Growth kinetics of Chlamydia trachomatis in primary human Sertoli cells

Chlamydia trachomatis (Ct) is the leading cause of bacterial sexually transmitted infections worldwide and has been associated with male infertility. Recently, it was hypothesized that Ct may infect the epithelium of the seminiferous tubule, formed by Sertoli cells, thus leading to impaired spermatogenesis. To date, there is a lack of data on Ct infection of the seminiferous epithelium; therefore, we aimed to characterize, for the first time, an in vitro infection model of primary human Sertoli cells. We compared Ct inclusion size, morphology and growth kinetics with those in McCoy cells and we studied F-actin fibres, Vimentin-based intermediate filaments and α-tubulin microtubules in Sertoli and McCoy cells. Our main finding highlighted the ability of Ct to infect Sertoli cells, although with a unique growth profile and the inability to exit host cells. Furthermore, we observed alterations in the cytoskeletal fibres of infected Sertoli cells. Our results suggest that Ct struggles to generate a productive infection in Sertoli cells, limiting its dissemination in the host. Nevertheless, the adverse effect on the cytoskeleton supports the notion that Ct may compromise the blood-testis barrier, impairing spermatogenesis.


Results
In vitro model of primary human Sertoli cells. We first analysed the growth behaviour of the human cell line selected for analysis. The primary human Sertoli cells were frozen at the 3 rd passage after isolation and were cultured up to the 9 th passage on 80 cm 2 or 175 cm 2 cell culture flasks. Cells were passaged every 7 days at a confluence level of 70-80%. The average yield on 80 cm 2 flasks was 7.25 × 10 5 ± 3.9 × 10 4 cells/flask up to the 6 th passage, where upon it decreased by half by the 9 th passage (data not shown). Therefore, all our experiments were carried on primary human Sertoli cells at the 5 th passage after isolation. Primary human Sertoli cells grown on coverslips possess a thin and wide cytoplasm with irregular morphology and are approximately ten times the size of McCoy cells.  www.nature.com/scientificreports www.nature.com/scientificreports/ Growth kinetics of C. trachomatis in primary human Sertoli cells. To characterize the growth phenotype of C. trachomatis in primary human Sertoli cells, we sampled multiple, regular time-points across the duration of chlamydial developmental cycle for infectivity, DNA replication and inclusion size.
The one-step infectivity growth curve showed a relatively long eclipse period for C. trachomatis in human Sertoli cells, with infectious EBs appearing exclusively after 36 hours post infection ( Fig. 2A). By contrast, chromosomal DNA replication, measured by qPCR, showed a constant and gradual increase in genome copy number over the course of the experiment (Fig. 2B), not correlating with the sharp increase in infectivity after 36 hpi. The generation time for C. trachomatis in primary human Sertoli cells was 5 hours and 34 minutes.
The 'inclusion size' growth curve showed chlamydial inclusions growing gradually up to 32 hpi, whereupon it showed a sudden increase in size after 36 hpi, as evidenced by Fig. 2C,D.

C. trachomatis lytic exit from primary human Sertoli cells.
To investigate the exit of C. trachomatis from primary human Sertoli cells, we observed infected cells every 24 up to 96 hpi.
Remarkably, C. trachomatis did not exhibit the ability to exit from host cells in primary human Sertoli cells, revealing the lack of cell lysis in Sertoli cells at 96 hpi (Fig. 3A). In fact, the one-step infectivity growth curve showed that the majority of infectious EBs were retained in the Sertoli cell monolayer (Fig. 3B). Furthermore, the number of infectious EBs decreased after 48 hpi in both Sertoli cell monolayer and culture medium (p < 0.01). By contrast, the qPCR assay demonstrated continuous chromosomal DNA replication up to 96 hours post infection.
In McCoy cells, C. trachomatis infectivity constantly increased from 48 to 96 hpi and infectious EBs were released into the culture medium, as shown in Fig. 3B. As expected, the qPCR assay showed the same trend as seen in primary human Sertoli cells.

