- Email: [email protected]
Parvovirus B19 capsid protein VP2 inhibits hematopoiesis in vitro and in vivo: implications for therapeutic use
Experimental Hematology 32 (2004) 1082–1087
Parvovirus B19 capsid protein VP2 inhibits hematopoiesis in vitro and in vivo: Implications for therapeutic use Oscar Norbecka, Thomas Tolfvenstama, Laurence E. Shieldsb, Magnus Westgrenc, and Kristina Brolidena a Division of Clinical Virology, Department of Laboratory Medicine and cCenter for Fetal Medicine, Department of Obstetrics and Gynecology, Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden; bDepartment of OB-GYN, Division of Maternal Fetal Medicine, University of Washington, Seattle, Wash., USA
(Received 17 May 2004; revised 21 July 2004; accepted 27 July 2004)
Objective. To evaluate the capacity of parvovirus B19 capsid protein VP2 to inhibit hematopoiesis in vitro and in vivo. If effective, a VP2-derived construct may have therapeutic and prophylactic utility in diseases associated to overproduction of hematopoietic cells. Methods. The effect on hematopoiesis in vitro of recombinant VP2, intact and enzymatically fragmented, was evaluated in a colony formation assay, using cells from fetal liver and macaque bone marrow. VP2 was also administered intravenously in macaques and hematological parameters as well as the ex vivo colony formation were assayed during a follow-up period of 33 days. Results. VP2 inhibited BFU-E colony formation by about 55%. CFU-GM and CFU-GEMM colony formation was also affected. Fragmented VP2 retained the inhibitory effect. The ex vivo colony-forming capacity of macaque bone marrow cells was lower in animals that received VP2 injections, and a drop in hematocrit values was noted in one animal. Conclusion. VP2 has an inhibitory effect on hematopoiesis in vitro and in vivo. An active region within VP2 is implied, which would be a strong candidate for use as a medicament in diseases such as polycytemia vera. 쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.
Introduction Parvovirus B19 (B19) is a human pathogen associated with various clinical conditions, ranging from mild symptoms (e.g., erythema infectiosum) to more serious and even lethal diseases in persons who are immunocompromised or suffer from hemolytic anemia. Fetal hydrops and intrauterine fetal death are well-known complications of B19 infection during pregnancy [1]. The cellular receptor for B19 has been identified as globoside, known as the blood group P-antigen (P-ag) [2]. P-ag is found predominantly on erythroid cells and their progenitors, but also on early myeloid precursors, megakaryoblasts, and megakaryocytes [3]. A co-receptor involved in cell entry, α5β1 integrin, has also recently been described [4]. B19 is extraordinarily tropic for hematopoietic tissue, and reticulocytopenia and anemia, as well as neutropenia and thrombocytopenia, have been shown in experimental human infections [5]. Structurally, B19 is very
Offprint requests to: Oscar Norbeck, M.D., Karolinska Institutet, Division of Clinical Virology, F68, Karolinska University Hospital, SE-141 86 Stockholm, Sweden; E-mail: [email protected]
0301-472X/04 $–see front matter. Copyright doi: 1 0 .1 01 6 /j.ex p h e m.2 0 0 4. 07 .0 1 6
simple. The icosahedral capsid has a diameter of 18 to 26 nm, and is composed of 60 capsid proteins, of which approximately 95% are the 58-kDa major capsid proteins (VP2). The remainder are the 83-kDa minor capsid proteins (VP1), which differ from VP2 by an additional 227 amino acids at the amino terminus, referred to as the VP1 unique region. Recombinant B19 capsid proteins, produced in a baculovirus system, spontaneously form virus-like particles and electron microscopic analyses have revealed that they are structurally similar to plasma-derived virions [6]. The atomic structure of VP2 empty capsids has been resolved to 0.8 nm and the capsids have been visualized by cryo-electron microscopy [7,8]. We have investigated the possibility to exploit the properties of the B19 capsid proteins to develop novel medicaments that interfere with the hematopoiesis. In an earlier report we showed that recombinant B19 capsids composed of both VP1 and VP2 proteins (VP1/2) inhibited colony formation of BFU-E cells in vitro [9]. However, in the context of a medicament to be administered to patients, it is desirable to define a smaller product with preserved activity but with
쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
reduced immunogenicity and with practical advantages regarding production. Therefore, we now investigated the effect on hematopoiesis of B19 capsids composed of only VP2, and fragments of such capsids. We also evaluated the effect of VP2 in vivo, in a pilot animal study.
Methods Origin and collection of cells Human fetal liver tissue was obtained, with informed consent, from legal abortions in gestational ages ranging from 6 to 12 weeks. Fetal liver was dissected under sterile conditions and passed through a vinyl mesh to form a single-cell suspension. Bone marrow cells were obtained from a patient with polycytemia vera (PV). Nonhuman primate bone marrow for in vitro experiments was obtained from adult macaques (M. nemestrina). Suspensions of fresh cells were separated on Lymphoprep (Nycomed Pharma, Asker, Norway), counted and frozen according to standard procedures, and stored until use. Recombinant viral proteins and B19 isolate Recombinant VP2 capsids used in experiments on human cells and in toxicity studies were generously provided by Dr. K. Hedman (University of Helsinki, Finland). A second batch of comparable VP2 capsids, used for intravenous injection in macaques, was provided by Biotrin (Biotrin Inc., Dublin, Ireland). Canine parvovirus (CPV) capsids were obtained from Dr. C. Oker-Blom (University of Jyva¨skyla¨, Finland). Recombinant proteins were expressed and purified by using baculovirus-infected Spodoptera frugiperda (Sf-9) cells, as described elsewhere [10]. Total protein concentration was determined by bicinchoninic acid protein assay reagents (Pierce, Rockford, IL, USA). Purity for VP2 was more than 90%, as determined by SDS-PAGE with subsequent silver staining (Novex SilverXpress silver staining kit, Helixx Technologies Inc., Toronto, ON, Canada) and by densitometry (Gel Doc 2000 gel documentation systems with Quantity One quantitation software, Bio-Rad, Hercules, CA, USA). A B19 isolate was derived from a patient with acute infection and B19 DNA content in plasma was determined by quantitative PCR (TaqMan). Plasma was diluted 10,000-fold to 106 genome equivalents per mL (geq/mL), and stored at ⫺70⬚C until use. Colony formation assay Two comparable commercial culture media, “Stem Cell CFU kit” (Invitrogen, Carlsbad, CA, USA) and “MethoCult” (StemCell Technologies, Vancouver, BC, Canada), were used for the human colony formation assays. The latter was used when the manufacturer no longer provided the first product. Iscove’s modified Dulbecco’s medium (Invitrogen, Garlsbad, CA, USA) was used as dilution medium for cell suspensions. The recombinant proteins or B19 isolate were diluted in a buffer (20 mM Tris, 0.5 M NaCl, pH 8.5) and 25 µL of each dilution were added to 25,000 cells in 100 µL of dilution medium and incubated for one hour at 4⬚C. The mixtures were then plated in triplicate in four-well Nunclon Delta Surface incubation dishes (Nalge Nunc Intl., Rochester, NY, USA), and culture medium was added to a final volume of 0.5 mL per well. After 11 days of incubation in a humidified atmosphere at 37⬚C and 5% CO2, cultures were scored for BFUE and in some experiments also for CFU-GM and CFU-GEMM.
