Whole population cell kinetics of jejunal and colonic epithelium in lactating dams.код для вставкиСкачать
THE ANATOMICAL RECORD 228:262-266 (1990) Whole Population Cell Kinetics of Jejunal and Colonic Epithelium in Lactating Dams HAZEL CHENG AND MATTHEW BJERKNES Department of Anatomy, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada M5S 1A8 ABSTRACT Previous studies make it likely that the response of the intestinal epithelium as a whole to lactation is different from that observed in the crypt population alone. We confirm this difference by whole population cell kinetics measurements of jejunal and colonic epithelium in mice that have been suckling pups for various lengths of time. We found that the fraction of cells in S phase in jejunal epithelium was significantly increased after only 1week of lactation, maintained this elevated level after 2 weeks of lactation, but returned to normal during the third week of lactation. The cell number density in jejunum was also significantly higher after 2 and 3 weeks of lactation before returning to normal by 4 weeks. In the colonic epithelium no changes were found in the distribution of cells in G,, S, and G2+ M phases. However, a significant increase in cell number density was observed after 2 weeks of lactation, followed by a sharp decrease to a level significantly below that of normal mice after 3 and 4 weeks of lactation. We conclude that the observed significant increase in the fraction of S phase in jejunal epithelium of lactating mice is probably due to a smaller relative expansion of the villus population when compared with the expansion of the crypt population. Our data also indicate that a number of cell kinetic parameters in the intestinal epithelium of lactating mice are changing throughout the period of lactation. Thus the intestinal epithelium is probably not in a steady state during lactation. Lactation has been shown to be a good physiological Results of Harding and Cairnie (1975) make it likely model to study small intestinal adaptation (Saunders that k , > k , (see Appendix I), and therefore that 1; > and Morgan, 1957; Fell e t al., 1963; Boyne et al., 1966; 1,. Cairnie and Bentley, 1967; Craft, 1970; Harding and To test the hypothesis that the whole population laCairnie, 1975; Elias and Dowling, 1976). Previous cell beling index 1, changes during lactation (in contrast to kinetic studies of the response of the epithelium to lac- the crypt labeling index 1, which does not appear to tation have concentrated on the crypt population. As a change during lactation), we studied the effects of lacresult, little is known about the response of the epithe- tation on whole population cell kinetics in jejunal epilium as a whole but there is some evidence suggesting thelium of lactating dams using flow cytometry and that the response of the epithelium as a whole may be Coulter particle counting. We also looked at the colonic different from that observed in the crypt population. epithelium because it is not known whether lactation In virgin mice, the labeling index in the crypt is ap- affects the colonic epithelium. In addition, since previproximately 1, = n,ln,, where n, is the number of cells ous studies have been limited to a single time point in S phase, and n, is the number of cells in a n average after lactation, little is known about the time course of crypt. After 15 days of lactation, n, has increased but epithelial response to lactation. We therefore also the crypt labeling index and the cell cycle characteris- made a study of the time course of the response of the tics of the epithelial cells do not change significantly intestinal epithelium to lactation. (Harding and Cairnie, 1975). Therefore, it is likely that In mouse jejunal epithelium, we found significant both n, and n, have increased by a constant factor k,. differences in the distribution of cells in S and G, Similarly, let I , be the whole epithelial fraction of cells phases after 1and 2 weeks of lactation. Thus, the whole in S phase, which in virgin mice will be approximately population labeling index changed during lactation I , = mn,/(mn, + n,) where m is the average number of and 1, # 1;. Significant effects on cell number density crypts per villus and n, is the number of epithelial cells were not observed until after 2 weeks of lactation. In in a n average villus. Given t h a t the villus population the colonic epithelium, large fluctuations in cell numincreases by a factor k2 after 15 days of lactation while, as stated above, n, and n, each increases by a factor k,, the whole epithelial labeling index after 15 days of lactation is Received October 9, 1989; Accepted March 20, 1990. Address reprint requests to H. Cheng, Department of Anatomy, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada. M5S 1A8. 