Inhibitory effect of granulocyte-macrophage colony-stimulating factor therapy on the generation of natural killer cells.

We investigated the effects of recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) and recombinant human granulocyte colony-stimulating factor (rhG-CSF) therapy on the natural killer (NK) cell lineage in patients with aplastic anemia and myelodysplastic syndrome. Selected bone marrow (BM) cells were prepared by the elimination of nylon wool-adherent cells and mature T and NK cells from BM cells. The frequency of BM NK progenitors relative to BM cells selected was significantly decreased 4 weeks after the start of rhGM-CSF therapy (P less than .01), while the peripheral blood NK cell count and NK activity were also significantly decreased (P less than .05). A return to the pretreatment levels was seen 4 weeks after the cessation of treatment in all cases. No suppressive effect was noted in the patients who received rhG-CSF therapy. These results suggest that rhGM-CSF therapy suppresses the generation of NK cells from human BM NK progenitors.     

Reduced TGF-beta1 in patients with aplastic anaemia in vivo and in vitro

Transforming growth factor beta (TGF-beta) 1 is a ubiquitous bifunctional cytokine implicated in the regulation of haemopoietic stem cells and bone marrow stromal cells. We analysed sera from 63 patients with aplastic anaemia and describe a significant reduction of TGF-beta1 that was directly related to their treatment status. Untreated patients (n = 35), patients who did not respond (n = 15) and those with a partial response (n = 23) to treatment had significantly lower TGF-beta1 than the normal control group (n = 55), P < 0.0001, P < 0.0001 and P = 0.002 respectively. Patients in complete remission (n = 15) exhibited TGF-beta1 serum levels comparable to the control group. In addition, there was a correlation (r = 0.83, P < 0.0001) between serum TGF-beta1 and platelet count at time of sample. We have demonstrated that the primary source of TGF-beta1 in peripheral blood mononuclear cell (PBMC) cultures was not CD3-positive cells. These data indicate aplastic anaemia is associated with a decreased TGF-beta1 expression in peripheral blood circulation, which may be a direct consequence of thrombocytopenia. In vitro stromal layers grown from aplastic patient bone marrow (n = 14) produced significantly lower levels of TGF-beta1 (P = 0.02) when compared to normal stroma (n = 15). In the aplastic anaemia bone marrow compartment we postulate that accessory cells down-regulate TGF-beta1 expression to allow stem cell cycling to counteract hypoplasia. As TGF-beta1 is important in the regulation of haemopoiesis, dysregulation of this cytokine in combination with previously described abnormal cytokine expression may contribute significantly to the pathophysiology of aplastic anaemia by exacerbating primary stem cell defects.     

Intracellular Fas ligand is elevated in T lymphocytes in severe aplastic anaemia

Increased expression of Fas receptor by haemopoietic progenitors in aplastic anaemia (AA) suggests that excessive apoptosis contributes to multilineage bone marrow (BM) failure. To investigate the role of Fas ligand (FasL) in triggering progenitor cell death, we examined FasL levels in T lymphocytes of patients with severe untreated AA (n = 8). Expression of FasL on the surface of CD3+ cells was not detectable. However, flow cytometric analysis of saponin-permeabilized cells demonstrated higher levels of intracellular FasL in AA than in normal T cells (P < 0.005), both prior to and following activation with phytohaemagglutinin. Confocal microscopy revealed that FasL-specific signals overlapped with cathepsin D staining, indicating that intracellular FasL is stored in lysosomal granules. Levels of intracellular FasL in patients examined 1 month after immunosuppression with antilymphocyte globulin and cyclosporin A were lower than prior to treatment. The caspase inhibitors, DEVD and zVAD, enhanced colony formation and prolonged survival of AA BM cells in liquid cultures by about 10-fold (P < 0.05). Taken together, these data provide further evidence that apoptosis by the Fas/FasL system plays a role in the depletion of stem cells in AA.     

