Diminished response of granulocyte-macrophage colony-stimulating factor (GM-CSF) in mice after sensitisation with bacterial cell-wall components.

The secondary induction of serum granulocyte-macrophage colony-stimulating factor (GM-CSF) by structurally unrelated and chemically highly purified bacterial cell-wall components (BCWC) was studied. Homologous challenge of mice 7 days after treatment with lipid A, lipoprotein or murein failed to increase serum GM-CSF levels and the extent of the decreased responsiveness was dependent upon the dose used for the initial injection. Lipid A-induced decreased responsiveness took 48 hours to develop and remained fully expressed approximately up to day 7 following injection. Then responsiveness reappeared gradually and was virtually normal 4 weeks after injection. Lipoprotein-induced decreased responsiveness developed in a similar manner but peristed maximally over the whole 8 week period studied. The decreased responsiveness induced by the injection, 7 days previously of either lipid A or lipoprotein was not specific as cross-challenge also failed to elevated GM-CSF to normal levels. On the other hand, 7 weeks after priming the lipoprotein induced decreased responsiveness was found to be specific. Mixing experiments failed to show increased levels of GM-CSF inhibitors in the serum from mice injected 7 days previously with lipid A and decreased responsiveness could not be transferred with serum to normal recipients. Similarly the lowered GM-CSF response to lipoprotein could not be transferred wth serum collected 7 weeks after primary injection. Medium conditioned by spleens from mice injected with lipid A contained less detectable GM-CSF than medium conditioned by normal spleens, but a variety of other organs did not show this difference. Mixing experiments failed to show significant differences in GM-CSF inhibitory activity between the two types of spleen conditioned media.

Humoral regulation of splenic hemopoiesis in mice.

The effects on splenic hemopoiesis were investigated of injecting the bacterial cell wall component, lipid A, or post-lipid A serum (PLAS) into mice. Both lipid A and syngeneic PLAS caused an increase in splenic numbers of multipotential hematopoietic stem cells (CFUS), committed hemopoietic progenitor cells (CFC) of different hemopoietic lineages and morphologically recognizable hemopoietic cells of various lineages. C57BL/6 Sld/Sld PLAS had a lower stimulating effect on hemopoiesis than C57BL/6 +/+ PLAS. PLAS from pre-irradiated mice was as active as normal PLAS in elevating splenic CFC levels. Serum from mice which received irradiation only, had no stimulating effect on splenic hemopoiesis. Medium conditioned by the myelomonocytic leukemia cell line WEHI-3B elevated splenic CFC numbers of similarly to PLAS, but supernatants from long-term marrow cultures or serum of mice treated with latex or phenylhydrazine did not have such an effect. Despite its marked effect on splenic numbers of nucleated erythroid cell precursors, PLAS did not contain elevated levels of erythropoietin.

Hemopoietic effects in mice of a lipid A-associated protein.

The effects of the bacterial cell-wall components (BCWC) lipid A and lipid A-associated protein (LAP) on humoral and cellular hemopoietic parameters were investigated in mice. Both lipid A and LAP increased serum levels of granulocyte-macrophage colony-stimulating factors (CSF) in C57BL/6 mice. In C3H/HeJ mice the CSF responses to lipid A and LAP were 7 and 3 fold less than the corresponding CSF responses found in C3H/GSF mice. Both BCWC increased the numbers of splenic multipotential hemopoietic stem cells (CFUS) as well as colony-forming cells (CFC) for neutrophilic granulocytes, macrophages, eosinophils and megakaryocytes. Lipid A but not LAP caused a marked decrease in the femoral numbers of B lymphocyte colony-forming cells (BL-CFC). The Bl-CFC incidence in the spleen or in the mesenteric lymph node changed little if at all after injection of either of the two BCWC. Morphological analysis of marrow cells showed an increase in the proportion of myeloid cells and a concomitant decrease in the proportion of erythroid precursor cells after injection of both BCWC. In the spleen, lipid A but not LAP caused an increase in the proportion of myeloid cells, erythroid precursor cells and plasma cells. In all experiments where both BCWC showed activity, lipid A was more potent than LAP on a weight basis.

The effect of two different types of colony-stimulating factor on the expression of aminopeptidase on marrow-derived murine macrophages.

