Persistent vaginal cornification in mice treated with estrogen prenatally.

Twelve of 14 female mice of the ICR strain which had received a single injection of 50 mug estradiol-17beta on day 17 of fetal life exhibited irreversible cornification or stratification of the vaginal epithelium which persisted after ovariectomy until sacrifice performed 42-48 days later. Eight of the 12 mice had corpora lutea in their ovaries removed at 3-5 months of age. A similar injection of estradiol on day 15 of fetal life induced irreversible cornification or stratification of the vaginal epithelium in 6 of 12 females and only one of the 6 had corpora lutea in its ovaries when removed at 3-5 months. Mice given the same dose of estradiol on the day of birth or at 3 days of postnatal age invariably had ovaries bearing follicles of varying sizes and hypertrophied interstitial tissue but no corpora lutea. Changes in the vaginal epithelium in these animals were less remarkable as compared to that in prenatally treated mice.     

Effects of gonadotrophin on the ovary of Apodemus flavicollis in winter anoestrus

Wild-caught female Apodemus flavicollis were given daily subcutaneous injections of PMSG + HCG on two consecutive days (2xPMSG + HCG), HCG on two days (2 X HCG), PMSG + HCG until their vagina became perforate (Perforate), and PMSG + HCG until they became perforate and were left undisturbed for another five days (Perforate + 5). Untreated females served as controls. The ovaries and uteri increased in weight as a result of the treatment. The number of healthy follicles increased in "perforate" females and decreased in "2 X HCG". Luteinization of the follicles and possibly of the interstitial tissue was seen in "2 X HCG". In this group, the mean diameter of the corpora lutea increased, and it decreased in "perforate" and "perforate + 5". The mean number of corpora lutea was unaffected by the treatment. The activity of 3 beta hydroxysteroid dehydrogenase (3 beta HSD) was strong in the corpora lutea of the untreated females and remained so throughout treatment. The interstitial tissue of the untreated females showed only weak 3 beta HSD activity, but the activity was strong in all the experimental groups. One female ovulated, but it is not clear if it was in response to treatment.     

3 Alpha-hydroxysteroid dehydrogenase and 3 beta-hydroxysteroid dehydrogenase in the ovary of young Mongolian gerbils

The ovaries of sexually mature, pregnant mare serum gonadotropin (PMSG) stimulated, 12 week old Mongolian gerbils were investigated morphologically and enzyme histochemically for the appearance of the 3 alpha-hydroxy-steroid and the 3 beta-hydroxysteroid dehydrogenase during the estrous cycle. Up to ovulation, on day 3 of the estrous cycle, the number of vesicular follicles increases continuously. Primarily atretic follicles can be seen on day 4. On day 5 corpora lutea appear, but they degenerate already by day 6. During the entire estrous cycle, 3 alpha-hydroxysteroid dehydrogenase and 3 beta-hydroxysteroid dehydrogenase activity can be found in the theca of tertiary follicles and in the interstitial cells, whereas the theca of secondary follicles and the granulosa of healthy follicles do not exhibit any enzyme activity. The activity decreases from day 1 till day 6. The granulosa of atretic follicles and the cells of corpora lutea show only weak activity. It may be significant that the intensity of enzyme activity in the ovary and the estrogen level in the plasma are differently correlated to the estrous cycle.     

The effects of an aromatase inhibitor and a 5 alpha-reductase inhibitor upon the occurrence of polyovular follicles, persistent anovulation, and permanent vaginal stratification in mice treated neonatally with testosterone