Infective complications of aplastic anaemia

Patients with aplastic anaemia have a relatively specific immune defect--leucopenia with neutropenia. We have carried out a retrospective analysis of infective episodes in 11 patients with aplastic anaemia. 5723 follow-up days accrued and 29 infective episodes were documented. Overall the number of infective episodes was significantly associated with the mean white cell and monocyte counts (r = 0.59, 0.02 less than P less than 0.05) but not with length of follow-up, presentation values of white cell, neutrophil and monocyte counts, or mean neutrophil count. The patients appeared to divide clinically into two groups, those at low risk (seven patients) and those high risk (four patients) of infection. Patients in the high risk group had significantly more infections (P = 0.01) and significantly lower monocyte counts (0.02 less than P less than 0.05) than patients in the low risk group. These results are in contrast to similar studies in patients with chemotherapy-induced neutropenia; in our patients the overall rate and severity of infection was low, Gram negative infections were uncommon and monocytopenia appeared to be of greater importance than neutropenia in determining susceptibility to infection.     

Phenotypic and functional analysis of bone marrow progenitor cell compartment in bone marrow failure

SUMMARY. Many laboratory findings have demonstrated that the haemopoietic stem cell compartment is defective in aplastic anaemia (AA). AA bone marrow (BM) and peripheral blood (PB) are profoundly deficient in colonyforming cells, and AA progenitors fail to proliferate in longterm assays even in the presence of an intact stroma. Our study was designed to characterize some quantitative and qualitative aspects of the progenitor cell defect in AA. Using flow cytometric analysis of BM from new AA patients and from those recovering after immunosuppressive therapy, we determined that the numbers of CD34⁺ and CD33⁺ cells were markedly decreased in AA. Although PB neutrophil counts did not correlate with BM CD34⁺ cell numbers in acute disease, there was an association between the overall severity of the disease and the degree of CD34⁺ cell reduction. A decrease in BM CD33⁺ cells was a common finding in MDS patients, but reduction in CD34⁺ cells was found only in some hypoplastic MDS cases. Sorting experiments demonstrated lower plating efficiency for purged CD34⁺ cells from AA BM in comparison to controls. Thus, diminished colony formation of total BM appeared to result from both quantitative and qualitative defects. Based on the association between increased cycling and c-kit receptor expression on CD34⁺ cells, we found that the mitotically active CD34⁺ cells bearing the c-kit antigen were reduced in AA. With clinical improvement, CD34⁺ and CD33⁺ cells increased in correlation with PB parameters, but they did not return to normal values. Sorted CD34⁺ cells from recovered patents showed improved plating efficiency. In patients with aplastic anaemia, use of CD34 antigen as a phenotypic marker of progenitor cells may be helpful for the analysis of the early haemopoietic cell compartment and BM recovery.     

Enrichment of interleukin-2-responsive natural killer progenitors in human bone marrow

Natural killer (NK) cells can be cultured in interleukin-2 (IL-2)- containing medium from selected human bone marrow (BM) cells obtained after the elimination of mature T and NK cells. To isolate and characterize IL-2-responsive NK progenitors in the selected BM cells, we investigated the expression of IL-2 receptors (IL-2R) on these cells. Neither CD25 (IL-2R alpha) nor IL-2R beta antigen was observed on the selected BM cells before culture. However, CD25+ cells without detectable levels of IL-2R beta antigen appeared 24 hours after culture in IL-2-containing medium. Cells were sorted from each fraction of the selected BM cells 24 hours after culture after staining with anti-CD33, anti-CD34, and anti-CD25 monoclonal antibodies. The generation of NK cells (CD3- CD56+ cells) and NK activity were observed only from the CD33-/CD34-/CD25+ cell fraction after culture in IL-2-containing medium. The frequency of IL-2-responsive NK progenitors relative to the fraction was 1/231 (95% confidence range, 1/156 to 1/289), which corresponded to the frequency relative to the total number of selected BM cells when the frequency relative to the CD33-/CD34-/CD25+ cell- fraction was converted according to the percentage of these cells in the total number of selected BM cells. These results indicated that IL- 2-responsive NK progenitors were enriched in the CD33-/CD34-/CD25+ cell fraction.     