1083
As controls, triplicates of cells and dilution medium only (medium control) were included in each experiment. Mean values of colony counts based on triplicates were expressed as percentages of the colony count in the medium control of the respective experiments. For animal cells, a comparable colony formation assay was used, as previously described [11,12]. Toxicity studies VP2 was added in serial dilutions to cultures of undifferentiated MB02 or K562 cells to a final concentration of 1.7 µg/mL. Parallel cultures with bovine serum albumin (BSA) added instead of VP2 were used for comparison with untreated cells. The cells were evaluated after 1, 4, and 8 days, for cell growth and viability through cell counting and trypan blue staining. All samples were processed in duplicate. Cleavage of VP2 VP2 was digested by three endoproteinases, and the resulting mixtures of VP2 protein fragments were used in the colony formation assay. Enzymes used were endoproteinases Lys-C, Arg-C, and Glu-C, which specifically cleave peptide bonds C-terminally at lysine, arginine, and glutamic acid residues respectively (all from Roche, Basel, Switzerland). Lyophilized enzymes were reconstituted and used for protein cleavage according to the manufacturer’s instructions. Total fragmentation was confirmed by showing shorter fragments, but no intact VP2, on a SDS-PAGE and subsequent silver staining. Solutions with only enzyme were included in the assays as controls. In vivo experiments The ability of VP2 to inhibit hematopoiesis in vivo was tested using neonatal M. nemestrina. Two experimental animals received 1.72 mg/kg of VP2 given intravenously, divided into five equal dosages on five consecutive days (days 0–4). A third animal received no injections and was used as a control. Peripheral blood and bone marrow were collected on days 5, 9, 12, 15, 19, and 33. Peripheral blood was evaluated for hemoglobin level, hematocrit value, reticulocyte count, and white blood cell count. Serology for B19 anti-VP1 IgG and anti-VP2 IgM antibodies was tested using a commercially available enzyme-linked immunosorbent assay (IBLHamburg, Germany). Care was taken to limit the total blood and bone marrow collections to less than 2 mL/sample to minimize the effect from repeated collections, and similar total amounts relative to body weight were sampled from all three animals. Statistical analysis For calculation of p-values, unpaired t-tests were performed on colony counts in individual test vs control cultures. Paired t-tests were employed when means from multiple experiments were analyzed. InStat software from GraphPad was used for statistical calculations. Ethical permissions The Ethical Committee at the Karolinska Institutet gave ethical permissions for the human sample collections. The collection of nonhuman primate marrow and blood and the administration of VP2 to nonhuman primates was an approved protocol through the University of Washington Animal Care Committee.
Results VP2 was shown to inhibit erythroid colony formation by approximately 55% at a concentration of 1.9 µg/mL, and
1084
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
the effect was dose-dependent (Fig. 1). Inhibition of the myeloid lineages CFU-GM and CFU-GEMM was also demonstrated by VP2 (Fig. 2). BFU-E was also inhibited by the B19 isolate cultured at a concentration of 1.3 x 104 geq/mL, but CPV capsids showed no inhibitory effect in cultures with human fetal liver cells at a concentration that was inhibitory for VP2. In an experiment with bone marrow cells from a PV patient, VP2 inhibited BFU-E colony formation by about 30% at a concentration of 1.9 µg/mL. Formation of BFU-E and CFU-GM in cultures of macaque bone marrow was also inhibited by VP2 by about 55% and 40%, respectively, at a concentration of 5 µg/mL. The cleavage products from VP2 digested by endoproteinases were also capable to inhibit BFU-E. The cleavage products resulting from digestion by an endoproteinase that cleaves at arginine showed inhibition similar to intact VP2, and colony counts were significantly lower than in a control with only the enzyme, whereas cleavage at lysine and glutamic acid residues resulted in less active fragments (Fig. 3). By silver gel staining, the enzymatically treated preparations were shown not to contain any remaining intact VP2. VP2 was tested for general toxicity against cell lines permissive for B19. On day one following addition of the capsids, the cultures appeared under the microscope to have a number of “clusters” of cells, possibly due to agglutination. Otherwise, no changes in cell growth, morphology, or frequency of cells staining with trypan blue
Figure 1. Inhibitory effect of VP2 on BFU-E. Human fetal liver cells incubated with VP2 at different concentrations resulted in a dose/response effect. The mean values of BFU-E counts in 6 separate experiments are shown and expressed as percent of medium control. Colony counts for the two highest VP2 concentrations were significantly lower than the medium control, p ⫽ 0.018 and 0.023, respectively. Error bars ⫽ SEM.
were observed at any time point, as compared to untreated cells (data not shown). Because of the demonstrated effect of VP2 on colony formation in cultures of macaque bone marrow in vitro, we next evaluated the effect on hematopoiesis in vivo using three macaques. The two experimental animals did not have preexisting B19 antibodies, whereas the control animal was B19 IgG⫹. However, the serology should be taken with precaution since the kit has not been evaluated for testing of animal antibodies. VP2 administered intravenously was well tolerated and no side effects were noted during followup. Cells from consecutive bone marrow specimens showed decreased colony-forming capacity ex vivo, relative to the first sampling on day five (Fig. 4). The drop did not recover during the follow-up period and was markedly greater in the two experimental animals than in the control animal,
Figure 2. Inhibitory effect of different preparations. Tested preparation, cell origin, and evaluated colony type is given in the x-axis labels. All mean colony counts are presented as percentages of the medium control in the respective experiments. From left to right the bars represent: CPV at 2.5 µg/mL; B19 isolate at 1.3 geq/mL; mean based on 6 experiments of VP2 at 1.9 µg/mL; mean based on 3 experiments of VP2 at 1.0 µg/mL, evaluated for CFU-GM and CFU-GEMM; VP2 at 1.9 µg/mL tested on BM from a PV patient (PV-BM); and the last two bars are macaque bone marrow (mBM) cultured with VP2 at 5 µg/mL. p-values for each comparison to respective medium control are given. FL ⫽ fetal liver. Error bars ⫽ SEM.
