THE ANATOMICAL RECORD 220~138-142(1988) Removal of Collagen Bundles in Murine Uterus During Postpartum Involution KIYOSHI SHIMIZU AND MASAMI HOKANO Department of Anatomy, Tokyo Medical College, Tokyo 160, Japan ABSTRACT The localization of collagenolytic activity within the tissue compartments of the mouse uterus was investigated during postpartum involution. The rate of collagenase activity was measured by analysis of tissue levels of hydroxyproline from the day of parturition to the 10th postpartum day. Collagen bonding was analyzed by viewing birefringence induced by the picrosirius red-binding technique. An attempt was made to interrelate quantitative analysis with the histologic distribution of collagen during postpartum days 1-10. Histologic and quantitative evidence indicated the following: 1)The collagenous compartments of the endometrium and myometrium differ in their response to the postpartum rise in collagenase activity; collagen degradation occurs primarily in the endometrium, that is, the myometrial collagen remains, but much of the endometrial collagen is removed. 2) Endometrial collagen is degraded particularly in the immediate subluminal compartment. Following parturition, both the weight and collagen content of the uterus show a rapid decrease, which is known as postpartum involution. A conspicuous feature in postpartum involution is collagen removal, since this represents one of the fastest known rates in any connective tissue (Harkness, 1964). Regional differences in the amount and degradation of collagen have not been analyzed. Because fuchsin in van Gieson’s trichrome method fades after a few months, Sweat et al. (1964) performed tests on a series of dyes to avoid this problem and introduced sirius red (SR). Later, Constantine and Mowry (1968) observed that collagen fibers stained with picroSR show high birefringence when observed with polarized light. Sirius red binds interstitial collagens and not proteoglycans (Junqueira, et al., 1979). Although SR also stains some basic proteins, these proteins are not birefringent (Junqueira et al., 1979). The picrosiriuspolarization method (picro-SR pol method) shows complete and detailed distribution of interstitial collagen bundles in tissue sections (Junqueira et al., 1979; PerezTamayo and Montfort, 1980). With the picro-SR pol method, we observed postpartum uteri to determine where collagenolytic activity within the tissue compartment is most evident. MATERIALS AND METHODS The animals used were female mice of the N C S strain. They were reared under a 12-hr light : 12-hr dark regimen and given food and water ad libitum. At 8 weeks of age they were mated. Successful mating was validated by vaginal plug. Since collagen degradation starts after the day of parturition in rodents (rat: Harkness, 1964; mouse: Shimizu and Hokano, 19871, the animals were killed on the day of parturition, or 1, 2, 3, 5 or 10 days post partum. All animals were allowed to suckle their 0 1988 ALAN R. LISS. INC. pups. After sacrifice, the uterine horns were dissected and weighed. The uterine horns taken from 10-weekold diestrous virgin mice served a s controls. A whole uterine horn was frozen at -20°C to determine hydroxyproline (Hyp). A whole horn was put into one test tube with a screw cap, and 2 ml 6 N HC1 was added. The samples were hydrolyzed a t 130°C for 3 hr in a n electric drying oven (FS-42D, Advantic, Tokyo). Hydroxyproline levels were determined in these hydrolysates according to the method of Woessner (1961).The results were represented as total Hyp content per pair of uterine horns. Several portions of uterine horn were fixed with 10% buffered formalin for 24 hr. They were then dehydrated in alcohol, passed through xylene, and embedded in paraffin. Sections 3 pm thick were deparaffinized and stained for 1 h r in picro-SR solution (0.1% Sirius red FB200 in saturated aqueous picric acid; Sweat et al., 1964). Since the results produced by SR FB200 were similar to results for SR F3BA (Zambrano et al., 19821, we used SR FB200 (Polysciences, Warrington, PA, USA). As collagen molecules are rich in basic groups, this protein reacts with acidic dyes. When collagen is stained with the elongated strongly acidic dye SR, collagen birefringence increases considerably because the axes of both the collagen molecules and the dye bonds are parallel. As collagen is the only oriented basic protein so far described in animal tissues, this enhancement in birefringence is specific for oriented collagen (Junqueira Received February 3, 1987; accepted July 28, 1987. Address reprint requests t o K. Shimizu Ph.D., Department of Anatomy, Tokyo Medical College, 6-1-1Shinjuku, Shinjuku-ku, Tokyo 