Confocal analysis of Sertoli cell cytoskeleton following C. trachomatis infection. In Sertoli cells,
we found that the morphology of stress fibres and cortical actin in infected cells was not overly affected by C. trachomatis, whereas we observed the re-organization of Vimentin-based IFs and microtubules in thick fibres surrounding the inclusion (Fig. 4A).
By contrast, in McCoy cells we observed the recruitment of F-actin fibres to form a ring-shaped structure around the inclusion, as well as a similar restructuring of Vimentin-based IFs (Fig. 4B). However, the microtubule network was not perturbed by the presence of C. trachomatis inclusions. Infected primary human Sertoli cells were removed for analysis at 4 hourly intervals and then titrated to assess the quantity of IFUs and the relative number of chlamydial genomes by qPCR. Values are expressed as means ± SD of four replicates from two independent experiments. Chlamydial inclusions were visualized by indirect immunofluorescence with species-specific anti-MOMP monoclonal antibodies (Mab6ciii). Representative images of ten chlamydial inclusion per time-point are shown. Mean and standard deviation of inclusion size, expressed as square pixels, were measured from fluorescence micrographs using ImageJ software.

Discussion
In this study we characterized a novel in vitro infection model of C. trachomatis in primary human Sertoli cells. Our main finding demonstrated the ability of C. trachomatis to infect and replicate in primary human Sertoli cells, although with a much lower efficiency than in McCoy cells, which are considered to be the gold standard of Chlamydia laboratory culture because of their permissiveness to all strains of C. trachomatis 21 .
More importantly, the growth kinetics of C. trachomatis in primary human Sertoli cells growing under optimal conditions revealed a distinct profile, namely the very long eclipse period and the late appearances of infectious EBs toward the end of the developmental cycle. This was accompanied by a steady increase in genomic DNA copy number during the entire duration of the C. trachomatis developmental cycle, as evidenced by the qPCR analysis. By contrast, in McCoy cells the exponential phase of the infectivity growth curve is known to happen much earlier, at 20-22 hours post infection 22,23 . A possible explanation of this phenomenon may lie in a delayed differentiation of chlamydial RBs, that keep replicating as suggested by the qPCR, possibly the consequence of a hostile cellular environment that may partially arrest or contain chlamydial intracellular growth.
Further evidence that C. trachomatis struggles to infect and develop inside human Sertoli cells is also provided by the absence of cell lysis at the end of the developmental cycle, as evidenced by the retention of the majority of EBs within host cells, and by the constant decrease in the number of infectious EBs after 48 hours up to 96 hours post infection. This is intriguing and may suggest that C. trachomatis infection in Sertoli cells might lead to a persistent or chronic state.
All the above evidence suggests that C. trachomatis, despite being able to generate a productive infection of human Sertoli cells, is limited in its further dissemination within the human testis and potentially leads to long-term damage of Sertoli cell function and structure. www.nature.com/scientificreports www.nature.com/scientificreports/ In fact, by analysing the subcellular localization and integrity of the main components of cell cytoskeleton, namely F-Actin fibres, Vimentin-based intermediate filaments and α-tubulin microtubules, we observed that the development of a C. trachomatis inclusion into a human Sertoli cell led to alteration in its cytoskeleton. Recent studies have shown that actin and microtubule-based cytoskeleton in Sertoli cells plays a crucial role to preserve the homeostasis of the BTB 20 . Our observation that Vimentin-based intermediate filaments and α-tubulin microtubules were re-organized in thick fibres surrounding the chlamydial inclusion, hints at the intriguing possibility that C. trachomatis might adversely affect the integrity of the BTB, and, hence, impair the spermatogenesis. Nevertheless, to confirm our preliminary observations, a polarized three-dimensional cell culture model of human Sertoli cells infected with C. trachomatis is needed for a more in-depth investigation of the different junction types forming the BTB.
The main strength of our study lies in the design of an in vitro infection model that utilizes a primary human Sertoli cell line, for better mimicking the physiology of their in vivo counterpart, infected with C. trachomatis serovar D, one of the most prevalent strains in men with urethritis [24][25][26] . In addition, to address potential variations between different cell preparations, we utilized a well characterized and commercially available primary human Sertoli cell line with optimised growth media and culture conditions. To minimize the variability of results, all our experiments were performed at the 5 th passage after isolation, following the observation that primary human Sertoli cells underwent significant morphological changes after the 7 th or 8 th passage.
Our future goal will be to set up a C. trachomatis infection model using a permanently established human Sertoli cell line to be compared to primary cells; this cell line will be widely shared to encourage research in this field. Furthermore, it will be helpful to investigate the morphology and development of C. trachomatis inclusions in Sertoli cells through more advanced imaging techniques such as scanning electron microscopy, to clarify whether the slow developmental cycle and the inability to exit host cells might be a mechanism of persistence.  www.nature.com/scientificreports www.nature.com/scientificreports/ The primary human Sertoli cell line that we utilized was purchased from Lonza, USA (product code MM-HSE-2305, lot number 360806141) and was harvested on 14th June 2011. Sertoli cells were cultured in Dulbecco's Modified Eagle Medium/Ham's Nutrient Mixture F12 (1:1), with L-glutamine and HEPES (DMEM/ F12, Gibco ™ , USA), supplemented with 10% (v/v) FCS at 37 °C in humidified atmosphere with 5% CO 2 .

