Evidence that hematopoiesis may be a stochastic process in vivo.

To study the behavior of hematopoietic stem cells in vivo, hematopoiesis was simulated by assuming that all stem cell decisions (that is, replication, apoptosis, initiation of a differentiation/maturation program) were determined by chance. Predicted outcomes from simulated experiments were compared with data obtained in autologous marrow transplantation studies of glucose 6-phosphate dehydrogenase (G6PD) heterozygous female Safari cats. With this approach, we prove that stochastic differentiation can result in the wide spectrum of discrete outcomes observed in vivo, and that clonal dominance can occur by chance. As the analyses also suggest that the frequency of feline hematopoietic stem cells is only 6 per 10(7) nucleated marrow cells, and that sem cells do not replicate on average more frequently than once every three weeks, these large-animal data challenge clinical strategies for marrow transplantation and gene therapy.

Feline leukemia virus integrase and capsid packaging functions do not change the insertion profile of standard Moloney retroviral vectors.

Adverse events linked to perturbations of cellular genes by vector insertion reported in gene therapy trials and animal models have prompted attempts to better understand the mechanisms directing viral vector integration. The integration profiles of vectors based on MLV, ASLV, SIV and HIV have all been shown to be non-random, and novel vectors with a safer integration pattern have been sought. Recently, we developed a producer cell line called CatPac that packages standard MoMLV vectors with feline leukemia virus (FeLV) gag, pol and env gene products. We now report the integration profile of this vector, asking if the FeLV integrase and capsid proteins could modify the MoMLV integration profile, potentially resulting in a less genotoxic pattern. We transduced rhesus macaque CD34+ hematopoietic progenitor cells with CatPac or standard MoMLV vectors, and determined their integration profile by LAM-PCR. We obtained 184 and 175 unique integration sites (ISs) respectively for CatPac and standard MoMLV vectors, and these were compared with 10 000 in silico-generated random IS. The integration profile for CatPac vector was similar to MoMLV and equally non-random, with a propensity for integration near transcription start sites and in highly dense gene regions. We found an IS for CatPac vector localized 715 nucleotides upstream of LMO-2, the gene involved in the acute lymphoblastic leukemia developed by X-SCID patients treated by gene therapy using MoMLV vectors. In conclusion, we found that replacement of MoMLV env, gag and pol gene products with FeLV did not alter the basic integration profile. Thus, there appears to be no safety advantage for this packaging system. However, considering the stability and efficacy of CatPac vectors, further development is warranted, using potentially safer vector backbones, for instance those with a SIN configuration.

Retrovirus-induced feline pure red cell aplasia. Hematopoietic progenitors are infected with feline leukemia virus and erythroid burst-forming cells are uniquely sensitive to heterologous complement.

Feline leukemia virus subgroup C/Sarma (FeLV-C) induces pure red cell aplasia (PRCA) in cats. Just before the onset of anemia, erythroid colony-forming cells (CFU-E) become undetectable in marrow culture, yet normal frequencies of erythroid burst-forming cells (BFU-E)- and granulocyte-macrophage colony-forming cells (CFU-GM) persist. To determine if erythroid progenitors were uniquely infected with retrovirus, marrow mononuclear cells from cats viremic with FeLV-C were labeled with monoclonal antibodies to gp70 and then analyzed with a fluorescence-activated cell sorter. Both erythroid and granulocyte-macrophage progenitors were among cells sorting positively, suggesting that infection of BFU-E alone did not result in PRCA. The results were confirmed by complement (C') lysis studies using baby rabbit or guinea pig sera as sources of C'. These studies also suggested that BFU-E from cats with PRCA were unusually sensitive to C' alone, without the addition of antibody. In further studies, we demonstrated that C' activation was via the classical pathway and that C' sensitivity was unique to BFU-E and not a property of CFU-E, CFU-GM, or progenitors that were capable of giving rise to BFU-E in suspension culture. As BFU-E from cats viremic with FeLV-A/Glasgow-1 or the Rickard strain of feline leukemia virus were not sensitive to C', this finding may relate to the pathogenesis of feline PRCA. We hypothesize that, in cats viremic with FeLV-C, the abnormal C' sensitivity of BFU-E leads to the absence of CFU-E and anemia.

Pancytopenia as a clonal disorder of a multipotent hematopoietic stem cell.