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
1085
during the follow-up in any of the three animals (data not shown).
Figure 3. Effect on BFU-E of intact and fragmented VP2. Human fetal liver cells were incubated with intact VP2 at 0.9 µg/mL (Intact), and endoproteinase-digested VP2 at a concentration corresponding to 0.9 µg/mL intact VP2. Enzymes used were endoproteinases that cleave at lysine (LYS), glutamic acid (GLU), or arginine (ARG) residues, respectively. Mean colony counts for intact VP2 are presented relative to the medium control, while mean colony counts for fragmented VP2 are presented relative to a control containing the respective endoproteinase. p-values for each comparison to respective control are given. Error bars ⫽ SEM of triplicate wells.
even though a decrease was also noted in the control animal. In one of the two experimental animals, a drop in hematocrit value by 26% at day 12 was noted that recovered within the follow-up period (Fig. 5). Hemoglobin level, reticulocyte count, and white blood cell count did not change significantly
Figure 4. Ex vivo colony formation of macaque bone marrow after VP2 injections. Experimental animals 1 and 2 received intravenous VP2 (1.72 mg/kg divided into 5 daily doses), whereas the control animal received no injections. Samples were collected during follow-up and the capacity of CFC (BFU-E ⫹ CFU-GM) formation is plotted as means of 5 wells relative to the capacity of colony formation at day 5. Error bars ⫽ SEM.
Discussion We have previously reported that VP1/2 specifically inhibits BFU-E formation of cells from bone marrow, cord blood, and fetal liver [9]. We suggest that this effect may be exploited into a medicament for bone marrow suppression; a smaller capsid-derived agent would, however, be more advantageous. In the present report, we show that less complex capsids of only VP2 possess similar inhibitory capacity as VP1/2. Besides the effect on BFU-E, we showed that CFU-GM and CFU-GEMM are also subject to this inhibition. Although myeloid cells were initially reported to be resistant to B19, infection of myeloid precursors is now demonstrated, with inhibition of colony formation as a result [13–17]. Recent data suggest that early myeloid precursors also express P-ag [3,18]; however, the presence of the newly discovered co-receptor remains to be investigated. It thus appears reasonable that myeloid, as well as erythroid, colony formation is affected by VP2. The possibility that the observed inhibition was due to a toxic effect was excluded by toxicity studies of VP2, and by the fact that CPV empty capsids produced in the same system as VP2 lacked inhibitory capacity. When VP2 was enzymatically cleaved, the resulting mix of protein fragments retained the inhibitory capacity, which suggests that the inhibition of hematopoiesis is achieved by a region within VP2, which should be defined further. The inhibitory capacity depended on the amino acid specificity of the cleaving enzyme in these experiments, indicating that the active region was differently well preserved by the enzymes. Another biologically active region of the B19 capsid is described in the unique region of VP1, which possesses an intrinsic phospholipase A2 activity believed to be important for viral entry in the cell nucleus [19].
Figure 5. Hematocrit values after VP2 injections. Hematocrit levels were evaluated at intervals after intravenous administration of VP2 to the experimental animals.
1086
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
The major advantage of defining the smallest active compound, besides a more convenient production, is that it would probably be less immunogenic and be less likely to be neutralized by a preexisting immunity to B19. Neutralizing antibodies are mainly directed to the VP1 unique region, and the few neutralizing epitopes localized in VP2 seem conformationally dependent on sites in the VP1 unique region [1]. The immune response to VP1 appears to be critical for protective immunity and VP1 must be present in recombinant capsids to elicit a neutralizing antibody response in animals [20,21]. A modified B19 capsid optimized for generating an efficient neutralizing response, for use as a vaccine candidate, has been defined and consists of 25% VP1 and 75% VP2 [22]. In a human vaccine trial, the immunogenicity was highly dependent on the adjuvant used, indicating that the capsid formation on its own elicits a poor neutralizing antibody response. Thus, at least theoretically, a VP2-derived construct should not elicit neutralizing antibodies, and an optimal drug would have the parts of the protein deleted that do not contribute to the biological activity. The active part could then ideally be produced as a much smaller protein, linear fragment, or maybe even a peptide. In the above-mentioned vaccine trial, a maximal dose of 25 µg capsid was given intramuscularly, which could theoretically result in a maximal systemic capsid concentration of 5 x 10-3 µg/mL (even distribution in 5000 mL of blood volume is assumed). This concentration was well tolerated and no severe side effects were noted, nor were any alterations in hematological parameters reported [22]. This is consistent with our results since we detected only a negligible effect on BFU-E at comparable VP2 concentrations. B19 infectious particles are potent inhibitors of erythropoiesis [2,23,24]. The B19 isolate we used resulted in a 40% reduction in BFU-E formation at 1.3 x 104 geq/mL, which represents the same number of capsids. Similar inhibition was seen at a VP2 concentration of about 1 µg/mL, which can be calculated to represent about 1.7 x 1011 particles. This 7-log difference in concentration may appear large, but other mechanisms such as viral replication and cell-to-cell infection come into play when infectious virus is used. A drug that specifically inhibits hematopoiesis may be beneficial for a broad spectrum of hematological proliferative disorders and possibly also useful in prenatal stem cell transplantations [25]. However, PV is perhaps the most obvious disorder matching the target cell spectrum of VP2. PV arises as a monoclonal and apparently neoplastic disorder, and the increase in red cell mass is often accompanied by an elevation of granulocyte and platelet counts as well [26]. PV appears insidiously, usually in late middle age, with an incidence of 1 to 2 per 100,000 and is a lethal disease within 1 to 3 years if untreated. No curative treatment is available, and current intervention to suppress the overproduction of bone marrow cells is not satisfactory. Aiming at keeping the hematocrit level below about 45% in PV patients, venesection is often used that typically reduces the
