Long-Term In Vivo Studies in Mice Unexpectedly Reveal That Prolonged Growth Hormone (GH) and Insulin/Insulin-Like Growth Factor-1 (IGF-1) Signaling Have a Negative Effect on Normal Bone Marrow Hematopoiesis – Implications for GH-Based “rejuvenation” Therapies,

Growth hormone (GH) has been proposed as a rejuvenation factor that delays aging in older patients. The long-term results of such therapy, however, have recently become controversial. It is known that GH action is mediated by release of insulin-like growth factor-1 (IGF-1, also known as somatomedin C) from the liver. IGF-1 is an important factor affecting proliferation of several types of cells, including malignant hematopoietic blasts. A role for the GH-IGF-1 axis in normal hematopoiesis still remains controversial and some stimulatory effects to the clonogenicity of normal hematopoiteic stem/progenitor cells (HSPCs) have been described. We have previously reported that IGF-1 does not directly stimulate proliferation of normal clonogenic hematopoietic progenitors (J Clin Invest. 1994;94:320); however, it may promote erythroid differentiation and hemoglobinization of erythroblasts (Leukemia 1998;12:371). Aim of the study: Since long-term studies on the effect of GH-IGF-1 signaling on normal hematopoiesis in vivo have not been performed yet, we investigated the influence of these factors on the hematopoietic system in appropriate mouse models that express low or high levels of plasma GH and IGF-1. Experimental approach: We employed two mutant mouse strains that have low circulating plasma levels of IGF-1 due to a GH-receptor mutation (Laron dwarf mice) or to a defect in GH synthesis in the pituitary gland (Ames dwarf mice). We also employed transgenic mice that express the bovine GH gene (bGH), which results in a constitutively high level of circulating plasma IGF-1. In some experiments, normal wild type or Laron and Ames dwarf mice, with low circulating plasma IGF-1 levels, were injected for a prolonged period of time with recombinant GH or IGF-1. All these animals were analyzed by FACS for the presence of Sca-1⁺Lin^–CD45⁺ HSPCs in bone marrow (BM) and the clonogenic growth of progenitors from all hematopoietic lineages monitored in standard methylcellulose (CFU-GM and BFU-E) and plasma clot (CFU-Meg) cultures. We also analyzed changes in peripheral blood count by Hemavet. The mice employed in our studies were at different ages (2 months, 6 months, and 2 years), and both female and male mice were analyzed. Results: We observed that Laron and Ames dwarf mice, with low circulating plasma levels of IGF-1, at the age of 1–2 years had ∼3–5 times more Sca-1⁺Lin^–CD45⁺ in the BM than normal control littermates of the same age. Similarly, these mice also had ∼3–5 times more CFU-GM, BFU-E, and CFU-Meg in BM. The number of Sca-1⁺Lin^–CD45⁺ cells and clonogenic progenitors from all lineages declined ∼5 fold in these animals (at 2 months and 8 months of age) after 4 weeks of treatment by daily injection of GH or IGF-1. However, interestingly, if this treatment was ceased the number of these cells showed over time a tendency to normalize. In another set of experiments, we evaluated the number of HSPCs in BM of 6-month- and 1-year-old mice that express the bovine GH transgene. We found that these mice have an ∼6-fold and ∼12-fold lower number in the BM of Sca-1⁺Lin^–CD45⁺ and clonogenic progenitors, respectively, than control transgenic littermates expressing empty vector. Overall, a decrease in the number of HSPCs was accompanied by microcytic anemia. Conclusions: Our data, obtained from several in vivo animal models, show that prolonged exposure to GH-IGF-1 signaling depletes the number of HSPCs in the BM over time. This observation is highly relevant to GH rejuvenation therapies currently employed in humans. Our data indicate that this may lead to the premature depletion of HSPCs, and we also envision that prolonged stimulation of HSPCs by GH-IGF-1 may result in development of leukemias. In support of this notion, Laron dwarf mice and Laron dwarf patients do not develop hematological malignancies, while leukemias are often seen in bGH transgenic mice and in acromegalic patients.

No relevant conflicts of interest to declare.