Abnormal canine trophoblast physiology in late gestation can result in postpartum complications including subinvolution of placental sites and placental retention. Understanding normal late gestation canine trophoblast physiology provides a basis for future studies aimed at characterizing postpartum placenta-related disorders. The objective of the current study was to characterize term canine trophoblast migration, invasion, and proliferation in vitro. Further, the effects of interleukin-8 (IL-8) and tissue inhibitor of metalloproteinase-2 (TIMP-2) on trophoblast physiology were examined. Both IL-8 and TIMP-2 are known to play key roles in tissue remodeling in late gestation in several species. IL-8 stimulated term canine trophoblast migration by 35% compared to the control (P<0.01), while TIMP-2 had no significant effect on migration (P=0.38). TIMP-2 decreased term canine trophoblast invasion by 57% (P<0.05). However, IL-8 had no significant effect on term canine trophoblast invasion (P=0.42) or proliferation (P=0.18). This was the first investigation of in vitro canine trophoblast physiology. This study revealed that both IL-8 and TIMP-2 play a role in regulating trophoblast function at the end of gestation in the dog.
Keywords: Chorioallantois Dog; Matrix-Metalloproteinases; Placenta; Wound Healing Assay
Abnormal trophoblast physiology in late canine gestationcan lead to postpartum complications including subinvolutionof placental sites and placental retention. Subinvolution of placental sites is a pathologic condition in dogs resultingfrom excessive trophoblast invasion . This disorder often results in postpartum hemorrhaging, a delay in uterineinvolution, and increased maternal morbidity. Retained placentas can also cause severe problems in the bitch,especially when accompanied by retained fetuses or infection . Successful release of the placenta requiressynchronization of cellular and molecular interactionsbetween trophoblasts and the endometrium . Disruption of these interactions can result from abnormal trophoblastmigration, invasion, or proliferation at the end of gestation. Investigating normal trophoblast physiology in vitroprovides a unique insight into the physiology of placenta-related disorders and may pave the way forfuture therapeutic studies.
Matrix-metalloproteinases (MMPs) and their tissueinhibitors (TIMPs) are key regulators in signaling atthe trophoblast-endometrial interface. The MMP-TIMP interaction controls trophoblast invasion into the endometriumin humans . MMPs are zinc-dependent endopeptidases important in the degradation of thematernal endometrium . High levels of MMP-2 expressionhave been observed in the canine endometrium and myometrium at implantation . MMP-2 has also beendetected in the canine pre-implantation embryo 10-12 days after mating  and in the canine pre-term,pre-labor, and parturient placenta . TIMP-2 is a 21.8 kDa protein that preferentially binds to and inhibitsMMP-2. TIMP-2 has been documented in the primatetrophoblast and endometrium  and a decrease inTIMP-2 expression during late gestation is a necessary step in the sequence of successful placental release in cattle . Canine peripartum trophoblasts also express TIMP-2, however, its expression does not significantly decrease prior to successful placental release . Therefore, theinfluence of TIMP-2 on canine trophoblasts warrants further investigation to determine its role in placentalphysiology at the end of gestation.
Interleukins are dependent on MMP activity . Interleukin-8 (IL-8/CXCL8), an 8 kDa inflammatory chemokine, mediates its effect through the IL-8 receptors, CXCR1 and CXCR2 . Human trophoblasts constitutively secrete IL-8 throughout the entire duration of pregnancy  and express CXCR1 [13,14]. Studies using first trimester human trophoblasts show that IL-8 stimulates in vitro proliferation, migration, and invasion through the upregulation of MMP-2 . The canine pre-implantation embryo also expresses IL-8 , but its role in late gestation has not been examined.
Recombinant human IL-8 was obtained from Peprotech (Rocky Hill, NJ, USA) and was used at a concentration of 10 ng/ml for each assay. Recombinant human TIMP-2 (Peprotech) was used at a concentration of 0.5 μg/ml for each assay. Concentrations selected for IL-8 and TIMP-2 were based on dosages in previous studies found to elicit maximum effects with human trophoblasts [14,15].
Term canine placentas were collected following elective C-sections (n=5 dogs) at prepartum luteolysis, the period preceding the onset of the first stage of labor at which there is a marked decrease in progesterone below 2.5 ng/ml. Following tissue collection, the villous chorioallantois was dissected from placental tissues. Villous chorioallantois was minced and washed in a cell isolation medium as previously described  until further processed forprimary trophoblast isolation.
Isolation of TCT and cell culture
Term canine trophoblasts (TCT) were isolated from the villous chorioallantois as previously described . Briefly, following serial collagenase and trypsin tissue digestions, cells were separated by Percoll density gradient centrifugation. The trophoblast-containing gradient layer was washed twice and re-suspended in cell culture medium. The cells were identified using immunostaining cytokeratin- 7 to observe cellular morphology as previously described . The cell culture conditions were the same for all assays at 37oC in 5% CO2.
The migration (wound healing) assay was performed as previously described . Briefly, primary TCT (1000X103 cells/well) were suspended in 2 ml of a Dulbecco’s Modified Eagle’s Medium (DMEM) containing 10% FBSa and cultured in 12-well tissue culture plates. Cells were grown to a confluent monolayer. A wound was added with a sterile 200μl pipette tip. The monolayer was rinsed with the same media to remove loosened cells. Media with no factor (control), with IL-8 or with TIMP-2 was then added. Photomicrographs of the wound were taken with phase-contrast microscopy (100X magnification) after 8 h in culture (Fig. 1). The area of the wound was measured using ImageJ v.1.34 software . The experiment was performed in triplicate.