Female C57BL/Tw mice were given 5 daily injection of 20 micrograms testosterone (T), 100 micrograms 4-hydroxy-4-androstene-3, 17-dione (4-HA), 100 micrograms 6-methylene-4-pregnene-3, 20-dione (6-MP), 4-HA + 6-MP, T + 4-HA, T + 6-MP, and T + 4-HA + 6-MP starting on the day of birth. The animals were ovariectomized at 30 days or 90 days of age and were killed at 150 days. The incidence of polyovular follicles (PF) at 30 days of age was significantly increased by neonatal treatment with T. By contrast, the PF incidence was significantly reduced by injections of 4-HA given simultaneously with T. Neonatally T- or T + 6-MP-treated 90-day-old mice had ovaries containing follicles and hypertrophied interstitial cells but no corpora lutea. By contrast, T + 4-HA (64%)- and T + 4-HA + 6-MP (82%)-treated mice had ovaries with corpora lutea. In the T + 4-HA + 6-MP (82%)-treated mice had ovaries with corpora lutea. In the T + 4-HA + 6-MP-treated, 150-day-old, ovariectomized mice, the number of mice showing vaginal epithelial stratification was significantly decreased as compared with T-treated mice. There were no significant differences in the number of layers, thickness, and mitotic rate of vaginal epithelium of T-treated mice compared with mice treated with T + 4-HA, T + 6-MP, or T + 4-HA + 6-MP. The present results indicate that development of PF and persistent anovulation are due to the direct action of estrogen (E) derived from T upon neonatal ovarian follicles and the neonatal hypothalamo-hypophysial system, and that T itself can induce ovary-independent vaginal changes, although 5 alpha-reduced androgen and estrogen derived from T seem to be more effective in this regard.     

Immunocytochemical localization of aromatase in immature rat ovaries treated with PMSG and hCG,and in pregnant rat ovaries

Summary Immunocytochemical localization of aromatase cytochrome P-450 was examined in immature rat ovaries treated with pregnant mare serum gonadotropin (PMSG) and human chorionic gonadotropin (hCG), and in pregnant rat ovaries. It is well known that PMSG and hCG treatments induce ovulation about 12 h after hCG injection.At 24 h after hCG injection, many antral follicles were recognized in immature rat ovaries and only the granulosa cells in the antral follicles were stained weakly with the anti-aromatase antibody. At 0 to 9 h after hCG injection, in addition to the antral follicles, some large Graafian follicles could be observed in the rat ovaries, and the granulosa cells of these follicles were positively stained for aromatase. Each follicle was surrounded by the basal lamina which shows lineally distinct positive reaction against anti-laminin antibody. At 12 h after hCG injection, some large Graafian follicles without oocyte were weakly positive to the anti-aromatase antisera, and the outline of their basal lamina stained with anti-laminin antibody became irregular in shape and fragmentous. At 15 to 18 h after hCG injection, the luteinized cysts could be seen, and the granulosa-lutein cells of these cysts were almost negative for aromatase. Fragmentous reaction to the anti-laminin antibody was observed around the luteinized cysts.In the ovaries of day 4 in pregnancy, only the granulosa cells of the large antral follicles were weakly stained, but corpora lutea negatively reacted to the anti-aromatase antibody. At 7 to 19 days in gestation, both the granulosa cells of antral follicles and pregnant luteal cells were positively stained against aromatase antisera. The luteal cells were increased in size during pregnancy. And weakly positive reaction was detected on day 7 of pregnancy, then the immunoreaction became stronger in the corpora lutea on day 15 and 19 of pregnancy.The localization of aromatase was immunocytochemically examined in immature rat ovaries treated with PMSG and hCG injection, and the reaction of the granulosa cells of the antral follicles against anti-aromatase antibody became strongly positive about 12 h before ovulation and the became very weak suddenly after ovulation. In rat-ovaries, the pregnant corpora lutea was positively stained for aromatase after day 7 of pregnancy.This study was supported by Grants from the Ministry of Education, Science and Culture, Japan, and from USPHS Research Grants HD04945, USA     

Studies of the genital organs of gilts culled for anoestrus

The genital organs from 54 gilts, slaughtered because of failure to exhibit oestrus, were examined. The mean age of the gilts at slaughter time was 10.8 months. The ovaries of 19 gilts (35.2%) contained no luteal tissue. The ovaries of the other gilts contained luteal tissue as solid corpora lutea only or in combination with cystic corpora lutea and/or luteinized cysts. The average age of the gilts was highest in the latter group and lowest for gilts with ovaries containing only small follicles. Bacteriological and histopathological examinations did not indicate an infectious cause of the condition.     

