Hepatic regeneration and erythropoietin production in the rat.

The regenerating liver produces erythropoietin in response to hypoxia. The amounts of erythropoietin produced in animals subjected to hepatectomy are significantly higher than those observed in sham-operated animals. Hepatic erythropoietin production appears to be dependent upon the stage of regeneration with the highest levels being produced during the period of greatest proliferation and increase in liver mass.

Phenylhydrazine-induced leukocytosis in the rat.

Rats were injected with phenylhydrazine (PHZ) for periods of up to 6 months, during which time a marked leukocytosis was induced. The highest leukocyte counts occurred within 4-5 days following injection. An initial injection of 4 mg/100 gm body weight evoked a mean total leukocyte count of 129 X 10(3) cells/microliter. Successive weekly injections of 2 mg/100 gm resulted in a mean total leukocyte count of 70 X 10(3) cells/microliter compared to a mean total leukocyte count of 12.5 X 10(3) cells/microliter in saline-injected rats. Lymphocytes and monocytes accounted for approximately 75% of the total cell counts in both the PHZ-treated and control rats. The presence of increased numbers of mononuclear cells was confirmed by Percoll gradient separation and by phase-contrast microscopy. Although a leukocytosis was evident when using the automated Coulter electronic cell counter, it was not discernible when blood samples were counted manually in a hemocytometer by light microscopy. Histological examination of the thymus, lymph nodes, and spleen of the PHZ-treated rats indicated that lymphocytes and monocytes were mobilized from these sites. Lymphocyte depletion was evident, and germinal centers were found in all these lymphoid organs, indicating that PHZ induced a lymphopoietic response. A possible autoimmune etiology for PHZ-induced red blood cell destruction is discussed.

Mechanisms of extrarenal erythropoietin (Ep) production.

Evidence is presented for production, by the subtotally hepatectomized (hepx) rat, of a factor which induces morphological and physiological hyperactivity of the Kupffer (K) cells and increased formation of Ep by the regenerating liver. This factor, originally termed hepatopoietin (Hp), and more recently the hepatic erythropoietic factor (HEF), is detectable in higher concentration in the hepatic venous blood than in blood draining other organs, thus supporting its hepatic origin. Production of the HEF is best related to hyperactivity of the K cells and not to the parenchymal (P) cells. The HEF can be demonstrated by administering serum from hepx donors to normal rats which respond with increased production of Ep when nephrectomized (nephrx) and rendered hypoxic. Removal of the kidneys from hepx donors further augments the Ep response to this serum in recipients. Subtotal hepx also evokes the production of a renal inhibitory factor (RIF) which reduces the ability of the liver to function as an extrarenal source of Ep. This inhibitor is found in renal venous blood and not in blood draining other organs. It is suggested that the RIF reduces the hepatic Ep response to hypoxia by diminishing the production and/or activity of the HEF. The RIF possesses no anti-Ep activity and its appearance and actions are not influenced by accumulation of metabolic wastes (as in the nephrx or ureterally-ligated rat). Erythroblastic nests have been observed in regenerating livers at 24-48 hr after subtotal hepx. It would seem that removal of a considerable part of the liver, which stimulates hepatic regeneration, confers upon this organ an increased ability to produce Ep and to function as a hematopoietic inductive microenvironment (HIM) for erythropoiesis.

Granulopoiesis and colony stimulating factor production in regenerating liver.

Partial hepatectomy (hepx) of the adult animal induces a substantial hepatic erythropoietin (Ep) response following hypoxia and results in a transient reinitiation of hepatic erythropoiesis. This phenomenon is associated with an elevation in the number and activity of the hepatic macrophages or Kupffer cells. Kupffer cells have been implicated in the production not only of Ep, but of the moiety which regulates granulocyte growth in vitro, Colony Stimulating Factor (CSF). Histological examination of the regenerating liver revealed some foci of granulopoiesis, but not to the extent noted for erythropoiesis. Granulopoiesis was not associated with islet conformations involving central reticular or macrophagic cells. In addition, serum derived from the hepatic venous drainage of hepx animals possessed moderate levels of CSF, as determined by its ability to promote the growth of granulocytic-macrophagic colonies in agar culture.

The effects of interferon on murine erythropoiesis.

