G00052Isolation of Lymphocytes from Ovine Uterine Tissue - A Comparison of Methods
S.L.Bowland(1), P.E.Shewen(1), J.R.Papp(1), and B.A.Croy(2)
Departments of Pathobiology(1) and Biomedical Sciences(2), University of Guelph, Ontario, Canada
Abstract:
Reproductive disease results in major production losses for the sheepindustry. One approach to gaining insight into the pathogenesis of reproductive disease affecting female animals, is to investigate the local uterine cellular immune response to a particular pathogen. Commonly, uterine lymphocytes are obtained by digestion of endometrialtissue with enzymes, resulting in a cell suspension from whichlymphocytes are isolated. The potential drawbacks of enzymatic methods include possible cleavage of cell surface antigens, inavailability of lymphocytes below the exposed tissue surface, and expense. Described here is a comparison of an enzymatic and amechanical method to obtain lymphocytes from ovine uterus. Each uterus (n=3) was split into its corresponding uterine horns. One hornwas digested using a mixture of collagenase, hyaluronidase, and DNase;the other was disaggregated mechanically using a 'Stomacher', a machine commonly used to obtain cells for primary cell culture. Regardless of the method used for isolation, an approximately equalyield of lymphocytes per gram of tissue was recovered from resulting cell suspensions by density gradient centrifugation using Ficoll-Hypaque (d=1.077g/mL). Flow cytometry, following staining with monoclonal antibodies recognizing ovine CD8, gamma/delta TcR, and CD45R(A), consistently revealed higher numbers of CD8+ and CD45R(A)+cells in samples mechanically processed compared to those receiving enzymatic treatment. Thus the use of mechanical means to obtain ovineuterine lymphocytes proved equal, if not superior to enzymatic digestion, based on CD marker detection, lymphocyte yield, labour, andexpense.
Introduction:
Abbreviations: BSA - bovine serum albumin; EDTA -
(ethylenedinitro)-tetraacetic acid disodium salt; FITC - fluorescein
isothiocyanate; HBSS - Hank's balanced salt solution; mAb- monoclonal
antibody; MNL - mononuclear leukocytes; NMS - normal mouse serum; PBS -
phosphate buffered saline; PE - phycoerythrin; TcR - T cell receptor
The advent of flow cytometry and the general availability of mAbs has made it possible to characterize lymphocyte subpopulations in many species and in various tissues. Identification of uterine lymphocyte subpopulations may yield information regarding host response to reproductive disease or pregnancy. It is of particular importance when identifying cell populations derived from tissue, to choose a method of cell recovery that balances yield with viability and cell surface integrity. Enzymatic digestion and mechanical disaggregation are two means by which cells are routinely recovered from tissue. Although enzymatic digestion is the most frequently reported method (1), it risks alteration of surface markers. Mechanical means of obtaining cells from tissue have a broad range of potential effects on recovered cells, depending on the technique used. An inexpensive and relatively innocuous tool for mechanical tissue disaggregation is a ‘Stomacher’. This machine uses two stainless steel paddles to alternately apply evenly distributed pressure to a tissue sample contained in a polyethylene bag held in place by a door. *See Figure 1.*
The objective of this study was to compare MNL yield and selected cell surface marker detection by flow cytometry for ovine uterine MNL recovered through either enzymatic digestion or mechanical tissue disaggregation.