Immunocytochemical studies of relaxin in ovaries of pregnant and cycling mice

Immunocytochemical staining for relaxin in ovarian sections of pregnant mice from day 11 through day 18 of gestation revealed that only corpora lutea (CL) of pregnancy are stained. Evaluation of serial sections of ovaries from a day 16 pregnant mouse revealed that the only luteal structures present are CL of pregnancy. The number of CL present in each ovary equaled the number of implantation sites in each related horn (7 on the right side and 8 on the left side). These large CL varied in shape, being round in some profiles to very elongate in others. All CL were immunochemically stained for relaxin using the peroxidase-antiperoxidase method of L. Sternberger (Immunocytochemistry, 2nd ed. Wiley, New York, 1979). The intensity of the strain varied from cell to cell within each CL. Small luteal structures that were observed to be immunochemically stained for relaxin were demonstrated to represent the periphery of CL of pregnancy. No luteinized follicles were observed and interstitial cells and follicles were not immunochemically stained in any of the day 16 serial ovarian sections or in any of the ovarian sections from pregnant mice on the other days of gestation studied. CL of previous cycles were not observed to be present in the ovaries at days 15, 16, or 18 of gestation. However on day 14 and before, CL of previous cycles were observed and they did not exhibit any relaxin immunostaining. Immunocytochemical studies using the biotin-avidin system revealed that no relaxin immunostaining could be demonstrated in the ovaries of cycling mice at any stage of the estrous cycle. In conclusion, this study revealed that the only ovarian structures demonstrating relaxin immunocytochemical staining in the mouse were CL of pregnancy.     

Localization of some steroidogenic enzymes in the ovary of the rabbit

The distribution of delta 5-3 beta-HSD, peroxidase and cytochrome oxidase in immature, sexually mature and pregnant rabbit ovary has been studied histochemically. Corpora lutea are found only in pregnant rabbits. delta 5-3 beta-HSD is present in the theca interna of mature follicles, corpora lutea and interstitial gland cells but is absent in the granulosa cells of both developing and mature follicles. The granulosa cells of mature and developing follicles, hypertrophied theca interna and the luteal cells all show intense cytochrome oxidase activity. Peroxidase is present in the corpora lutea only. It is suggested that delta 5-3 beta-HSD in the theca interna and interstitial gland cells is the enzyme responsible for steroid synthesis in the ovaries of immature as well as sexually mature rabbits, while peroxidase and delta 5-3 beta-HSD present in the corpora lutea together regulate luteal steroidogenesis during pregnancy. The intense cytochrome oxidase activity together with peroxidase and delta 5-3 beta-HSD confirms the observations that this tissue is a site of intense oxidative activity.     

Immunocytochemical localization of oxytocin and neurophysin in human corpora lutea

Corpora lutea, corpora albicantia, and ovarian stroma from normal human premenopausal ovaries were examined for the presence of oxytocin and neurophysin by using highly specific antisera and peroxidase-antiperoxidase light-microscopic immunohistochemistry. Oxytocin and neurophysin immunoreactivity was found in some but not all cells of the corpora lutea obtained on days 19 to 24 of the menstrual cycle. Stromal tissue and corpora albicantia did not give a positive reaction for either of these peptides, and negative results were also obtained with corpora lutea of mid- and term-pregnancy and preovulatory follicles. Specificity of the immunohistochemical reaction was confirmed by immunoabsorption tests. The specific localization of immunoreactive oxytocin and neurophysin in corpora lutea of the human menstrual cycle directly demonstrates the presence of oxytocin- and neurophysin-positive cells within the human corpus luteum.     

Relationships of liver weight, cholesterol, albumin and alpha2-macroglobulin concentrations with ovarian function in swine

In two genetic swine models selected for diversity in ovulation rates (White composite controls and ovulation rate selection line, n = 131; 1/2 White composite: 1/2 Meishan crossbreds, n = 387), a positive relationship was established with liver weight and ovulation rate (P < 0.01). Serum changes of cholesterol, albumin and alpha2-macroglobulin were monitored during various stages of the luteal phase and follicular phase (days 17 and 19 of the estrous cycle; 1/2 White composite: 1/2 Meishan gilts). Serum cholesterol concentrations increased with liver weights (r = 0.19; P < 0.01) and corpora lutea numbers (r = 0.14; P < 0.01). Albumin concentrations were negatively correlated with corpora luteal numbers (r = -0.3; P < 0.01) but had no relationship with liver weight. Serum concentrations of alpha2-macroglobulin were not related to liver weight or corpora lutea numbers. Circulating concentrations of cholesterol and alpha2-macroglobulin increased with day of the estrous cycle (P < 0.01). Testosterone concentrations were inversely related to circulating cholesterol concentrations during the estrous cycle, but testosterone concentrations on day 17 or 19 of the cycle were unrelated to corpora lutea numbers. Concentrations of estrone on day 17 or 19 (as an index of follicles destined to ovulate) were also not related to numbers of corpora lutea. Many interactions between liver and ovarian function involving metabolic and endocrine systems are plausible, but defined mechanisms resulting in coordinate increases in liver weight and ovulation rates are presently unelucidated.     

