Chlamydia Serine Protease Inhibitor, targeting HtrA, as a New Treatment for Koala Chlamydia infection

The koala, an iconic marsupial native to Australia, is a threatened species in many parts of the country. One major factor in the decline is disease caused by infection with Chlamydia. Current therapeutic strategies to treat chlamydiosis in the koala are limited. This study examines the effectiveness of an inhibitor, JO146, which targets the HtrA serine protease for treatment of C. pecorum and C. pneumoniae in vitro and ex vivo with the aim of developing a novel therapeutic for koala Chlamydia infections. Clinical isolates from koalas were examined for their susceptibility to JO146. In vitro studies demonstrated that treatment with JO146 during the mid-replicative phase of C. pecorum or C. pneumoniae infections resulted in a significant loss of infectious progeny. Ex vivo primary koala tissue cultures were used to demonstrate the efficacy of JO146 and the non-toxic nature of this compound on peripheral blood mononuclear cells and primary cell lines established from koala tissues collected at necropsy. Our results suggest that inhibition of the serine protease HtrA could be a novel treatment strategy for chlamydiosis in koalas.

has been observed clinically have little detrimental effects on the koala's specialized gut microflora compared to other medications. When administered at a dose rate of 60 mg/kg once each day for 45 days, this product results in cessation of shedding of Chlamydia within two weeks of treatment 13 . However, production of this product is limited, and recent shortages and sporadic manufacturing have resulted in the inability to treat some animals presenting to animal hospitals with chlamydial infections, with many requiring euthanasia as a consequence of disease progression (Personal Communication, Claude Lacasse, Veterinary Services Manager, Australian Wildlife Hospital).
All chlamydiae share a developmental cycle that consists of transitions between two main cell types, the infectious non-dividing extracellular form, known as the elementary body (EB), and an intracellular replicative form, called the reticulate body (RB) 16,17 . EBs are responsible for dissemination of infection by attaching to and invading susceptible cells. Upon infection, EBs are internalized in membrane bound vacuoles termed inclusions. EBs differentiate into RBs, the metabolically active form that repeatedly replicate before differentiating back to the infectious EB form. The host cell then lyses, releasing EBs that infect neighbouring cells. If this developmental cycle is disrupted under stressful growth conditions, such as immunological responses, antibiotics or nutrient deprivation, a long term growth that consists of aberrant RB's that neither replicate, nor differentiate into EBs can result in a process termed persistence (reviewed in Abdelrahman & Belland 17 ).
High temperature requirement A (HtrA), is a serine protease, which has been demonstrated to be critical for virulence and intracellular survival of many bacteria [18][19][20] . We recently demonstrated that C. trachomatis HtrA (CtHtrA) is essential for the replicative phase of this organism 21 . Utilising a screening strategy, we identified a serine protease inhibitor (JO146) that binds to HtrA and inhibits the protease activity 21 . The addition of JO146 to the mid-replicative phase of the C. trachomatis developmental cycle was lethal in vitro and no obvious toxicity was detected in mouse or human cells. The compound was also found to be effective against C. trachomatis in a broad range of host cell types 21 . Similar results for JO146 were also obtained for several other species in the genus Chlamydia 21 . In this study, we provide the basis for a novel therapeutic strategy for the treatment of chlamydiosis in koalas using the HtrA protease inhibitor, JO146. Our results suggest that this could be an effective target and that further optimization of JO146 could provide valuable new drugs for use in treating the koala.

Results
Characterization of the in vitro growth characteristics of koala C. pecorum urogenital and ocular isolates. Our previous work suggested that JO146 was most efficacious when added during the mid-replicative growth phase of C. trachomatis 22 . However, there are no published growth curves for koala Chlamydia isolates so we first needed to profile the growth of these isolates in order to subsequently evaluate inhibitor activity. Growth curves were conducted for urogenital isolates C. pecorum MarsBar and C. pecorum DBDeUG, and ocular isolate C. pecorum IPTaLE 23 . Formation of infectious units were measured at 20, 28, 36, 44, 52 and 60 h PI for C. pecorum MarsBar and C. pecorum IPTaLE (Fig. 1), and a selection of these time points were used for C. pecorum DBDeUG for confirmation of growth similarities for urogenital strains ( Supplementary Fig. S1). All isolates demonstrated characteristic bi-phasic chlamydial growth patterns with infectious elementary body (EB) production not detected until 20 h PI and these reached a peak at 36 h PI (Fig. 1). Approximately one-log difference was observed between C. pecorum MarsBar and C. pecorum IPTaLE EB yield, when measured at 52 h PI (2.85 × 10 7 IFU/ml and 7.32 × 10 6 IFU/ml, respectively). The two urogenital isolates C. pecorum IPTaLE and C. pecorum DBDeUG were consistent in terms of EB yield at 52 h PI (7.32 × 10 6 IFU/ml and 3.46 × 10 6 IFU/ml, respectively) (C. pecorum DBDeUG data shown in Supplementary Fig. S1).