160, Japan. COLLAGEN BUNDLES IN POSTPARTUM UTERUS 139 On postpartum day 5, collagen bundles reappeared in the subluminal compartment of the endometrium, where most of collagen bundles had been removed until postpartum day 3 (Figs. 4 and 10). The distribution of collagen bundles in the endometrium on postpartum day 10 (Fig. 5) was similar to the pattern of nulliparous animals RESULTS (Fig. 1). Organ Weight and Chemical Analysis Although muscle fascicles became thin during the Both the organ weight and total Hyp content de- postpartum period (compare Fig. 2 with Fig. 3), removal creased to the nulliparous level on day 10 post partum of collagen bundles around muscle fascicles was not (Table 1). Total Hyp content decreased more rapidly evident. than did organ weight during postpartum involution. DISCUSSION Decrease of organ weight was high from the day of The rate of collagenase turnover was measured by parturition to postpartum day 2 (Fs = 216, df 2,66, P < 0.01). The overall variance for the organ weight was analyzing tissue levels of Hyp from the day of parturisignificant (Fs = 247, df5, 118, P < 0.01). The weight tion to postpartum day 10. Collagenase activity was decreased by 60% on postpartum day 2 compared to the high during the first two postpartum days. This result day of parturition. After postpartum day 3 the weight agrees with biochemical studies (Jeffrey et al., 1971; Halme and Woessner, 1975). On the day of parturition, decreased gradually (Fs = 133, df 2,52, P < 0.01). Hydroxyproline content decreased by 90% on postpar- there is little collagenase, the majority being procollatum day 2 compared to the day of parturition (Fs = 162, genase that is not yet activated (Woessner, 1979; Roswit df 2, 66, P < 0.01). The overall variance for the Hyp et al., 1983).The principal cellular origin of procollagecontent was very significant (Fs = 227, df 5, 118, P < nase is obscure. Procollagenase synthesized by fibroblasts 0.01). After postpartum day 3 relatively little Hyp was is similar in structure to procollagenase in the postpartum uterus (Woessner, 1977). However, with regard to removed (Fs = 47, df2, 52, P < 0.01). this similarity, given simply that the structure of an enzyme is similar in various organs, it cannot be known Histological Analysis whether the enzyme is produced by the same cells in In the uterus of nulliparous animals, collagen bundles each organ. Fibroblasts synthesize collagen and prodistributed in the whole endometrium and around the collagenase simultaneously (Birkendal-Hansen et al., muscle fascicles in the myometrium (Fig. 1). 1976). As collagen molecules polymerize to form fibers, On the day of Parturition, collagen bundles existed in procollagenase may be woven into the collagen fibers. It the immediate subluminal compartment of the endome- is probable that there will be collagen-bound collagenase trium (Fig. 6). Although much Hyp was removed from in the postpartum uterus (Ryan and Woessner, 1971; the day of parturition to postpartum day 1, removal of Woessner and Ryan, 1973; Perez-Tamayo, 1975) under collagen bundles was not evident histochemically (Figs. these conditions. Immunohistochemical study (Montfort 2 and 7). On postpartum day 2, when most of Hyp was and Perez-Tamayo, 1975) shows that collagenase distribremoved, disappearance of collagen bundles was ob- utes in the whole uterus on the day of parturition. served in the immediate subluminal compartment of the Procollagenase is activated by collagenase-activating endometrium (Fig. 8). Removal of collagen bundles was proteases derived from serum under physiological conmore apparent in the immediate subluminal compart- ditions (Eeckhout and Vaes, 1977; Werb et al., 1977). ment than in the deep compartment of the endome- Passage of collagenase-activating proteases to uterine trium. On postpartum day 3, a majority of the collagen tissue from serum is induced by high vascular permebundles in the subluminal compartment of the endome- ability on the day of parturition (Shimizu and Hokano, trium were removed (Figs. 3 and 9). Some collagen bun- 1987). On the proestrous day, plasminogen activator acdles remained in the deep compartment of the tivity is high in the subluminal compartment of the endometrium (Fig. 3). endometrium but not in both the deep compartment of the endometrium and the myometrium (Kwaan and Albrechtsen, 1966). Estradiol in the plasma on the day of parturition is similar in concentration to that of the TABLE 1. Organ weight (mg) and total hydroxyproline proestrous day in the cyclic mouse (McCormack and content (mg) in nulliparous and postpartum uteri Greenwald, 1974). Conversion from plasminogen to plas(mean SE) min by plasminogen activator occurs most effectively in Organ Total the subluminal compartment of the endometrium. Thus, N weight % hydroxyproline % the discrete collagenolytic response of the suluminal compartment of the endometrium may reflect a n amount Nulliparous 10 55 & 2 (10) 3.6 f 0.3 (4) of plasmin that activates procollagenase. Whether other (postpartum) Parturition 29 576 f 23** (100) 92.8 f 4.0** (100) collagenase-activating proteases participate in activation of procollagenase during postpartum collagen degDav 1 2 1 379 f 21** (66) 50.3 f 3.1** (54) 7.9 f 0.3** (9) radation is obscure. 