Vaginal Bacteria Elicit Acute Inflammatory Response in Fallopian Tube Organoids.

The study was approved by the institutional review board at University of Washington. We obtained informed written consents from each patient before her surgical procedure. FT tissues from 4 patients were collected in the operating room during salpingectomy for benign gynecological indications. None of these patients had overt signs of infections or tubal diseases at the time of surgery. The FT tissues were minced and digested with collagenase type I (Sigma) for one hour at 37°C. The dissociated mucosal cells were seeded and cultured to organoids in Matrigel following the published protocols with modifications 4 . Briefly, the cells were cultured in 2D culture media with progenitor growth factors as previously published 4 at 37°C, 5% CO2 in a humidified incubator until 70–80% confluent. The epithelial cells were then detached and seeded in 75–80% Matrigel (Corning). The Matrigel domes were overlaid with 3D expansion medium as previously described 4 and incubated at 37°C, 5% CO2 in a humidified incubator until the time for passage which on average takes 14–21 days. For the passage of organoids, after the Matrigel was dissolved with cold PBS, mechanical fragmentation was achieved by vigorous pipetting. Sheared organoids were centrifuged and resuspended in cold Matrigel and replated at the desired density. Organoid reformation were usually observed within 2–3 days after the passage. The culture medium was changed every 3–4 days. IHC was perfomed as described previously 7 . FT organoids were released from the Matrigel with cold 4% paraformaldehyde (Fisher Scientific) and fixed for 2 hours before being embedded in Specimen Processing HistoGel ™ (Thermo Fisher). The organoids encapsulated in the Histogel were dehydrated in an ascending series of alcohol, followed by isopropanol and acetone at room temperature. The dehydrated organoids were paraffin-embedded and cut into 5 μm sections on a microtome. Hematoxylin & Eosin (H&E) staining and IHC were carried out following standard procedures. The primary antibodies used in IHC included: Mouse anti-E Cadherin (BD, 610181); Rabbit anti-Pax8 (ProteinTech, 10336-1-AP); Mouse anti-acetylated Tubulin (Sigma, T7451). Lactobacillus crispatus DNF00082 and Fannyhessea vaginae DNF00720 were inoculated from frozen stock onto Brucella agar supplemented with 5% laked sheep blood, hemin and Vitamin K (BRU) (Hardy Diagnostics, Santa Maria, CA) and grown at 37°C in anaerobic conditions (90% N2, 5% H2, 5% CO2). Isolates were subcultured twice on BRU or in modified peptone-yeast-glucose medium supplemented with 1% yeast extract and 1% glucose (PYG-mod-YG). A final culture for each isolate of 20mL PYG-mod-YG was centrifuged and pelleted and the pellets were reconstituted with 400 to 500 μL of PYG-mod-YG. Serial dilutions were made from 10 μL of reconstituted pellet and plated onto BRU to count colony forming units (CFU). The remaining reconstituted pellet was frozen with glycerol to make a 10% glycerol stock solution. CFU were counted once colonies had grown. Based on the final CFU count, portions of the 10% glycerol stock solutions were diluted with PYG-mod-YG to make 12 μL aliquots of 10 6 CFU each. The aliquots were kept frozen at −80°C until they were inoculated into the organoids. Uninoculated aliquots of 12 μL PYG-mod-YG were also frozen with the isolate aliquots as negative controls. The 10% glycerol stock solutions were plated again for CFU counts on the same day as organoid inoculation to verify isolate viability after freezing. The FT organoid culture media was changed to antibiotics-free media 24 hours before the acute infection experiment, and the media remained antibiotics-free during the following experiment. The FT organoids from each patient (n=4 patients) was inoculated with 10 6 CFU of Lactobacillus cripspatus or Fannyhessea vaginae per well (n=3 replicates for each bacteria strain for 3 patients, n=1 for each strain for the first patient). Three negative controls (n=3 replicates for each control) included: PBS inoculation, sterile bacterial suspension media innoculation, and organoids alone without any innoculation. After 24 hours of incubation, the organoids were released from Matrigel, and dissociated into single cells as described below before RNA extraction. Total RNA was extracted using Trizole (Life Technologies) according to the supplier’s protocol. The RNA quality was assessed using Tapestation and the RNA quantity was measured with Qubit fluorometer. All samples had RNA integrity number (RIN) > 