The Chlamydia occupy a unique intracellular vacuole termed the inclusion vacuole. The morphology of this vacuole was analysed by confocal microscopy on immuno-labelled cultures at selected time points during the growth curve. Both urogenital strains showed similar inclusion formation and shape (Fig. 2). Initially, inclusions were small and round, located next to the nucleus and increased substantially in size by 36 h PI. Multiple inclusions within a single cell were observed for both strains (Fig. 2). Inclusions were observed to occupy substantial cellular space around the nucleus at 36 h PI onwards. A variety of inclusion morphologies were observed, most often crescent shaped, around the host cell nucleus. Similarly the ocular isolate C. pecorum IPTaLE formed small round inclusions at 20 h PI with an increase in inclusion size observed between 20 and 36 h PI (Fig. 2). In vitro susceptibility testing of koala C. pecorum and C. pneumoniae isolates to chloramphenicol and tetracycline. We measured the in vitro susceptibility of koala C. pecorum and C. pneumoniae isolates to chloramphenicol and tetracycline. The MICs of tetracycline and chloramphenicol were comparable for both koala C. pecorum and C. pneumoniae isolates, with all isolates being susceptible to ≤ 0.25 μ g/ml tetracycline and ≤ 1.00 μ g/ml chloramphenicol (Table 1). Both antibiotics were chlamydicidal for all isolates at their determined MIC values were all within one two-fold dilution across all isolates tested (Table 1). JO146 addition during the replicative phase of the Chlamydia developmental cycle results in significant loss of infectious progeny. As described previously, JO146 is effective in the mid-replicative phase of the Chlamydia developmental cycle 21 . Mid-replicative phase for the C. pecorum isolates was demonstrated to be 16 h PI (see Fig. 1). This was estimated based on the observed peak in EB production at approximately 36 h. Mid-replicative phase for the C. pneumoniae LPCoLN was determined from data previously published 24 . Infectious progeny (or inclusion forming units; IFU/ml) was determined by passaging the cultures at completion of the developmental cycle after treatment. As shown in Fig. 3, JO146 treatment resulted in complete or highly significant (p < 0.0001) loss (up to 10 3-4 log) of progeny (IFU) when added during the replicative phase. In addition, there was a considerable amount of variability in susceptibility of the isolates at higher doses of JO146. For example, C. pecorum MarsBar had 10 2 IFU/ml at 150 μ m JO146, while a similarly urogenital clinical isolate C. pecorum AWH4 had a 10 5 IFU/ml at 150 μ m. The activity was most effective at higher doses, with 150 μ M being lethal to C. pecorum AWH7 (Fig. 3A), C. pecorum PM15 (Fig. 3B), and C. pneumoniae LPCoLN (Fig. 3C). JO146 treatment is most effective when maintained in the culture throughout the replicative and transition to infectious phase of the developmental cycle. In order to understand the chlamydial developmental cycle factors that impact on treatment efficacy, we conducted further in vitro experiments varying the length of JO146 treatment. Firstly, we tested short exposures limited to the peak replicative phase, and secondly, extended exposures to ensure the compound is not simply slowing the developmental cycle. Previously, treatments of C. trachomatis D/UW-3/Cx showed JO146 to be most effective when maintained in the culture throughout the replicative and transition to infectious developmental cycle phase (elementary bodies) 21 .