Impaired natural killer cell recycling in childhood chronic neutropenia with morphological abnormalities and defective chemotaxis of neutrophils

Natural killer (NK) cell activity was measured by a 51Cr-release assay using K562 target cells in 12 neutropenic children. NK cell activity was depressed in four patients who had childhood chronic neutropenia with abnormal neutrophil morphology and chemotaxis. The percentage of lysis at a 40:1 effector-target ratio was 28.4% to 42.1% (P less than .001) of the normal lymphocyte value during the study period (32 to 40 months). NK cell activity was normal in the other eight children with chronic neutropenia without any of these neutrophil abnormalities: lazy leukocyte syndrome, Shwachman syndrome, or dysgammaglobulinemia type I with neutrophil defects. NK cell activity of the four patients was depressed at 5:1 to 40:1 effector-target ratios. The NK cells responded to in vitro interferon (IFN)-alpha and interleukin 2, as did normal lymphocytes, but the activated levels were still lower than those of normal lymphocytes (P less than .01). Because NK cells kill a target through recognition, binding, killing, and detaching, and they repeat this lytic sequence (ie, recycling), the localization of the NK cell defect was further analyzed in the four patients using both 51Cr- release and single cell-in-agarose assays. The patients' NK cells were normal in recognizing, binding, and killing a target but were defective in recycling; the estimated maximum recycling capacity (MRC) values in a four-hour assay were 1.8 to 2.4 (P less than .01), as compared with the normal lymphocyte value of 5.5 +/- 0.6 (mean +/- SD). The stimulation of the effector cells with 1,000 U/mL IFN-alpha did not significantly increase the estimated MRC. These results demonstrate that NK cells are defective in recycling in some type of childhood chronic neutropenia with abnormal neutrophil morphology and chemotaxis. The NK cell deficiency is of clinical interest in terms of its relationship to the recurrent infections, development of malignancy, and dysgranulopoiesis in the disorder.     

Deficient natural killer (NK) cells in paroxysmal nocturnal haemoglobinuria (PNH): studies of lymphoid cells fractionated by discontinuous density gradient centrifugation

Non-adherent, non-B lymphoid cells from six patients with PNH and six healthy subjects were fractionated by Percoll discontinuous density gradient centrifugation (DDGC). The cell distribution pattern, NK cell activity (NKA), large granular lymphocytes (LGL) count and surface marker phenotypes were studied. The distribution patterns of patients' cells did not significantly differ from the controls. The peak of the NKA was found in low density fractions where the maximum counts of LGL, Leu-7+2- cells and Leu-11+ cells were present. The NKA and the proportion of Leu-7+2- cells and Leu-11+ cells were significantly lower in patients with PNH (P less than 0.001 for NKA and surface phenotypes; P less than 0.02 for LGL counts). NKA in the Percoll fractions was correlated with the counts of LGL (r=0.69, P less than 0.001), Leu-7+2- cells (r = 0.75, P less than 0.001) and Leu-11+ cells (r = 0.89, P less than 0.001). Therefore, we concluded that NKA is deficient in PNH because of decreased NK cell counts.     

Reduced function of the multidrug resistance P-glycoprotein in CD34+ cells of patients with aplastic anaemia

Drug exposure is implicated in the aetiology of some cases of acquired aplastic anaemia (AA), but the reason for this susceptibility remains unclear. We previously demonstrated that P-glycoprotein (P-gp) function, a drug efflux pump, is decreased in AA lymphocytes. To further evaluate whether P-gp activity is also abnormal in AA stem cells, we examined bone marrow (BM) CD34+ cells from newly diagnosed AA patients (AA-d, n = 25), after immunosuppression (AA-IST, n = 13) and after BM transplantation (AA-BMT, n = 8). Of the AA patients with autologous haematopoiesis (AA-d + AA-IST), 15 had drug-induced AA. Thirty-two BM donors were studied as controls. P-glycoprotein function was assessed by the rhodamine 123-efflux assay. P-glycoprotein function in CD34+ cells was reduced in AA-d patients (17.8%, 0-67.7) compared with controls (42.5%, 13.4-57.4; P < 0.001), as well as in AA-IST (20.3%, 1.2-32.0; P < 0.001), but not in AA-BMT (40.9%, 19.0-55.9). P-gp function was reduced more in drug-induced AA (14.5%, 0-27.4) than in the other cases (26.1%, 0-67.7; P = 0.04), but it did not correlate with disease severity. These results indicate that P-gp function is defective in AA CD34+ cells, pointing to a role of P-gp in increased cell susceptibility to xenobiotics in AA.     

