PMC:5608921 / 4665-14613 JSONTXT

{"project":"2_test","denotations":[{"id":"28846079-12525422-79081914","span":{"begin":4639,"end":4641},"obj":"12525422"},{"id":"28846079-18548112-79081915","span":{"begin":5892,"end":5894},"obj":"18548112"},{"id":"28846079-18548112-79081916","span":{"begin":7261,"end":7263},"obj":"18548112"},{"id":"28846079-18548112-79081917","span":{"begin":8501,"end":8503},"obj":"18548112"},{"id":"28846079-18548112-79081918","span":{"begin":9365,"end":9367},"obj":"18548112"}],"text":"Results\n\nGrowth induction without changing the estrous cycle-associated pattern in Tg(K6b-E6/E7) cervix\nKeratin gene expression in the cervix of wild-type (WT) mice significantly changed during the estrous cycle (Figure 1a). Except for K5, the genes encoding all keratins tested increased during the diestrus–proestrus transition. K14 and K10 expression increased about fivefold, whereas K6b and K16 expression increased between 10- and 20-fold; the lowest mRNA levels were reached at metestrus. Similar pattern was observed in Tg(K6b-E6/E7) mice but upregulation was much higher than that determined in WT mice reaching up to 100-fold for K6b and K16 and 5- to 10-fold for K14 and K10 (Figure 1a); in the cervix of Tg(K6b-E6/E7) mice, mRNA levels at metestrus were higher than those in equivalent samples of WT mice. As expected, E6/E7 expression correlated with the pattern observed for K6b (Figure 1b).\nThe E6/E7 expression pattern along the estrous cycle and distribution of E7 protein in the cervix at proestrus was similar to that of K6b and restricted to suprabasal layers in either outer or inner cervix (Figure 1c); oncogene proteins were more abundant in the most suprabasal layers. Distribution of K5, the K14 companion keratin usually located in the basal layer, was not altered by oncogenes in the outer cervix, where it showed an evident basal–suprabasal gradient (Figure 2a). However, this gradient was only detected in the thicker than WT transformation zone and inner cervix epithelia of Tg(K6b-E6/E7) mice (Figure 2a).\nIn agreement with the expected effect and expression pattern of HPV oncogenes in the cervix of Tg(K6b-E6/E7) mice, the increase in proliferating cells was restricted to suprabasal layers (Figures 2a and b; see also Figures 3e and f). Actually no Ki67+ cell was found in the suprabasal layer of WT mice and the few BrdU+ cells found (1 for every 10 basal) could be recently differentiated cells. These observations correlated with a higher growth of cervical epithelium of Tg(K6b-E6/E7) in comparison with that of WT mice, evident during proestrus, where epithelial protrusions into the stroma were noted (Figure 2c). Despite these abnormalities, the epithelium of cervix established the major characteristics of each estrous phase (Figure 2c).\n\nIncreased number of cycles and longer proestrus–estrus phase in Tg(K6b-E6/E7) mice\nIn order to obtain indications of estrous cycle progression, we compared Tg(K6b-E6/E7) and WT mice regarding the number of cycles, as determined by the frequency of sequential proestrus–estrus phases, and the time staying at the growth phases, as determined by the frequency of proestrus and estrus, both within a defined number of days (19 days). Notably, the frequency of cycles (Figure 3a) and, apparently, the incidence of growth phases (Figures 3b–d) were higher in Tg(K6b-E6/E7) than in WT mice. Nonetheless, the initiation of each cycle was still dependent on estradiol as cycling was blocked in both Tg(K6b-E6/E7) and WT mice treated with raloxifene (Figures 3c and d).\nCytological analysis after a short-term β-estradiol (E2) treatment showed a tendency of mice to stay in the growth phases of the estrous cycle (Figures 3c and d). This result suggests a possible effect of estradiol on epithelium during the growth phases. To test this hypothesis, mice were treated with raloxifene once proestrus had initiated and, then, cell proliferation evaluated by the presence of Ki67 (Figure 3e). Despite the higher number of suprabasal proliferating cells due to oncogene expression, reduction in the number Ki67+ cells was observed in both WT and Tg(K6b-E6/E7) mice with a similar effect in basal/parabasal and suprabasal cells (Figure 3f); no indication of cell death was detected in the presence of raloxifene (Supplementary Figure S1). These results confirm the critical role of estradiol in epithelial growth and support the cooperation with oncogenes during the proestrus–estrus phase.\nIn contrast with short-term E2 treatments, proliferating cells in suprabasal layers of WT cervical epithelia were detected (about 28% of total Ki67+ cells) after long-term treatments (\u003e1.5 months) and, as expected, many more of those were detected (about 44% of total Ki67+ cells) in the cervical epithelia of Tg(K6b-E6/E7) mice (Figure 4a). Accordingly, long-term E2 treatments thickened the epithelia of the transformation zone of WT cervix (n=4), though a more pronounced growth in this region occurred in Tg(K6b-E6/E7) mice (n=3; Figure 4b). Interestingly, as a possible indication of a precancerous condition, koilocyte-like cells (i.e., squamous cells with a clear cytoplasm22) were found in the grown cervical epithelia of both WT and Tg(K6b-E6/E7) mice, with an evident increased number in the latter (Figure 4c). In addition, a high frequency of bi-nucleated cells was found in vaginal smears of E2-treated Tg(K6b-E6/E7) mice (Figure 4d). In WT mice, even after 9 months of E2 treatment, hyperplasia was the major phenotype of cervical epithelium. In contrast, in Tg(K6b-E6/E7) mice, hyperplasia and an irregular epithelium resembling a dysplasic tissue were frequently observed even in the absence of the hormone, as it was described above, but E2 treatment notoriously exacerbated the hyperplastic phenotype showing deep dysplasic protrusions and epithelial islands that commonly characterize the carcinoma in situ (Figures 4c and e).