We added JO146 to cultures of C. pecorum IPTaLE and C. pneumoniae LPCoLN at 16 h PI and subsequently removed it from the culture by extensive washing at 24 h PI. We determined Chlamydial viability at 44 h PI (C. pecorum IPTaLE) and 72 h PI (C. pneumoniae LPCoLN; Fig. 4) and found that JO146 was only slightly effective on C. pecorum IPTaLE when washed out 8 h after treatment (24 h PI) with only a 1-log reduction in viability. The compound had a greater effect on C. pneumoniae LPCoLN when washed out 8 h after treatment (24 h PI), with a 2-log reduction in viability. The ability of these isolates to recover after treatment with JO146 for 8 h demonstrates a need for the compound to be maintained in culture throughout the replicative phase and into the infectious phase of the developmental cycle.
To examine the long term stability of the drug, we determined if infectivity can be rescued by extended culturing after exposure to the drug. We added JO146 to cultures of C. pecorum IPTaLE and C. pneumoniae LPCoLN at 16 h PI and measured infectious progeny at 48, 68, and 88 h PI (C. pecorum IPTaLE) and 70, 90 and 110 h PI (C. pneumoniae LPCoLN; Fig. 5). JO146 treatment at 16 h PI was still effective for C. pecorum IPTaLE after extended culturing and was not notably different between the 48-88 h PI exposures. JO146 treatment at 16 h PI for C. pneumoniae LPCoLN however, was lethal for chlamydial infectivity, even after extended culturing.

JO146 treatment blocks chlamydial inclusion vacuole expansion.
Using confocal microscopy, we monitored inclusion vacuole size and inclusion numbers after 150 μ M JO146 was added to McCoy B grown C. pecorum IPTaLE and HEp-2 grown C. pneumoniae LPCoLN cultures at 16 h PI. As shown in Fig. 6, C. pecorum IPTaLE and C. pneumoniae LPCoLN treated cultures did not increase in size compared to control (DMSO) cultures, and at 40 h PI, C. pecorum IPTaLE inclusions treated with 150 μ m JO146 were 62% smaller than the control while C. pneumoniae LPCoLN inclusions treated with 150 μ m JO146 at 50 h PI were 83% smaller than control (Fig. 6). In addition, the number of inclusions in the JO146 treated cultures was fewer than that of control for both isolates, suggesting treatment induces loss of inclusions from the cultures (Fig. 6).
In addition, we examined C. pecorum cultures by western blot for the HtrA protein compared to β -actin controls. Neither the JO146 treatment or DMSO control treatments had any impact on β -actin levels. However, there was less HtrA protein in the JO146 treatment compared to the DMSO controls 12 (28 h PI) and 26 hours (40 h PI) after the 16 h PI addition of JO146 (see Supplementary Fig. S2).

JO146 is not cytotoxic for koala peripheral blood mononuclear cells. Previous studies by our lab-
oratory demonstrated that JO146 had no detectable toxicity as measured by cell lysis or metabolic turnover in mouse or human cell lines 21 . To ensure this was also the case for koala cells, we treated fresh koala peripheral blood mononuclear cells (PBMCs) with 100 μ M JO146 or DMSO control and assayed cytotoxicity. Cell lysis was assessed using the lactate dehydrogenase assay and metabolic turnover was monitored using the MTS

JO146 is not cytotoxic for koala endometrial and conjunctival cells when cultured ex vivo.
To determine if JO146 could be a potential therapeutic treatment for koala chlamydiosis, it is essential that the cytotoxicity of the compound be examined directly on koala epithelial cells. As no stable koala cell line currently exists for use in in vitro studies, we established primary cell lines for use in this study. We isolated koala primary endometrial and conjunctival cells from epithelium of the koala endometrium and conjunctiva, respectively. We assayed these primary epithelial cells for cytotoxicity after treatment with 100 μ M JO146 and DMSO controls and found that JO146 was not cytotoxic to koala primary endometrial cells as evident by the lack of cell lysis compared to control. Metabolic turnover was reduced by approximately 23% for koala primary endometrial cells treated with 100 μ M JO146 while only 10% reduction in metabolic activity was detected for the DMSO treated control (Table 2). Similarly for the koala primary conjunctival epithelial cells we observed no toxicity (no cell lysis), but metabolic turnover was reduced by approximately 4% (100 μ M JO146 and DMSO control was reduced by 1%) ( Table 2).