Generation of human natural killer cells from immature progenitors does not require marrow stromal cells

Human natural killer (NK) cells comprise 10% to 15% of peripheral blood mononuclear cells and have an important role in immune responses against tumors, viral infections, and graft rejection. NK cells originate in bone marrow (BM), but their progenitors and lineage development have not been completely characterized. We studied the generation of NK cells from purified CD34+HLADR- and CD34+HLADR+ BM progenitors and the influence of various cytokines on their production. We show that CD3-CD56+ cytotoxic NK cells can develop from both progenitors populations when interleukin-2 (IL-2) is present in an in vitro suspension culture system containing IL-1 alpha and stem cell factor. Up to 83.8% and 98.6% CD3-CD56+ cells were detected in CD34+HLADR- and CD34+DR+ cultures, respectively, after 5 weeks of culture; significant numbers of NK cells were first detected after 2 weeks. Cytotoxic activity paralleled NK cell numbers; up to 70% specific lysis at an effector:target ratio of 10:1 was observed at 5 weeks. IL-7 also triggered development of CD3-CD56+ cells from these immature progenitors (up to 24% and 55% appeared in CD34+HLADR- and CD34+HLADR+ cultures, respectively). Our data suggest that BM stromas are not necessary for NK cell development and that IL-2 remains essential for this lineage development and differentiation.     

Prognosis in aplastic anaemia: a review of 138 cases

The outcome of 138 patients with aplastic anaemia was studied to ascertain whether prognostic factors at the onset of the disease are predictive. Applying four classic prognostic criteria, two groups of patients were identified: (1) those with an absolute granulocyte count of less than 0.5 × 10⁹/l, platelet count less than 20 × 10⁹/l, reticulocyte count less than 15 × 10⁹/l, and non-myeloid marrow cellularity greater than 75% who had 100% mortality in the first three months from the onset of the disease, (ii) patients who fulfilled three or fewer of these criteria who had a 20% three month mortality. The survival of these groups differed significantly (P = less than 0.001). The variable that most closely correlated with prognosis was the number of non-haematopoietic cells in the bone marrow aspirate. When Lynch's prognostic formula for patients with aplastic anaemia was applied to the same group of patients, the correlation with prognosis was less close and the calculations were tedious. The identification of a group of patients with severe aplastic anaemia is important because these patients do not benefit from conventional forms of therapy and alternative approaches such as transplantation and immunosuppression are justified.     

Regeneration of peripheral blood cells following ABO incompatible allogeneic bone marrow transplantation for severe aplastic anaemia

The regeneration of peripheral blood cells in six patients receiving ABO incompatible bone marrow transplants for severe aplastic anaemia has been studied. The results are compared with those of 18 similar aplastic patients treated with ABO compatible transplants. Reticulocyte recovery in the ABO incompatible cases was significantly delayed (P less than 0.01) and associated with greatly prolonged red cell transfusion requirements (P less than 0.001). Peripheral blood regeneration of neutrophils, lymphocytes and platelets was also significantly delayed (P less than 0.01). Platelet support was required for longer in the ABO incompatible cases (P less than 0.001) but there was no increase in the incidence or severity of infections in the post transplant period.     