9.0. Each sample was run on a lane of nCounter Inflammation Panel (Nanostring). The data was analyzed using the Nanostring nSolver analysis software 4.0 following the manufacturer’s instructions. After being released from the Matrigel with cold PBS, the FT organoids were dissociated into single cells with dispase (at 37°C for 30 minutes) and 0.25% trypsin (at 37°C for 2 minutes) digestions. After viability check, the cell suspension was filtered through 40 μm filter to get single cells for flow cytometry. Control samples were obtained from previously collected peripheral blood mononuclear cells (PBMC) which were isolated by Ficoll Histopaque (Pharmacia Biotech) gradient centrifugation at a density of 1.077 g/mL and cryopreserved in dimethylsulfoxide (DMSO). Dissociated organoids and control PBMCs were labeled with a fluorescent marker for viability (LIVE/DEAD Violet, Invitrogen) and anti-human CD45 (FITC conjugated, BD Biosciences cat#555482). The cells were fixed (Fix/Perm Buffer, BD Biosciences) and analyzed by flow cytometry using a FACS Canto II. The frequency of the CD45+ population was determined by percentage of total live cells (FlowJo).

From FT tissues collected from consented patients at the time of surgery, we have established robust FT organoid cultures which can be repeatedly passaged for months while maintaining the same morphological and histological appearance ( Figure 1 ). Hematoxylin and eosin (H&E) staining and immunohistochemistry demonstrated luminal structures with single-layer columnar epithelium consisting of ciliated and secretory cells (positive for Tubulin and PAX8 cell markers, respectively) ( Figure 1 ). The ciliated cells are apical with the cilia facing the lumen, away from the basal membrane (positive for E-cadherin) ( Figure 1 ). We introduced acute infection in the FT organoid culture system by inoculating the organoid culture media with two common vaginal bacterial species, Lactobacillus crispatus ( L. crispatus ) and Fannyhessea vaginae (F. vaginae) . After 24 hours of culturing with or without either bacterial species, no notable morphological changes were observed among the three groups ( Figure 2 ). RNA was extracted from the organoids and the human gene expression profiles were analyzed using Nanostring Inflammatory panels which quantified the expression levels of 249 genes involved in the inflammatory process ( Figure 2 ). Among the 249 genes analyzed, 162 genes were expressed by all organoid samples and 71 genes were not expressed by any organoid sample. Organoids from each individual showed different inflammatory responses, however, the differential expression pattern between the experimental groups and in comparison with the controls can be clearly observed regardless of the inter-individual heterogeneity ( Figure 3 ). Compared to the negative controls that were not cultured with any bacteria, the organoids cultured with either bacterial species showed multiple differentially expressed inflammatory genes ( Figure 4 ), indicating the epithelial cells in the organoids can respond to acute bacterial infection and generate inflammatory responses. Notably, multiple members of the CC-chemokine ligand (CCL) family were upregulated in both F. vaginae and L. crispatus infected organoids, while multiple genes in the mitogen-activated protein kinase (MAPK) family are downregulated in L. crispatus infected organoids ( Figure 4 ). When the gene expression profiles were compared between the two experimental groups, marked differences were noted between L. crispatus infected organoids and those infected by F. vaginae ( Figure 4 ). Among the most highly differentially expressed genes, transcripts from the CXCL family, including CXCL1,2,3,5,6,10, were highly enriched in F. vaginae infected organoids, which may contribute to leukocyte recruitment. Several other genes, such as TGFB1 or TRADD, are upregulated by L. crispatus ( Figure 4 ). With flow cytometry, we assessed the presence of immune cells during the FT organoid culturing period. On the day of tissue collection, 22% of cells were CD45+ immune cells. On the next day after 24 hours of culture in 2D growth media, only 11% of cells were immune cells. The CD45+ immune cells disappeared rapidly within the first five days of 3D culture to ≤1% and were undetectable at later timepoints (Day 6–8 of 3D culture, <1% of cells were CD45+) ( Figure 5 ). The lack of immune cells in mature organoids (on average 14 days in 3D culture) indicated the inflammatory response observed with bacterial culture was generated by the epithelial cells in the organoids.