19 224 I 9 * * (39) D& 2 (28) 7.1 0.5** (8) Day 3 16 162 f 5** Most of the collagen bundles are extracellularly de(19) 5.0 0.3** (5) Day 5 19 110 f 5** graded (Shimizu and Maekawa, 1983). Procollagenase (4) (11) 3.4 f 0.2 20 62 3 Day 10 on the surface of collagen bundles may be activated by **P < 0.01 compared with nulliparous animals. Percentage of the plasmin initially. Activated collagenase may cleave colparturition value is represented in parentheses. lagen molecules in quick succession from the surface of et al., 1979). The stained sections were viewed by a light microscope with a polarizing filter (Olympus, BH-2, Tokyo). Data were analyzed by one-way analysis of variance or Student’s t test. 140 K. SHIMIZU AND M. HOKANO Figs. 1-5. Collagen bundle distribution. Picrosirius red polarization. ~ 5 0 E, . endometrium; L, uterine lumen; IM, inner muscle layer; OM, outer muscle layer. Fig. 3. Postpartum day 3, when collagen degradation has finished (Table 1).Removal of collagen bundles is most evident in the immediate subluminal compartment of the endometrium. Although muscle fascicles have thinned, removal of collagen bundles is not apparent. Fig. 1. Nulliparous animals. Collagen bundles distributed in the whole endometrium and around the muscle fascicles. day 5. Collagen bundles have reappeared in the Fig. 4, immediate subluminal compartment of the endometrium. Fig. 2. Postpartum day 1. Collagen bundles distributed similarly as in nulliparous animal. Fig. 5. Postpartum day 10. In the endometrium, distribution of collagen bundles is similar to the pattern of nulliparous animals Pig. 1). COLLAGEN BUNDLES IN POSTPARTUM UTERUS 141 Fig. 6-10. Collagen bundle distribution. Picrosirius red polarization. x260. EP, luminal epithelium; UG, uterine gland. Fig. 8. Postpartum day 2. Collagen bundles are degraded from the immediate subliminal compartment of the endometrium. Fig. 6. Day of parturition. Collagen bundles occur in the subluminal compartment of the endometrium. Fig. 9. Postpartum day 3.Most of the collagen bundles in the immediate subluminal compartment of the endometrium have disappeared. Fig. 7. Postpartum day 1. Although collagen degradation is high, removal of collagen bundles is not apparent. Fig. 10. Postpartum day 5. Collagen bundles have reappeared in the immediate subluminal compartment of the endometrium. 142 K. SHIMIZU AND M. HOKANO collagen bundles. Since small peptides derived from the collagen degradation are rapidly removed by the blood stream (Shimizu et al., 1986), this hypothesis would provide a mechanism for the extremely rapid and extensive (90%)breakdown of collagen during postpartum uterine involution. After most of the collagen bundles are removed from the postpartum uterus, collagenase activity is low (Woessner, 1977, 1979). The reappearance of collagen bundles in the subluminal compartment of the endometrium indicates that collagen synthesis exceeds its degradation. On postpartum day 10, the distribution of collagen bundles became similar to the pattern of nulliparous animals. Small peptides derived from the collagen molecule are known to be chemotactic for human monocytes in vitro (Postlethwaite and Kang, 1976). Such endogenous antigenic stimulation by fragments of collagen molecules would be local, strong, and temporary and might lead to the cellular manifestations of transient inflammation and/or humoral immune response (Padykula, 1976). For example, macrophages disappear from the endometrium after collagen degradation has finished (Padykula, 1976; Padykula and Campbell, 1976; Padykula and Taylor, 1976). The cellular and tissue manifestations of inflammation and local humoral response might represent normal regulatory mechanisms involved in the extracellular compartment of the endometrial stroma. Moreover, these regulatory mechanisms are most likely influenced by plasma and tissue levels of estrogen and progesterone (Padykula, 1981). LITERATURE CITED Birkendal-Hansen, H., C.M. Cobb, R.E. Taylor, and H.M. Fullmer (1976) Synthesis and release of procollagenase by cultured fibroblasts. J. Biol. Chem., 251~3162-3168. Constantine, V.S., and R.W. Mowry (1968) The selective staining of human dermal collagen. 