SEARCH

SEARCH BY CITATION

Content

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

The aim of this study was to localize and evaluate the role of Toll-like receptor 2 (TLR2) in the endometrium and cervix of bitches at different stages of the oestrous cycle and in bitches with pyometra. Sixty-seven nulliparous dogs, ranging in age from 1 to 13 years, were allocated amongst five groups (pro-oestrus; n = 7, oestrus; n = 10, dioestrus; n = 16, anoestrus; n = 11, pyometra; n = 23). Blood samples were collected for the measurement of progesterone concentration. The mean progesterone concentration was analysed as a parameter for validating the stage of the oestrous cycle in bitches. Tissues collected from uterine horn and cervix were fixed in 4% paraformaldehyde for immunohistochemical examination of TLR2. The expression of TLR2 was assessed semi-quantitatively. No pathological changes were found in the uterine samples of healthy dogs. In bitches with pyometra, the glandular epithelium expressed TLR2 more intensely than the surface epithelium. The expression of TLR2 in the glandular epithelium was also significantly higher in healthy dogs at oestrus, dioestrus and dogs with pyometra compared with anoestrous dogs (p < 0.01). The expression of TLR2 in the stroma was not observed in the group of healthy dogs at all stages. The surface epithelium of cervix in dogs with pyometra expressed TLR2 significantly more intensely than did the stoma, whereas the expression of TLR2 during oestrus and dioestrus was absent in the stroma of cervix. This study provides the first report of immunohistochemical localization of TLR2 in the canine reproductive tract. In the present study, TLR2 was expressed in endometrial epithelium but was absent in the endometrial stroma of healthy dogs at all oestrous cycle stages. These findings suggest differential expression of TLR in endometrial cells. On the other hand, the lack of TLR2 in the stroma of healthy uteri of dogs may predispose to infection from the invading pathogens once the epithelial cells have been destroyed by the pathogens, especially Gram-positive bacteria.


Introduction

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

The canine female reproductive tract is exposed to the outside environment and must have sufficient immunological defence mechanisms to avoid infection (Aflatoonian and Fazeli 2008). Rapid innate immune defence mechanisms against infection involved the recognition of invading pathogens by pattern recognition receptors (PRRs) attributed to the Toll-like receptors (TLRs). The member of the TLR family recognize distinct pathogen-associated molecular patterns (PAMPs) produced by various bacterial, viral and fungal pathogens (Janeway and Medzhitov 2002; Aflatoonian and Fazeli 2008) and play an important role in activating intracellular signalling pathways leading to the direct killing of pathogens (Takeda et al. 2003; Pasare and Medzhitiv 2004). Pyometra (chronic uterine inflammation with accumulation of pus in the uterus) is the most important pathological condition of the uterus resulting from uterine infection in dog (Ishiguro et al. 2007). Of the 13 TLRs identified, TLR4 and TLR2 are the best understood in terms of responses to Gram-negative bacteria and Gram-positive bacteria, respectively (Darville et al. 2003). In the previous study, TLR4 (which recognizes Gram-negative bacteria) has been reported in healthy and infected canine endometrium (Chotimanukul and Sirivaidyapong 2011). However, the bacterial species frequently isolated from uterus of dog are not only the Gram-negative Escherichia coli, but also Gram-positive bacteria. Accordingly, TLR2 (which is responsive to Gram-positive bacteria) was investigated in this study.

Materials and Methods

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

Animals

In total, 67 nulliparous bitches, ranging in age from 1 to 13 years, were submitted for ovariohysterectomy at the Small Animal Teaching Hospital, Faculty of Veterinary Science, Chulalongkorn University, Bangkok. All bitches were post-pubertal and none had received hormonal treatment. The bitches were allocated amongst five groups. Group 1–4 were healthy bitches that were classified depending on the stage of the oestrous cycle: group 1, bitches in pro-oestrus (n = 7); group 2, bitches in oestrus (n = 10); group 3, bitches in dioestrus (n = 16); group 4, bitches in anoestrus (n = 11). Group 5 were bitches with pyometra (n = 23). All healthy bitches were examined for vulvar swelling and oestrous behaviour, and blood was collected.