The expression of aminopeptidase, a surface-membrane-bound enzyme, on macrophages formed in liquid cultures of hemopoietic progenitor cells was studied over a period of 20 days. The cultures were stimulated by two biochemically distinct types of colony-stimulating factor (CSF) derived from mouse-lung-conditioned medium (MLCM) and L-cell-conditioned medium (LCCM), respectively. The enzyme content of single cells was determined microphotometrically after staining with Fast Blue B salt and leucine 4-methoxy-2-naphthylamide. In LCCM-stimulated cultures the number of cells expressing aminopeptidase, the enzyme content per cell and the enzyme concentration increased markedly from day 10 onward, while macrophages from MLCM-stimulated cultures only showed borderline yet significantly positive aminopeptidase levels. Maximum enzyme concentrations were found earlier than maximum enzyme content indicating an early local increase in the aminopeptidase concentration on the membrane and subsequently a more uniform distribution over the cell surface. The two types of CSF differ not only in their effect on macrophage production but also in their influence on the expression of the surface enzyme aminopeptidase on these cells.

The response of murine splenic granulocyte-macrophage colony-forming cells to lipid A in vivo.

The physical and biological properties of murine splenic granulocyte-macrophage colony-forming cells (GM-CFC) were analyzed after the injection of the splenic hemopoiesis stimulating agent lipid A. In continuous gradients of Percoll, the majority of the splenic GM-CFC of untreated mice peaked at a buoyant density of 1.090 g/cm3, while a small second GM-CFC peak could be detected at 1.065 g/cm3. One day after the injection of lipid A, the splenic GM-CFC were almost equally distributed among these two density peaks. This altered proportion was still detectable 72 to 96 h later, although to a smaller extent. No difference in the responsiveness to the colony-stimulating factor (GM-CSF) from mouse-lung conditioned medium (MLCM) was observed between these two density subpopulations. The differentiation pattern of splenic GM-CFC was altered after the injection of lipid A. However, this altered pattern was the same in both density subpopulations. The percentage of splenic GM-CFC as well as the percentage of multipotent hemopoietic stem cells (CFUs) in DNA synthesis were markedly elevated after the injection of lipid A. A striking difference in the proliferative activity was found between high- and low-density GM-CFC in post-lipid A spleens. 24 h after the injection of lipid A, 43% of the high-density GM-CFC subpopulation was found in S according to the suicide technique using tritiated thymidine, whereas in the low-density fraction only 10% of the population was killed. This finding allows alternative interpretations.

Biological and biochemical properties of a serum factor that stimulates splenic hemopoiesis in mice.

Some biological and biochemical properties of a distinct hemopoietic factor that stimulates splenic hemopoiesis in mice are described. This factor can be detected by measuring the increase in the number of in vitro hemopoietic colony-forming cells (CFCs) in the spleens of mice after transfer of serum from syngeneic donors that have been treated previously with the bacterial cell-wall components: lipid A or lipoprotein. Serum collected 5 min after the IV injection of lipid A contained almost no splenic hemopoiesis-stimulating factor (SHSF). The highest serum levels of the factor were found between 30 min and 3 h after lipid A was injected IV. The residual levels of lipid A or lipoprotein in the serum of treated mice were too low to account for their splenic hemopoiesis-stimulating effects. A component of SHSF in both post-lipid-A serum (PLAS) and postlipoprotein serum (PLPS) bound to concanavalin A (Con A)-Sepharose and could be eluted by alpha-methyglucopyranoside (0.05 M). Partial fractionation of PLAS using Con A-Sepharose and gel filtration (Sephacryl S-200) indicated that the SHSF glycoprotein had an apparent molecular weight of 30,000 daltons. SHSF was detected in serum in response to lipid A and lipoprotein, but this was separable (by gel filtration) from the major form of granulocyte-macrophage colony-stimulating factor (GM-CSF) in PLAS.

The responses of hemopoietic precursor cells in mice to bacterial cell-wall components.

The influence upon different cellular and humoral parameters of hemopoiesis of three structurally unrelated, highly purified bacterial cell-wall components (BCWC) was investigated. The spleens of C57BL/6 mice assayed 6 days after the injection of either lipid A or outer-membrane lipoportein, but not murein, showed a marked increase in granulocyte-macrophage, eosinophil, and megakaryocyte progenitor cell levels. The number of pluripotent hemopoietic stem cells (CFU-S) also increased in the spleens of mice treated with either lipid A or lipoprotein. Similar results were obtained following the injection of lipoprotein or lipid A into CBA or C57BL/6.nu mice. Genetically anemic Wf/Wf mice were found to have spontaneously elevated numbers of splenic progenitor cells, which increased further after the injection of lipid A. The proportions of the different splenic progenitor cell types were similar in both untreated and lipid A treated Wf/Wf mice, and in normal littermate controls. When tested in vitro, unfractionated or partially purified post-lipid A serum was found to stimulate the growth of granulocyte-macrophage progenitor cells (GM-CFC), but no detectable stimulation of eosinohphil, megakryocyte, or erythroid progenitor cells was observed. The data suggest that the rise in splenic levels of the different progenitor cells is not mediated by the corresponding types of CSF, but more likely by proliferation and differentiation of CFU-S.