Hematopoiesis was investigated in a 14-yr-old girl who had a 2-yr history of stable asymptomatic pancytopenia and who was also heterozygous at the structural locus for glucose-6-phosphate dehydrogenase (G-6-PD). There was no morphologic or cytogenetic evidence for preleukemia and no suggestion of Fanconi anemia. In the skin and sheep erythrocytes-rosetted T lymphocytes, the ratio of G-6-PD A/B activities was 1:1. However, only type B activity was found in peripheral blood erythrocytes, granulocytes, and platelets. Most erythroid bursts and all granulocyte/macrophage colonies formed in methylcellulose culture were derived from the abnormal clone. These findings demonstrate that (a) some cases of pancytopenia are stem cell diseases that apparently develop clonally; (b) circulating differentiated cells originate from this clone; (c) despite a hypoproliferative anemia, the in vivo expression of presumably normal (nonclonal) progenitors is suppressed. In this patient, the relationship between clonal dominance and possible malignancy may be assessed prospectively.

Feline glucose-6-phosphate dehydrogenase cellular mosaicism. Application to the study of retrovirus-induced pure red cell aplasia.

Neoplasms result from the uncontrolled proliferation of abnormal or transformed cells. The early stages of this process are difficult to study because of the lack of sensitive and specific markers of clonal evolution in an experimental system. We have developed a cat model using cellular mosaicism for glucose-6-phosphate dehydrogenase (G-6-PD). Our findings confirm that the structural locus for feline G-6-PD is on the X-chromosome and demonstrate that it is randomly inactivated in somatic cells. Heterozygous cats have balanced ratios of G-6-PD enzyme types in peripheral blood cells and hematopoietic progenitors that remain stable over time. In our initial studies, we used the model to analyze the events surrounding marrow failure experimentally induced by selected strains of feline leukemia virus (FeLV). Two G-6-PD heterozygous cats, one F1 male hybrid and one domestic cat were infected with FeLV (C or KT) and developed pure red cell aplasia (PRCA). Colonies arising from the more mature erythroid colony-forming cell were not detected in marrow culture of anemic animals although erythroid bursts persisted, suggesting that the differentiation of early erythroid progenitors (BFU-E) was inhibited in vivo. The ratio of G-6-PD types in hematopoietic progenitors and peripheral blood cells from the heterozygous cats did not change when the animals developed PRCA. Thus, the anemia did not result from the clonal expansion of a transformed myeloid stem cell. With this experimental approach, one may prospectively assess clonal evolution and cellular interactions in other FeLV-induced diseases.

Arylsulfatase B activities and glycosaminoglycan levels in retrovirally transduced mucopolysaccharidosis type VI cells. Prospects for gene therapy.

Mucopolysacchariodosis type VI (MPS VI) is the lysosomal storage disorder caused by the deficient activity of arylsulfatase B (ASB; N-acetylgalactosamine 4-sulfatase) and the subsequent accumulation of the glycosaminoglycan (GAG), dermatan sulfate. In this study, a retroviral vector containing the full-length human ASB cDNA was constructed and used to transduce skin fibroblasts, chondrocytes, and bone marrow cells from human patients, cats, or rats with MPS VI. The ASB vector expressed high levels of enzymatic activity in each of the cell types tested and, in the case of cat and rat cells, enzymatic expression led to complete normalization of 35SO4 incorporation. In contrast, overexpression of ASB in human MPS VI skin fibroblasts did not lead to metabolic correction. High-level ASB expression was detected for up to eight weeks in transduced MPS VI cat and rat bone marrow cultures, and PCR analysis demonstrated retroviral-mediated gene transfer to approximately 30-50% of the CFU GM-derived colonies. Notably, overexpression of ASB in bone marrow cells led to release of the enzyme into the media and uptake by MPS VI cat and rat skin fibroblasts and/or chondrocytes via the mannose-6-phosphate receptor system, leading to metabolic correction. Thus, these studies provide important rationale for the development of gene therapy for this disorder and lay the frame-work for future in vivo studies in the animal model systems.

Lymphocytes and antibody in retrovirus-induced feline pure red cell aplasia.

