The Forgotten Lactogenic Activity of Growth Hormone: Important Implications for Rodent Studies

Abstract

Studies of the effects of GH and the mechanisms of its actions frequently use rats or mice and various recombinant human GH preparations. Authors of many of these studies appear unaware of the fact that, in rodents, human GH signals through both GH and prolactin (PRL) receptors; thus, treatment with human GH is equivalent to a combined treatment with GH and PRL. GH receptors and PRL receptors are present in multiple cell types. Importantly, PRL exerts major effects on brain neuroendocrine action, female and male reproduction, metabolism, body composition, immune responses, and a host of other functions; thus, treatment of rodents with recombinant human GH could affect these important physiological parameters.

Lactogenic activity of human GH (hGH) in rodents was already known in the 1960s (1, 2) and contrasted with the purely somatotropic activity of the bovine, ovine, and porcine GH. Readers interested in history of endocrinology may be intrigued by the fact that, based on the difficulties of extracting biologically active prolactin (PRL) from human pituitaries collected postmortem and the luteotropic and lactogenic activity of hGH in rodents, it was believed that humans may be unique among mammals by not producing PRL and instead having a pituitary hormone that serves both somatotropic and lactotropic functions. In a review article on the discovery of human prolactin, Friesen (3) stated, “Although prolactin was discovered in the early 1930s in sheep, cows, birds etc, no human form had been purified because it was thought to be identical to human growth hormone. In fact, prior to 1970, most endocrinologists doubted human prolactin even existed.” This doubt was obviously discarded after successful identification and purification of human PRL in the early 1970s (4–6).

Many years later, several laboratories produced transgenic mice that expressed a variety of cloned GH genes. Remarkable increase in adult body size was a common characteristic of animals expressing human, bovine, ovine, or rat GH. However, transgenic mice expressing hGH exhibited various characteristics not seen in transgenics that expressed other species of GH (7–11). Prominent among them was stimulation of mammary glands leading to lactation. In a line of mice expressing high levels of hGH under the control of rat phosphoenolpyruvate carboxykinase promoter/enhancer, females experienced repeated pseudopregnancies consistent with the luteotropic activity of PRL receptor (PRLR) ligands in rodents, and virgin hGH transgenic females readily accepted foster pups and produced enough milk to raise normal-size litters to weaning (12). In a different line of transgenic mice with a lower level of hGH expression, activation of tuberoinfundibular dopaminergic neurons (a well documented target of PRL action in the brain) interfered with endogenous PRL release, leading to the infertility of females (13). Interestingly, some of the effects of expression of hGH in transgenic mice on hypothalamic neurotransmitter turnover, pituitary morphology, and hormone levels were opposite to the effects of expression of bovine GH on the same parameters (8, 15). Additionally, consistent with lactogenic activity of hGH in rodents, transgenic mice expressing hGH exhibit high frequency of mammary tumors (16, 17).

Evolutionally, GH and PRL are closely related as members of the same family and are similar in terms of size and structure. Both GH and PRL act primarily (and likely exclusively) by binding to dimeric cell surface receptors and activating the JAK2-STAT5 (Janus kinase-signal transducer and activator of transcription) signaling pathway to modulate the expression of numerous genes in target tissues. However, they exert many hormone-specific (and often also species specific) actions. For example, PRL provides the key luteotropic stimulus in rodents and carnivores (but not in ruminants or primates) and is importantly involved in the control of seasonal transitions between reproductive activity and quiescence (18, 19), whereas there is no evidence for GH sharing any of these actions. However, GH and PRL actions can also overlap. For example, hepatic IGF-1 expression, circulating IGF-1 levels and somatic growth are primarily regulated by GH, but PRL may also enhance IGF-1 expression, at least in fetal tissues (20), and stimulate somatic growth.

Although neuroendocrine control of the release of GH and PRL from the pituitary is very different involving distinct neuronal groups, neuropeptides, and neurotransmitters, in many species both PRL and GH are released in response to stress.

Cloning of GH and PRL genes from various species and elegant crystallographic studies elucidated details of three-dimensional structure of GH and PRL molecules and their interaction with the corresponding receptors (21, 22). These studies identified molecular regions responsible for the lactogenic (PRL-like) activity of hGH and for differences between human, as compared with ruminant, porcine, or rodent GHs (21, 22)

For example, nonprimate GHs are unable to bind to the human GH receptor (GHR), whereas primate GHs can bind to nonprimate GHRs and PRLRs with high affinity (21, 23–25). It has been suggested that nonprimate GHs are unable to bind to the human GHRs because of amino acid differences between the two hormones. In particular, Cunningham and Wells (22) have shown that nonprimate GHs possess substitutions of at least 10 of the 25 hGH amino acid-binding determinants (Ile4, Metl4, Ser62, Asn63, Arg64, Lys70, Tyr164, Asp171, Phe176, Ile179). These changes explain the lack of binding of nonprimate GH to human GHRs. An extension of this work by Souza et al (26) suggested that species-specific amino acid residues, primarily the substitution of Asp171 for His in all nonprimate GH relative to hGH, is responsible for this effect. Similarly to the GH induced intracellular signaling, effects of PRL vary with the species of hormone and cognate receptor origin. Utama et al (27) described major differences in the biological activities of PRLs derived from seven different mammalian species on human and rat PRLR. In the same study, mouse PRL was shown to be a competitive inhibitor of human PRL binding to human PRLR and to exhibit a partial antagonist activity in this system (27).

Major species differences in biological activity also extend to placental hormones from the same family, which exhibit various combinations of somatotropic and lactogenic activity. Bovine and porcine placental lactogen have been shown to bind to the somatogen receptor (28–31). Additionally, human placental GH, similarly to the human pituitary GH (and recombinant hGH), is lactogenic in rodents (32, 14).

We want to finish this brief article with an appeal to investigators, both authors and reviewers (and particularly those now entering this field), to carefully consider the biological differences between GHs derived from different species in designing their experiments and interpreting their results. Attention to the well-documented and powerful lactogenic effects of hGH in rodents and high affinity of this hormone for the rodent PRLR is of particular importance (Figure 1).

Figure 1.

Actions of homologous (mouse, rat), ruminant (bovine, ovine), porcine or human GH in rodents.