Effect of recombinant human granulocyte-macrophage colony-stimulating factor administration on the lymphocyte subsets of patients with refractory aplastic anemia

Human recombinant granulocyte-macrophage colony-stimulating factor (rhGM-CSF) was administered to 14 patients with refractory aplastic anemia (AA). The effect of rhGM-CSF therapy on the lymphocyte phenotype; on the proliferative responses to the mitogen phytohemagglutinin, Candida albicans, and tetanus toxoid antigens; and on the natural killer (NK) activity of the circulating lymphocytes was studied. Samples were collected before (baseline) and twice during the rhGM-CSF administration. The absolute number of circulating lymphocytes remained relatively constant during the first period, but experienced a significant increase (P less than .001) during the second period. The increase was most prominent in the B cells (P less than .001), but the T cells (P less than .016) also increased. Detailed investigation of lymphocyte subsets showed an increase of the markers CD38 (Leu17), HLA-DR, and the transferrin receptor throughout the treatment course giving evidence of lymphoid cell activation. The NK cell activity was suppressed (P less than .008) throughout the treatment. However, proliferative responses to phytohemagglutinin, Candida antigen, and tetanus toxoid were unaffected. Although the mechanism is not yet defined, GM-CSF does induce activation and increase in absolute lymphoid cell number, especially B cells, together with a decrease in NK cytotoxicity. The implication of these immune cell changes in relation to host resistance to microorganisms remains to be established.     

Stimulatory serum factors in aplastic anaemia

Human serum contains an activity which enhances the release of colony stimulating and burst promoting activity (CSA and BPA). It is low in the majority of normal sera and was found elevated in patients with aplastic anaemia. The patient's 'response' to autologous serum releaser activity was measured by comparing CSA and BPA release by patient cells in percentage of release by normal cells. This 'response' was negatively correlated with serum releaser activity (P = 0.0035 for CSA-release, P less than 0.0001 for BPA-release), i.e. releaser activity was high, when factor release was low. This inverse relationship between releaser activity and the patient's response was also observed in four patients with pancytopenia of causes other than aplastic anaemia. We conclude that elevated serum releaser activity reflects a repair mechanisms which operates when CSA- and BPA-production is inadequate. Thus, releaser activity is likely to be a haemopoietic regulator.     

"Stem cell" origin of the hematopoietic defect in dyskeratosis congenita

We have used the long-term bone marrow culture (LTBMC) system to analyze hematopoiesis in three patients with dyskeratosis congenita (DC), two of whom had aplastic anemia, and the third had a normal blood count (apart from mild macrocytosis) and normal BM cellularity. Hematopoiesis was severely defective in all three patients, as measured by a low incidence of colony-forming cells and a low level of hematopoiesis in LTBMC. The function of the marrow stroma was normal in its ability to support the growth of hematopoietic progenitors from normal marrows seeded onto them in all three cases, but the generation of hematopoietic progenitors from patients marrow cells inoculated onto normal stromas was reduced, thus suggesting the defect to be of stem cell origin. The parents and unaffected brother of one of the families have also been studied in LTBMC and all showed normal hematopoietic and stromal cell function. From this study we speculate that there are some similarities between DC and the defect in the W/Wv mouse.     

Natural and activated cytotoxic lymphocytes reactivity to human hepatocellular carcinoma cell lines in hepatocellular carcinoma patients

The status of cellular cytotoxic activity in Hepatocellular Carcinoma (HCC) patients was compared to that in normal individuals by testing the cytotoxicity against K562 and five established HCC cell line targets. Natural killer (NK) activity of fresh peripheral blood mononuclear (PBM) cells in HCC patients to K562 cell line target was lower than that in normal donors. NK activity of unstimulated PBM cells from either source was minute against all five HCC cell line targets. Three different activation systems were employed to examine the cellular cytotoxicity of activated PBM cells: (1) conventional mixed lymphocyte culture (MLC), (2) allogeneic mixed lymphocyte tumor culture (MLTC), and (3) lymphokine-activated killer (LAK) cell culture. The cytotoxic effects of PBM cells in all three activation conditions were significantly lower in HCC patients than in normal donors (P less than 0.05 to P less than 0.01). These results suggest that, in addition to naturally present NK cells, the degree of in vitro activation of PBM cells may also have decreased in HCC patients.     

