Bone Marrow Suspensions Research Articles

Purification and functional evaluation of mature neutrophils from human bone marrow

Human myeloid maturation proceeds within the bone marrow and results in a mature neutrophil that is released into the peripheral circulation. Previous reports have indicated that neutrophils from bone marrow demonstrate decreased adherence, impaired phagocytosis, and decreased nitroblue tetrazolium dye reduction when stimulated. Due to lack of a suitable method for isolating purified bone marrow neutrophils, these studies have been performed by microscopic techniques. We now report a method for isolating 1 X 10(8) neutrophils [bands plus polymorphonuclear leukocytes (PMNs)] from 10 mL of bone marrow aspirate sample. By means of a discontinuous Percoll-gradient centrifugation through densities of 1.085, 1.095, and 1.10 g/mL a leukocyte-rich suspension of bone marrow can be separated into three leukocyte layers. By combining the lower two leukocyte layers (M2/3), a population of neutrophils consisting of 26% bands and 63% PMNs is seen. When compared with peripheral blood PMNs, these bone marrow neutrophils had a lower alkaline phosphatase activity, decreased ingestion of Oil Red O-coated particles, impaired superoxide release on stimulation with the chemotactic peptide Fmet-leu-phe (FMLP) or the tumor promotor phorbol myristate acetate (PMA), and less activity of the NADPH-dependent oxidase. These results indicate that morphologically mature neutrophilic cells within the bone marrow exist in a still functionally immature state.

Purification and Functional Evaluation of Mature Neutrophils From Human Bone Marrow

Human myeloid maturation proceeds within the bone marrow and results in a mature neutrophil that is released into the peripheral circulation. Previous reports have indicated that neutrophils from bone marrow demonstrate decreased adherence, impaired phagocytosis, and decreased nitroblue tetrazolium dye reduction when stimulated. Due to lack of a suitable method for isolating purified bone marrow neutrophils, these studies have been performed by microscopic techniques. We now report a method for isolating 1 X 10(8) neutrophils [bands plus polymorphonuclear leukocytes (PMNs)] from 10 mL of bone marrow aspirate sample. By means of a discontinuous Percoll-gradient centrifugation through densities of 1.085, 1.095, and 1.10 g/mL a leukocyte-rich suspension of bone marrow can be separated into three leukocyte layers. By combining the lower two leukocyte layers (M2/3), a population of neutrophils consisting of 26% bands and 63% PMNs is seen. When compared with peripheral blood PMNs, these bone marrow neutrophils had a lower alkaline phosphatase activity, decreased ingestion of Oil Red O-coated particles, impaired superoxide release on stimulation with the chemotactic peptide Fmet-leu-phe (FMLP) or the tumor promotor phorbol myristate acetate (PMA), and less activity of the NADPH-dependent oxidase. These results indicate that morphologically mature neutrophilic cells within the bone marrow exist in a still functionally immature state.

Number of fibroblast colony-forming cells in mouse bone marrow

Data is given on extraction of an additional number of fibroblast colony forming cells from mouse bone marrow tissue by trypsinization, over and above the number released by mechanical disaggregation of the hematopoietic tissue, and on the effect of feeder cells on the efficiency of fibroblast colony formation (EFCF). Feeder cells were added to some of the cultures after explanation: suspensions of bone marrow or spleen cells for guinea pigs or mice, irradiated in a dose of 60 Hz (/sup 60/Co). The effect of additional feeder on the efficiency of colony formation was determined in experiments with explanation of suspensions of bone marrow cells with the addition of different numbers of feeder cells. It is found that by trypsinization and the use of an irradiated feeder it is possible to increase EFCF for mouse bone marrow and substantially by comparison with EFCF values achieved previously.