hematocrit by 20 to 25%. Although speculative, assuming that the demonstrated effect of VP2 on BFU-E in vitro is translatable into hematocrit reduction in vivo, our data suggests that a therapeutic effect would be achieved by a VP2 concentration in the bone marrow of 0.2 µg/mL. However, optimization of a candidate drug as discussed earlier would ideally also improve the effect/concentration ratio. Interestingly, a recent report described anemia following B19 infection in a patient suffering from PV [27]. This not only shows that bone marrow cells in PV patients are susceptible to B19, but also that the concept of utilizing a VP2 derivate as a medicament for this disorder is theoretically sound. Indeed, we noted a 30% inhibition of bone marrow from a PV patient by VP2 in vitro, although the level of inhibition on these cells needs to be established in repeated experiments. Outbreaks of anemia in cynomologus monkeys as well as in rhesus monkeys and pig-tailed macaques have led to the identification of novel simian parvoviruses [28,29]. Infection of immunosuppressed macaques with simian parvovirus leads to persistent anemia, whereas in immunocompetent animals there is a temporal drop in reticulocytes [30]. These viruses have been proposed to be classified to the erythrovirus genus, together with B19, since they are similar at the genome level and in their predilection for bone marrow cells in vitro [31]. Encouraged by these similarities between simian and human parvoviruses, we tested VP2 on nonhuman primate hematopoietic cells. It was thus demonstrated in vitro that VP2 was capable to suppress BFU-E and CFUGM formation of cells derived from macaque bone marrow. The results from these experiments implicated that macaques were suitable for in vivo studies, and three animals were used for this purpose. Newborn animals were preferred due to a higher erythropoietic turnover than later in life, and a relatively low distribution volume. The dose given at each injection was based on the in vitro results and was estimated to result in comparable VP2 concentration in blood (and in the bone marrow). In the two experimental animals, there was a marked reduction over time of the total ex vivo hematopoietic colony formation relative to the control time point. A less pronounced reduction was also detected in the control animal, to which no explanation was found. The drop in hematocrit level around day 12 after the first injection, seen in one experimental animal, was reminiscent of the kinetics of reticulocyte numbers reported in experimentally infected human subjects [5]. Reticulocyte numbers fell to undetectable levels around day 10 postinfection, and started to recover about 6 days later. However, as compared to the intranasally administered infectious virus, our intravenously administered VP2 probably reached the bone marrow more rapidly. Clinically nonsignificant lymphopenia, neutropenia, and thrombocytopenia, with a brief overshoot prior to stabilizing at preincubation levels, occurred in the experimentally infected human subjects, whereas such effects on blood cell counts were not seen in our experimental animals. Even though an effect of VP2 in vivo was suggested, very few
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
animals were used in this pilot study. The in vivo results were therefore not conclusive and need to be repeated in a larger number of animals. Further characterization of the region within VP2 responsible for the observed inhibitory effect will provide new VP2-derived constructs that will be interesting to evaluate in future studies.
Acknowledgments This work was supported by the NIH (HL62422-02), European Comission Biomed 2 Programme (Eurofetus), the Swedish Medical Research Council (k2001-06x-11231-07B), the Swedish Research Council (K2002-16X-14189-01A), the Swedish Cancer Foundation, the Swedish Children’s Cancer Foundation, and the Tobias Foundation. We thank Drs. L. Goobar-Larsson, A. Vahlne, and E. Hellstro¨m-Lindberg, and Ms. L. Markling, C. Go¨therstro¨m, and L. Radler for expert technical assistance and discussion.
References 1. Heegaard ED, Brown KE. Human parvovirus B19. Clin Microbiol Rev. 2002;15:485–505. 2. Brown KE, Anderson SM, Young NS. Erythrocyte P antigen: cellular receptor for B19 parvovirus. Science. 1993;262:114–117. 3. Cooling LL, Zhang de S, Naides SJ, Koerner TA. Glycosphingolipid expression in acute nonlymphocytic leukemia: common expression of shiga toxin and parvovirus B19 receptors on early myeloblasts. Blood. 2003;101:711–721. 4. Weigel-Kelley KA, Yoder MC, Srivastava A. α5β1 integrin as a cellular coreceptor for human parvovirus B19: requirement of functional activation of β1 integrin for viral entry. Blood. 2003;102:3927–3933. 5. Anderson MJ, Higgins PG, Davis LR, et al. Experimental parvoviral infection in humans. J Infect Dis. 1985;152:257–265. 6. Kajigaya S, Fujii H, Field A, et al. Self-assembled B19 parvovirus capsids, produced in a baculovirus system, are antigenically and immunogenically similar to native virions. Proc Natl Acad Sci U S A. 1991; 88:4646–4650. 7. Agbandje M, Kajigaya S, McKenna R, Young NS, Rossmann MG. The structure of human parvovirus B19 at 8 A resolution. Virology. 1994;203:106–115. 8. Chipman PR, Agbandje-McKenna M, Kajigaya S, et al. Cryo-electron microscopy studies of empty capsids of human parvovirus B19 complexed with its cellular receptor. Proc Natl Acad Sci U S A. 1996; 93:7502–7506. 9. Lindton B, Tolfvenstam T, Norbeck O, et al. Recombinant parvovirus B19 empty capsids inhibit fetal hematopoietic colony formation in vitro. Fetal Diagn Ther. 2001;16:26–31. 10. Kaikkonen L, Lankinen H, Harjunpaa I, et al. Acute-phase-specific heptapeptide epitope for diagnosis of parvovirus B19 infection. J Clin Microbiol. 1999;37:3952–3956. 11. Shields LE, Andrews RG. Gestational age changes in circulating CD34⫹ hematopoietic stem/progenitor cells in fetal cord blood. Am J Obstet Gynecol. 1998;178:931–937.