Cellular and molecular basis of the increased splenic hemopoiesis in mice treated with bacterial cell wall components.

An analysis was made of the mechanisms responsible for the increased splenic hemopoiesis occurring in mice after the injection of the bacterial cell wall components lipid A and outer membrane lipoprotein. No evidence was obtained for the presence of functional lipid A receptors on hemopoietic precursor cells. Serum from lipid A-injected mice, on injection into normal mice, induced in the spleen an increased content of all hemopoietic progenitor cells. The magnitude of the response was dependent on the dose of lipid A used and the volume of serum transferred to the recipients. C3H/HeJ mice unresponsive to lipid A exhibited similar spleen changes when injected with active post-lipid A sera. Progenitor cells of all hemopoietic lineages, including multipotential hemopoietic stem cells, were involved in the response. The results suggest that a humoral factor mediates the lipid A-induced increase of splenic hemopoiesis.

Serum of lipopolysaccharide-treated mice contains two types of colony-stimulating factor, separable by affinity chromatography.

Serum from mice treated with bacterial lipopolysaccharide (LPS) was fractionated by Con A-Sepharose affinity chromatography, and assayed in vitro for colony-stimulating factor (CSF) using mouse bone marrow cells. The CSF failing to bind to concanavalin A-Sepharose (pool A) had similar biological properties to the unfractionated serum, i.e., it stimulated the formation of about equal numbers of granulocytic, mixed granulocyte-macrophage and macrophage colonies. The fraction eluted from the Con A-Sepharose column with alpha-methyl-D-glucopyranoside (pool B) had a steeper dose-response curve than either the unfractionated serum or the pool A CSF and most of the colonies were composed of macrophages. A mixture of the pool A and pool B CSFs stimulated colonies in a similar way as unfractionated serum and poolA. The apparent molecular weights of the two types of CSF were determined by two different gel-filtration procedures. Sephacryl S-200 gel-filtration suggested an apparent molecular weight of 85,000 for pool A CSF and 180,000 for pool B CSF. Gel-filtration on Sepharose CL-6B in the presence of guanidine hydrochloride (6M) yielded an apparent molecular weight of approximately 23,000 for pool A CSF and 33,000 for pool B CSF. The colony-forming cells (CFC) responding to pool B CSF were found to have a relatively high sedimentation velocity (peak sedimentation velocity 5.6--6.2 mm/hr) compared to the CFC responding to mouse-lung conditioned medium (MLCM) whose peak sedimentation velocity was between 4.0--4.5 mm/hour. The CFC responding to pool A CSF had an intermediate sedimentation velocity (peak 4.6--5.2 mm/hour). A time-course analysis of the morphology of clones or colonies in cultures stimulated with either MLCM or pool B CSF showed that the proportion of different colony types depends significantly on the incubation period and suggested that pool tb csf induced an early commitment of CFC towards macrophages differentiation.

Lysolecithin analogs as adjuvants in delayed-type hypersensitivity in mice. II. Studies on the mode of action.

Lysolecithin analogs (LLA) possess adjuvant activity in delayed-type hypersensitivity (DTH). To elucidate their mode of action, we studied the influence of LLA on the recruitment of hemopoietic progenitor cells and on the induction and transfer of DTH. Whereas the number and composition of colony-forming cells (CFC) in the bone marrow remained unaltered in LLA-treated mice, the number of CFC in the spleen was augmented severalfold, and the content of macrophage progenitors was remarkably increased. DTH was induced with macrophages from peritoneal cells (PC) pulsed with keyhole limpet hemocyanin (KLH). A more marked DTH response was obtained with KLH-pulsed PC from LLA-treated mice than from mice given phosphate-buffered saline (PBS). On the other hand, there was no difference in the DTH response following adoptive transfer of sensitized cells to naive mice pretreated either with PBS or LLA. It is concluded that the adjuvant effect of LLA can be accounted for partly in terms of their effects on macrophage progenitors and partly in terms of a more effective T cell activation by antigens associated with LLA-derived macrophages.