By inoculating the culture media with two common vaginal bacterial species, we were able to elicit acute inflammatory responses from the epithelial cells in the FT organoids. Our results indicate the divergence in the inflammatory responses provoked by pathogenic and non-pathogenic vaginal bacterial species. We demonstrate rapid disappearance of immune cells in the FT organoid culture, so this inflammatory response was likely generated directly by the FT epithelial cells. Among the genes upregulated in F. vaginae infected organoids in comparison to the negative controls and to the L. crispatus infected organoids, multiple genes encoding C-X-C motif chemokine ligand (CXCL) family were represented. The CXCL family widely participates in immune cell recruitment, specifically, attracts monocytes and neutrophils, and is involved in the pathogenesis of inflammatory diseases such as endometriosis and its associated infertility 16 – 18 as well as angiogenesis and tumor progression in ovarian and cervical cancer 19 , 20 . C5 that is highly upregulated in the F. vaginae infected organoids is a chemotactic molecule that directly attracts neutrophils, eosinophils, basophils, and monocytes to the sites where complement is activating 21 . Therefore, it is not surprising for the FT organoid epithelial cells to increase the production of these chemokines in reaction to pathogenic bacteria such as F. vaginae . Multiple members of the CC-chemokine ligand (CCL) family were upregulated in both F. vaginae and L. crispatus infected organoids. These chemokines attract monocytes, T cells and eosinophils to inflammatory sites. In a previous publication, when these bacterial species were incubated with vaginal epithelial cell aggregrates, it was found that F. vaginae led to a strong cytokine response including upregulation of CCL20 as seen in our study with FT organoids; however, L. crispatus at a lower concentration elicited much lower and in some cases unmeasurable cytokine release from the vaginal epithelial cells 22 . Even though L. crispatus is a non-pathogenic bacteria in the vagina, it does not normally reside in the FT in high concentration, which may explain why L. crispatus elicited inflammatory response when incubated with the FT organoids in our study, but minimal repsonse when incubated in low concentration with vaginal epithelial cells in previous study. Interestingly, TGFB1, the gene that encodes TGF-β1, is upregulated by L. crispatus , but not by F. vaginae , and TGF-β1 plays an important role in the maintenance of immune homeostasis and self-tolerance by inducing unresponsiveness in naïve T cells and preventing lymphocyte activation 23 . This may indicate a potential mechanism that L. crispatus utilizes to promote the tolerance of normal microbiome in the female genital tract. Another interesting finding was the downregulation of MAPK pathway genes in L. crispatus infected organoids, as MAPK genes were found to be some of the key genes in novel pathways identified in PID associated infertility 24 . To our knowledge, no previous publication has studied the inflammatory response of FT organoids to vaginal bacteria. Our study indicates that FT organoids can be a useful in vitro model system to study the host-pathogen interaction during bacterial infection which may facilitate future mechanistic investigations in PID and its contribution to tubal factor infertility. Some of the technical strengths of this study include: use of three different types of negative controls; little intra-individual variations among replicates; high RNA integrity observed across the samples (RIN > 9); and processing of all samples by the same individuals and on the same Nanostring panels at the same time, thus minimizing batch effects. Inter-individual variations were noted in the inflammatory response to either bacteria; while commonly observed in other organoids and used as a tool for precision medicine 25 – 27 , a much larger sample size would be required to confirm whether consistent individual heterogeneity exists in FT organoid inflammatory responses. Building on the feasibility established in this pilot study, we are developing a more complex FT organoid culture system to better investigate the host-pathogen interaction, such as incorporating the immune cells into the culture, co-culturing multiple bacterial species together to better represent the normal or pathological genital tract microbiome, and evaluating the genome-wide changes in gene expression rather than focusing on the inflammatory genes only. In addition, we are exploring direct microinjection of bacteria into the organoid lumen so that we can better mimic the in vivo FT bacterial exposure. These efforts will lead to a better understanding of the local microenvironmental changes during acute and chronic infections which may lead to tissue damage and subsequent infertility.