The effects of interferon (IF) on erythropoietin (Ep) action and production were studied in mice. In comparison to control animals, Ep action in exhypoxic, polycythemic mice was significantly decreased (p less than 0.05) following two low dose injections of IF (2.9-3.5 X 10(4) units). In addition, renal Ep production in normal intact mice was also significantly decreased (p less than 0.01) following a single injection of IF (5.4-6.3 X 10(4) units) and hypoxic exposure.

Rat bone-marrow cell proliferation and differentiation as an index of the effects of xenobiotics in vitro.

A three-dimensional bone-marrow culture system was utilized for assessing the toxicity of various chemotherapeutic agents. This model, which exhibits multilineage haematopoiesis and promotes the growth of progenitor cells for extended periods, was treated with various concentrations of beta-d cytosine arabinofuranoside (Ara-C), cyclophosphamide (CP), methotrexate (MTX) or 5-fluorouracil (5-FU). The effects of these agents on the phenotypes of adherent zone cells was ascertained by labelling with monoclonal antibodies against rat leucocytes and evaluating by flow cytometry. Adherent zones also were assayed for content of colony-forming unit culture (CFU-C) and cellular viability after drug exposure was measured using the neutral red (NR) assay. The results indicated that Ara-C, CP, MTX and 5-FU treatment caused dose-dependent decreases in the CFU-C concentration of adherent zones in cultures of various ages. These agents also altered the phenotypic distribution of adherent zone cells and displayed differential lineage specificities. A dose-related decrease in viability also was observed with the NR assay, albeit at higher drug doses than those which induced measurable CFU-C and phenotypic alterations. These three-dimensional cultures may prove to be ideal substrates for toxicity testing as they contain all of the cell types present in vivo, are physiological with respect to their growth patterns, are easily manipulable, and can be maintained for extended periods.

Temporal relation between a hepatic erythropoietic factor and the site of rat erythropoietin production.

Serum borne hepatic erythropoietic factor (HEF), which can stimulate hepatic erythropoietin (Ep) production in the adult rat, is found at elevated levels in the serum of partially hepatectomized rats and of rats subjected to hepatotoxic injury. It is also detected in sera of patients with liver disease. The purpose of the present study was to determine whether or not HEF activity is increased in the serum of the normal neonatal rat at a time when the liver is the primary site of Ep production. Our results show significantly increased HEF activity in the serum of young rats during the second to fifth weeks of life. Negligible activity was detected in rats over five weeks of age. In the rat, the kidney is reported to begin producing Ep by the third week of life and by the eighth week the kidney is the major site of synthesis with liver production at this age significantly diminished. Thus, our findings show a temporal relation between HEF activity in the serum and the reported transition from liver to kidney production of Ep.

Phenylhydrazine is a mitogen and activator of lymphoid cells.

Rats were administered a single injection of phenylhydrazine (PHZ) which induced an hemolytic anemia that reached maximal levels two to four days following injection. This was accompanied by a leukocytosis which was most pronounced four to six days after injection; lymphocytes and monocytes accounted for 75 percent to 80 percent of the leukocyte count, respectively. All peripheral blood cell values, including the red cell count and hematocrit, returned to their pre-injection levels by the 11th post-injection day. Analysis by flow cytometry of peripheral blood mononuclear cells (PBMC) isolated from PHZ-treated rats by Ficoll-Hypaque gradient separation showed a marked increase in the B cell population of the peripheral blood. This was also seen in cultures of PBMC obtained from untreated rats following incubation with PHZ. Cultures of PBMC obtained from rats four to five days after PHZ injection which were incubated with pokewood mitogen (PWM) or phytohemagglutinin (PHA) showed significant increases in blastogenesis as indicated by [3H] thymidine incorporation when compared to cultures of PBMC obtained from untreated rats incubated with these mitogens. Incubation of cultures of PBMC obtained from untreated rats with PHZ significantly increased blastogenesis in cultures of five day duration. Atypical and blastic lymphoid cells were evident in cytosmears of PBMC isolated from PHZ-treated rats and also in sections of PBMC pellets studied using the transmission electron microscope. Serum of the PHZ-treated rats contained elevated immunoglobulin titers as measured by radial immunodiffusion. The results show that PHZ stimulates lymphoid cell blastogenesis and can sensitize circulating lymphoid cells to PHA and PWM indicating that PHZ is capable of stimulating the immune system of the rat.

