It was just last month that the first case of a woman spreading Zika sexually to her male partner was documented, and around the same time that a study found the virus in a woman’s cervical mucus 11 days after infection. Before that, most of the talk about sexual transmission of the normally mosquito-borne virus focused on semen, where Zika can survive for a long time. (A recent report puts it at up to six months.)

But it’s starting to look like the vagina is very hospitable to Zika as well. Which is alarming, given the severity of the birth defects Zika can cause. So far there are only case studies in humans, but a new study in mice found that the female reproductive tract is “a highly susceptible site of ZIKV replication.”

The thing about mice is, they’re not very susceptible to Zika generally. If you take what researchers call “wild-type mice,” meaning they haven’t been genetically altered, and inject them with the virus like a mosquito would, “[it] doesn’t replicate very robustly,” says Laura Yockey, the lead author on the new study and a graduate student at Yale University.

Normally, a mouse’s natural immune response will take care of Zika. Not the adaptive immune response, which is when a body develops antibodies to a specific pathogen, but the innate immune response, which is a more generalized response triggered by interferons—proteins that signal cells to get into fighting mode. In humans, Zika appears to mess with the interferon response, but not in mice. So most research on Zika in mice has had to be done with mice that are genetically altered to have diminished interferon responses.

Yet in this new study, when Yockey and her colleagues infected wild-type mice vaginally, Zika was able to replicate in the vagina for four or five days. (It replicated for seven days in the vaginas of mice with deficient interferon responses.) That suggests that something is different in the vagina, that the mouse’s body is less able to defend itself from Zika when infected this way.

And vaginal infection affected the fetuses of pregnant mice. In the ones with “interferon knockout,” Yockey says, “when I infect them very early in pregnancy—four days—all of the fetuses were resorbed.” That’s analogous to the early-to-mid first trimester in humans. At eight and a half days, or the equivalent of the early second trimester, the fetuses had restricted growth. Even in the wild-type mice, if they were infected early in pregnancy, “there was a mild but significant growth restriction,” Yockey says.

We can’t generalize findings in mice to humans, of course, but given the evidence that Zika can survive in the human vagina, and the fact that humans are generally more susceptible to the virus than mice, the researchers speculate that Zika “introduced into the human vagina is likely to replicate more robustly than in the vaginal cavity of [wild-type] mice.”

So here is yet another open question for researchers, one that reproductive-age women urgently need answered. Yockey wonders “if vaginal exposure versus exposure by a mosquito might have different effects—on the fetus, on pregnant women, or on women in general.”