In this pilot study, we demonstrate that epithelial cells in patient tissue derived FT organoids respond to acute bacterial infection with upregulation of inflammatory genes specific to each vaginal bacterial species in culture: non-pathogenic L. crispatus or pathogenic F. vaginae. FT organoids is a useful in vitro model system to study the host-pathogen interaction during bacterial infection, such as in PID. Further studies are ongoing to develop a more sophisticated FT organoid culture system to better mimic the FT microenvironmental changes during acute and chronic infections.

The fallopian tube (FT) is responsible for gamete and early embryo transportation, and is the site of fertilization and preimplantation embryonic development in the first 3 to 4 days of human life. As the conduit between the peritoneal and uterine cavity, the FT is a site of pelvic inflammatory disease (PID) in which ascending vaginal bacterial infections affect the uterus, FT and ovaries. Untreated PID can lead to infertility due to tubal damage and occlusion, as well as tubal ectopic pregnancy. Repeated PID increases the likelihood of tubal factor infertility 1 and is also is a risk factor for developing ovarian cancer 2 . Establishing an in vitro model system for FT infection will facilitate mechanistic studies in PID and subsequent tubal factor infertility. Organoids have been shown to preserve the original tissue characteristics in many different types of tissues and have emerged to serve as in vitro models for many diseases 3 . FT epithelial organoids have been derived from FT tissues 4 or differentiated from induced pluripotent stem cells 5 , in which the secretory and ciliated cells are self organized into a single layer of columnar epithelium surrounding a lumen, mimicking the FT epithelium (FTE). Several studies have been published on FT organoids 4 – 7 but few have utilized FT organoids to study infectious diseases, although one study with the emphasis on cancer transformation did establish chronic infection of FT organoids with Chlamydia trachomatis and showed a molecular phenotype associated with aging in epithelial cells after prolonged culture 6 . We sought to establish a FT organoid infection model to study common vaginal bacteria and their interaction with the FT epithelial cells in the organoids. The reasons for choosing vaginal bacteria to study include: 1) In our previous studies, we have shown that vaginal bacteria can ascend into the uterus and become part of the uterine microbiome 8 ; 2) Recent studies demonstrated that vaginal bacteria were identified in the FT in women without overt signs of infection 9 , 10 ; and 3) Some vaginal bacterial species and genera that are associated with bacterial vaginosis (a common vaginal infection caused by bacterial overgrowth of anaerobes) including Fannyhessea vaginae , Sneathia , and Megasphaera , have been isolated from the upper genital tract of people with PID including the FT 11 – 14 . Taken together, this evidence indicates potential physiological or pathological significance of vaginal bacterial species in the upper female genital tract including the FT. Two common vaginal bacterial species were chosen for study: Lactobacillus crispatus ( L. crispatus ) which represents part of the healthy vaginal microbiota, and Fannyhessea vaginae (F. vaginae), which is a pathogen linked to PID 12 and bacterial vaginosis 15 . In this pilot study, we introduced acute infection using non-pathogenic L. crispatus and pathogenic F. vaginae and studied the inflammatory response of the FT organoids, demonstrating the feasibility of establishing an infectious disease model in the FT organoids and the divergence of some of the host genes whose expression is influenced by these two bacterial species.