In vivo study of prolactin (PRL) intracellular signalling during lactogenesis in the rat: JAK/STAT pathway is activated by PRL in the mammary gland but not in the liver

The rat prolactin receptor (PRL-R) exists in two forms, which differ in the length of the cytoplasmic domains, tissue distribution, and biological activity. The short form predominates in liver while the long form is prevalent in mammary gland. We have compared activation by PRL of the JAK2-STAT pathway (protein tyrosine phosphorylation and STAT5 activation) in mammary gland and liver in an in vivo rat model of induction of lactogenesis by PRL injections, and we have studied the relative proportion of both forms of the receptor in these tissues by reverse transcription-polymerase chain reaction. Rats were ovario-hysterectomized on Day 19 of pregnancy, treated with bromocriptine, subsequently injected with 250 mu g ovine PRL i.p. on Day 20, and killed 0-12 h after. Western blots of solubilized mammary gland and liver membranes immunoprecipitated with anti-PRL-R or anti-JAK2 antibodies showed that the PRL-R is constitutively associated with JAK2 and that the long form of the PRL-R is present in both tissues, while the short form was detected only in liver. Phosphorylated proteins corresponding to the long form of PRL-R and JAK2 appeared 15-60 min after ovine PRL injection in mammary extracts but not in liver. At these same times, an electrophoretic mobility shift assay, using a rat p-casein probe specific for STAT5 binding, showed activated STAT5 in mammary gland cytosol and nuclear extracts. In the liver, low levels of activated STAT5 were detected in nontreated animals, which were not modified by PRL. Quantitative RT-PCR of liver and mammary PRL-R mRNA showed that the amount of the long form of PRL-R mRNA is roughly comparable in both tissues, while the short form is predominant in liver and in a minority in mammary tissue. Both forms were down-regulated by PRL only in mammary glands. Thus, during lactogenesis, mammary tissue responds to PRL by activation of JAK2 and STAT5, while the liver does not respond to PRL in spite of the presence of PRL-R associated with JAK2 and pre-existing activated STAT5. Thus, liver tissue may lack a critical component for activation of the PRL pathway, or the large quantities of the short form of the PRL-R may associate with the long form to constitute inactive heterodimers.