Discussion
This study presents the first stage of development of a new therapeutic for koala chlamydiosis. With an increasingly limited availability of the front-line antibiotic commonly used to treat koala chlamydiosis, a new treatment is urgently needed. We have demonstrated significant reduction (and in some cases, lethality) in infectious progeny of Chlamydia from koala clinical cases of disease using both in in vivo and ex vivo cell cultures.
In vitro susceptibility patterns of antibiotics across chlamydial spp. and hosts appear uniform [25][26][27] , and indeed, our results for susceptibility of koala C. pecorum isolates appear consistent with activity at or below the range previously reported (1-2 μ g/ml) 28 . These results indicate that the plasma concentrations of chloramphenicol reached in the koala (reaching a peak plasma concentration of 3.02 μ g/ml on day 1 of treatment, and 4.82 μ g/ml by day 15 12 ) are sufficient for treatment. However, for koalas affected by severe chlamydial disease, antibiotics alone are not sufficient to cure the clinical symptoms 12 . In addition, production of the commonly used antibiotic chloramphenicol is limited. Although there is a potential commercial alternative treatment (Florfenicol), pharmacokinetics, efficacy or safety studies for this drug in koala chlamydiosis has not been carried out 28 . Therefore, alternative treatment strategies are required that not only treat the disease, but also limit side effects given the unique biology of the koala.
Very little in vitro growth data is available for koala C. pecorum or any strains of this species. Prior to assessing the efficacy of our therapeutic agent, it was necessary to analyse the growth characteristics of ocular and urogenital koala C. pecorum strains. Growth curves have previously demonstrated that human C. trachomatis ocular and genital strains have quite distinct kinetics, with the ocular strains generally considered to be slow growing strains 29 . We tested three koala C. pecorum strains from ocular and urogenital sites and found that there were differences in the observed growth kinetics. Specifically, the C. pecorum IPTaLE (ocular) isolate has a longer lag phase and lower overall yield of infectious progeny (EBs) compared to the others. Genetically, the three strains are > 98% identical with variances limited to two regions; the plasticity zone and an area encoding polymorphic membrane proteins, where SNPs were found to cause premature stop codons in six C. pecorum genes 23 . Expanded in vitro and genetic studies of additional clinical isolates will be required to establish whether genital and ocular koala C. pecorum strains share similar tropisms to their human C. trachomatis counterparts.
As a first step to evaluating the efficacy of our HtrA protease inhibitor JO146, we treated cultures during the replicative phase (estimated to be ~16 h PI) of the koala C. pecorum and C. pneumoniae developmental cycle. This treatment resulted in significant loss of infectious progeny for many koala C. pecorum isolates and lethality for two koala C. pecorum isolates and koala C. pneumoniae. Analysis of the chlamydial cell morphology during treatment supported this loss of infectious progeny, with inclusion size failing to progress (and in some cases diminishing) within the treated cultures. As previously shown by our laboratory, mid-replicative phase addition of JO146 is detrimental to many species of Chlamydia including C. trachomatis D, C. trachomatis L2, C. pecorum IPA, C. suis and C. caviae and in many host cell types 21,30 . Here we show that JO146 treatment of C. pecorum IPTaLE and C. pneumoniae LPCoLN was most effective when maintained in the culture throughout the replicative and transition to infectious phase of the developmental cycle. This indicates that lead drug candidates will need to be optimized for tissue stability and bio-availability with overall extended half-life in vivo.
The removal of JO146 at 24 h PI (8 h after treatment) showed only a 1-log (C. pecorum IPTaLE) and 2.5-log (C. pneumoniae LPCoLN) reduction in infectious progeny. However, when the compound was maintained in the media throughout the replicative phase, a 3-log (C. pecorum IPTaLE) and complete loss (C. pneumoniae LPCoLN) of infectious progeny resulted. Importantly, with extended culture of C. pecorum IPTaLE (68 and 88 h PI), there was no rescue of infectious progeny and C. pneumoniae LPCoLN was similarly not able to recover at 90 and 110 h PI. This result suggests a sustained effect of JO146 treatment and further supports its activity as being primarily bactericidal. This result could also suggest that the targets of JO146 (HtrA most likely but possible other serine proteases) are required for not only the replicative phase but also for the differentiation from reticulate bodies to the infectious elementary body form of Chlamydia.