0 1990 WILEY-LISS, INC. 263 LACTATION AND INTESTINAL CELL KINETICS TABLE 1. Whole population kinetic parameters of mouse intestinal epithelium % G, phase' X t SEM Weeks of lactation Normal 1 2 3 4 + M phase X t SEM % S phase X t SEM lo6 cells/cm2 % G, X t J C J C J C J 85.5 t 0.92 82.2 t 0.44 79.9 t 0.32 86.3 t 0.91 84.1 2 1.32 84.3 t 0.12 83.3 t 0.46 81.9 t 0.86 80.7 t 1.75 82.7 k 1.35 10.2 2 0.78 13.9 t 0.40 16.6 2 0.35 10.4 2 0.74 11.6 t 0.88 11.7 t 0.23 12.6 t 0.55 13.6 t 0.92 14.0 t 1.49 13.0 t 1.52 4.31 2 0.43 3.89 t 0.12 3.49 2 0.09 3.26 2 0.21 4.31 t 0.44 4.04 t 0.24 4.11 k 0.14 4.55 t 0.33 5.37 2 0.76 4.25 t 0.50 16.7 t 0.60 21.6 t 2.33 27.2 t 1.51 24.6 t 1.72 21.0 t 1.13 SEM C 11.8 t 0.81 13.5 t 1.20 16.1 t 0.66 6.98 t 0.75 6.05 t 0.43 'This fraction includes all cells with 2N DNA, i.e., cells in GI phase of the cycling population as well a s terminally differentiated noncycling cells. ber density were observed after 2, 3, and 4 weeks of lactation, but no changes were observed in the distribution of cells through the cell cycle. RESULTS The mean percentages of epithelial cells in G,, S, and G, + M phases,' and the number of epithelial cells per cm2 of intestine in normal mice and during lactation are shown in Table 1. In the jejunum, we found signifMATERIALS AND METHODS icant differences in the fraction of epithelial cells in G, Pregnant CD-1 mice were purchased from Charles and S phases, a s well a s in the number of epithelial River Canada and housed in animal facilities with a 12 cells per cm2 of intestine. There were no significant h light/dark cycle (light 7:OO-19:00), and fed ad libi- differences in the fraction of epithelial cells in G2 + M tum on mouse breeder diet (10% fat) (Teklad, Madison, phase. In the colon, we found significant differences in WI). The pups were distributed so that each litter con- the number of epithelial cells per cm2 of intestine while tained eight pups. Since there are significant estrous- no significant differences were found in the fraction of cycle-related variations in mouse intestinal epithelium epithelial cells in G,, S, and G 2 + M phases. (Cheng and Bjerknes, 1988) we used only mice that Figures 1-4 show the 95% confidence limits of the were in diestrus. The stages of estrous cycle were de- measurements from the jejunal epithelium. After 1 termined by analysis of vaginal smears (Rugh, 1968). week of lactation the fraction of cells in S phase had Mice were killed after 1, 2,3, and 4 weeks of lactation. increased significantly from the level found in normal Five mice were used for each time interval. Another mice. After 2 weeks of lactation both the fraction of five virgin female CD-1 mice in diestrus were used as cells in S phase and the number of cells per cm2 of normal controls. All mice were killed between 11:OO jejunum increased significantly while the fraction of and 12:OO. cells in G, phase decreased significantly relative to The entire epithelium was isolated from a 5 cm seg- normal. After 3 weeks of lactation, only the number of ment of jejunum (16 cm distal to the pylorus) and a 5 epithelial cells per cm2 of jejunum remained significm segment of colon (proximal to the rectum) by intra- cantly different. After 4 weeks of lactation all measureventricular perfusion of EDTA followed by vibration ments were not significantly different from those of into calcium-magnesium-free saline (Bjerknes and normal mice. Cheng, 1981). An aliquot of the isolated epithelial Figures 5-8 show the 95% confidence limits of the preparation was used for flow cytometry to determine colonic epithelial measurements. Significant differthe fraction of epithelial cells in G,, S, and G,+M ences were seen only in the number of epithelial cells phases of the cell cycle. Single-cell suspensions were per cm2 of colon. There were significantly more cells prepared from the epithelial preparations by serial pas- per cm2 in the colonic epithelium after 2 weeks of lacsage through 20 pm Nitex filters (Cheng and Bjerknes, tation relative to normal. However, after 3 and 4 weeks 1982). The cells were stained with the DNA fluoro- of lactation the number of cells per cm2 was significhrome 4'-6'-diamidino-2-phenylindole(DAPI) and cantly less than that found in normal mice. analysed with a flow cytometer (ICP 22A, Ortho Diagnostic Systems, Inc.). Another aliquot of the epithelial DISCUSSION preparation was used to measure the number of epitheIt is well established that in lactating rats and mice, lial cells per unit area by Coulter particle counting the intestine increases in weight, surface area, villus (Cheng and Bjerknes, 1983). The data from each animal were