Diagnostic criteria

The stage of the oestrous cycle was determined by vaginal cytology, gross inspection of the ovaries and serum progesterone concentrations. None of bitches presented abnormal clinical signs, and the blood profile was normal. In bitches with pyometra, the diagnosis of pathology was based upon clinical signs, haematology and usually confirmed by radiography and/or ultrasonography.

Hormonal analysis

A blood sample was collected from the cephalic vein of bitches before ovariohysterectomy. For serum production, blood samples were centrifuged for 5 min at 2500 × g. Peripheral blood serum progesterone concentrations were measured using chemiluminescent microparticle immunoassay (CMIA). For quality control, an internal and an external control were used. The internal control was performed by using a single sample of all control levels tested once every 24 h each day to ensure that the assay control values were within the concentration ranges. The external control was performed every 2 weeks with Randox International Quality Assessment Scheme (RIQAS).

Tissue collection

Uterine tissues were collected from each group of bitches undergoing ovariohysterectomy. Each tissue sample was divided into the horn and cervix of the uterus. Tissue from the left uterine horn was collected from the middle portion. Full thickness segments of the uterus, approximately 1 cm in length, were removed. Tissue samples were fixed in 4% paraformaldehyde for immunohistochemical examination (Chotimanukul and Sirivaidyapong 2011).

Determination of the stage of the oestrous cycle

The stage of the oestrous cycle of healthy bitches was confirmed by vaginal cytology using cotton swab technique and vaginal speculum, staining with modified Wright–Giemsa stain or Diff-Quick (pro-oestrus: mixed types of epithelial cells, red blood cells and white blood cells may be present in early to mid-pro-oestrus; oestrus: >90% cornified cells with fewer red blood cells than pro-oestrus, few to no white blood cells; dioestrus: >50% parabasal and intermediate cells on first day of dioestrus, white blood cells may be present with fewer red blood cells than pro-oestrus; anoestrus: >90% parabasal and intermediate cells with few white blood cells, fewer bacteria), by gross aspect of the ovaries (pro-oestrus and oestrus: presence of follicles; dioestrus: large corpora lutea; anoestrus: regressed corpora lutea) and by serum progesterone concentration (anoestrus: ≤0.5 ng/ml; pro-oestrus: <1–2 ng/ml; oestrus: >2–15 ng/ml; dioestrus: >1 ng/ml) (Van Cruchten et al. 2004; Kida et al. 2010).

Immunohistochemical staining

Samples fixed in 4% paraformaldehyde were embedded in paraffin and cut into 4-μm sections. Sections were mounted on positively charged slides and dried overnight at 37°C. After deparaffinization in xylene and rehydration in a graded series of ethanol, slides were boiled in citrate buffer (10 mm, pH 6.0) at 95°C for 40 min for antigen retrieval and then cooled down at room temperature for 20 min. Sections were immersed in 3% hydrogen peroxide in methanol for 30 min to block endogenous peroxidase activity. Non-specific binding was blocked in 3% bovine serum albumin (BSA) for 30 min. Slides were incubated in a humidified chamber overnight at 4°C with the primary antibody, mouse anti-human TLR2 (TL2.1; eBioscience, San Diego, CA, USA) in a dilution of 1 : 100. The sections were then incubated with horseradish peroxidase (HRP)–conjugated anti-mouse immunoglobulin (Dako Cytomation, Denmark) for 30 min and washed three times with PBS. Bound antibody conjugates were visualized using NovaRED peroxidase substrate as a chromogen, and sections were counterstained with haematoxylin and mounted with glycerine gelatine. In each run, negative controls (the substitution of the primary antibody with PBS) were run along with the samples (Chotimanukul and Sirivaidyapong 2011).