Methods
Propagation and titration of C. trachomatis. C. trachomatis serovar D strain UW3 ATCC VR-855 was propagated in McCoy cells as previously described 27 . Briefly, confluent McCoy cell monolayers grown in 25 cm2 flasks were infected with chlamydial EBs by centrifugation at 754xg for 30 min and harvested by scraping after 36 hours of incubation. The suspension containing Chlamydial EBs was, then, added to equal volumes of 4X Sucrose Phosphate (4SP) buffer and stored at −80 °C.
For measuring C. trachomatis infectivity, confluent McCoy cell monolayers were infected with 10-fold dilutions of C. trachomatis EB suspension and, 36-hours post-infection, were fixed in 96% ice cold methanol. Chlamydial inclusions were blue-stained by using a mouse monoclonal antibody against genus-specific chlamydial LPS (Mab29, The Chlamydia Biobank, UK) followed by an anti-mouse antibody conjugated with β-galactosidase (Millipore, USA) and incubated with a X-Gal staining solution as previously described 27 . Blu-stained inclusions were visualized and counted by using a bright-field inverted microscope (200X magnification). Infectivity assay. To assess infectivity at each time-point for the one-step growth curve and the inclusion lysis time-course in primary human Sertoli cells, C. trachomatis EB suspensions were titrated as above described.

Efficiency of C. trachomatis infection in Sertoli and
Real-time quantitative PCR analysis. Chromosomal DNA quantification at each time point of the one-step growth curve and the inclusion lysis time-course was determined as previously described 28 . The primers CM_ omcB_F (5′-GGAGATCCTATGAACAAACTCATC-3′), CM_omcB_R (5′-TTTCGCTTTGGTGTCAGCTA-3′) and the probe CM_omcB_Probe (5′-FAM-CGCCACACTAGTCACCGCGAA-TAMRA-3′) were used for amplifying a conserved region in the omcB gene. The qPCR was performed in a VIIA7 Real-Time PCR system (Applied Biosystem).

Confocal microscopy. Primary Sertoli cells grown at 60% confluence on glass coverslips in 24 well trays
were infected with C. trachomatis at a MOI = 1.0 in the absence of cycloheximide, fixed in 4% paraformaldehyde and permeabilised using saponin buffer. Then, cell monolayers were stained for confocal microscopy as previously described 27  Cat. No. A12381) was used for high-affinity labelling of F-Actin microfilaments. Nuclei were counterstained with 1 µg/mL DAPI (Fisher Scientific, USA). Images were captured using a Leica TCP SP8 confocal microscope at 1000X magnification. Confocal image processing and inclusion size measurement from fluorescence microscope images (µm2) were executed in ImageJ software (NIH, USA, version 1.8.0_112).

Statistical analysis.
All values are expressed as means ± standard deviation (SD) of two to four replicates from at least two independent experiments. Comparisons of means were performed by using a two-tailed Student t-test for independent samples. The single or multiple inference significance level was set to 5%. The C. trachomatis infection efficiency in Sertoli and McCoy cells, the one step growth curves, the chlamydial inclusion lysis time-course and the chromosomal replication graphs, as well as all statistical calculations, were produced in GraphPad Prism software (GraphPad Software, USA, version 7.0.3.0).

Data Availability
The datasets generated and/or analysed during the current study are available from the corresponding author on request.