Evidence for a hepatic-renal antagonism in the production of hepatic erythropoietin.

A tandem perfusion of the liver and kidneys of hepatectomized (hepx) rats was performed to ascertain the precise levels and time of appearance of the previously reported hepatic erythropoietic factor (HEF) and its antagonist, the renal inhibitory factor (RIF). Hepatic erythropoietic factor is produced by the hepx animal and is capable of augmenting the hepatic erythropoietin (Ep) response to hypoxia when administered to normal rats, whereas RIF acts to diminish this Ep response. Renal inhibitory factor does not directly affect Ep but appears to alter its synthesis owing to an inhibition to HEF production and/or action. In the present study, levels of HEF and RIF were determined at different intervals following hepx by in situ perfusion of the livers and kidneys of these animals. Levels of HEF are significantly higher than controls in rats at 24 to 72 hrs post hepx whereas its antagonist, the RIF, is not detectable until between 48 and 72 hrs after hepx. Inhibition of RIF of the hepatic Ep response is nearly total by 96 hrs following surgery. In previous studies, an increase in the hepatic Ep response to hypoxia was noted from 24 to 72 hrs after hepx. Peak hepatic Ep levels occurred at 72 hrs post hepx and declined thereafter, manifesting near normal hepatic Ep levels at 96 hrs after hepx. The present study provides an explanation for these findings by demonstrating that production of Ep by the hepx animal is dependent on the relative levels of a stimulatory (HEF) and an inhibitory (RIF) factor.

Recovery of an erythropoietin inducing factor from the regenerating rat liver.

The perfusion of livers of partially hepatectomized (H mean) rats lends support to earlier findings that the regenerating rat liver is the source of an erythropoietin-inducing hepatic factor (Ep-IHF) which stimulates hepatic production of extrarenal erythropoietin (Ep). Blood plasma was collected from perfused livers of rats that were partially hepatectomized (H mean) 48 to 72 hrs prior to perfusion with whole blood from normal rats. This plasma, when injected into normal rats which were nephrectomized (N mean) and rendered hypoxic 18 hrs after injection, evoked a significant increase in Ep values when compared to blood plasma collected from perfused livers of normal rats. Ep values were significantly higher when regenerating livers were perfused with blood collected from nephrectomized (N mean) rats than when such livers were perfused with blood of normal rats. The highest Ep values resulting from the liver perfusions were obtained when the liver donor and blood donor rats were both H mean and N mean. The results demonstrate that the liver is the principal source of an Ep-inducing factor since perfusion of the liver eliminated other potential tissue sources of activity in the rat. This was achieved by perfusing the livers directly through the portal vein and collecting the perfusate from the hepatic vein, thereby eliminating potential contributions from organs draining into other parts of the systemic circulation. In addition, it was shown that the kidney inhibits the activity and/or production of the Ep-IHF which is evoked by H mean.

A stereotypic, transplantable liver tissue-culture system.

A method of coculturing adult rat hepatic parenchymal cells (PC) and stromal cells in a three-dimensional framework of nylon filtration screens or biodegradable polymer meshes was developed in our laboratory. Rat liver stroma, which includes vascular and bile duct endothelial cells, fat-storing cells, fibroblasts, and Kupffer cells, were isolated by gradient centrifugation after in situ liver perfusion and expanded in monolayer culture prior to seeding onto nylon screens or bioresorbable polyglycolic acid (PGA) polymers oriented into a felt-like construct. A second inoculum of freshly isolated PC was applied after the stromal cells became established. Histological analyses revealed that PC proliferation occurred until all available space for expansion within the template was exhausted. These cells retained their rounded morphology, and after 4-5 wk 7-9 "layers" of PC filled the 140-microns deep template. Dioxin-inducible cytochrome P450 activity was detected for up to 58 d in culture, and albumin, fibrinogen, transferrin, and soluble fibronectin were detected in the medium by enzyme-linked immunosorbent assay (ELISA) for 48 d in vitro. Immunohistochemical analysis of sections through the cultures confirmed the presence of these proteins as well as cytokeratin at the cellular level; the extracellular matrix stained for both collagen type III and laminin. Long-term PC proliferation and function were enhanced by the presence of stromal cells as well as by a meshwork template whose geometry allows the interaction of PC with stroma and matrix on several different planes. To permit transplantation, cocultures of hepatic PC and stromal cells were established on PGA felt constructs instead of nylon screens. After approximately 24 d in vitro, these constructs were grafted into sites in the mesentery, omentum, and subcutaneous tissues of adult Long-Evans rats. The growth of hepatocytes after 30 d in situ was evident by histological analysis; grafts of cocultures regenerated a liver-like architecture consisting of sinusoids and putative biliary structures. In addition, PC at these extrahepatic graft sites were positive for albumin, transferrin, and fibrinogen synthesis by immunohistochemistry. Graft survival was enhanced when recipients were subjected to approximately 40% hepatectomy. Hepatic PC:stromal cell cocultures may prove useful in the restoration of liver function either by direct transplantation using PGA or similar templates, or as extracorporeal devices, using nylon screens.