Materials and Methods:
Three grossly normal ovine uteri were obtained within 5 hours of slaughter from an abattoir. All sheep were in seasonal anestrus, but no information was available regarding age, breed, or health status. Following removal of uterine tubes, cervix, and excess connective tissue each uterus was split into left and right halves. Two left and one right halves were randomly assigned to be digested enzymatically, while the opposite two right and one left halves were subjected to mechanical tissue disaggregation. *See Figure 2.*
Each intact uterine horn receiving enzymatic treatment(2) was cut into thin strips (5 mm x 3 cm), weighed, and incubated for 30 minute intervals in 3 changes of approximately 100 ml of 5 mM EDTA (Fisher Chemical, Fairlawn, New Jersey, USA) at room temperature with continual stirring. Tissue was rinsed three times using HBSS, then cut into very small pieces (< 2 mm), and incubated for 60 minutes at 37C with solution #1 containing 30,000 U of collagenase (Type 4; Worthington Biochemical Corp., Freehold, New Jersey, USA), 30,000 U hyaluronidase (Worthington Biochemical Corp.), 100 microlitres of 1 g/ml DNase (Boehringer Mannheim, Montreal, Quebec, Canada), and 1% BSA (Sigma Chemical Co., St.Louis, Missouri, USA) in 75 ml of HBSS. Following treatment #1, supernatant containing dispersed cells was decanted and the remaining tissue was incubated for 60 minutes at 37C with solution #2; containing the same ingredients as solution #1 except 15,000 U of collagenase and 15,000 U of hyaluronidase. During the second incubation, the decanted supernatant from the first treatment was centrifuged at 300 x g for 10 minutes, the pellet resuspended in PBS, and centrifugation was repeated. The resulting pellet, containing dispersed cells from the first treatment, was resuspended in 3-5 ml of PBS and maintained on ice. Following the second treatment, fluid surrounding the remaining uterine tissue was strained through cheesecloth and centrifuged at 300 x g for 10 minutes. The pellet was washed once and resuspended in 3-5 ml PBS. The two cell suspensions were pooled, diluted to 35 ml with PBS, and underlaid with 15 ml Ficoll-Hypaque solution (density 1.077 g/ml). Following 15 minutes of centrifugation at 850 x g, leukocytes were recovered from the interface, diluted 50/50 with PBS, and centrifuged at 300 x g for 10 minutes. The pellet was resuspended in 2-3 ml of PBS and the live MNL yield per gram of uterine tissue was calculated using Trypan Blue exclusion and a hemacytometer.
Recovery of MNL using mechanical tissue disaggregation proceeded concurrently. The opposite uterine horn was cut into < 1 cm squares, weighed, and washed for 10 minutes using approximately 300 ml PBS and continual stirring, in a flask packed with ice. Excess PBS was decanted. The remaining tissue was cut into very small pieces (2-3 mm squares), distributed evenly between four sterile polyethylene sample bags (FisherBrand, Fisher Scientific, Toronto, Ontario, Canada), and surrounded with approximately 75 ml of cold PBS, to ‘buffer’ cells from direct force as they were dislodged from the tissue. Each sample was ‘stomached’ four times for 1 minute intervals. Cold PBS (75 ml) was added and each sample was ‘stomached’ for an additional minute. Samples were maintained on ice when not in the Stomacher. Following mechanical disaggregation, the resulting cell suspension was strained through cheesecloth. Isolation of MNL and calculation of live MNL yield proceeded as outlined above.
MNL were then stained using a panel of mAbs which would be expected to stain T cells in ovine uterine (endometrial) tissue(3): mouse anti-ovine CD8 alpha-chain (CACT80C, VMRD, Pullman, Washington, USA), mouse anti-ovine gamma/delta TcR (86D, VMRD) biotinylated (using biotin (long arm) N-hydroxysuccinimidyl 6-biotinamido hexanoate, Dimension Laboratories Inc., Mississauga, Ontario, Canada), and mouse anti-ovine CD45R(A) (73B1, VMRD), a portion of which was biotinylated. PBS was both the wash buffer and diluant for all solutions. The volume of solutions added to each well, except washes, was always 25 microlitres, and incubations were at 4C.
For each uterine half, approximately 0.5-1.0 x 10^6 cells were aliquoted to five wells of a 96 well round-bottomed polystyrene microtitre (Dynatech Laboratories, Chantilly, Virginia, USA). Cells were washed using 100 microlitres of PBS, followed by centrifugation at 300 x g for 2 minutes. The resulting supernatant was discarded and the first mAb solution at a dilution of 1/1000 was added. Following a 45 minute incubation, cells were washed twice, and the secondary goat anti-mouse Ig gamma-chain antibody conjugated to FITC (Jackson ImmunoResearch Laboratories, Inc., West Grove, Pennsylvania, USA) was applied (1/200) for 45 minutes. The cells were washed, and a 10% solution of NMS (Cedarlane Laboratories Limited, Hornby, Ontario, Canada) was added. Following a short incubation (10 min), cells were washed, and incubated for 45 minutes with the biotinylated second mAb at a dilution of 1/100. Cells were washed twice and streptavidin conjugated to PE (1/200; Jackson ImmunoResearch Laboratories, Inc.) was applied for 15 minutes. Following a wash, cells were resuspended in 100 microlitres of 0.5% paraformaldehyde solution, transferred to Falcon Tubes (Becton Dickinson Labware, Becton Dickinson & Co., Lincon Park, New Jersey, USA), and diluted with approximately 300 microlitres of PBS for flow cytometry. Controls included unstained samples and samples incubated with the secondary antibody conjugated to FITC, 10% NMS, and streptavidin conjugated to PE.
