosal immunisation against Chlamydia spp. results in stronger vaccine-induced immune stimulation and greater reductions in chlamydial shedding compared to parenteral vaccination in mice^64,66 and pigs⁶⁷. While urogenital administration of vaccination has not been trialled in koalas, its potential merits warrant consideration.

As an alternative to magnitude, it could be that the immune pathways stimulated by this vaccine do not confer benefits. Our understanding of the immune responses that facilitate chlamydial clearance or contribute to pathology in koalas is limited and is predominantly drawn from research in mice and humans, which involves different species of Chlamydia, potentially limiting its applicability to koalas. For example, the assumption that an effective vaccine for C. pecorum in koalas should elicit a predominant CD4 Th1 response, is based on strong evidence that this response is crucial for infection clearance and protection against C. trachomatis and C. muridarum infection in mice^68,69. However, C. pecorum and genital C. trachomatis strains^70,71 have mechanisms to circumvent the action of IFNγ^72,73 the main mechanism through which a CD4 Th1 response provides protection^74,75. Unlike humans⁷⁶ and koalas^77,78 experimental mice lack a microbiome at the site of infection that allows for tryptophan synthesis. There has been no association of IFNγ responses in koalas to reduced chlamydial disease^44–46 so it remains possible that evidence from murine research may not be directly applicable to koalas, as has been the case for humans with C. trachomatis infection^79–83. Stimulating the host’s immune response through a chlamydial vaccine requires a comprehensive understanding, given the significant influence the host immune response has on chlamydial pathology development^84–89. Vaccination in this study stimulated IL-17 and TNFα -associated pathways. There is some evidence associating these responses with the development of clinical chlamydiosis in koalas following natural infection^44,45. Enhancing a chlamydial vaccine for koalas hinges on identifying the immune pathways that effectively reduce chlamydial shedding and pathology in koalas. Enhanced comprehension of these pathways could facilitate the development of a vaccine tailored to target these responses specifically, through exploration of various adjuvants^90,91.

Through implementation of a blinded, placebo-controlled approach across diverse populations, this study has advanced understanding of the efficacy of a C. pecorum rMOMP vaccine in managing chlamydiosis within certain wild koala population scenarios. Despite stimulating immune responses such as systemic anti-MOMP IgG and greater IL-17 and TNFα expression from rMOMP stimulated leukocytes, this vaccine did not reduce chlamydial shedding, prevent the development of chlamydiosis or prevent new infections compared to a placebo-control. These findings underscore the need to improve our understanding of what constitutes a protective immune response in koalas, to inform the design of an effective vaccine. As our study was designed to detect the estimated desired effect of vaccination based on modelling studies, we were unable to demonstrate that this vaccine would be an effective management strategy in these population scenarios. It remains possible, however, that vaccination may have a modest effect and offer benefits to koala populations with a lower prevalence of disease than what was examined in this study.

Author contributions

S.S project design, veterinary examinations, field work, laboratory work, data analysis and writing. D.H project design, field work, data analysis and review. P.T vaccine design, project design and review. A.K field work, laboratory work and review. M.C, field work, data analysis and review. V.M field work, data analysis and review. S.P vaccine design and review. M.K project design, field work, data analysis and review.

Data availability

The data that support the findings of this study is available from the corresponding author upon request. There are no restrictions on data availability.

Declarations

Competing interests

The authors declare no competing interests.