Restricted Expression of a New Acute Myelogenous Leukemia-Associated Antigen (NHL-30.5) on Normal Hemopoietic Progenitor Cells

Abstract We have previously reported the isolation of a monoclonal antibody (NHL- 30.5) that reacts with an antigen expressed on a substantial proportion of marrow and blood cells of most patients with newly diagnosed or relapsing acute myeloid leukemia. This antigen is also found on several cell lines derived from myeloid malignancies of human origin. It is not present on mature hemopoietic cells or on the majority of differentiating bone marrow cells. In order to determine whether the NHL-30.5 antigen may, nevertheless, be expressed on low-frequency primitive normal hemopoietic cells, not detected in standard antibody screening procedures, its expression was studied on clonogenic erythropoietic and granulopoietic cells. Light-density (less than 1.077 g/mL) suspensions of normal or chronic myelogenous leukemia bone marrow and peripheral blood cells were stained with NHL-30.5 and fluorescein isothiocyanate labeled second antibody and then sorted into two fractions using the fluorescence-activated cell sorter. The first contained the top 5% of cells with the highest fluorescence intensity. The remainder were collected in the second fraction. Colony assays of both fractions showed the first to be enriched in CFU-E, BFU-E, and CFU- C content (fourfold to 17-fold). The second fraction was correspondingly depleted of these progenitors. These findings reveal NHL-30.5 antigen expression to be a transient event during normal hemopoiesis that characterizes primitive hemopoietic cells on several pathways. Subsequent experiments showed that the presence of up to 10 micrograms/mL of purified NHL-30.5 antibody in colony assay cultures neither inhibited nor stimulated colony formation. Marrow fibroblasts (subcultured marrow adherent cells) were NHL-30.5 negative. Immunoprecipitation studies showed that the antigen detected by NHL- 30.5 is clearly distinct from that identified by My-10, another monoclonal antibody that has previously shown some similarities to NHL- 30.5. It thus appears that the NHL-30.5 antibody reacts with a new myeloid differentiation antigen of as yet unidentified function that is normally restricted in its expression to early stages of hemopoiesis.

Restricted expression of a new acute myelogenous leukemia-associated antigen (NHL-30.5) on normal hemopoietic progenitor cells

We have previously reported the isolation of a monoclonal antibody (NHL-30.5) that reacts with an antigen expressed on a substantial proportion of marrow and blood cells of most patients with newly diagnosed or relapsing acute myeloid leukemia. This antigen is also found on several cell lines derived from myeloid malignancies of human origin. It is not present on mature hemopoietic cells or on the majority of differentiating bone marrow cells. In order to determine whether the NHL-30.5 antigen may, nevertheless, be expressed on low-frequency primitive normal hemopoietic cells, not detected in standard antibody screening procedures, its expression was studied on clonogenic erythropoietic and granulopoietic cells. Light-density (less than 1.077 g/mL) suspensions of normal or chronic myelogenous leukemia bone marrow and peripheral blood cells were stained with NHL-30.5 and fluorescein isothiocyanate labeled second antibody and then sorted into two fractions using the fluorescence-activated cell sorter. The first contained the top 5% of cells with the highest fluorescence intensity. The remainder were collected in the second fraction. Colony assays of both fractions showed the first to be enriched in CFU-E, BFU-E, and CFU-C content (fourfold to 17-fold). The second fraction was correspondingly depleted of these progenitors. These findings reveal NHL-30.5 antigen expression to be a transient event during normal hemopoiesis that characterizes primitive hemopoietic cells on several pathways. Subsequent experiments showed that the presence of up to 10 micrograms/mL of purified NHL-30.5 antibody in colony assay cultures neither inhibited nor stimulated colony formation. Marrow fibroblasts (subcultured marrow adherent cells) were NHL-30.5 negative. Immunoprecipitation studies showed that the antigen detected by NHL-30.5 is clearly distinct from that identified by My-10, another monoclonal antibody that has previously shown some similarities to NHL-30.5. It thus appears that the NHL-30.5 antibody reacts with a new myeloid differentiation antigen of as yet unidentified function that is normally restricted in its expression to early stages of hemopoiesis.

Thymic stem cells: their interaction with the thymic stroma and tolerance induction.

There are a number of lines of indirect evidence which suggest that lymphoid stem cells which migrate into the thymus have considerable proliferative potential. The first indication that this is the case was provided by Wallis et al. (1975) who derived estimates of the number of syngeneic bone marrow stem cells which repopulate the thymus of mice given a potentially lethal dose of irradiation. By injecting various mixtures of two syngeneic but chromosomally distinguishable bone marrow suspensions, they were able using probability analysis to estimate that fewer than 10 stem cells colonize a single thymus lobe. More recently, Ezine et al. (1984), using irradiated mice reconstituted with limited numbers of mixed bone marrow cells from congeneic donors differing at the Thy-1 locus, have demonstrated in frozen sections that patterns of lymphoid repopulation suggest that distinct cell clones develop within the thymus.KeywordsTolerance InductionFetal Liver CellPancreas GraftThymic EpitheliumThymus LobeThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Standardization and characterization of the procedure for in vitro treatment of human bone marrow with cyclophosphamide derivatives.