1087
12. Shields LE, Gaur LK, Gough M, Potter J, Sieverkropp A, Andrews RG. In utero hematopoietic stem cell transplantation in nonhuman primates: the role of T cells. Stem Cells. 2003;21:304–314. 13. Kurtzman GJ, Ozawa K, Cohen B, Hanson G, Oseas R, Young NS. Chronic bone marrow failure due to persistent B19 parvovirus infection. N Engl J Med. 1987;317:287–294. 14. McClain K, Estrov Z, Chen H, Mahoney DH Jr. Chronic neutropenia of childhood: frequent association with parvovirus infection and correlations with bone marrow culture studies. Br J Haematol. 1993;85:57–62. 15. Ozawa K, Kurtzman G, Young N. Productive infection by B19 parvovirus of human erythroid bone marrow cells in vitro. Blood. 1987;70: 384–391. 16. Potter CG, Potter AC, Hatton CS, et al. Variation of erythroid and myeloid precursors in the marrow and peripheral blood of volunteer subjects infected with human parvovirus (B19). J Clin Invest. 1987;79: 1486–1492. 17. Srivastava A, Bruno E, Briddell R, et al. Parvovirus B19-induced perturbation of human megakaryocytopoiesis in vitro. Blood. 1990; 76:1997–2004. 18. Weigel-Kelley KA, Yoder MC, Srivastava A. Recombinant human parvovirus B19 vectors: erythrocyte P antigen is necessary but not sufficient for successful transduction of human hematopoietic cells. J Virol. 2001;75:4110–4116. 19. Zadori Z, Szelei J, Lacoste MC, et al. A viral phospholipase A2 is required for parvovirus infectivity. Dev Cell. 2001;1:291–302. 20. Saikawa T, Anderson S, Momoeda M, Kajigaya S, Young NS. Neutralizing linear epitopes of B19 parvovirus cluster in the VP1 unique and VP1-VP2 junction regions. J Virol. 1993;67:3004–3009. 21. Bansal GP, Hatfield JA, Dunn FE, et al. Candidate recombinant vaccine for human B19 parvovirus. J Infect Dis. 1993;167:1034–1044. 22. Ballou WR, Reed JL, Noble W, Young NS, Koenig S. Safety and immunogenicity of a recombinant parvovirus B19 vaccine formulated with MF59C.1. J Infect Dis. 2003;187:675–678. 23. Takahashi T, Ozawa K, Takahashi K, Asano S, Takaku F. Susceptibility of human erythropoietic cells to B19 parvovirus in vitro increases with differentiation. Blood. 1990;75:603–610. 24. Mortimer PP, Humphries RK, Moore JG, Purcell RH, Young NS. A human parvovirus-like virus inhibits haematopoietic colony formation in vitro. Nature. 1983;302:426–429. 25. Shields LE, Lindton B, Andrews RG, Westgren M. Fetal hematopoietic stem cell transplantation: a challenge for the twenty-first century. J Hematother Stem Cell Res. 2002;11:617–631. 26. Spivak JL. Polycythemia vera: myths, mechanisms, and management. Blood. 2002;100:4272–4290. 27. Kaptan K, Beyan C, Cetin T, et al. Anemia following human parvovirus B19 infection in a patient with polycythemia vera. Am J Hematol. 2002;69:296–297. 28. O’Sullivan MG, Anderson DC, Fikes JD, et al. Identification of a novel simian parvovirus in cynomolgus monkeys with severe anemia. A paradigm of human B19 parvovirus infection. J Clin Invest. 1994;93: 1571–1576. 29. Green SW, Malkovska I, O’Sullivan MG, Brown KE. Rhesus and pigtailed macaque parvoviruses: identification of two new members of the erythrovirus genus in monkeys. Virology. 2000;269:105–112. 30. O’Sullivan MG, Anderson DK, Goodrich JA, et al. Experimental infection of cynomolgus monkeys with simian parvovirus. J Virol. 1997; 71:4517–4521. 31. Brown KE, Young NS. The simian parvoviruses. Rev Med Virol. 1997;7: 211–218.
Parvovirus B19 capsid protein VP2 inhibits hematopoiesis in vitro and in vivo: Implications for therapeutic use Oscar Norbecka, Thomas Tolfvenstama, Laurence E. Shieldsb, Magnus Westgrenc, and Kristina Brolidena a Division of Clinical Virology, Department of Laboratory Medicine and cCenter for Fetal Medicine, Department of Obstetrics and Gynecology, Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden; bDepartment of OB-GYN, Division of Maternal Fetal Medicine, University of Washington, Seattle, Wash., USA
(Received 17 May 2004; revised 21 July 2004; accepted 27 July 2004)
Objective. To evaluate the capacity of parvovirus B19 capsid protein VP2 to inhibit hematopoiesis in vitro and in vivo. If effective, a VP2-derived construct may have therapeutic and prophylactic utility in diseases associated to overproduction of hematopoietic cells. Methods. The effect on hematopoiesis in vitro of recombinant VP2, intact and enzymatically fragmented, was evaluated in a colony formation assay, using cells from fetal liver and macaque bone marrow. VP2 was also administered intravenously in macaques and hematological parameters as well as the ex vivo colony formation were assayed during a follow-up period of 33 days. Results. VP2 inhibited BFU-E colony formation by about 55%. CFU-GM and CFU-GEMM colony formation was also affected. Fragmented VP2 retained the inhibitory effect. The ex vivo colony-forming capacity of macaque bone marrow cells was lower in animals that received VP2 injections, and a drop in hematocrit values was noted in one animal. Conclusion. VP2 has an inhibitory effect on hematopoiesis in vitro and in vivo. An active region within VP2 is implied, which would be a strong candidate for use as a medicament in diseases such as polycytemia vera. 쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.
Introduction Parvovirus B19 (B19) is a human pathogen associated with various clinical conditions, ranging from mild symptoms (e.g., erythema infectiosum) to more serious and even lethal diseases in persons who are immunocompromised or suffer from hemolytic anemia. Fetal hydrops and intrauterine fetal death are well-known complications of B19 infection during pregnancy [1]. The cellular receptor for B19 has been identified as globoside, known as the blood group P-antigen (P-ag) [2]. P-ag is found predominantly on erythroid cells and their progenitors, but also on early myeloid precursors, megakaryoblasts, and megakaryocytes [3]. A co-receptor involved in cell entry, α5β1 integrin, has also recently been described [4]. B19 is extraordinarily tropic for hematopoietic tissue, and reticulocytopenia and anemia, as well as neutropenia and thrombocytopenia, have been shown in experimental human infections [5]. Structurally, B19 is very
Offprint requests to: Oscar Norbeck, M.D., Karolinska Institutet, Division of Clinical Virology, F68, Karolinska University Hospital, SE-141 86 Stockholm, Sweden; E-mail: [email protected]
0301-472X/04 $–see front matter. Copyright doi: 1 0 .1 01 6 /j.ex p h e m.2 0 0 4. 07 .0 1 6
simple. The icosahedral capsid has a diameter of 18 to 26 nm, and is composed of 60 capsid proteins, of which approximately 95% are the 58-kDa major capsid proteins (VP2). The remainder are the 83-kDa minor capsid proteins (VP1), which differ from VP2 by an additional 227 amino acids at the amino terminus, referred to as the VP1 unique region. Recombinant B19 capsid proteins, produced in a baculovirus system, spontaneously form virus-like particles and electron microscopic analyses have revealed that they are structurally similar to plasma-derived virions [6]. The atomic structure of VP2 empty capsids has been resolved to 0.8 nm and the capsids have been visualized by cryo-electron microscopy [7,8]. We have investigated the possibility to exploit the properties of the B19 capsid proteins to develop novel medicaments that interfere with the hematopoiesis. In an earlier report we showed that recombinant B19 capsids composed of both VP1 and VP2 proteins (VP1/2) inhibited colony formation of BFU-E cells in vitro [9]. However, in the context of a medicament to be administered to patients, it is desirable to define a smaller product with preserved activity but with
쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
reduced immunogenicity and with practical advantages regarding production. Therefore, we now investigated the effect on hematopoiesis of B19 capsids composed of only VP2, and fragments of such capsids. We also evaluated the effect of VP2 in vivo, in a pilot animal study.