Quantification of immunohistochemical staining

The immunostaining was evaluated under a light microscope. The expression of TLR2 was analysed blind by one experienced scorer. The expression was assessed semi-quantitatively incorporating both the proportion of cells stained positively and the intensity of specific staining. Scoring of the uterine horn was evaluated separately in three different layers (surface epithelium, stroma and glandular epithelium). And scoring of the cervix in two different layers (surface epithelium and stroma) was evaluated separately. The proportionate area in each tissue later showing a positive staining was taken into account to the nearest 5% and defined as the percentage expression of the cell layer. The intensity of staining was visually classified on a scale of 1–3 (1 = weak, 2 = moderate and 3 = strong). These scores were then used to generate an expression index by multiplying the percentage expression with the average intensity score (AIS). The expression index was derived by multiplying percentage expression (PE) with an AIS, and this index was subsequently used for the statistical analyses.

Statistical analysis

The mean expression index at each tissue layer of uterus for each group of dog was calculated. Multiple analysis of variance using SAS was used to compare the differences in protein expression level between groups (pro-oestrus, oestrus, dioestrus, anoestrus and pyometra) and tissue layers (surface epithelium, stroma and glandular epithelium). Differences with p < 0.05 were regarded as statistically significant, p < 0.01 as highly statistically significant. Results are shown as mean ± SEM.

Results

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

No pathologic changes of the uterine horns were detected in any of the healthy dogs. The mean progesterone concentration in each group was 1.07 ng/ml (pro-oestrus), 8.04 ng/ml (oestrus), 25.93 ng/ml (dioestrus), 0.48 ng/ml (anoestrus) and 11.17 ng/ml (pyometra). In the pyometra group, the glandular epithelium expressed TLR2 more intensely than did the surface epithelium (p < 0.05). The expression of TLR2 in the glandular epithelium was significantly higher in healthy dogs at oestrus, dioestrus and dogs with pyometra compared with anoestrous dogs (p < 0.01) (Figs 1 and 2), while no expression of TLR2 in the stroma was observed in the group of healthy dogs at all stages of the oestrous cycle. The immunostaining of TLR2 in all groups and layers revealed differential expression (Fig. 3).

image

Figure 1. Immunohistochemistry showing the intense staining of Toll-like receptor 2 (TLR2) in glandular epithelium (a, b, black arrow) and weak staining in surface epithelium of uterine horn in pyometra dog (c, open arrow head)

Download figure to PowerPoint

image

Figure 2. Immunohistochemistry showing the intense staining of Toll-like receptor 2 (TLR2) in glandular epithelium of uterine horn at oestrus (a, b, black arrow), dioestrus (c, d, black arrow) and weak staining at anoestrus (e, f, black arrow). The immunostaining of TLR2 in immune cells at anoestrus (f, open arrow head) can be noted. Negative control is shown in (g)

Download figure to PowerPoint

image

Figure 3. The mean expression index (±SEM) for Toll-like receptor 2 (TLR2) in tissue layers (SE, surface epithelium; GE, glandular epithelium; S, stroma) of endometrium (horn part) in bitches during pro-oestrus, oestrus, dioestrus, anoestrus and in bitches with pyometra. The letters ‘a’ and ‘b’ indicate differences in the expression between groups within a similar tissue layer. The letters ‘A’ and ‘B’ indicate differences in the expression between tissue layers within a similar group. Different letters indicate a significant difference (p < 0.05)

Download figure to PowerPoint

The surface epithelium of cervix in dogs with pyometra expressed TLR2 significantly more intensely than did the stroma (p < 0.01). The expression of TLR2 for dogs at oestrous and dioestrous stages was absent in the stroma of cervix. The immunostaining of TLR2 in all groups and layers of cervix were differentially expressed. Furthermore, the expression of TLR2 in the surface epithelium of the cervix was significantly higher than in the uterine horn (p < 0.01) (Fig. 4) when compared with other uterine regions.

image

Figure 4. The mean expression index (±SEM) for Toll-like receptor 2 (TLR2) in surface epithelium in different regions of the reproductive tract (uterine horn, uterine body and cervix) in bitches during pro-oestrus, oestrus, dioestrus, anoestrus and in bitches with pyometra. The letters ‘a’ and ‘b’ indicate differences in the expression between groups within a similar tissue layer. Different letters indicate a significant difference (p < 0.05)

