A persisting enigma is why gap junctions become The hepatocytes in the mature normal liver are tightly more sparse during liver regeneration. 10-17 coupled through gap junctions, except during compen-
The growth-control hypothesis proposed by Loewenstein satory hyperplasia (regeneration) after partial hepatecand Rose 18 states that gap junctional communication exerts tomy when the gap junctions become down-regulated.
a negative control over proliferation, in particular that a brief The significance of this down-regulation has been a longdown-regulation just before the G1/S phase can trigger replistanding enigma. The present study of hepatocytes in cation. This hypothesis is compatible with the inverse relaprimary culture and in the regenerating liver aimed at tionship between gap junctions and proliferation noted in the defining the relationship, if any, between hepatocyte regenerating liver. It could also explain why glucocorticoid gap junctional communication and proliferation. Gap and cAMP inhibit hepatocyte DNA replication 20 because both junctional down-regulation in the regenerating liver appeared to be a specific phenomenon because desmo-retard the down-regulation of gap junctions in primary culsomes and the surface contact area between neighture of freshly isolated adult hepatocytes. [22] The hypotheboring hepatocytes remained constant. All agents and sis is further supported by the inverse relationship noted conditions (dexamethasone in vivo; dexamethasone, cybetween growth and gap junctional expression in embryonic clic adenosine monophosphate, serum, and high cell mouse hepatocytes grown at various densities, 14 in hepatodensity in vitro) delaying gap junctional down-regulacytes treated with the mitogen hepatocyte growth factor tion also increased the lag before the cells reached com-(HGF), and in some tumorigenic cell lines transfected with petence to enter S phase. This raised the possibility that Cx complementary DNA (cDNA). [25][26][28] On the other hand, reshepatocyte DNA replication was inhibited through prestoration of gap junctional communication to hepatoma cells, ervation of gap junctions. However, we disproved this rat glioma cells, or HeLa cells failed to affect their assumption by showing that the DNA replication (more growth, 25,28, and fibroblasts with enhanced gap junctional specifically the G1/S transition rate constant) was inhibcommunication still responded to mitogens. It appears from ited even in hepatocytes completely devoid of gap juncthe above that the inverse relationship between gap junctional communication. The teleological advantage of tional communication and proliferation has been seen most linking gap junctional down-regulation to hepatocyte G1 consistently in primary hepatocytes. In-depth investigation progression therefore may not be to trigger DNA replicaof the relationship between gap junctions and proliferation tion but to ensure that proliferating hepatocytes and in such cells was therefore considered important not only to hepatocytes responsible for liver-specific metabolic learn more about hepatocyte biology, but also to evaluate the functions maintain separate pools of metabolites and gap junction growth-control hypothesis in general. signaling molecules. (HEPATOLOGY 1997;25:847-855.) If gap junctions inhibit hepatocyte proliferation, their down-regulation must precede increased proliferation under various experimental conditions. This was probed in the pres-The adult liver is remarkable because its highly differentient study. Hepatocyte proliferation was modulated in vitro ated cells are able to proliferate. In fact, during the compenby glucocorticoid, cAMP, and various seeding densities and satory hyperplasia occurring after partial surgical hepatecin vivo by continuous intraperitoneal infusion of the synthetic tomy 1 the liver remnant doubles its mass in a few days. 2,3 glucocorticoid dexamethasone during liver regeneration. The hepatocytes in the normal adult mammalian liver have Whether gap junctional down-regulation correlated with any well-developed gap junctional communication. Gap junctions particular proliferation-associated parameter, such as the are composed of members of the connexin (Cx) multigene G1/S cell-cycle phase transition probability or the lag period family. They allow the intercellular transfer of low-molecubefore the cells became competent to enter S phase, was also lar-weight metabolites and signaling substances such as tested. Another approach was to culture primary hepatocytes Ca 2/ , phosphoinositol, and cyclic adenosine monophosphate until their gap junctions had become completely down-regulated and then expose them to inhibitors of proliferation. In this way it could be determined whether the ability of cAMP and glucocorticoid to retard the down-regulation of gap junc-Abbreviations: Cx, connexin; cAMP, cyclic adenosine monophosphate; HGF, hepatocyte tions 8,21-23 was relevant for their ability to inhibit hepatocyte growth factor; cDNA, complementary DNA; EGF, epidermal growth factor; SDS, sodium proliferation. Finally, we were interested in whether the indodecyl sulfate; mRNA, messenger RNA.