Contact between bone marrow (BM) hematopoietic stem cells (HSC) and osteoblast/stromal (OS) cells has been shown to be crucial in the regulation of hematopoiesis. However, very little is known about the regulatory mechanisms of direct cell-to-cell communication in the hematopoietic microenvironment. Gap junctions (GJs) represent the best described intercellular communication (IC) system, and they are characterized by the existence of plaques of narrow channels between contacting cells. Each cell contributes with one hemichannel, which is composed of six proteins, called connexins. Connexin 43 (Cx43) is expressed by BM OS cells. Cx43 has been associated with the cadherin/β-catenin signaling pathway, recently reported as relevant in the OS/HSC interaction at the stem cell niche. Multiple osteogenic defects have been reported in human Cx43 mutations and Cx43 has been shown to be essential in controlling osteoblast functions. BM Cx43 expression is up-regulated up to 100-fold by 5-fluorouracil (5-FU) treatment. Due to the perinatal death of Cx43 germline null mice, a conditional genetic approach was employed to study the role of Cx43 in stem cell proliferation and differentiation. The interferon-inducible Mx1 gene is expressed by both hematopoietic and stromal BM cells. Therefore, we crossed Cx43flox/+ mice with Mx1-Cre transgenic (Mx1-CreTg/−) mice. Cx43+/+:Mx1-CreTg/−and Cx43flox/flox:Mx1-CreTg/− littermates have been analyzed. Gene deletion was induced in vivo by injecting the interferon-inducer polyI:C (8 injections of 300 μg every other day), generating control (Cx43+) and Cx43-deficient (Cx43KO) mice. After one week, Cx43+ and Cx43KO mice were injected with 5-FU (150 mg/Kg i.v.). Cx43 expression in Cx43KO BM was markedly reduced (>80%) as analyzed on day +14 post-5-FU treatment. Cx43 deficiency did not induce a significant change in peripheral blood counts before 5-FU treatment, but the hematopoiesis recovery after 5-FU treatment was severely impaired as demonstrated by absence of recovery of peripheral blood counts, including profound neutropenia, anemia with reticulocytopenia, thrombocytopenia and a 5 to 8-fold decrease of cellularity and hematopoietic progenitor content (granulomacrophagic colony-forming-units -CFU-GM-, erythroid burst forming units (BFU-E) and mixed colony forming units -CFU-mix-) in BM and spleen on day +14 post-5-FU treatment. However, the femoral content of Lin/c-kit+/Sca1+ cells in Cx43KO BM was maintained when compared to Cx43+ BM (139±19 vs 117±32 x 103 Lin/c-kit+/Sca1+ cells per femur, respectively). Short-term competitive repopulation ability of Cx43KO BM cells was diminished as compared to Cx43+ mice (5.9±0.35% vs 22.2±4.6%, respectively, p<0.01), specifically for myeloid (9.2±1.4% vs 35.8±4.3%, respectively, p<0.001) and B lymphoid (0.4±0.2% vs 3.0±1.0%, respectively, p<0.01) cells, but showed spared long-term (6-month) competitive repopulation ability with roughly normal hematopoietic differentiation. Altogether, these data suggest that hematopoietic regeneration after cycle-specific chemotherapy is blocked in Cx43-deficient mice at the long-term HSC repopulating level. Cx43 expression within the BM appears to be crucial in the development of an efficient response to hematopoietic stress.

Author notes

Corresponding author