kept separate. The height, and crypt depth (Saunders and Morgan, 1957; data from the five groups were compared by using a Fell et al., 1963; Boyne et al., 1966; Cairnie and Bentone-way analysis of variance. Multiple comparisons of ley, 1967; Craft, 1970; Harding and Cairnie, 1975; results from different groups were made by analysis of Elias and Dowling, 1976). In addition, after 15 days of variance followed by Tukey's T method (Sokal and lactation, both the number of cells per crypt as well as Rohlf, 1981). A significance level of 5% was used. The the number of cells in S phase per crypt increase rela95% confidence limits of the mean value presented for 'We use G, to indicate all cells with 2N DNA including cycling cells each group were determined from the analysis of vari- in GI phase as well as terminally differentiated noncycling cells (e.g., villus cells). ance with Tukey's T method. 264 H. CHENG AND M. ELJERKNES t 4 N 1 2 3 4 Lactation Weeks o f Weeks of Fig. 1. 1 Fig. 3. T II N \ .d Ln Lactation '"1 T 10 d 1 0 N 1 2 Weeks of 3 4 Lactation Weeks o f Lactation Fig. 2. Fig. 4. Figs. 1-4. Graphs showing whole population measurements of mouse jejunal epithelium in normal control (N) and lactating mice. The data were compared with one-way analysis of variance. Each data point represents the mean of five animals. The vertical bars associated with each data point represent the 95%confidence limits derived from an analysis of variance with Tukey's T method. tive to the corresponding levels found in virgin mice. Nonetheless, the fraction of crypt cells in S phase in lactating mice is similar to that found in virgin mice because the proliferating population in the crypt expands at the same rate a s the crypt as a whole (Harding and Cairnie, 1975). A different picture emerges when the fraction of epithelial cells in S phase is determined relative to the epithelium a s a whole. The whole population studies carried out here showed significant increases in the percentage of epithelial cells in S phase in mice that have been suckling pups for 2 weeks when compared with normal mice (Fig. 2). The observed significant in- crease in the fraction of S phase cells in jejunal epithelium of lactating mice is probably due to a smaller relative expansion of the villus population when compared with the expansion of the crypt population (see Appendix I and Introduction). This shows that the results of kinetic studies on the crypt population do not necessarily reflect the behavior of the intestinal epithelium as a whole. Thus, whole population measurements are needed to gain a complete picture of the behavior of the intestinal epithelium. When we looked a t the time course of the jejunal epithelial response during lactation, we found that 1 week of lactation is sufficient to induce a significant 265 LACTATION AND INTESTINAL CELL KINETICS 100 90 c *-r( I T “G 70 H I 2 3 4 of Lactation Weeks I H I I I 2 1 3 I 4 of Lactation Weeks Fig. 5. Fig. 7. tt I I 0 H I 2 Weeks o f 3 4 Lactation Fig. 6. N Weeks 1 2 3 4 of Lactation Fig. 8. Figs. 5-8. Graphs showing whole population measurements of mouse colonic epithelium in normal control (N) and lactating mice. The data were compared with one-way analysis of variance. Each data point represents the mean of five animals. The vertical bars associated with each data point represent the 95% confidence limits derived from an analysis of variance with Tukey’s T method. increase in the fraction of epithelial cells in S phase (Fig. 2 ) . By 3 weeks the pups are beginning to sample solid food and most likely are beginning to depend less on suckling. At this time, the fractions of cells in GI and S phases had returned to the level found in normal mice (Figs. 1 and 2), but the number of epithelial cells per cm2 of intestine was still significantly higher than normal (Fig. 4). By 4 weeks the pups had ceased suckling and all aspects of the epithelium measured had returned to normal. Our data indicate that some of the measured parameters, such as fraction of GI and S phase cells, and cell number density of mouse jejunal epithelium are changing throughout lactation. The same is true of the cell number density of mouse colonic epithelium (Fig. 8). Thus the intestinal epithelium is probably not in a steady state during lactation. This is of concern because cell kinetic studies of the effects of lactation generally assume that the intestinal epithelium is in a steady state. Our results indicate that it is risky to assume steady-state status during lactation and that further study should be carried out to determine, during lactation, whether the intestinal epithelium is ever in a steady state. The fraction of cells in GI, S, and G2+ M phases of the mouse colonic epithelium did not differ significantly from normal throughout lactation (Figs. 5-7). 