The results of the present study indicate that time-dependent expression of Hsp105 in the uterine luminal, glandular epithelium and stromal cells during periimplantation period might be essential for regulation of embryo implantation. Our data also show that the presence of embryo in uterus as a stimulus may be important for increasing Hsp105 expression. If Hsp105 is involved in regulation of endometrial differentiation for embryo implantation, one may expect that a reduction of its expression could prevent acquisition of receptive state of the endometrium leading to a failure of implantation. Therefore, we designed an experiment with
Hsp105 antisense oligodeoxynucleotides directly injecting into pregnant rat uterus at early pregnancy, which allowed us to investigate a function of this protein in the process of implantation. Technically, it would be important to do such an experiment to know the transient nature of
Hsp105 gene expression in the uterus. In order to select an appropriate time window of ODNs administration for blockage of Hsp105 expression, we reasoned that the time window should be immediately preceding that of Hsp105 induction, i.e., between days 3 and 6 of gestation. The precise half-life of the
Hsp105 mRNA or its protein in uterus has not yet been determined, nevertheless, the modified
Hsp105 ODNs are known to have a half life of 24–48 hours in certain tissues [
37]. Therefore, ODNs were designed for injection in the afternoon of day 3 of pregnancy, one may expect the tissue on observation to survive for the subsequent 3–4 days of gestation, for an effective suppression of the surge of Hsp105 expression. Because rat embryos were observed to be also capable of expressing Hsp105 (unpublished data), we examined a potential effect of ODNs on embryonic development by observing its normality. Therefore we selected a much later time point (Day 9) to count and examine the embryos. The statistical analysis of the difference of the numbers of implanted embryo between the antisense- and the sense ODNs-treated groups indicated that embryo implantation was indeed prohibited by the antisense ODNs (P < 0.01). These results together with the other observations suggest that treatment with antisense
Hsp105 ODNs, but not with complementary sense ODNs, could severely impair the process of embryo implantation, but no effect on the normality of implanted embryos was observed on day 9 of pregnancy. However, Nakamura et al. just recently generated the Hsp105 knockout mice which did not appear a problem with reproduction [
38], implying that Hsp105 may be not the necessary gene required for implantation in mouse. However, the authors did not specifically pay attention to examine if the animals had any implantation defect present. Hsp105 family has another two members, APG1 and APG which have shown a similar function with Hsp105, and may rescue its function in the absence of Hsp105. We have demonstrated previously that plasminogen activator is important in ovulation of rat and monkeys both in vivo and in vitro[
39], however, double knockout of tPA and uPA in mice showed only 26% inhibition of ovulation could be observed[
40]. Our further studies showed that mouse ovary produces not tPA and uPA, but also MMPs which have also shown to play a role in ovulation. It is possible that MMPs could rescue the function in absence of tPA or uPA. Implantation is a very complex event, which involves various processes, such as blastocyst adhesion, trophoblast invasion, decidualization and cell-to-cell interaction, controlled by a variety of molecules produced by endometrium, embryo and ovary [
41]. During mammalian implantation stroma of the endometrium undergoes severe remodeling, involving apoptosis, proteolysis and angiogenesis [
41,
42]. Endometrial cells rapidly proliferate and differentiate to form the decidua tissue which accommodates and protects implanted embryos [
43]. In our previous reports, analysis of the endometrium of both rhesus monkey and human during peri-implantation period has demonstrated that a relatively high frequency of apoptosis occurs in the secretory endometrium and is correlated to increased expression of apoptosis related molecules [
25,
26], while only limited numbers of the apoptotic cells were observed in the other phases of the cycle. It appears that endometrial apoptosis and the cyclic changes in endometrial growth and regression during establishment of implantation window might be regulated precisely and coordinately, not only by
Fas,
FasL,
BcL-2 and
Bax [
17], but also by
Hsp105, because the profile of these molecules is well correlated with that of the Hsp105 expression in rat uterus as demonstrated in the present study. Evidence has shown that Hsp105 is capable of enhancing cell apoptosis in mouse embryonal F9 cells[
15,
44] and murine embryos during embryogenesis[
9,
45]. On the contrary, the Hsp protein was also observed to inhibit cell apoptosis in rat testis and some experimental cell models [
16,
46‐
48]. These observations suggest that Hsp105 may be involved in regulation of murine uterine cell apoptosis. Since the cell types, species used in the individual studies were different, some unknown factors as well as cellular environment present in the various studies might determine an inhibitory or a promotional effect of the Hsp protein on cell apoptosis. However, the molecular mechanism of Hsp105 in regulating uterine cell apoptosis during rat periimplantation period remains to be further investigated. In summary, our data have demonstrated a significant increase in Hsp105 expression on day 5. It seems that the protein might be able to induce luminal cell apoptosis which in turn destabilizes epithelial barrier at implantation site and facilitates trophoblast invasion and implantation. On D6 of pregnancy Hsp105 expression was down-regulated in luminal epithelium, and upregulated in stroma cells adjacent to the embryo attachment. The dynamic changes in the level of expression of this protein during early implantation period suggest involvement of this protein in certain cellular events in luminal epithelial and stromal cells, which are essential for implantation.