Thirty human bone marrow (BM) suspensions from patients with acute leukaemia patients in remission were processed with the Haemonetics 30 flow cell separator in order to separate buffy-coats and to treat them in vitro with a derivative of cyclophosphamide (ASTA Z 7557). After processing, the volumes of BM suspensions were reduced to 25%. Recoveries of leucocytes, CFUc and BFUe were respectively 62, 85 and 84%. In vitro treatment with doses of ASTA Z ranging from 50 to 140 micrograms/2 X 10(7) leucocytes (according to the CFUc sensitivity of each patient) destroyed 95 +/- 5% of initial CFUc. After freezing and thawing, recovery of CFUc from treated BM was poor (24%) in comparison to that obtained with untreated BM (79%).

Differentiation of normal marrow and HL60 cells induced by antithymocyte globulin.

Antithymocyte globulin (ATG) therapy is an important treatment alternative for patients with acquired aplastic anemia. The mechanism by which it exerts its effects on hematopoiesis is unknown. In this report, we describe the ability of horse ATG to induce growth and differentiation of normal bone marrow. A single cell suspension of normal human bone marrow was cultured in methylcellulose medium and examined for the growth and maturation after incubation with ATG (10 micrograms/ml). After 3-4 days of culture, spherical colonies containing mature myeloid elements were found in cultures containing ATG but not in cultures containing medium or preimmunization horse IgG. The addition of 10% colony-stimulating factor increased growth by 40%. The number of spherical colonies is not dependent on the presence of macrophages or T lymphocytes. This property of ATG may be relevant to the mechanism behind the hematologic recovery in some patients with acquired aplastic anemia. We also describe the ability of ATG to induce terminal differentiation in the HL60 leukemic cell line. ATG binds to HL60 cells and at concentrations between 10 and 100 micrograms/ml, 50% of the cells become mature granulocytes, acquire the ability to reduce nitroblue tetrazolium, and lose their proliferative capacity in the clonogenic assay. These new observations of ATG-induced differentiation of normal marrow myeloid elements and terminal differentiation of the HL60 cell line point to different avenues for future search of differentiation-inducing agents.

Characterization of subsets of bone marrow-derived macrophages by flow cytometry analysis.

Normal C3H bone marrow cells were grown 7 days in medium containing L cell-derived colony stimulating factor-1 (CSF-1). During the first 4 days of culture, erythroid and granulocytic cells decreased while macrophages increased exponentially with a doubling time of about 31 hr. Only 0.3% of all cells in the initial bone marrow suspension formed discrete colonies of mononuclear phagocytes, but by day 6 60% of the nonadherent cells were capable of forming macrophage colonies, representing a 200-fold enrichment of the original progenitor population. Using flow cytometry, mononuclear phagocytes obtained after 4 days of culture were separated into two distinct phenotypes based on their autofluorescence. Nonadherent cells were a discrete population of small cells exhibiting low autofluorescence, and the adherent cells were a broad heterogeneous population of large cells exhibiting high autofluorescence. A panel of currently available rat monoclonal antibodies (MABs) against murine hematopoietic cells were used to determine whether unique subsets of macrophages could be resolved. The MABs RA 31B6 and H-11 stained virtually all the nonadherent cells but not adherent cells. The MABs E-2 and 11-4.1 (anti-H-2Kk) stained almost all the adherent cells and demonstrated no significant staining of nonadherent cells. Nearly all the nonadherent and adherent cells were stained by the MABs DNL 4.4 and MAC-1. Additionally, the data suggest that the epitopes for MAC-2 and MAC-3 and gamma 2a Fc receptors develop late in nonadherent progenitor cells as they mature into adherent macrophages.