The possible role of antibody and T-lymphocytes was investigated in the pure red cell aplasia (PRCA) associated with feline leukemia virus, subgroup C (FeLV-C), infection. In previous studies, erythroid colony-forming cells were undetectable in marrow culture of cats with PRCA. Yet erythroid burst-forming cells (BFU-E) remained, suggesting that BFU-E were able to differentiate in vitro but not in vivo. It was inferred that immunologic suppression may contribute to the pathogenesis of feline PRCA, and the interactions of antibody and T-lymphocytes with erythroid and granulocyte-macrophage progenitors were studied. Incubation of normal or PRCA marrow cells with PRCA serum or IgG concentrated from this serum and then complement (C') failed to decrease hematopoietic colony growth when compared to the results obtained with cultures of marrow cells incubated with C' alone. In crossover coculture studies, T-cells from Safari cats with PRCA had no inhibitory effect on colony growth from normal or autologous PRCA marrow cells. For the determination of whether feline PRCAs were associated with a clonal T-cell process, lymphocytes were obtained periodically from glucose-6-phosphate dehydrogenase (Glc-6-PD) heterozygous cats following FeLV-C infection and were expanded with a crude preparation of interleukin-2. The ratios of Glc-6-PD enzyme types in these samples did not change as cats developed anemia, suggesting that the inhibition of erythropoiesis was not associated with the clonal expansion of T-cells. These studies, therefore, do not support the premise that feline PRCA results from the interaction of antibody or T-cells with erythroid progenitors.

Feline lymphomas: immunological and cytochemical characterization.

The immunological and cytochemical phenotypes of five primary feline lymphomas and six feline lymphoma lines are reported. Thymic lymphomas induced by the Rickard strain of FeLV (FeLV-R) are of prothymocyte or (immature) cortical thymocyte origin, as these express terminal deoxynucleotidyl transferase, the guinea pig erythrocyte rosette receptor, Ia antigens, partial cortisone sensitivity, and nonspecific esterase. Lymphomas associated with other strains of FeLV form rosettes with guinea pig erythrocytes, frequently have Ia antigens and cytoplasmic nonspecific esterase, and probably originate from helper T-cells, monocyte/macrophages, or null cells. These data belie previous conclusions that FeLV leukemogenesis is restricted to mature T-cells; rather, the considerable heterogeneity in the surface and cytochemical phenotype of feline lymphomas probably reflects transformation of multipotent lymphoid or monocytoid precursors in the bone marrow by FeLV.

In vivo infection of marrow stromal fibroblasts by feline leukemia virus.

Marrow stromal fibroblasts (FBs) likely play an important role in the regulation of hematopoiesis within the marrow microenvironment. Infection of these cells by feline leukemia virus (FeLV) might not only contribute to the pathogenesis of FeLV-induced hematologic diseases, but could provide a reservoir for virus in the infected cat. To determine the frequency of FeLV infection among marrow FB precursor cells (fibroblast colony-forming units, CFU-F) of cats viremic with FeLV-C/Sarma and FeLV-A/61E, marrow FBs and FB cell clones were isolated and assayed for expression of FeLV gag protein. From 30% to 86% and 64% to 88% of marrow FB precursors were infected with FeLV-C/Sarma and FeLV-A/61E, respectively. CFU-F from a cat viremic with FeLV-A/61E were not affected by exposure to antibody against FeLV envelope glycoprotein gp70 and heterologous complement, whereas similarly treated hematopoietic progenitors (erythroid colony-forming units, CFU-E; erythroid burst-forming units, BFU-E; and granulocyte-macrophage colony-forming units, CFU-GM) and culture-propagated, FeLV-infected marrow FBs were effectively lysed, suggesting that infected CFU-F within the marrow microenvironment do not express a significant amount of gp70 on their cell membranes. Thus, marrow FB precursor cells appear to be a major target for FeLV in vivo. Furthermore, the low level of gp70 antigen expression on the surface of these cells in vivo may allow them to escape immune surveillance and provide a reservoir of virus during active or latent infection.

Cyclic hematopoiesis in dogs: studies of erythroid burst-forming cells confirm an early stem cell defect.