Erythropoietin production by macrophages in the regenerating liver.

The site of erythropoietin (Ep) production and/or storage in the rat liver was determined. A guinea pig anti-Ep was produced against purified rat Ep (64,096 +/- 4j064 IU/mg). This antibody was found to be highly specific using rocket immunoelectrophoresis, Ouchterlony gel diffusion methods, and immunoprecipitin reactions as well as Ep neutralization tests (capable of completely neutralizing up to 2,000 IU Ep/mg). This anti-Ep was labeled with either fluorescein for light microscopic study or ferritin for electron microscopy. Kupffer cells showed varying degrees of labeling after hepatectomy alone or hepatectomy combined with nephrectomy and/or hypoxia. Greatest labeling was seen in Kupffer cells of rats that were nephrectomized 48 hr posthepatectomy and kept at ambient pressure. No labeling of hepatocytes or vascular and bile duct endothelium was noted.

Hematopoiesis on nylon mesh templates. I. Long-term culture of rat bone marrow cells.

Rodent hematopoietic cells have been perpetuated in long-term bone marrow culture (LTBMC), provided that they were plated onto a pre-established layer of bone marrow stromal cells (fibroblasts, reticular cells, adipocytes, macrophages, etc.). This monolayer-type system supports the self-renewal of murine pluripotent stem cells and produces substantial numbers of progenitors and mature cells of the myeloid lineage. In an effort to increase the growth potential of cells in LTBMC, an adherent matrix of stromal cells was established on a pretreated nylon screen template. Subsequent seeding of hematopoietic cells onto this matrix occurred both on the surface of the adherent layer and in the interstices formed by the developing stromal cells. A three dimensional growth pattern of hematopoietic colonies and clusters was observed. Mature cells and late stage precursors of the myeloid and erythroid series were observed in the non-adherent layer for the duration of the experiment (39 weeks). Cells similar in appearance to small lymphocytes were also seen in both the non-adherent and adherent layers. Cells associated with the adherent layer of this LTBMC system displayed the ability to reconstitute hematopoiesis in the irradiated host. It is suggested that because of its three-dimensional nature the nylon mesh LTBMC system possesses a greater hematologic potential per unit area than monolayer-type LTBMC.

A three-dimensional culture system for the growth of hematopoietic cells.

A physiological three-dimensional culture system was developed for the growth of human bone marrow. Bone marrow stromal cells were established on a nylon filtration screen template, suspended in liquid medium and grown to 70% confluence, and inoculated with hematopoietic cells. An intricate microenvironment is established to support hematopoiesis, which proceeds in a three-dimensional orientation. Analysis of the adherent zone of these cultures with flow cytometry and progenitor cell assays reveals multilineage hematologic expression and active proliferation of immature cells for the 12 week experimental period. Similar results were obtained with rat bone marrow cultures using this methodology. The suspended nylon mesh system is novel in that it supports the growth of several hematologic lineages concurrently. This system may lend itself to the growth of purged or untreated bone marrow for transplantation.

Stromal cells derived from spleen or bone marrow support the proliferation of rat natural killer cells in long-term culture.