Methods Origin and collection of cells Human fetal liver tissue was obtained, with informed consent, from legal abortions in gestational ages ranging from 6 to 12 weeks. Fetal liver was dissected under sterile conditions and passed through a vinyl mesh to form a single-cell suspension. Bone marrow cells were obtained from a patient with polycytemia vera (PV). Nonhuman primate bone marrow for in vitro experiments was obtained from adult macaques (M. nemestrina). Suspensions of fresh cells were separated on Lymphoprep (Nycomed Pharma, Asker, Norway), counted and frozen according to standard procedures, and stored until use. Recombinant viral proteins and B19 isolate Recombinant VP2 capsids used in experiments on human cells and in toxicity studies were generously provided by Dr. K. Hedman (University of Helsinki, Finland). A second batch of comparable VP2 capsids, used for intravenous injection in macaques, was provided by Biotrin (Biotrin Inc., Dublin, Ireland). Canine parvovirus (CPV) capsids were obtained from Dr. C. Oker-Blom (University of Jyva¨skyla¨, Finland). Recombinant proteins were expressed and purified by using baculovirus-infected Spodoptera frugiperda (Sf-9) cells, as described elsewhere [10]. Total protein concentration was determined by bicinchoninic acid protein assay reagents (Pierce, Rockford, IL, USA). Purity for VP2 was more than 90%, as determined by SDS-PAGE with subsequent silver staining (Novex SilverXpress silver staining kit, Helixx Technologies Inc., Toronto, ON, Canada) and by densitometry (Gel Doc 2000 gel documentation systems with Quantity One quantitation software, Bio-Rad, Hercules, CA, USA). A B19 isolate was derived from a patient with acute infection and B19 DNA content in plasma was determined by quantitative PCR (TaqMan). Plasma was diluted 10,000-fold to 106 genome equivalents per mL (geq/mL), and stored at ⫺70⬚C until use. Colony formation assay Two comparable commercial culture media, “Stem Cell CFU kit” (Invitrogen, Carlsbad, CA, USA) and “MethoCult” (StemCell Technologies, Vancouver, BC, Canada), were used for the human colony formation assays. The latter was used when the manufacturer no longer provided the first product. Iscove’s modified Dulbecco’s medium (Invitrogen, Garlsbad, CA, USA) was used as dilution medium for cell suspensions. The recombinant proteins or B19 isolate were diluted in a buffer (20 mM Tris, 0.5 M NaCl, pH 8.5) and 25 µL of each dilution were added to 25,000 cells in 100 µL of dilution medium and incubated for one hour at 4⬚C. The mixtures were then plated in triplicate in four-well Nunclon Delta Surface incubation dishes (Nalge Nunc Intl., Rochester, NY, USA), and culture medium was added to a final volume of 0.5 mL per well. After 11 days of incubation in a humidified atmosphere at 37⬚C and 5% CO2, cultures were scored for BFUE and in some experiments also for CFU-GM and CFU-GEMM.
1083
As controls, triplicates of cells and dilution medium only (medium control) were included in each experiment. Mean values of colony counts based on triplicates were expressed as percentages of the colony count in the medium control of the respective experiments. For animal cells, a comparable colony formation assay was used, as previously described [11,12]. Toxicity studies VP2 was added in serial dilutions to cultures of undifferentiated MB02 or K562 cells to a final concentration of 1.7 µg/mL. Parallel cultures with bovine serum albumin (BSA) added instead of VP2 were used for comparison with untreated cells. The cells were evaluated after 1, 4, and 8 days, for cell growth and viability through cell counting and trypan blue staining. All samples were processed in duplicate. Cleavage of VP2 VP2 was digested by three endoproteinases, and the resulting mixtures of VP2 protein fragments were used in the colony formation assay. Enzymes used were endoproteinases Lys-C, Arg-C, and Glu-C, which specifically cleave peptide bonds C-terminally at lysine, arginine, and glutamic acid residues respectively (all from Roche, Basel, Switzerland). Lyophilized enzymes were reconstituted and used for protein cleavage according to the manufacturer’s instructions. Total fragmentation was confirmed by showing shorter fragments, but no intact VP2, on a SDS-PAGE and subsequent silver staining. Solutions with only enzyme were included in the assays as controls. In vivo experiments The ability of VP2 to inhibit hematopoiesis in vivo was tested using neonatal M. nemestrina. Two experimental animals received 1.72 mg/kg of VP2 given intravenously, divided into five equal dosages on five consecutive days (days 0–4). A third animal received no injections and was used as a control. Peripheral blood and bone marrow were collected on days 5, 9, 12, 15, 19, and 33. Peripheral blood was evaluated for hemoglobin level, hematocrit value, reticulocyte count, and white blood cell count. Serology for B19 anti-VP1 IgG and anti-VP2 IgM antibodies was tested using a commercially available enzyme-linked immunosorbent assay (IBLHamburg, Germany). Care was taken to limit the total blood and bone marrow collections to less than 2 mL/sample to minimize the effect from repeated collections, and similar total amounts relative to body weight were sampled from all three animals. Statistical analysis For calculation of p-values, unpaired t-tests were performed on colony counts in individual test vs control cultures. Paired t-tests were employed when means from multiple experiments were analyzed. InStat software from GraphPad was used for statistical calculations. Ethical permissions The Ethical Committee at the Karolinska Institutet gave ethical permissions for the human sample collections. The collection of nonhuman primate marrow and blood and the administration of VP2 to nonhuman primates was an approved protocol through the University of Washington Animal Care Committee.
Results VP2 was shown to inhibit erythroid colony formation by approximately 55% at a concentration of 1.9 µg/mL, and
1084
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
the effect was dose-dependent (Fig. 1). Inhibition of the myeloid lineages CFU-GM and CFU-GEMM was also demonstrated by VP2 (Fig. 2). BFU-E was also inhibited by the B19 isolate cultured at a concentration of 1.3 x 104 geq/mL, but CPV capsids showed no inhibitory effect in cultures with human fetal liver cells at a concentration that was inhibitory for VP2. In an experiment with bone marrow cells from a PV patient, VP2 inhibited BFU-E colony formation by about 30% at a concentration of 1.9 µg/mL. Formation of BFU-E and CFU-GM in cultures of macaque bone marrow was also inhibited by VP2 by about 55% and 40%, respectively, at a concentration of 5 µg/mL. The cleavage products from VP2 digested by endoproteinases were also capable to inhibit BFU-E. The cleavage products resulting from digestion by an endoproteinase that cleaves at arginine showed inhibition similar to intact VP2, and colony counts were significantly lower than in a control with only the enzyme, whereas cleavage at lysine and glutamic acid residues resulted in less active fragments (Fig. 3). By silver gel staining, the enzymatically treated preparations were shown not to contain any remaining intact VP2. VP2 was tested for general toxicity against cell lines permissive for B19. On day one following addition of the capsids, the cultures appeared under the microscope to have a number of “clusters” of cells, possibly due to agglutination. Otherwise, no changes in cell growth, morphology, or frequency of cells staining with trypan blue
Figure 1. Inhibitory effect of VP2 on BFU-E. Human fetal liver cells incubated with VP2 at different concentrations resulted in a dose/response effect. The mean values of BFU-E counts in 6 separate experiments are shown and expressed as percent of medium control. Colony counts for the two highest VP2 concentrations were significantly lower than the medium control, p ⫽ 0.018 and 0.023, respectively. Error bars ⫽ SEM.