Studies of early erythropoiesis in dogs have been hampered by the lack of an efficient assay for canine erythroid burst-forming cells (BFU-E). We have developed a methyl-cellulose culture system for hematopoietic progenitors in dog marrow with a plating efficiency for BFU-E of 98 +/- 26 (SD) per 10(5) marrow mononuclear cells. We then applied this assay to the study of cyclic hematopoiesis in grey collie dogs, and regular fluctuations of colonies derived from erythroid colony-forming cells (CFU-E), BFU-E, and granulocyte-macrophage colony-forming cells (CFU-GM) were seen at 12- to 13-day intervals. Defining Cycle Day 1 as the first day that the granulocyte count falls below 1000/mm3, we found that the peak frequency of BFU-E (Cycle Day 10) always preceded the peak frequency of CFU-E (Cycle Day 12), which preceded that of the reticulocyte count (Cycle Day 3). The peak frequency of CFU-GM (Cycle Days 9-2) and neutrophil-containing colonies in agar culture (Cycle Days 1-2) preceded the peak of granulocytes (Cycle Day 6). The percentage of the various progenitors in the DNA synthetic phase of the cell cycle was similar in grey collies and normal dogs and showed no cyclic fluctuations. These data indicate that cyclic hematopoiesis results from a defect in a hematopoietic stem cell more primitive than BFU-E and CFU-GM. Cycling appears to be due to the commitment of this primitive cell to differentiation only at discrete intervals. The cell cycle kinetics and differentiation of subsequent cells appear normal.

Feline leukemia virus infection downmodulates the production of growth-inhibitory activity by marrow stromal cells.

To determine the role of infected marrow accessory cells in the pathogenesis of viral-associated hematologic disorders, we evaluated whether feline leukemia virus (FeLV) infection alters the cytoadhesive properties of long-term marrow culture (LTMC) stromal cells, the support of stromal-associated progenitors in LTMCs, and the production of progenitor growth-promoting and -inhibiting activities by marrow stromal cells. Our previous studies demonstrated that LTMCs containing FeLV-infected stromal cells generated two- to three-fold higher numbers of total nonadherent cells and nonadherent granulocyte-macrophage progenitors (CFU-GM) compared with uninfected LTMCs. In the present studies, CFU-GM and primitive erythroid progenitors (BFU-E) bound equivalently to FeLV-infected or uninfected LTMC stromal cells in a 2-hour adherence assay. In recharge LTMC studies, the numbers of adherent CFU-GM maintained in cultures containing stromal cells infected with FeLV-A/61E were not significantly different from controls (range 84-191% of uninfected control cultures, p > 0.1); however, the percentages of adherent CFU-GM in S phase of the cell cycle were consistently increased (range 42-62% compared with controls, range 5-23%). FeLV infection had no significant effect on the cell-cycle status of the nonadherent CFU-GM in LTMCs. Agar co-culture assays revealed that multilineage colony-stimulating activity was constitutively and equivalently produced by feeder cell layers consisting of either uninfected or FeLV-infected irradiated heterogeneous LTMC stromal cells, homogeneous marrow stromal fibroblasts, or a fibroendothelial marrow stromal cell line. However, FeLV infection significantly attenuated the soluble progenitor growth-inhibitory activity associated with higher densities of these stromal cells. Assays of conditioned medium from cultures of irradiated stromal cells demonstrated that FeLV infection or hydrocortisone exposure decreased the utilization of glucose, the production of acidic metabolic products, and the constitutive production of active and latent transforming growth factor beta (TGF-beta) bioactivity and TGF-beta 2 immunoreactivity. Levels of macrophage inflammatory protein 1 alpha (MIP-1 alpha) and tumor necrosis factor alpha (TNF-alpha) were undetectable and unchanged in CM samples. Together, these observations suggest that downmodulation of TGF-alpha and/or the basal metabolic status of stromal cells may be responsible for the high basal proliferative activity of adherent CFU-GM in FeLV-infected LTMCs, and by extension, that retroviral infection in vivo could alter hematopoiesis by perturbing the progenitor growth-regulatory and -supportive function of marrow stromal cells.

Stem cell factor production by human marrow stromal fibroblasts.