内毒素性ＤＩＣと骨髄細胞の組織トロンボプラスチン活性

1. Endotoxin injection caused significant increase in procoagulant activity (PCA) of bone marrow cells. Endotoxin also caused an increase in PCA of supernatant (sup) obtained from bone marrow suspension.2. Both activities were destroyed by heating at over 56°C. The activity of mouse brain thromboplastin used as reference preparation was destroyed at 60°C 'but stable at 80°C.3. PCA of bone marrow cells decreased after incubation with Concanavalin A (Con A). The activity recovered by addition of α-methyl-D-glucoside.4. Ploymyxin (PxB) inhibited the PCA of either cells (20%-inhibition) or sup (50%-inhibition).5. Addition of phospholipid (PL) to the cells or sup caused enhancement of PCA.6. Degrees of heat inactivation, Con A- or PxB-inhibition, and PL-enhancement of PCA were higher in sup than those in the bone marrow cells.Procoagulants of the cells and sup were shown to be tissue thromboplastin (Tp) which consists of glycoprotein-phospholipid complex functioning as active sites. Membrane perturbation due to endotoxin would initiate the increase in exposure and/or contents of the active site (PL) on the membrane surface and would participate in the manifestation of Tp activity. As a result of membrane alteration, release of Tp from the cell surface into circulation may also be caused in endotoxemia.

Protective effect of bone marrow and spleen suspensions on radiation-induced leukemogenesis in C57BL/6 mice.

The present experiments are an attempt to precise the type and localization of the cells involved in the protective effect of hemopoietic suspensions against the radiation-induced thymic lymphosarcoma (TLS) of C57BL/6 mice. Inocula containing variable numbers of BM or spleen CFUs from 60-day-old and 360-day-old donors were tested. According to their origin, the suspensions differed with respect to the CFU replication rate, the CFU ability to differentiate towards the T lineage and the content of the suspensions in thymic precursors. Two levels of inhibition were observed: BM suspensions from 60-day-old donors containing 1,500 CFUs had the best protective effect: 14.5% of TLS; 1,500 CFUs from 360-day-old donors were slightly but not significantly less efficient (28.5%). The second level of inhibition (36-46% of TLS) was obtained with all the following inocula: a) 1,200 and 300 spleen CFUs or 300 and 95 BM CFUs from 60-day-old donors, b) 1,500 spleen CFUs from aged donors. Seventy-six spleen CFUs from 60-day-old donors, 120 BM or 175 spleen CFUs from aged donors had no effect. These results suggest that in addition to the high replication rate of the BM CFUs as compared with spleen CFUs, cells endowed with an optimal protective effect are present in BM suspensions and are either absent or present in very small amount in spleen suspensions. These cells which induce an early repopulation of the thymus might correspond to thymic precursors.

Proliferation of antibody-forming cells within the lymphoid system.

Mice were given subcutaneous priming injections of horse radish peroxidase (HRP) in the hind leg foot pads, and ten weeks later intravenous booster injections. Cell suspensions of popliteal lymph nodes, spleen and bone marrow were obtained at various time intervals after boosting and incubated in vitro for 1 h in the presence of 3H-thymidine (3H-TdR). In each suspension the number of cells showing simultaneously anti-HRP antibody and incorporation of 3H-TdR was determined. Whereas in spleen and bone marrow only very few 3H-TdR labelled antibody-forming cells were observed, a clear proliferative response was noticed in the popliteal lymph nodes. Already on the first day after booster injection proliferation of cells showing incipient antibody synthesis was demonstrable in lymph nodes. It is strongly suggested that most - if not all - antibody-forming cells in these animals are derived from proliferating cells in the local draining lymph nodes. A relationship between the route of primary immunization and nature and location of the secondary proliferative response is discussed.

Evaluation of erythropoiesis in long-term hamster bone marrow suspension cultures: absence of a requirement for adherent monolayer cells

In long-term hamster bone marrow cultures, proliferation and differentiation of hemopoietic stem cells occurs for several months without need for hydrocortisone or adherent stromal elements, which are requirements for bone marrow growth in all other species studied. Only the most primitive erythroid progenitors (BFU-E) are produced in the cultures. Following treatment of the cells with erythropoietin, these progenitor cells undergo differentiation into mature hemoglobinized red blood cells. Concomitant addition of erythropoietin (Epo) and prostaglandin-E1 (PGE1) results in the production of large numbers of maturing red blood cells. In cultures stimulated with Epo and PGE1, as many as 70% of the cells are benzidine-positive, while Epo alone stimulated as many as 45% of the cells to become erythroid. Epo and PGE1 do not have any apparent deleterious effect on the continuous hemopoiesis occurring in these cultures. Under identical conditions, syngeneic adherent cell cultures do not produce any erythroid elements. The development of mature red blood cells from primitive erythroid precursors occurs in the presence of Epo alone and without any apparent need for adherent stromal elements. These cultures provide a useful in vitro model for dissecting the positive and negative signals that regulate erythropoiesis.