266 H. CHENG AND M. BJERKNES However, a more complex picture emerged from measurements of colonic epithelial cell number density during lactation. As in the jejunum, the colonic epithelium initially responded with a n increase in the number of cells per cm2 of intestine (Fig. 8). But then, in contrast to the jejunal epithelium which gradually returned to the normal levels after 3 and 4 weeks of lactation, cell number density in the colonic epithelium decreased sharply to a level significantly below normal (Fig. 8). The significance of the sharp decrease in colonic epithelial cell number density after 3 and 4 weeks of lactation is not clear and warrants future investigation. The observed changes in whole population cell kinetics in the mouse jejunal and colonic epithelium during lactation may reflect the direct or indirect influence of lactation hormones such a s prolactin and oxytocin on the intestinal epithelium. Other potential influences are the increased volume of food processed or the increased metabolic stress that accompanies lactation. These questions could not be answered by the present investigation and should be addressed in future studies. APPENDIX I Let n, be the number of cells in a n average villus in virgin mice and ni be the number of cells in a n average villus in lactating mice. Assuming steady state, the number of cells in a villus in virgin and lactating mice may be estimated a s follows: n, n; = mnptT (1) m(klnp)t; (2) = where m is number of crypts per villus (unchanged after 15 days of lactation; Harding and Cairnie, 19751, npis. the number of cells produced per crypt per hour in virgin mice, k , is the relative increase in the crypt cell population after 15 days of lactation,2and tT and tT are the transit times of villus cells in virgin and lactating mice, respectively. Solving Eq. (1)for mnp and substituting the result into Eq. (21, we have 'Since neither the crypt labeling index nor the cell cycle characteristics of epithelial cells change significantly after 15 days of lactation (Harding and Cairnie, 1975) k,n is the number of cells produced per crypt per hour in mice after 15 Bays of lactation. n, = -kltT nv ' tT Defining k , to be the relative increase in the number of cells per villus, we have n; = kzn, and therefore Since in both rats and mice t i is about 8 h shorter than tT (Cairnie and Bentley, 1967; Harding and Cairnie, 1975), it follows that k , must be smaller than k,. Using Harding and Cairnie's data (1975) on mice, k , may be estimated to be about 2.3 while k2 is about 1.9. ACKNOWLEDGMENTS We thank Dr. J. Totafurno for helpful comments on the manuscript, and R. Kuk for technical assistance. This work was supported by grants from the Medical Research Council of Canada. LITERATURE CITED Bjerknes, M., and H. Cheng 1981 Methods for the isolation of intact epithelium from the mouse intestine. Anat. Rec., 199t565-574. Boyne, R., B.F. Fell, and I. Robb 1966 The surface area of the intestinal mucosa of the lactating rat. J. Physiol., 183.570-575. Cairnie, A.B., and J . Bentley 1967 Cell proliferation studies in the intestinal epithelium of the rat. Hyperplasia during lactation. Exp. Cell Res., 46t428-440. Cheng, H., and M. Bjerknes 1982 Whole population cell kinetics of mouse duodenal, jejunal, ileal, and colonic epithelia as determined by radioautography and flow cytometry. Anat. Rec., 203: 251-264. Cheng, H., and M. Bjerknes 1983 Cell production in mouse intestinal epithelium measured by stathmokinetic flow cytometry and Coulter particle counting. Anat. Rec., 207t427-434. Cheng, H., and M. Bjerknes 1988 Variation of mouse intestinal epithelial whole population cell kinetics during the estrous cycle. Anat. Rec., 220r397-400. Craft, I.L. 1970 The influence of pregnancy and lactation on the morphology and absorptive capacity in the rat small intestine. Clin. Sci., 38t287-295. Elias, E., and R.H. Dowling 1976 The mechanism for small bowel adaptation in lactating rats. Clin. Sci. Mol. Med., 51r427-433. Fell, B.F., K.A. Smith, and R.M. Campbell 1963 Hypertrophic and hyperplastic changes in the alimentary canal of the lactating rat. J. Pathol. Bacteriol., 85r179-189. Harding, J.D., and A.B. Cairnie 1975 Changes in intestinal cell kinetics in the small intestine of lactating mice. Cell Tissue Kinet., 8t135-144. Rugh, R. 1968 The Mouse: Its Reproduction and Development. Burgess Publishing Go., Minneapolis, MN, pp. 38-43. Saunders, H.J., and A.F. Morgan 1957 Weight and composition of organs during the reproductive cycle in the rat. Am. J . Physiol., 191: 1-7. Sokal, P.R., and F.J. Rohlf 1981 Biometry, 2nd ed. Freeman, San Francisco, pp. 242-247.