To characterize the production of stem cell factor (SCF, the ligand for the c-kit receptor protein) and its regulation by inflammatory cytokines and glucocorticoids, primary marrow stromal fibroblasts were isolated from normal individuals and two patients with Diamond-Blackfan anemia. Unstimulated normal marrow stromal fibroblasts constitutively expressed a low level of SCF mRNA (9 +/- 2 copies/cell [mean +/- SEM]), continually secreted soluble SCF into the supernatant of 1- to 5-day-old cultures (0.16 +/- 0.02 to 0.73 +/- 0.04 ng/mL per 10(6) cells, respectively), and expressed membrane-bound SCF. Stimulation with interleukin-1 beta (IL-1 beta) only modestly increased SCF mRNA levels, soluble SCF production at 24 hours, and membrane-bound SCF. In comparison, hydrocortisone or tumor necrosis factor alpha (TNF-alpha) exposure increased SCF mRNA levels 3.5- to four-fold above controls, but with different kinetics. The peak TNF-alpha effect was at 6 hours, with return to near control levels at 24 hours, whereas hydrocortisone induced maximal mRNA increases at 12 to 18 hours, and the levels remained high at 24 hours. Similarly, a sustained increase in soluble SCF production was detected during 1 to 5 days of hydrocortisone exposure (0.27 +/- 0.03 to 1.10 +/- 0.08 ng/mL per 10(6) cells), while TNF-alpha stimulation modestly increased the production of soluble SCF in 24-hour cultures only. Unstimulated normal marrow fibroblasts expressed predominantly the long species of alternatively spliced SCF mRNA, and the relative amounts of long and short mRNAs did not change after stimulation with IL-1 beta, hydrocortisone, or TNF-alpha. SCF production by marrow stromal fibroblasts from a symptomatic patient with Diamond-Blackfan anemia was equivalent to simultaneously studied normal marrow fibroblasts. In contrast, marrow fibroblasts from a Diamond-Blackfan anemia patient in untreated hematologic remission constitutively expressed high levels of SCF mRNA (21 +/- 4 copies/cell) and soluble protein (0.40 ng/mL per 10(6) cells at 24 hours). Together, these observations suggest that SCF is constitutively produced by fibroblasts in the human marrow microenvironment and that hydrocortisone induces a modest but sustained increase in SCF gene expression and protein production, compared to only a transient increase induced by TNF-alpha. In addition, these findings support the hypothesis that endogenous or corticosteroid-induced increases in the production of SCF could play a physiologic role in the clinical improvement of congenital anemia.

Agreement between laboratory tests and self-reports of alcohol, tobacco, caffeine, marijuana and other drug use in post-partum women.

The agreement between self-reported consumption of several drugs and laboratory tests used to detect their use is examined. Post-partum women (N = 108) enrolled in a research study participated in a detailed interview covering alcohol and caffeine ingestion, tobacco smoking and use of marijuana and other psychoactive drugs. They also kept a 4-day record of their use of these substances. Blood and urine samples were taken and a physical exam done at the close of the record period. Laboratory tests to detect use of alcohol, tobacco, caffeine, marijuana and other drugs were carried out and the results compared to self-reported drug use in the interview and the record. The degree of agreement depended on the drug taken, the test used and the pattern of drug use in the sample. Sporadic or infrequent consumption related poorly to laboratory tests, especially those that were designed as screening tools. Regular consumption could be identified with greater accuracy. However, the group associations evident between self-reports of drug use and laboratory results were not sufficient to guarantee that subjects were correctly classified. Error in both self-report and the decision made from laboratory values must be taken into account in determining the confidence that should be placed in the data and the conclusions drawn from it.

Severe neutropenia associated with griseofulvin therapy in cats with feline immunodeficiency virus infection.

Griseofulvin administration was associated with the development of absolute neutropenia in six of seven (86%) cats with feline immunodeficiency virus (FIV) infection. The neutropenia was severe (less than 400 neutrophils/microliter) in four of the six affected cats, and one cat died from sepsis. Neutrophil counts returned to baseline values within 15 days after drug withdrawal in all surviving cats. No symptoms or hematologic abnormalities were observed in four normal (FIV-seronegative) cats treated with the same lot of griseofulvin at equivalent doses. Neutropenia recurred in two of two FIV-seropositive cats upon griseofulvin rechallenge. Cats with FIV infections appear to be at increased risk for griseofulvin-associated neutropenia. This phenomenon may be analogous to the increased frequency of antibiotic-induced neutropenias observed in humans infected with the human immunodeficiency virus.

Prospective hematologic and clinicopathologic study of asymptomatic cats with naturally acquired feline immunodeficiency virus infection.