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

This study provides the first report of the immunohistochemical localization of TLR2 in the canine reproductive tract. In the present study, TLR2 was expressed in endometrial epithelium but absent in the endometrial stroma of healthy dogs at all stages. These specific expressions of TLR in different tissue layers have been reported in earlier studies. For example, TLR3 expression was present in epithelial cells to combat the viral infection. Meanwhile, TLR4 was highly expressed in the stromal cells to prevent hyper-responsiveness to microorganisms and commensal bacteria (Hirata et al. 2007). These findings may indicate differential expression of TLR in a different endometrial cell types. However, the expression levels of TLR2 in human endometrial stromal cells were comparable to those of endometrial epithelial cells (Compton et al. 2003; Lin et al. 2009). The lack of TLR2 expression in the stroma of healthy canine uteri may predispose to infection from invading pathogens once the epithelial cells have been destroyed by the pathogens, especially Gram-positive bacteria. Nevertheless, TLR2 would be inducible post-infection by TNF-α (Pioli et al. 2004). The engagement of TLR4 resulted in the production of TNF-α may lead to increased expression of TLR2 (Pioli et al. 2004). These observations may explain why TLR2 expression was found increased in the endometrial stroma of dogs with pyometra in this study.

Recently, in a study of TLR2 and TLR4 mRNA in canine pyometra, these genes were found to be significantly up-regulated in the endometrium of dogs with pyometra infected with E. coli when compared with healthy dogs in the dioestrous stage (Silva et al. 2010). However, in infected endometrium of dogs suffering from pyometra, TLR2 was not highly expressed in the endometrial surface epithelium as was found with TLR4 expression in the previous study (Chotimanukul and Sirivaidyapong 2011). The possibility is that the endometrium has increased expression of other receptors such as TLR4, which also increased the production of cytokines in response to infection (Darville et al. 2003). This is in accordance with the finding of TLR4 in the previous study (Chotimanukul and Sirivaidyapong 2011), which found an extremely high level of TLR4 in infected endometrium. This result may indicate a protective role for TLR2 in canine uterine infection, which can decrease the uterine pathology due to unfavourable inflammation and improved balance between protective and pathological inflammatory responses to maintain the uterine homoeostasis (Si-Tahar et al. 2009). In addition, the expression of TLR2 was more intense in the endometrial glandular epithelium of oestrus, dioestrus and dogs with pyometra compared with anoestrus dogs. In the previous study, significantly higher levels of expression of TLR2 in the human endometrium have been observed during the secretory phase than other phases of the menstrual cycle (Compton et al. 2003; Lin et al. 2009). However, TLR2 was also expressed dominantly in the uterine body at pro-oestrus compared with anoestrus (Chotimanukul and Sirivaidyapong, unpublished data). This finding may support the influence of not only progesterone but also oestrogen in the expression of TLR2 in canine endometrium.

In dogs with pyometra, the expression of TLR2 was significantly more intense in the surface epithelium of the cervix than that in the stroma. The surface epithelium of the cervix maintains a first line of defence for the lower female reproductive tract (Pioli et al. 2004). To protect the host from the infection, the recognition of the invading bacteria in dogs suffering from pyometra was shown to be mediated by TLR2 in the cervix, and this may result in an increased expression of TLR2 in the surface epithelium of these tissues.

Interestingly, the expression of TLR2 was absent in the stroma of the cervix at oestrus and dioestrus. In dogs, the cervix usually opens approximately 2 days before the LH peak, which is close to the time of mating in oestrous stage (Silva et al. 1995). In this period, bacteria could enter the uterus easily. Consequently, from the result of this study, if the epithelial barrier has been destroyed at this time, these pathogens may invade into the reproductive tract and cause the uterine infection in dogs.

Acknowledgement

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

This study was funded by The 90th Anniversary of Chulalongkorn University Fund (Ratchadaphiseksomphot Endowment Fund).

Conflicts of interest

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References

None of the authors have any conflicts of interest to declare.

References

  1. Top of page
  2. Content
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflicts of interest
  9. References