\n\nEar regeneration in adult mice is influenced by estradiol and E6/E7 oncogenes\nInduction of K6b, K16, K14, K5 and K10 expression during epidermal growth in regenerating as compared with intact ears was similar to that observed in the cervix during the diestrus–proestrus transition (Figure 5a). Induction was higher for K6b and K16 (around 20- to 50-fold) than for K14, K5 and K10 (2- to 5-fold). As in the cervix, E6/E7 oncogenes promote growth of epidermal tissue in regenerating ears,21 but induction of keratin gene expression remained in the same range as in regenerating WT ears (Figure 5a) and that of E6/E7 expression followed similar pattern (fourfold; Figure 5b). Distribution pattern of E7 protein during regeneration was found very similar to that of K6b (Figure 5c). Around the wound borders (dashed line), restriction to suprabasal layers was evident; similar pattern was observed in the growing area where the basal layer was well defined (square bracket) and E7 was better detected in the most upper layers (Figure 5c); in disorganized epidermis, commonly seen in the regenerating area, this restriction was not evident. K5, in contrast with the basal restriction in intact epidermis, was widely distributed in the growing area with a basal to suprabasal gradient pattern in both WT and Tg(K6b-E6/E7) mice (Figure 5c).\nRegeneration is expected to be more efficient in young animals; accordingly, most ear regeneration studies have been done with mice younger than 2 months. Therefore, it was unexpected to find that the hole in ears of female WT adult mice regenerated more efficiently than in those of young female animals (Figure 5d). A more pronounced effect was noted in ears of Tg(K6b-E6/E7) adult mice (Figure 4d). Regeneration efficiency was related with faster hole closure (Figure 5e). As previously demonstrated for young animals,21 closing of ear holes involved fast re-ephitelization and, as expected for regeneration rather than a wound repair with scar, formation of new cartilage and hair follicles was observed (Figure 5f).\nHole closure efficiency decreased in juvenile ovariectomized mice (Figure 6a); this effect on regeneration was evident in adult WT mice but not noted in Tg(K6b-E6/E7) mice (Figure 6b). Supporting estradiol as the relevant molecule removed by ovariectomy, the exogenous addition of E2 fully recovered the regeneration capacity of WT ovariectomized mice (Figure 6b). In addition, raloxifene produced the same effect as ovariectomy especially noted in Tg(K6b-E6/E7) (Figure 6b). Addition of E2 to WT or Tg(K6b-E6/E7) mice had no effect on ear hole closure (Figure 6c). However, exogenous E2 in males did produce a detectable improvement in regeneration (Supplementary Figure S2). This observation is relevant considering that an inhibitory factor appeared present in males, as demonstrated by the increased ear regeneration capacity of castrated males (Supplementary Figure S2).\n\nEstradiol promotes cell proliferation during ear regeneration\nIn young mice, E6/E7 improved regeneration capacity in association with increased cell proliferation.21 Accordingly, the number of proliferating cells increased in both WT and Tg(K6b-E6/E7) adult mice in the regenerating area but the latter, likely due to oncogene expression, showed a higher proportion of proliferating cells in suprabasal than in basal layers (Figure 6d). In agreement with a poor effect of exogenous E2 on regeneration, the change in the number of proliferating cells (Ki67+ cells) in regenerating ears of mice treated with E2 was not significant (Figure 6e). However, treatment with raloxifene caused a reduction in the number of proliferating cells in both WT and Tg(K6b-E6/E7) mice (Figure 6e). BrdU incorporation supports this conclusion and, in addition, the apparent accumulation of BrdU+ in the wound borders suggests that estradiol may also contribute to the important migration at early stages of regeneration (Supplementary Figure S3).21 Therefore, E6/E7 cooperate with estradiol to promote proliferation, and possibly also migration, of epidermal cells during regeneration. The number of proliferating cells determined by estradiol and/or oncogenes is the likely factor responsible of epidermal growth, since direct measurement of epidermal tissue in the growing area followed the same pattern (Figure 6f). The contribution of proliferating hair follicle cells to the regenerating area was observed in few instances (Supplementary Figure S4), but could not be considered a major source of cells for ear regeneration.\n"}