Data were analysed using Lysis II flow cytometry software and compared using the paired Student’s t-test. A p-value of <0.05 was considered significant.
Results:
The mean live MNL yields for enzymatically and mechanically processed ovine uterine tissue, 1.46E+05 MNL/g and 1.13E+05 MNL/g, respectively, were not significantly different. Results from single stain analysis are presented *See Figure 3.*, and from double stain analysis *See Figure 4.*.
The major subpopulation (63%) of gated lymphocytes recovered using mechanical tissue disaggregation were CD8+. Of those cells, approximately half were recognized as gamma/delta T lymphocytes. All gamma/delta TcR+ cells were CD8+ and CD45R(A)+, a finding supported by previous studies(3). The majority (87%) of the remaining half CD8+ gamma/delta TcR- (presumed to be alpha/beta TcR+) were CD45R(A)+, a phenotype representing naive T cells(4). Only 7% of CD45R(A)+ cells were not CD8+, possibly representing naive CD4+ cells considering that endometrium and subendometrial layers were used in this study.
Although not statistically significant (n=3), almost 20% fewerCD8+ cells were detected within the lymphocyte gate for uterine MNL recovered by enzymatic means. Of the 45% enzymatically obtained CD8+ lymphocytes, approximately 1/3 were gamma/delta TcR+. The total number of gamma/delta T cells detected did not differ from mechanically recovered samples; however, there were significantly fewer (p<0.022) CD45R(A)+ cells. It appeared that the decrease in detection of CD8, and in particular CD45R(A) cell surface antigen was across several cell subpopulations, most notably for lymphocytes that coexpressed CD8 and CD45R(A).
Discussion and Conclusions:
Although a limited panel of mAb were employed in this study, those used should identify the major lymphocyte subpopulations in ovine uterus. Results for mechanical separation were most similar to previous findings(3). Interestingly, Meeusen and collegues also used mechanical means to dislodge intraepithelial lymphocytes (IEL) from ovine endometrium; however, density gradient centrifugation was not used to isolate lymphocytes from endometrial cell suspensions on the grounds that the larger, more granular IEL were lost in the cell pellet. Preliminary assessment during development of techniques for this study, did not reveal MNL in the pellet resulting from density gradient centrifugation using Ficoll-Hypaque (density 1.077 g/ml). The generally high percentages of CD8, CD45R(A), and gamma/delta TcR detection may possibly have been due to the use of ‘intact’ uterine tissue versus endometrium, or may be related to the age or health status of the animals. The differences observed between enzymatic and mechanically processed tissue were most probably due to the method employed, considering that both techniques were carried out on uterine tissue from the same animals.
The decrease in detectable CD8+ and/or CD45R(A)+ cells in enzymatically treated tissue may result from a failure to recover cells below the surface exposed for enzymatic digestion, or possibly due to selective cleavage of cell surface markers. These results suggest that method for obtaining MNL from ovine uterine tissue may affect interpretation of phenotype of recovered cells. In particular, the decreased detection of CD45R(A) may have implications for studies of recirculation or response to uterine pathogens.
This work was supported by the Medical Research Council of Canada and the Ontario Ministry of Agriculture, Food, and Rural Affairs.
References:
(1) Pretlow II,T.G. and Pretlow, T.P. (eds). Cell Separation: Methods and Selected Applications (Vol.1). Academic Press, New York, NY. Copyright 1982.
(2) protocol adapted from Frayne, J. and Stokes, C.R. (1994) MHC Class II Positive Cells and T Cells in the Equine Endometrium Throughout the Estrous Cycle. Vet. Immunol. Immunopathol. 41: 55-75, and personal communication with H. Engelhardt.
(3) Meeusen, E., Fox, A., Brandon, M., and Lee, C.S. (1993) Activation of Uterine Intraepithelial gamma/delta T Cell Receptor Positive Lymphocytes During Pregnancy. Eur. J. Immunol. 23: 1112-1117.
(4) Mackay, C.R., Marston, W.L., and Dudler, L. (1990) Naive and Memory T Cells Show Distinct Pathways of Lymphocyte Recirculation. J. Exp. Med. 171: 801-817.
Comments:
Address questions and comments about this abstract to Sandi Bowland ( sbowland@ovcnet.uoguelph.ca).