were observed at any time point, as compared to untreated cells (data not shown). Because of the demonstrated effect of VP2 on colony formation in cultures of macaque bone marrow in vitro, we next evaluated the effect on hematopoiesis in vivo using three macaques. The two experimental animals did not have preexisting B19 antibodies, whereas the control animal was B19 IgG⫹. However, the serology should be taken with precaution since the kit has not been evaluated for testing of animal antibodies. VP2 administered intravenously was well tolerated and no side effects were noted during followup. Cells from consecutive bone marrow specimens showed decreased colony-forming capacity ex vivo, relative to the first sampling on day five (Fig. 4). The drop did not recover during the follow-up period and was markedly greater in the two experimental animals than in the control animal,
Figure 2. Inhibitory effect of different preparations. Tested preparation, cell origin, and evaluated colony type is given in the x-axis labels. All mean colony counts are presented as percentages of the medium control in the respective experiments. From left to right the bars represent: CPV at 2.5 µg/mL; B19 isolate at 1.3 geq/mL; mean based on 6 experiments of VP2 at 1.9 µg/mL; mean based on 3 experiments of VP2 at 1.0 µg/mL, evaluated for CFU-GM and CFU-GEMM; VP2 at 1.9 µg/mL tested on BM from a PV patient (PV-BM); and the last two bars are macaque bone marrow (mBM) cultured with VP2 at 5 µg/mL. p-values for each comparison to respective medium control are given. FL ⫽ fetal liver. Error bars ⫽ SEM.
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
1085
during the follow-up in any of the three animals (data not shown).
Figure 3. Effect on BFU-E of intact and fragmented VP2. Human fetal liver cells were incubated with intact VP2 at 0.9 µg/mL (Intact), and endoproteinase-digested VP2 at a concentration corresponding to 0.9 µg/mL intact VP2. Enzymes used were endoproteinases that cleave at lysine (LYS), glutamic acid (GLU), or arginine (ARG) residues, respectively. Mean colony counts for intact VP2 are presented relative to the medium control, while mean colony counts for fragmented VP2 are presented relative to a control containing the respective endoproteinase. p-values for each comparison to respective control are given. Error bars ⫽ SEM of triplicate wells.
even though a decrease was also noted in the control animal. In one of the two experimental animals, a drop in hematocrit value by 26% at day 12 was noted that recovered within the follow-up period (Fig. 5). Hemoglobin level, reticulocyte count, and white blood cell count did not change significantly
Figure 4. Ex vivo colony formation of macaque bone marrow after VP2 injections. Experimental animals 1 and 2 received intravenous VP2 (1.72 mg/kg divided into 5 daily doses), whereas the control animal received no injections. Samples were collected during follow-up and the capacity of CFC (BFU-E ⫹ CFU-GM) formation is plotted as means of 5 wells relative to the capacity of colony formation at day 5. Error bars ⫽ SEM.
Discussion We have previously reported that VP1/2 specifically inhibits BFU-E formation of cells from bone marrow, cord blood, and fetal liver [9]. We suggest that this effect may be exploited into a medicament for bone marrow suppression; a smaller capsid-derived agent would, however, be more advantageous. In the present report, we show that less complex capsids of only VP2 possess similar inhibitory capacity as VP1/2. Besides the effect on BFU-E, we showed that CFU-GM and CFU-GEMM are also subject to this inhibition. Although myeloid cells were initially reported to be resistant to B19, infection of myeloid precursors is now demonstrated, with inhibition of colony formation as a result [13–17]. Recent data suggest that early myeloid precursors also express P-ag [3,18]; however, the presence of the newly discovered co-receptor remains to be investigated. It thus appears reasonable that myeloid, as well as erythroid, colony formation is affected by VP2. The possibility that the observed inhibition was due to a toxic effect was excluded by toxicity studies of VP2, and by the fact that CPV empty capsids produced in the same system as VP2 lacked inhibitory capacity. When VP2 was enzymatically cleaved, the resulting mix of protein fragments retained the inhibitory capacity, which suggests that the inhibition of hematopoiesis is achieved by a region within VP2, which should be defined further. The inhibitory capacity depended on the amino acid specificity of the cleaving enzyme in these experiments, indicating that the active region was differently well preserved by the enzymes. Another biologically active region of the B19 capsid is described in the unique region of VP1, which possesses an intrinsic phospholipase A2 activity believed to be important for viral entry in the cell nucleus [19].
Figure 5. Hematocrit values after VP2 injections. Hematocrit levels were evaluated at intervals after intravenous administration of VP2 to the experimental animals.