Ovarian response to pregnant kare serum gonadotrophin and prostaglandin F(2) proportional, variant in Africander and Mashona cows

Oestrus was synchronised in ten Africander and eight Mashona mature dry cows by two injections of prostaglandin F(2) proportional, variant (PG) 11 days apart. Half the cows of each breed received an injection of 3000 i.u. pregnant mare serum gonadotrophin (PMSG) two days prior to the second PG injection. All cows were observed for the incidence of cestrus, and blood samples were taken at intervals for progesterone assay. Cows were slaughtered 11 days after the second PG injection and their reproductive tracts examined. Treatment with PMSG increased numbers both of corpora lutea and of follicles more than 10 mm in diameter. When numbers of corpora lutea and follicles were considered together, the response to treatment was significant in the Africanders (P<0,01) and markedly greater than that of Kashona cows. The concentration of progesterone in plasma on the day before slaughter was significantly correlated with the mass of corpora lutea (P<0,001), total mass of ovaries (P<0,001), but not with numbers of corpora lutea. It is suggested that generally Africander cows may secrete lower levels of follicle stimulating hormone and oestrogen than kashona cows during normal cyclic sexual activity.     

Ovarian activity and peripheral plasma levels of oestrogens and progesterone in the lactating sow

Ten gilts were examined for peripheral plasma levels of progesterone and oestrogens 3 weeks before and up to 8 weeks after parturition. The sows were slaughtered at different intervals after parturition and the ovaries were examined. Peripheral plasma levels of progesterone decreased dramatically from about 8ng/ml two days before parturition to about 2 ng/ml on the day before parturition. After parturition the mean progesterone level was about 1.5 ng/ml. Maximum oestrone levels of about 7 ng/ml were obtained two days before parturition. After parturition the level dropped to below 0.1 ng/ml. Three sows showed high levels of oestradiol (75–440 pg/ml) without signs of heat during the lactation. In no case were ovulated follicles or periodic corpora lutea registered.     

The expression of androgen receptor, cytochrome P450 aromatase and FSH receptor mRNA in the porcine ovary

The aim of this study was to visualize the expression of androgen receptor, cytochrome P450 aromatase and FSH receptor mRNAs in various structures of porcine ovary. Porcine ovaries were frozen in liquid nitrogen, and 8 microm sections were prepared for in situ hybridization. In the small, medium and large antral follicles as well as in early, midluteal and regressing corpora lutea, mRNAs for androgen receptor, P450 aromatase and FSH receptor were detected. In small antral follicles high levels of mRNAs for androgen and FSH receptors were observed, mainly in the granulosa layer, while mRNA expression for P450 aromatase was negligible. As follicles grew, amount of mRNAs for androgen receptor and FSH receptor decreased, and that for P450 aromatase increased. Small amounts of androgen receptor mRNA were also present in corpora lutea at all examined stages. P450 aromatase mRNA was not detected in early and midluteal corpora lutea. However, regressing corpus luteum showed a weak expression of aromatase mRNA.     