Our previous research using JO146 supported that HtrA is essential for Chlamydia trachomatis serovar D during the replicative phase of the developmental cycle 22 . These previous data suggested that Chlamydia HtrA functions in critical protein maintenance and assembly; possibly for vital outer membrane virulence factors, based on some of our biochemical data 31,32 . This means that for Chlamydia at least HtrA is an essential virulence factor in addition to a well-known protein stress response function in many bacteria. However, this is not necessarily unexpected 33 , as HtrA is also an essential virulence factor for Legionella 33 .
A key finding of our study was the ability of JO146 to kill Chlamydia in koala primary ex vivo-cell cultures. There are no published reports of culturing Koala clinical isolates of Chlamydia in primary ex vivo koala cells and this method represents an important step forward in the context of an urgent need for new antibiotics. JO146 significantly reduced the number of infected cells in cultures of C. pecorum isolates in koala primary endometrial cells. Unfortunately, the primary cultures were not able to be maintained and amplified sufficiently to passage the Chlamydia and measure the infectious progeny on repeat infections. Due to limits in the availability of freshly necropsied koala tissue, additional replicates of this experiment could also not be performed. Nevertheless, the results of this single ex vivo treatment of C. pecorum infections with JO146 are consistent with those obtained in vitro.
Importantly, cytotoxicity assays performed on freshly harvested koala PBMCs, endometrial and conjunctival cells confirmed overall low toxicity of JO146. We did observe a measurable reduction in proliferation induced by JO146 in all ex vivo primary cell cultures. However, it is hard to evaluate the implication of this result, given that the cells were not able to be maintained beyond three passages in all experiments attempted. This reduction may not occur in more stable primary ex vivo cells, which may or may not be relevant in vivo, hence further studies are required to assess the implications of this result.
Overall, our studies of JO146 have thus far shown that it can significantly reduce the infectious progeny of clinical isolates of koala chlamydial. The compound is lethal to many clinical isolates and induced several log reductions in the remaining isolates. The compound is non-cytotoxic to koala PBMCs, conjunctival, and endometrial primary ex vivo tissue, maintains stability in culture extending beyond the Chlamydia infectious cycle, and remained bactericidal throughout treatment. Although these results are promising, this compound in its current formulation, is not completely lethal to all isolates, therefore continued optimization will be required. Further studies are currently underway to generate more potent derivatives of JO146. This is the first report of a protease inhibitor successfully being applied in vitro and ex vivo to treat koala C. pecorum and C. pneumoniae, and validates JO146 or other inhibitors derived from JO146 as possible options for treating koala chlamydiosis. Here we have provided the basis for future studies investigating the efficacy of JO146 and further derivatives in vivo.

Methods
Ethics Statement. Koala Chlamydia pecorum strains isolated in this study were collected from koalas presenting to Australia Zoo Wildlife Hospital and Port Macquarie Koala Hospital as a part of their routine veterinary care and diagnosis. The use of these samples for Chlamydia culturing was considered by the University of the Sunshine Coast Animal Ethics Committee (USCAEC) and determined to be exempt from requiring further approval (AN/E/14/01). Koala tissues for primary epithelial cell culture were collected from two koalas,  Table 3 were all collected from captive and/or wild koalas located in Queensland and New South Wales, Australia. Swab samples were collected and stored in SPG transport media at point of care and frozen at − 80 °C. The koala C. pecorum isolates were propagated in McCoy B cells and passaged no more than five times. All isolates have been confirmed by 16S RNA PCR and sequencing to be positive for C. pecorum 8 .