11. The use of picrosirius red F3BA with polarization microscopy. J. Invest. Dermatol., 50~419-423. Eeckhout, Y., and G. Vaes (1977)Further studies on the activation of procollagenase, the latent precursor of bone collagenases. Effect of lysosomal cathepsin B, plasmin and kallikrein, and spontaneous activation. Biochem. J., I66:21-31. Halme, J., and J.F. Woessner, Jr. (1975) Effect of progesterone on collagen breakdown and tissue collagenolytic activity in the involuting rat uterus. J. Endocrinol., 66~357-362. Harkness, R.D. (1964) The physiology of the connective tissue of the reproductive tract. Int. Connect. Tissue Res., 2t155-211. Jeffrey, J.J., R.J. Coffey, and A.Z. Eisen (1971) Studies on uterine collagenase in tissue culture. I. Relationship of enzyme production of collagen metabolism. Biochim. Biophys. Acta, 252t136-142. Junqueira, L.C.U., G. Bignolas, and R.R. Brentani (1979) Picrosirius staining plus polarization microscopy, a specific method for collagen detection in tissue sections. Histochem. J., 11t447-455. Kwaan, H.C., and O.K. Albrechtsen (1966) Histochemical study of fibrinolytic activity in the uterus in normal and hormonally induced estrus. Am. J. Obstet. Gynecol., 95~468-473. McCormack, J.T., and G.S. Greenwald (1974) Progesterone and estradiol-17P concentration in the peripheral plasma during pregnancy in the mouse. J. Endocrinol., 62:lOl-107. Montfort, I., and R. Perez-Tamayo (1975) The distribution of collagenase in the rat uterus during postpartum involution. An immunohistochemical study. Connect. Tissue Res., 3:245-252. Padykula, H.A. (1976) Cellular mechanisms involved in cyclic stromal renewal of the uterus. 111. Cells of the immune reponse. Anat. Rec., 184~49-71. Padykula, H.A. (1981)Shifts in uterine stromal cell populations during pregnancy and regression. In: Cellular and molecular aspects of implantation. S.R. Glasser and D.W. Bullock, eds. Plenum, New York, pp. 197-216. Padykula, H.A., and A. Campbell (1976)Cellular mechanisms involved in cyclic stromal renewal of the uterus. 11. The albino rat. Anat. Rec., 184~2748. Padykula, H.A., and T.M. Taylor (1976) Cellular mechanisms involved in cyclic stromal renewal of the uterus. I. The oppossum, Didelphis uirgiana Anat. Rec., 184:5-25. Perez-Tamayo, R. (1975)Pathology of collagen degradation. A review. Am. J. Pathol., 92t509-566. Perez-Tamayo, R., and I. Montfort (1980)The susceptibility of hepatic collagen to homologus collagenase in human and experimental cirrhosis of the liver. Am. J. Pathol., 100~427442. Poslethwaite, A.E., and A.H. Kang (1976) Collagen and collagen peptide-induced chemotaxis of human blood monocytes. J. Exp. Med., 143~1299-1307. Roswit, W.T., J. Halme, and J.J. Jeffrey (1983) Purification and properties of rat uterine collagenase. Arch. Biochem. Biophys., 225285295. Ryan, J.N., and J.F. Woessner, Jr. (1971) Mammalian collagenase: Direct demonstration in homogenates of involuting rat uterus. Biochem. Biophys. Res. Commun., 44t144-149. Shimizu, K., and M. Hokano (1987) Collagen concentration and mast cell in perinatal murine uterus. Acta Anat., 129262-264. Shimizu, K., and K. Maekawa (1983) Collagen degradation in the mouse uterus during postpartum involution: extracellular pathway. Acta Anat., 117:257-260. Shimizu, K., K. Honda, S. Takabe, and M. Hokano (1986)Hydroxyproline concentration in soluble and insoluble material from seium treated with trichloroacetic acid in postpartum mice. Experientia, 42319-820. Sweat, F., H. Puchtler, and S.I. Rosenthal(1964) Sirius red F3BA as a stain for connective tissue. Arch. Pathol., 78t69-72. Werb, Z., C.L. Mainardi, C.A. Vater, and E.D. Harris, Jr. (1977)Endogenous activation of latent collagenase by rheumatoid synovial cells: Evidence for a role of plasminogen activator. New Engl. J. Med., 296:1017-1023. Woessner, J.F., Jr. (1961)The determination of hydroxyproline in tissue and protein samples containing small proportions of this imino acid. Arch. Biochem. Biophys., 93t440-447. Woessner, J.F., Jr. (1977)A latent form of collagenase in the involuting rat uterus and its activating by a serine protease. Biochem. J., 161535-542. Woessner, J.F., Jr. (1979) Total latent and active collagenase during the course of post-partum involution of the rat uterus. Effect of oestradiol. Biochem. J., 18Ot95-102. Woessner, J.F., Jr., and J.N. Ryan (1973) Collagenase activity in homogenates of the involuting rat uterus. Biochim. Biophys. Acta, 309t397-405. Zambrano, N.Z., G.S. Montes, K.M. Shigihara, E.M. Sanchez, and L.C.U. Junqueira (1982) Collagen arrangement in cartilages. Acta Anat., 113~26-38.