Characterization of the stage in natural killer cell development in 14.5-day mouse fetal liver using adult bone marrow stroma.

Nonstimulated fetal liver (FL) from 14.5-day gestation mice had no natural killer (NK) cell activity and <3% expressed NK1.1. Even after short-term (3-4 day) culture of FL with the late-acting cytokines, interleukin (IL)-15 or IL-2, little or no NK activity was detected. However, longer-term (13 day) culture with IL-2 plus stroma derived from bone marrow (BM) of adult mice, resulted in extensive proliferation and differentiation to mature NK cells. Cell numbers began to increase after 4 days, and by day 13, they had increased 40-fold and 69% of the cells were NK1.1+ with high NK activity and 5%-10% were NK1.1- B220+. With stroma, but no IL-2, equivalent proliferation occurred, but differentiated cells were predominantly NK1.1- B220+, not NK cells. Culture for 13 days without stroma, but with either IL-2, IL-15, FLTK3-ligand (L) or stroma-conditioned medium, resulted in less than fivefold expansion, and minimal NK activity. Culture with combinations of FLTK3-L or ckit-L plus IL-15 or IL-2 increased both cell number and NK activity, but the increase in cell number was less than that seen with stroma plus IL-2. By limiting dilution assay on stroma plus IL-2, the precursor frequency was 1/(2660+/-292) whole FL cells and the absolute number, but not the frequency, increased during culture on stroma without IL-2. The NK cell progenitors were found in sorted NK1.1- and Sca-1+ c-kit+ lineage- subpopulations at a frequency of 1/(156+/-52.5). Together, these data suggest that the NK lineage cells in FL are primarily in early stages of development. They are highly proliferative, respond to early acting cytokines and express stem cell markers.     

Cyclosporin A (CyA) does not enhance CFU-c growth in patients with severe aplastic anaemia

8 patients with severe aplastic anaemia (SAA) in remission following immunosuppres-sive therapy were studied for CFU-c growth from unfractionated or E-rosette depleted (E-) bone marrow (BM) cells. Cyclosporin A (CyA) was added to unfractionated BM cells at a concentration of 1000 ng/ml. The mean number of CFU-c/10⁵ BM cells plated was 6 ± 6 from undractionated BM cells, 28 ± 20 from E-BM cells, and 8 ± 7 from unfractionated BM cells supplemented with CyA. All patients had significant increase of CFU-c growth after E rosette depletion (overall P = 0.002). On the contrary, only 1 patient showed an increase of CFU-c growth after addition of CyA, and overall there was no difference between untreated and CyA treated BM cells (P = 0.7). These results suggest that addition of CyA to BM cells in vitro is not an effective means of enhancing CFU-c growth in SAA patients.     

Natural killer, lymphokine-activated killer and interferon-gamma producing activities of peripheral blood- and regional lymph node-mononuclear cells in 23 cases of colorectal cancer

Natural killer (NK) activity, lymphokine-activated killer (LAK) activity and interferon-gamma (IFN-gamma) producing activity of peripheral blood mononuclear cells (PBMC) and regional lymph node mononuclear cells (LNMC) were studied in 23 previously untreated cases of colorectal cancer. NK and LAK activities were significantly lower in LNMC than in PBMC. Patients showed depressed NK and LAK activities in PBMC. In the Dukes C group, especially, both NK activity and LAK activity decreased compared to control patients. NK and LAK activities of PBMC decreased as the grade of invasion to lymphatic channels progressed. LAK activity positively correlated with NK activity in PBMC. Patients with high LAK activity showed high IFN-gamma production in both controls and Dukes A . B patients. However, in Dukes C patients, no relationship between LAK activity and IFN-gamma production was observed. We conclude that the depressed NK and LAK activities of PBMC reflect the local lymphatic invasion and that IFN-gamma involvement in LAK cell generation is impaired in advanced cancer patients. More fundamental studies should be carried out before clinical trials of adoptive immunotherapy using LAK cells, because LAK activity is not induced sufficiently in advanced cancer.