Chlamydia culture. C. pecorum and C. trachomatis isolates were routinely cultured in McCoy B cells on DMEM, 10% Fetal calf serum (FCS), 0.1 mg/ml streptomycin, and 0.05 mg/ml gentamycin. C. pneumoniae was routinely cultured in HEp-2 cells on DMEM, 10% FCS, 0.1 mg/ml streptomycin, and 0.05 mg/ml gentamycin. Inhibitor experiments were routinely conducted in 96-well plates seeded with 20,000 host cells per well 24 h prior to chlamydial infection. All strains were cultured in 37 °C incubators with 5% CO 2 and, unless otherwise stated, all infections were routinely conducted at a Multiplicity of Infection (MOI) of 0.3. The Inclusion Forming Units (IFU) was determined from cultures harvested at the completion of the developmental cycle during which inhibitor treatment was conducted (time of harvest is indicated on the figure). Briefly, JO146 and DMSO control (JO146 solvent), at doses indicated on the graphs (always at a 1 in 1000 dilution factor), was added at 16 h post infection (PI) and left in the cultures until completion of the developmental cycle (or at time points indicated). Harvested cultures were lysed by vigorous pipetting and serially diluted onto fresh monolayers and fixed and stained at 30 h PI for enumeration of IFU/ml 21 . The compound JO146 was commercially sourced as previously described 30 . Antimicrobial Susceptibility Testing of Koala Chlamydia isolates. Antimicrobial susceptibility testing of koala C. pecorum and C. pneumoniae isolates in McCoy B and HEp-2 cell lines respectively, was performed as previously described 35 with chloramphenicol and tetracycline on 96-well microtiter plates without passage (MIC) and by one passage (MCC). The inoculum size of infectious Chlamydia for all MIC and MCC comparisons was 10,000 IFU/well. The MIC was defined as the concentration of antibiotics one two-fold dilution more concentrated than the MIC TP (where MIC TP is defined as the concentration of antibiotics where 90% of inclusions or greater displayed altered size or morphology). The MCC is defined as the lowest concentration of drug that produces no morphologically normal inclusions by one freeze-thaw passage in 96-well microtiter plates.
Microscopy. C. pecorum and C. pneumoniae cultures were examined using immunofluorescence with a Leica SP5 Confocal microscope with a 63× oil objective. Coverslip cultures were washed with PBS and fixed with 100% methanol for 10 min. Cultures were stained with fluorescein isothiocyanate (FITC)-labelled Chlamydia-specific anti-lipopolysaccharide (anti-LPS) monoclonal antibody and host cells stained with Evan's blue (Cell labs, Australia) and then mounted on slides in Prolong Gold (Invitrogen) prior to visualization.

Isolation of koala Peripheral Blood Mononuclear Cells. Koala Peripheral Blood Mononuclear Cells
(PBMC's) were isolated as previously described 36 . Briefly, the PBMC fraction was isolated by Ficoll-Paque density gradient centrifugation and washed twice with phosphate-buffered saline (PBS). The pellet was then resuspended in RPMI (10% FCS, 0.1 mg/ml streptomycin, and 0.05 mg/ml gentamycin) before seeding into 96-well plates 12 h prior to use in cytotoxicity assays. All wells were stimulated with a final concentration of 100 μ g/ml Phorbol Myristate Acetate (PHA) at this time.
Isolation of koala primary epithelial cells. Tissue from the conjunctiva and endometrium were collected from two koala's euthanized due to illness at Australia Zoo Wildlife Hospital. Harvested tissue was placed in HBSS media with 0.2% collagenase D and a homogenate prepared. The tissue was washed twice with HBSS and further  Table 3. Koala C. pecorum and C. pneumoniae isolates used in this study.
treated with 2 U/ml DNase, with 10% FCS added to stop the DNase activity. The tissue was then spun and the pellet resuspended in red blood cell lysis buffer before being washed twice with complete keratinocyte-SFM and finally resuspended in complete keratinocyte-SFM, 10% fetal calf serum, 0.1 mg/ml streptomycin, 0.25 mg/ml Fungizone and 0.05 mg/ml gentamycin. Cells were passaged no more than two times before use in cytotoxic assays and JO146 efficacy assays.

Cytotoxicity and Proliferation Assays. Koala PBMC's and primary conjunctival and endometrial cells
were assayed for cytotoxicity after treatment with JO146 and DMSO controls using the lactate dehydrogenase assay for cell lysis (CytoTox 96 ® Non-Radioactive Cytotoxicity Assay, Promega). Metabolic turnover was monitored using the MTS incorporation assay (CellTiter 96 ® AQueous Non-Radioactive Cell Proliferation Assay, Promega). PBMCs were collected from anticoagulated blood from four different koalas. Epithelial cells were extracted from endometrial tissue of two koalas for primary culture. Conjunctival epithelial cells were extracted from conjunctival tissue of one koala for primary culture. DMSO control was used at the same volume of DMSO as that added for 100 μ M JO146 treatments. PBMC's were stimulated with a final concentration of 100 μ g/ml PHA at 12 h prior to treatment with JO146.