Presence of mixed colony-forming cells in long-term hamster bone marrow suspension cultures: response to pokeweed spleen conditioned medium

In contrast to the murine system, long-term hamster bone marrow suspension cultures maintain proliferation of both pluripotent and committed stem cells in the absence of an adherent layer and without addition of exogenous factors, such as hydrocortisone. Addition of pokeweed-mitogen-stimulated hamster spleen conditioned medium (SCM) to these long-term suspension cultures produces an increase in the number of mixed colonies assayed in soft-agar, These mixed colonies, which contained four cell lineages--granulocytic, erythroid, megakaryocytic, and macrophage--could be generated from cells grown in suspension for over 6 mo. Addition of SCM also induces an initial rapid expansion of the myeloid compartment, and this expansion results in 70% of the cells being terminally differentiated granulocytes. In contrast, addition of SCM to hamster bone marrow cultures containing both adherent cells and hematopoietic stem cells produced no change in the number of mixed colonies generated in the culture. This system allows the in vitro study of the process of stem cell proliferation and differentiation and also provides a means to examine the relationship of adherent and supernatant bone marrow populations.

Presence of Mixed Colony-Forming Cells in Long-Term Hamster Bone Marrow Suspension Cultures: Response to Pokeweed Spleen Conditioned Medium

In contrast to the murine system, long-term hamster bone marrow suspension cultures maintain proliferation of both pluripotent and committed stem cells in the absence of an adherent layer and without addition of exogenous factors, such as hydrocortisone. Addition of pokeweed-mitogen-stimulated hamster spleen conditioned medium (SCM) to these long-term suspension cultures produces an increase in the number of mixed colonies assayed in soft-agar, These mixed colonies, which contained four cell lineages--granulocytic, erythroid, megakaryocytic, and macrophage--could be generated from cells grown in suspension for over 6 mo. Addition of SCM also induces an initial rapid expansion of the myeloid compartment, and this expansion results in 70% of the cells being terminally differentiated granulocytes. In contrast, addition of SCM to hamster bone marrow cultures containing both adherent cells and hematopoietic stem cells produced no change in the number of mixed colonies generated in the culture. This system allows the in vitro study of the process of stem cell proliferation and differentiation and also provides a means to examine the relationship of adherent and supernatant bone marrow populations.

Modulation of bone marrow heme and protein synthesis by trace elements

We examined the effect of heavy metals and trace elements, such as Cu, Au, Cd, Zn, and Pb, on heme synthetic and degradatory enzymes in rat bone marrow suspensions. Cu, Cd, and Au, at a low concentration of 5 × 10 −6 m, inhibited δ-aminolevulinic acid (δALA) dehydratase by 61, 47, and 47% respectively. A similar concentration of Zn did not inhibit the enzyme activity and, in fact, was able to protect against Au and Pb inhibition in the normal marrow. Cd was unable to recover any of the enzyme activity and, in fact, caused further inhibition. We determined the rate of cellular protein synthesis by measuring the rate of incorporation of radiolabeled precursor [ 14C]leucine. Cu and Au, at a concentration of 10 −5 m, inhibited leucine incorporation into marrow globin and nonglobin protein by 26 and 34%, respectively. Substantially higher concentrations of Zn, as well as Fe, were required to inhibit leucine incorporation into protein. These studies led us to postulate that the inhibition of δALA dehydratase by heavy metals was directly related to the inhibition of heme and protein synthesis in the bone marrow cell suspensions. No evidence of a significant change in heme oxygenase activity in vitro was observed after treatment with any of the heavy metals or trace elements used. These results suggest that trace elements by themselves or in a combination reduce the capacity of the erythroid cell to synthesize heme, the prosthetic group of many hemoproteins. In addition, we have shown that in an iron-overloaded rat the bone marrow contains approximately 45% more heme oxygenase activity than that of the control. Administration of Au alone did not increase heme degradation, but the combination of Fe and Au resulted in a marked induction of heme oxygenase. Finally, for the first time we were able to demonstrate that excess iron and/or decreased Zn concentration may predispose δALA dehydratase to inhibition by an interaction with other metals, such as Pb, by an undefined mechanism(s).