Characterization of integrin expression in the mouse ovary

Integrin alpha:beta heterodimers mediate cell contacts to the extracellular matrix and initiate intracellular signaling cascades in response to a variety of factors. Integrins interact with many determinants of cellular phenotypes and play roles in controlling the development, structural integrity, and function of every type of tissue. Despite their importance, little is known about the regulation of integrin subunits in the mammalian ovary and how they function in folliculogenesis. To determine their relevance to ovarian physiology, we have studied the expression of integrin subunit mRNAs by Northern blot analysis and in situ hybridization in ovaries of wild-type, growth differentiation factor 9 (Gdf 9) knockout, FSHbeta (Fshb) knockout, and inhibin alpha (Inha) knockout mice. Integrin alpha6 mRNA is expressed in oocytes and granulosa cells of single-layer follicles and in oocytes and theca cells of multilayer follicles. Integrin alpha6 is highly expressed in Gdf 9 knockout ovaries, which are enriched in oocytes and primary (single layer) follicles because of a block at this stage of follicular development. Integrin alpha(v) mRNA is most highly expressed in the granulosa cells of multilayer growing follicles, and therefore only low levels of expression are detectable in the Gdf 9 knockout ovaries. Integrin beta1 mRNA exhibits a broad expression pattern in ovaries, including oocytes, granulosa cells, theca cells, and corpora lutea. Integrin beta3 mRNA is expressed in theca and interstitial cells and is upregulated in corpora lutea. It is nearly undetectable in ovaries of Fshb knockout mice, which develop preantral follicles but have no luteal cells. Integrin beta5 mRNA is predominantly expressed in granulosa cells of multilayer follicles. It is expressed at high levels in the Fshb knockout mice and in a compartmentalized manner in the granulosa cell/Sertoli cell tumors that develop in the Inha knockout mice. Specific integrins are associated with ovarian cellular phenotypes in mice, which raises intriguing possibilities as to integrin functions in oocyte competence, follicular development, luteinization, and granulosa cell proliferation.     

Failure of two progesterone antagonists, mifepristone and onapristone, to affect luteal activity in lactating rats

The progesterone antagonists, mifepristone (RU-38,486) and onapristone (ZK-98,299), given as 2 mg daily, did not markedly affect lactation in rats. Both litter growth and time spent by 10-pup litters attached to their mothers were similar in antagonist-treated mothers and in solvent-treated controls. The progesterone antagonists did not affect the steroid content in corpora lutea remaining from the preceding pregnancy. Corpora lutea formed after post-partum ovulation also showed nearly normal function throughout the first 17 days of lactation. It is concluded that progesterone itself plays no role in the initiation or maintenance of luteal function when prolactin secretion is governed through an action independent of the ovaries, as through suckling. Antagonist-treated rats ovulated around Day 13 of lactation despite suckling. This ovulation was not associated with a decrease of progesterone production by the corpora lutea formed after post-partum ovulation. Apparently, elimination of progesterone action may protect corpora lutea from luteolysis. The latter finding indicates a possible role of progesterone in luteolysis and deserves further analysis.     

Effect of high doses of retinol acetate on the 3-beta-ol-steroid dehydrogenase and alkaline phosphatase content in mouse ovaries

The author reports on the effects of different doses of retinol acetate on ovarian steroidogenesis. Two groups of CBA/C57BL mice with a mean body weight of 18-20 g received 3.44% oily retinol acetate per os in daily doses of 50 000 and 80 000 IU for 10 days. After completion of the experiments the quick-frozen sections of the ovaries were subjected to a histochemical assay for the content of 3-beta-ol-steroid dehydrogenase and alkaline phosphatase. Administration of 50 000 IU vitamin A was found to stimulate ovarian steroidogenesis. The effect of vitamin A was the most demonstrable in the interstitial tissue, atretic corpora, and, in the internal theca of the follicles. Administration of 80 000 IU retinol acetate inhibited ovarian steroidogenesis. The estrous cycle in animals ceased. Administration of vitamin A (80 000) primarily affected the follicular apparatus of the ovaries, namely the epithelium of the follicles and yellow bodies. At the same time secretory function of atretic corpora and interstitial tissue remained within normal, which was regarded as a compensatory-adaptive mechanism under toxic hypervitaminosis A.     

Evidence for the existence of substance P in the prepubertal rat ovary. II. Immunocytochemical localization

Nerve fibers containing substance P (SP) were localized in ovaries from juvenile and peripubertal rats by immunofluorescence. These fibers were closely associated with the theca externa of antral follicles, as well as being in the interstitial tissue and within the tunica adventitia of small blood vessels, mostly arterioles. Consistently, the greatest amount of SP immunoreactivity was observed surrounding the ovarian vasculature. Substance P was not detected in cells or within the corpora lutea (CL). Additionally, the peripubertal animals seemed to have a greater concentration of ovarian SP than the juvenile animals. Possible functional roles for this peptide in the ovary are discussed.     

Changes in the expression of cytochrome P450c17 associated with ovarian cystic follicles. An immunocytochemical and enzymatic analysis of porcine ovaries.