Modulations of myelopoiesis in vivo by chemically pure preparations of cell wall components from gram-negative bacteria: effects at different stages.

Modulation of myelopoiesis by chemically pure preparations of different cell wall components from gram-negative bacteria was investigated in vivo. The effects of lipid A, outer membrane lipoprotein, and murein were evaluated at several distinct stages: induction of colony-stimulating activity (CSA) in the serum, increase in the number of committed splenic precursor cells (CFU-C) forming granulocyte-macrophage colonies in vitro, and triggering into the cell cycle of noncommitted hemopoietic stem cells (CFU-S) from bone marrow. The results reveal different patterns of activity of the bacterial cell wall components (BCWC) tested. (i) In C57Bl/6 mice and C3H/Bom mice, all three preparations were potent inducers of CSA. In C3H/HeJ mice, CSA was only induced by lipoprotein and murein and not by lipid A. After injection of lipid A or lipoprotein, but not murein, the number of CFU-C in spleens of C57Bl/6 mice was increased up to 100-fold. In C3H/Bom and C3H/HeJ mice, not only murein but also lipoprotein were much less potent in this respect. (iii) In C57Bl/6 mice, both lipid A and lipoprotein, but not murein, were capable of inducing the proliferation of CFU-S, as demonstrated by a hot thymidine cytocide technique. Thus, induction of CSA and changes in the pool size of splenic CFU-C after administration of BCWC may be unrelated events. On the other hand, the increase of CFU-C might reflect the mitogenicity of BCWC for CFU-S.

Cloning and frequency analysis of haemopoietic stem cells producing CFC over weeks in long-term marrow cultures.

Using adherent marrow cell layers devoid of stem cells, in vitro cloning and the determination of the frequency of murine haemopoietic stem cells were performed by means of limiting dilution. The presence of stem cells in individual microcultures was indirectly proven by the sustained presence of in vitro colony-forming cells (CFC). These cells increased in number as a function of the culture period, which seems to indicate that the CFC were, de novo, produced in vitro. Although stem cell frequencies comparable to the in vivo spleen colony assay were determined in some experiments, most of our frequency estimates suggested stem cell frequencies ranging from 5/10(6) to 90/10(6). After a period of approximately 4 weeks, the stem cell activity in the microcultures declined rapidly. This may have been caused either by an increased differentiation pressure governed by CSF and/or similar factors or by sub-optimal culture conditions.

Differential expression of lectin receptors during hemopoietic differentiation: enrichment for granulocyte-macrophage progenitor cells.

Molecular changes occur at the surface of hemopoietic cells during differentiation from progenitor cells to mature granulocytes and macrophages. The differential expression of surface carbohydrate residues has been probed using lectins and the results used to purify normal mouse granulocyte-macrophage progenitor cells. Ten different lectins were screened for selective interaction with mouse hemopoietic colony-forming cells (CFCs), using agglutination or a quantitative analysis of the number of fluoresceinated lectin molecules bound per cell using a fluorescence activated cell sorter (FACS). Pokeweed mitogen (PWM), Helix pomatia agglutinin (HPA), soybean agglutinin (SBA), and peanut agglutinin (PNA) preferentially bound to CFCs so that it was possible to enrich 4 to 10-fold for these progenitor cells by sorting for the highly fluorescent cells. Further analysis of the low and high angle light scattering characteristics of the CFCs indicated that these cells were polydisperse, but could be enriched ten-fold by selecting for cells with high intensity low angle (0 degrees) scatter and low intensity high angle (90 degrees) scatter. PWM gave the best enrichment (10 to 15-fold) for adult bone marrow CFCs, for CFCs from fetal sources (fetal liver, fetal blood), and for CFCs from the spleens of mice injected previously with outer membrane lipoprotein from E. coli. Three parameter sorting for CFC using the FACS (low angle scatter, high angle scatter, and PWM-fluorescence) resulted in large enrichment factors (16 to 50-fold) for CFCs from all the above sources. Over 7% of the cells sorted from bone marrow, 10% of the cells sorted from post-lipoprotein spleen, and 28% of the cells sorted from fetal peripheral blood were hemopoietic CFCs. Ninety percent of the cells in these fractions had the morphology of blast cells or myelocytes. Thus, it was possible to identify the morphological characteristics of the hemopoietic progenitor cells. Screening of other developmental systems using quantitation of fluorescence with lectins should prove of general value for the purification of selected differentiation states.