Rat nylon wool nonadherent bone marrow cells were propagated for up to 75 days in co-culture with stromal cells derived from either spleen or bone marrow. Interleukin (IL) 1 enhanced the ability of spleen stroma to support the long-term culture of natural killer (NK) cells, ostensibly by inducing these support cells to synthesize other cytokines. Flow cytometry studies indicated that the nylon wool separation procedure enriched the concentrations of mature NK cells from 7.9% to 38.1% for splenocytes and from 3.8% to 19.5% for bone marrow cells. Analyses of the adherent zones of suspended nylon screen NK cell cultures revealed substantial numbers of large granular lymphocytes that expressed NK 323+/MOM/3F12/F2- phenotypes. The presence of both mature and immature cells of the NK lineage in this matrix was inferred by the presence of both IL-2 receptor (IL-2R) positive and IL-2R negative, and OX-8+ and OX-8- NK 323+ cells over the greater than 4-month experimental period. Suspended nylon screen cultures displayed a greater potential for producing cytolytic cells than either co-cultures of bone marrow nonadherent cells on stroma monolayers or suspension cultures. The large granular lymphocytes produced in suspended nylon screen cultures could be transformed into active killers of YAC-1 targets by IL-2. In contrast to bone marrow nonadherent cells, more splenic nylon-wool-passed cells displayed a mature NK phenotype, but their proliferative potential and ability to be transformed into cytolytic cells by IL-2 decreased rapidly in culture. In the suspended nylon screen culture system, NK cells migrate from the underlying stroma in stages as they mature, retain their cytolytic potential, and manifest a capacity for self-renewal. Cultured cells were routinely dissociated into single cell suspensions via enzyme treatment and were reinoculated onto "fresh" nylon screen/stromal cell templates after passage through nylon wool columns. These co-cultures continued to generate cytolytic cells in numbers greater than those of the initial inoculum.

The influence of pancreatic hormones and diabetogenic procedures on erythropoietin production.

The influence of the pancreas on renal and extrarenal erythropoietin (Ep) production and on the elaboration of the hepatic erythropoietic factor (HEF) was studied in these experiments. Insulin was found to elevate Ep levels in the anephric hypoxic rat when compared to controls, whereas glucagon treatment augmented the hepatic Ep response to hypoxia in the subtotally hepatectomized (hepx) animal while lowering it in the renal intact rat. Production of experimental diabetes either through chemical induction by alloxan or following pancreatectomy diminished the Ep response in all groups tested. Treatment with antiglucagon caused an elevation in the Ep response to hypoxia in the intact rat but lowered Ep levels in the hepx animal. In addition, glucagon and a synthetic hepatotrophic agent (L-histidyl L-lysine acetate) stimulated HEF production in the hepx rat, although none of the agents tested were capable of enhancing HEF levels in the intact rat.

The effect of methylcellulose on extrarenal erythropoietin production.

The macromolecular polymer, methyl cellulose (MC), is a known hepatosplenomegalic agent which promotes a state of experimental hypersplenism in rats. This is characterized by massive splenomegaly, pancytopenia of the blood elements, medullary hypoplasia, and marked gross and histologic alteration of the liver, kidney, adrenals, and lungs. Massive splenomegaly results from storage of this inert material by splenic macrophages. In the present study, chronic MC administration in rats augmented the hepatic Ep response to hypoxia but did not appreciably affect renal production of Ep. Splenectomy resulted in a decrease in the extrarenal Ep response to hypoxia indicating a possible role of the massively enlarged spleen of these MC-treated rats in extrarenal Ep production. The augmentation of extrarenal Ep elaboration may be attributed to a stimulatory effect of MC on the hepatic and splenic macrophages.

Hepatic and renal erythropoietin production during various intervals of short-term hypoxia.

Hypoxia stimulates erythropoietin (Ep) production and subsequent stimulation of erythropoiesis. Chronically applied hypoxia results in a variety of physiological alterations including polycythemia, hemoglobinemia, and vascular and metabolic changes. As red cell mass increases to such an extent to allow for adequate tissue oxygenation, plasma Ep levels drop and the rate of erythropoiesis is lowered. In the present work, the precise contribution of the kidney and the liver to the Ep response to acutely administered hypoxia (6-24 hr/0.4 atm) was determined using an in situ tandem perfusion of the liver and kidneys. The results indicate that hepatic and renal Ep production during different intervals of acute hypoxia is not uniform. This alteration may be partially explained by the reported reduction in splanchnic blood flow and a possible redistribution of hepatic microcirculation, which may cause the liver to become more hypoxic than the kidney under the same ambient pO2. The role of the liver becomes more pronounced as the time of acute hypoxic exposure increases, whereas the Ep content of renal perfusates is not drastically altered. At certain intervals during hypoxia, the liver displays a greater potential to elaborate Ep.