Prospective studies were performed over a 28- to 77-month period (median, 66 months) on 5 cats with naturally acquired feline immunodeficiency virus (FIV) infection in an attempt to correlate hematologic and clinicopathologic changes with the emergence of clinical disease. On presentation, all cats were asymptomatic; free of opportunistic infections; and had normal complete blood counts, bone marrow morphologies, marrow progenitor frequencies, and progenitor in vitro growth characteristics. During study, 2 cats remained healthy, 2 cats showed mild clinical signs, and 1 cat developed a malignant neoplasm (ie, bronchiolar-alveolar adenocarcinoma). Although persistent hematologic abnormalities were not observed, intermittent peripheral leukopenias were common. In 3 of 5 FIV-seropositive cats, lymphopenia (< 1,500 lymphs/microL; normal reference range, 1,500 to 7,000 lymphs/microL) was a frequent finding and the absolute lymphocyte counts had a tendency to progressively decline. One of the other 2 cats had consistently low to low-normal absolute neutrophil counts (1,300 to 4,800 segs/microL; mean, 2,730 segs/microL; normal reference range, 2,500 to 12,500 segs/microL), and the remaining cat had consistently normal leukograms, except for a transient period (ie, 11 months) of benign lymphocytosis (7,200 to 13,430 lymphs/microL) early in the study. Periodic examinations of bone marrow aspirates revealed normal to slightly depressed myeloid-to-erythroid ratios with normal cellular morphology and maturation. Bone marrow abnormalities observed late in the study included mild dysmorphic changes (ie, megaloblastic features) in 2 cats, and a significant decrease (60% of controls, P < .001) in the frequencies of burst-forming units erythroid (BFU-E) in marrow cultures of FIV-seropositive cats compared with uninfected control cats.(ABSTRACT TRUNCATED AT 250 WORDS)

Quantitative studies of erythropoiesis in the clinically normal, phlebotomized, and feline leukemia virus-infected cat.

Erythropoiesis was evaluated in 5 cats at base line with normal PCV and then in the same cats with anemia induced by phlebotomy and in 5 other cats with nonregenerative anemia from community-acquired feline leukemia virus (FeLV) infection. The hematologic evaluation included complete blood cell and reticulocyte counts, marrow morphologic features, determination of serum erythropoietin concentrations by radioimmunoassay, ferrokinetic studies, and in vitro marrow culture of early erythroid progenitors (erythroid burst-forming units; BFU-E) and late erythroid progenitors (erythroid colony-forming units; CFU-E). Phlebotomized cats developed marrow erythroid hyperplasia and an increased reticulocyte count. Ferrokinetic studies revealed an increase in plasma iron turnover from 1.4 to 3.8 mg of Fe/dl of blood/day and RBC use from 50.4% to 78.5%. The mean CFU-E number and CFU-E/BFU-E ratio increased after phlebotomy, but the increase was not significant (P greater than 0.05). Serum erythropoietin values did increase significantly. In FeLV-infected cats, a nonregenerative anemia was demonstrated by marrow erythroid hypoplasia and a low total reticulocyte count. An increased percentage of rubriblasts and prorubricytes was observed in 4 of the 5 cats. Although serum erythropoietin values were high (321 +/- 123 mU/ml vs normal 14 +/- 1 mU/ml), ferrokinetic data revealed decreased erythropoiesis. Marrow culture studies in the FeLV-infected cats also revealed low numbers of BFU-E and CFU-E, but normal numbers of granulocyte-macrophage progenitors remained. Seemingly, the FeLV infection impaired the ability of feline marrow to respond physiologically to anemia.

Chronic leukopenia associated with feline immunodeficiency virus infection in a cat.

Leukopenia attributable to lymphopenia and neutropenia was detected over a 28-week period in a 12-year-old domestic cat infected with feline immunodeficiency virus (FIV). Mild normocytic, normochronic anemia also was evident. Platelet counts were normal, and serum biochemical values were unremarkable. Antibodies to FIV were detected in serum by use of immunofluorescence and immunoblot electrophoresis assays. Cytologic evaluation of bone marrow aspirates revealed normal cellular morphologic features, maturation, and myeloid-to-erythroid ratio. Normal marrow cellularity was determined histologically. There was, however, a significant (P less than 0.01) inhibition of colony-forming unit granulocyte/macrophage-derived progenitors when marrow cells were cultured in the presence of autologous serum, compared with that when marrow cells were cultured in the presence of serum obtained from clinically normal cats, thus suggesting the presence of a humoral inhibitory substance directed specifically at the granulocyte/macrophage lineage. These cell culture results were consistent with those reported for human beings with acquired immunodeficiency syndrome and neutropenia. Thus, FIV infection may be an excellent animal model in which to study human immunodeficiency virus and should be considered in the differential diagnosis of cats with chronic leukopenia.