We previously demonstrated that androgen insensitive female mice were less sensitive to phosphatase and tensin homolog (PTEN) inactivation induced uterine cancer suggesting significant role for androgens acting via androgen receptor (AR) expressed in epithelial and stromal uterine cells. We hypothesized that uterine gland specific AR inactivation will also reduce PTEN inactivation induced uterine pathology.
To test our hypothesis, we developed uterine gland specific PTEN and AR knockout mouse model (Cre/LoxP system). Development of uterine pathology was compared between wild-type (WT), gland specific PTEN knockout (utPTENKO) and combined PTEN and AR knockout (utPTENARKO) female mice.
Contrary to our hypothesis the simultaneous gland specific AR inactivation increased uterine pathology and induced glandular hyperproliferation in utPTENARKO. While 75% (6 out of 8) utPTENARKO mice developed microscopic uterine abnormalities, only 20% (1/5) of utPTENKO had abnormal uterus as detected age of 20 weeks (Fishers exact test, p=0.049). No abnormalities were found in WT females (n=5). The microscopic uterine abnormalities were manifest as significantly increased uterine weight in utPTENARKO [407±118mg (mean±SE); n=8] when compared to WT (92±14mg; n=5) and utPTENKO (121±40mg; n=5) (p=0.01).
Uterine morphology was further quantified by measuring total area, endometrial area and myometrial area. Total uterine area was significantly (p=0.019) increased in utPTENARKO (4.16±0.78mm2; n=8) compared to WT (1.93±0.22mm2; n=5) and non-significantly (p=0.057) compared to PTENKO (2.07±0.31mm2; n=5). Also, the endometrial area was significantly increased in utPTENARKO (2.39±0.4mm2) compared to WT (0.78±0.08mm2, p=0.003) and PTENKO (0.95±0.19mm2, p=0.013). Myometrial area was similar between the genotypes, leading to significantly increased endometrial to myometrial ratio in utPTENARKO (1.42±0.17) compared to WT (0.69±0.02, p=0.003) and PTENKO (0.85±0.1, p=0.028).
In conclusion, while systemic androgen deficiency inhibited PTEN knockout induced uterine pathology; the uterine gland specific AR inactivation had the opposite effect suggesting a complex role for androgens acting via AR in endometrial pathology. Therefore, further analysis on mechanism(s) for androgen actions in uterine cancers is warranted.