An invasion (Matrigel) assay was performed as previously described  using Matrigel invasion chambers (BD Biosciences, San Jose, CA, USA). Briefly, primary TCT (250X103 cells/well) were suspended in 0.5 ml of a modified DMEM media containing 10% FBSb with no factor (control), with IL-8 or with TIMP-2 and then incubated for 22 h. Non-invading cells were removed from the Matrigel-coated filter membrane and filters were fixed and stained with DiffQuik (Harleco®, EMD Chemicals, Inc. Gibbstown, NJ, USA). Cells from ten randomly selected non-overlapping fields were counted under light microscopy (400X magnification). The experiment was performed in triplicate.
The proliferation (((3-)4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide) (MTT) assay (Cell growth determination kit, MTT based, Sigma), was performed as previously described . Primary TCT (100X103 cells/ well) were suspended in 100 μl of a modified DMEM media with 10% FBSb with no factor (control), with IL-8 or with TIMP-2 and cultured in 96-well tissue culture plates for 28 h.) MTT 50 μg was added then to each well for 4 h. MTT and media was replaced with 100 μL 1-propanol and the plates were vigorously shaken for 5 min. Using a microplate reader (SpectraMax 190, Molecular Devices, LLC, Sunnyvale, CA, USA), absorbance was measured at 570 nm and 690 nm. Measurements were analyzed with SoftMax Pro 5.2 program SoftMax®
Pro Data Acquisition & Analysis Software (Molecular Devices, LLC, Sunnyvale, CA). The experiment was performed in quadruplicate.
The data were analyzed statistically by Repeated Measures or Two-Way ANOVA in PROC MIXED using SAS (Version 9.2, SAS Institute Inc., Cary, NC, USA) for all assays. For the migration assay, the mean of the control wound area at 0h was set to 0% wound closure and the data were expressed as the percent closure of the control wound area for each dog. For the invasion assay and proliferation assay, the mean of the controls was set to 100% and the data were expressed as the percentage of the control for each dog. For the proliferation assay, measurements at 690 nm were subtracted from measurements at 570 nm to account for background. In all experiments, data were expressed as mean ± SEM and significance was defined as P<0.05.
Table 1. Percent Migration, Invasion, and Proliferation of Control.
^TIMP-2 was not used for proliferation assay because it is not involved in cell proliferation
* indicates statistical significance
Figure 1. Representative wells illustrating the wound at time 0 and 8 h for the control, IL-8-treated, and TIMP-2-treated term canine trophoblast cells in primary culture (100X).
Figure 2. Representative fields of cell invasion for control, IL-8-treated, and TIMP-2-treated cells (400X). Arrows indicate the trophoblast nuclei and arrowheads indicate the filter membrane pores.
In vitro canine trophoblast physiology has not been previously described. Therefore, the objective of this study was to characterize normal in vitro term canine trophoblast (TCT) physiology in response to IL-8 and TIMP-2 using migration, invasion, and proliferation assays. Similar to observations in first trimester human trophoblasts (HTR-8/SVneo) where IL-8 increased migration , IL-8 increased TCT migration in the current study. Previous studies have also shown that IL-8 increases invasion and proliferation in first trimester human trophoblasts [13,14]. However in the current study, IL-8 did not increase invasion and proliferation in TCT primary cultures. It is possible that canine trophoblasts do not retain the IL-8 receptors (CXCR1 and CXCR2) at term. A decrease in expression of the IL-8 CXCR1 receptor has been established as a normal condition of human myometrial tissue during parturition . It is not known if the IL-8 receptors are similarly decreased in expression in the dog at term, but this warrants further investigation.
Whereas the effects of IL-8 are dependent on MMP activity , MMP activity is regulated by TIMPs in human trophoblast invasion [15,21]. Down-regulation of TIMP-2 mRNA increases migration and invasion of first trimester human trophoblasts (HTR-8/SVneo) . In the current study, TIMP-2 significantly decreases invasion, but not migration of normal TCT. This study has laid the groundwork for future investigations to examine responses to TIMP-2 in abnormal TCT conditions (e.g. subinvolution of placental sites, retained placenta). Studies in cattle have reported that a reduction of TIMP-2 at the end of gestation is essential for the separation of fetal membranes from the endometrium in normal parturition . Also in cattle, the expression of MMP-2 is higher in those with retained placenta compared to those with a normal placenta, further supporting the regulatory role of TIMP-2 .
In conclusion, this is the first study to examine canine trophoblast physiology in vitro. The current study provides evidence that IL-8 promotes migration, whereas TIMP-2 decreases invasion in normal primary TCT. This information provides a basis for in vitro investigations to understand trophoblast pathologies that occur during the canine postpartum period.
The authors would like to acknowledge Drs. Cheryl Lopate, Hernán Montilla, Charles Estill, and Chris Holenstein for providing placental tissue. We are grateful to Dr. Fred Menino and Caitlin Donovan for helping with statistical analysis. This work was supported by the Department of Animal and Rangeland Sciences at Oregon State University.
a86.5% DMEM with 1% Hepes (Life Technologies), 1% Glutamine Plus, 1% Penicillin/Streptomycin, 0.5% Gentamycin, and 10% FBS.
b96.5% DMEM with 1% Hepes, 1% Glutamine Plus, 1% Penicillin/ Streptomycin, and 0.5% Gentamycin