The steroidogenic activity of normal preovulatory and cystic follicles, and corpora lutea of porcine ovaries was investigated by immunocytochemical and radioenzymatic techniques. Using a specific antibody to porcine cytochrome P450c17, immunocytochemical staining was specifically localized in the theca interna layer of normal follicles and undetectable in the granulosa layer. The theca interna layers of non-luteinized cystic follicles were immunoreactive while those of luteinized follicles were not. Corpora lutea cells were essentially negative. The 3 beta-hydroxysteroid dehydrogenase/delta 5-delta 4 isomerase activity was similar in luteinized cystic follicular and corpora lutea tissues, which had 8 times higher activity than found in normal preovulatory follicles. The formation of either corpora lutea or luteinized cysts led to a profound decline (12- to 15-fold) in 17 alpha-hydroxylase and 17,20 lyase activities compared to normal preovulatory follicles. In agreement with these enzyme findings, radioimmunoassays revealed very high levels of progesterone with nearly undetectable levels of androgens in the luteinized cysts. These studies demonstrate the functional similarities between cells of luteinized cysts and those of normal corpora lutea and suggest a pathology associated suppression of P450c17 expression in porcine cystic follicles.     

Fibrinolytic activity in ovaries of rats and golden hamsters after gonadotrophic stimulation and hypophysectomy

The fibrinolytic activity (FA) in immature ovaries of rats was compared with that of golden hamsters on days 3 and 4 after stimulation with pregnant mares' serum gonadotrophin (PMSG). In addition, FA of immature PMSG-stimulated rats was compared with that of adult rats 18 days after hypophysectomy. By means of the fibrin slide method, FA was localized in serial sections and evaluated semiquantitatively. On PMSG day 3, FA occurred in preovulatory follicles and developing corpora lutea. On PMSG day 4, FA was mostly localized in medullary vessels. This was also the case for hypophysectomized rats. While both species showed a comparably high FA on PMSG day 3, hamsters had a lower FA than rats on PMSG day 4. Similar low FA values were obtained in hypophysectomized rats. Non-vascular FA localization may be attributed to granulosa and thecal cells of preovulatory follicles and to luteinizing cells of developing corpora lutea. Vascular FA localization is ascribed to capillary sprouting in these two structures and to medullary vessels.     

The effect of pre-ovulatory anaesthesia on ovulation in laparoscopically inseminated domestic cats.

Laparoscopic intrauterine artificial insemination (AI) of electroejaculated spermatozoa was used to compare embryo development and conception rates in domestic cats inseminated either before or after ovulation. Females were given a single (100 iu) injection of pregnant mares' serum gonadotrophin (PMSG) followed by either 75 or 100 iu human chorionic gonadotrophin (hCG) 80 h later. Cats were anaesthetized (injectable ketamine HCl/acepromazine plus gaseous halothane) 25-50 h after administration of hCG for laparoscopic assessment of ovarian activity and for transabdominal AI into the proximal aspect of the uterine lumen. At the time of AI, 23 cats were pre-ovulatory (25-33 h after hCG injection) and 30 were post-ovulatory (31-50 h after hCG injection). Pre-ovulatory females produced 10.5 +/- 1.1 follicles and no corpora lutea compared with 1.9 +/- 0.5 follicles and 7.5 +/- 0.9 corpora lutea for the post-ovulatory group (P < 0.05). Six days later, the ovaries of nine pre-ovulatory and 12 post-ovulatory females were re-examined and the reproductive tracts flushed. On this day, pre-ovulatory cats produced fewer corpora lutea (2.8 +/- 1.5; P < 0.05) and embryos (0.4 +/- 0.3; P < 0.05) than post-ovulatory females (18.9 +/- 3.3 corpora lutea; 4.6 +/- 1.2 embryos). Two of the 14 cats (14.3%) inseminated before ovulation and not flushed became pregnant compared with 9 of 18 cats (50.0%) inseminated after ovulation and up to 41 h after hCG injection (P < 0.05). These results indicate that ovulation in cats is compromised by pre-ovulatory ketamine HCl/acepromazine/halothane or laparoscopy or by both and that electroejaculated spermatozoa deposited by laparoscopy in utero, after ovulation, result in a relatively high incidence of pregnancy. Because ovulation usually occurs 25-27 h after injection of hCG, the lifespan for fertilization of the ovulated ovum appears to be at least 14 h in vivo in cats.