Distribution of nervous system-specific forms of enolase in peripheral tissues

The distribution of 3 forms of rat enolase (αα, αγ and γγ forms), including nervous system-specific forms (αγ and γγ), was determined in various tissues with a sensitive enzyme immunoassay system. The brain and spinal cord contained more than 100 pmol/mg protein of the αα, αγ and γγ enolases. Organs such as the lungs, heart, spleen, liver and kidney contained similarly high levels of αα enolase, but these tissues contained αγ and γγ enolases at levels less than 1% of the central nervous tissues. High levels of the αγ (10pmol/mg) and the γγ (1.5pmol/mg) forms were found in rectum, bladder, and uterus. In gut, major portions of the nervous system-specific forms were localized in the muscle layers. Skeletal muscle and diaphragm, which are composed of striated muscle, contained low levels of the 3 forms of enolase. Megakaryocytes separated from the suspension of bone marrow contained 11.3, and 0.53 amol/cell of the αα and γγ enolases, respectively, with little, if any, of the αγ form.

Effect of X Irradiation on Release of Prostaglandin E from Marrow Stromal Cells in Culture

GIBSON, D. P., DEGOWIN, R. L., AND KNAPP, S. A. Effect of X Irradiation on Release of Prostaglandin E from Marrow Stromal Cells in Culture. Radiat. Res. 89, 537-545 (1982). Earlier we showed that increasing the inoculum of mouse bone marrow cells to form the underlayer of a split-phase culture first enhanced and then suppressed erythroid colony (BFUE, CFU-E) formation related to increasing concentrations of prostaglandin E (PGE) released from the marrow stromal cells (MSC). These nonhemopoietic adherent stromal cells from suspensions of bone marrow have been associated with the repopulation of irradiated murine bone marrow in vivo. Studies were initiated to determine whether incremental doses of X rays to cultures of replicating and static MSC affected the release of PGE and erythroid colony formation. The latter was suppressed with doses of 100-800 rad which inhibited MSC colony formation and PGE release by a factor of 7. Static or confluent cultures of MSC in an underlayer produced no PGE and suppressed erythroid colony formation until cytotoxic doses of 10,000-20,000 rad released stimulatory concentrations of PGE. Production of PGE by replicating MSC was directly related to the size of the bone marrow stromal cell inoculum, whereas confluent cultures with markedly diminished tritiated-thymidine([3H]TdR) labeling indices released no detectable PGE. The results of this study emphasize the importance of knowing the proliferative status of the MSC underlayer and the PGE levels in evaluating hemopoiesis in culture.

Elimination of T-lymphocytes from human bone marrow with monoclonal T-antibodies and cytolytic complement.

Complement-mediated cytolysis has been used to remove T-lymphocytes from suspensions of human peripheral blood and bone marrow. Selective T-cell removal was investigated by three monoclonal antibodies. OKT3, MBG6 and OKT11A. All three removed greater than 90% of T-cells but combinations were necessary to kill greater than 99% of T-cells in vitro. The macrophage-granulocyte and erythroid colony forming cells of the bone marrow were spared. The method can be applied on bulk BM samples during clinical BM transplantation and will be useful to establish whether the virtually complete removal of T-lymphocytes totally prevents transplant associated graft-versus-host disease in man.

Evidence that tissue mast cells derive from mononuclear phagocytes.

A new method has been developed that allows in vitro differentiation of rat mast cells from precursor cells in the absence of feeder layers. In the present investigation, the precursor cells are further characterized as to their nature, state of activation and distribution in different organs of the rat. The data support our previously published evidence, based on morphology and enzyme-cytochemical reaction patterns that the precursor cells are mononuclear phagocytes. Peritoneal macrophages elicited by mineral oil, thioglycollate or proteose-peptone have lost the ability to differentiate into mast cells. Single cell suspensions of spleen and bone marrow contain mast cell precursors, while cells from mesenteric lymph nodes and thymus or purified lymphocytes never yield mast cells in this culture system.