1086
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
The major advantage of defining the smallest active compound, besides a more convenient production, is that it would probably be less immunogenic and be less likely to be neutralized by a preexisting immunity to B19. Neutralizing antibodies are mainly directed to the VP1 unique region, and the few neutralizing epitopes localized in VP2 seem conformationally dependent on sites in the VP1 unique region [1]. The immune response to VP1 appears to be critical for protective immunity and VP1 must be present in recombinant capsids to elicit a neutralizing antibody response in animals [20,21]. A modified B19 capsid optimized for generating an efficient neutralizing response, for use as a vaccine candidate, has been defined and consists of 25% VP1 and 75% VP2 [22]. In a human vaccine trial, the immunogenicity was highly dependent on the adjuvant used, indicating that the capsid formation on its own elicits a poor neutralizing antibody response. Thus, at least theoretically, a VP2-derived construct should not elicit neutralizing antibodies, and an optimal drug would have the parts of the protein deleted that do not contribute to the biological activity. The active part could then ideally be produced as a much smaller protein, linear fragment, or maybe even a peptide. In the above-mentioned vaccine trial, a maximal dose of 25 µg capsid was given intramuscularly, which could theoretically result in a maximal systemic capsid concentration of 5 x 10-3 µg/mL (even distribution in 5000 mL of blood volume is assumed). This concentration was well tolerated and no severe side effects were noted, nor were any alterations in hematological parameters reported [22]. This is consistent with our results since we detected only a negligible effect on BFU-E at comparable VP2 concentrations. B19 infectious particles are potent inhibitors of erythropoiesis [2,23,24]. The B19 isolate we used resulted in a 40% reduction in BFU-E formation at 1.3 x 104 geq/mL, which represents the same number of capsids. Similar inhibition was seen at a VP2 concentration of about 1 µg/mL, which can be calculated to represent about 1.7 x 1011 particles. This 7-log difference in concentration may appear large, but other mechanisms such as viral replication and cell-to-cell infection come into play when infectious virus is used. A drug that specifically inhibits hematopoiesis may be beneficial for a broad spectrum of hematological proliferative disorders and possibly also useful in prenatal stem cell transplantations [25]. However, PV is perhaps the most obvious disorder matching the target cell spectrum of VP2. PV arises as a monoclonal and apparently neoplastic disorder, and the increase in red cell mass is often accompanied by an elevation of granulocyte and platelet counts as well [26]. PV appears insidiously, usually in late middle age, with an incidence of 1 to 2 per 100,000 and is a lethal disease within 1 to 3 years if untreated. No curative treatment is available, and current intervention to suppress the overproduction of bone marrow cells is not satisfactory. Aiming at keeping the hematocrit level below about 45% in PV patients, venesection is often used that typically reduces the
hematocrit by 20 to 25%. Although speculative, assuming that the demonstrated effect of VP2 on BFU-E in vitro is translatable into hematocrit reduction in vivo, our data suggests that a therapeutic effect would be achieved by a VP2 concentration in the bone marrow of 0.2 µg/mL. However, optimization of a candidate drug as discussed earlier would ideally also improve the effect/concentration ratio. Interestingly, a recent report described anemia following B19 infection in a patient suffering from PV [27]. This not only shows that bone marrow cells in PV patients are susceptible to B19, but also that the concept of utilizing a VP2 derivate as a medicament for this disorder is theoretically sound. Indeed, we noted a 30% inhibition of bone marrow from a PV patient by VP2 in vitro, although the level of inhibition on these cells needs to be established in repeated experiments. Outbreaks of anemia in cynomologus monkeys as well as in rhesus monkeys and pig-tailed macaques have led to the identification of novel simian parvoviruses [28,29]. Infection of immunosuppressed macaques with simian parvovirus leads to persistent anemia, whereas in immunocompetent animals there is a temporal drop in reticulocytes [30]. These viruses have been proposed to be classified to the erythrovirus genus, together with B19, since they are similar at the genome level and in their predilection for bone marrow cells in vitro [31]. Encouraged by these similarities between simian and human parvoviruses, we tested VP2 on nonhuman primate hematopoietic cells. It was thus demonstrated in vitro that VP2 was capable to suppress BFU-E and CFUGM formation of cells derived from macaque bone marrow. The results from these experiments implicated that macaques were suitable for in vivo studies, and three animals were used for this purpose. Newborn animals were preferred due to a higher erythropoietic turnover than later in life, and a relatively low distribution volume. The dose given at each injection was based on the in vitro results and was estimated to result in comparable VP2 concentration in blood (and in the bone marrow). In the two experimental animals, there was a marked reduction over time of the total ex vivo hematopoietic colony formation relative to the control time point. A less pronounced reduction was also detected in the control animal, to which no explanation was found. The drop in hematocrit level around day 12 after the first injection, seen in one experimental animal, was reminiscent of the kinetics of reticulocyte numbers reported in experimentally infected human subjects [5]. Reticulocyte numbers fell to undetectable levels around day 10 postinfection, and started to recover about 6 days later. However, as compared to the intranasally administered infectious virus, our intravenously administered VP2 probably reached the bone marrow more rapidly. Clinically nonsignificant lymphopenia, neutropenia, and thrombocytopenia, with a brief overshoot prior to stabilizing at preincubation levels, occurred in the experimentally infected human subjects, whereas such effects on blood cell counts were not seen in our experimental animals. Even though an effect of VP2 in vivo was suggested, very few
O. Norbeck et al. / Experimental Hematology 32 (2004) 1082–1087
animals were used in this pilot study. The in vivo results were therefore not conclusive and need to be repeated in a larger number of animals. Further characterization of the region within VP2 responsible for the observed inhibitory effect will provide new VP2-derived constructs that will be interesting to evaluate in future studies.
Acknowledgments This work was supported by the NIH (HL62422-02), European Comission Biomed 2 Programme (Eurofetus), the Swedish Medical Research Council (k2001-06x-11231-07B), the Swedish Research Council (K2002-16X-14189-01A), the Swedish Cancer Foundation, the Swedish Children’s Cancer Foundation, and the Tobias Foundation. We thank Drs. L. Goobar-Larsson, A. Vahlne, and E. Hellstro¨m-Lindberg, and Ms. L. Markling, C. Go¨therstro¨m, and L. Radler for expert technical assistance and discussion.
References 1. Heegaard ED, Brown KE. Human parvovirus B19. Clin Microbiol Rev. 2002;15:485–505. 2. Brown KE, Anderson SM, Young NS. Erythrocyte P antigen: cellular receptor for B19 parvovirus. Science. 1993;262:114–117. 3. Cooling LL, Zhang de S, Naides SJ, Koerner TA. Glycosphingolipid expression in acute nonlymphocytic leukemia: common expression of shiga toxin and parvovirus B19 receptors on early myeloblasts. Blood. 2003;101:711–721. 4. Weigel-Kelley KA, Yoder MC, Srivastava A. α5β1 integrin as a cellular coreceptor for human parvovirus B19: requirement of functional activation of β1 integrin for viral entry. Blood. 2003;102:3927–3933. 5. Anderson MJ, Higgins PG, Davis LR, et al. Experimental parvoviral infection in humans. J Infect Dis. 1985;152:257–265. 6. Kajigaya S, Fujii H, Field A, et al. Self-assembled B19 parvovirus capsids, produced in a baculovirus system, are antigenically and immunogenically similar to native virions. Proc Natl Acad Sci U S A. 1991; 88:4646–4650. 7. Agbandje M, Kajigaya S, McKenna R, Young NS, Rossmann MG. The structure of human parvovirus B19 at 8 A resolution. Virology. 1994;203:106–115. 8. Chipman PR, Agbandje-McKenna M, Kajigaya S, et al. Cryo-electron microscopy studies of empty capsids of human parvovirus B19 complexed with its cellular receptor. Proc Natl Acad Sci U S A. 1996; 93:7502–7506. 9. Lindton B, Tolfvenstam T, Norbeck O, et al. Recombinant parvovirus B19 empty capsids inhibit fetal hematopoietic colony formation in vitro. Fetal Diagn Ther. 2001;16:26–31. 10. Kaikkonen L, Lankinen H, Harjunpaa I, et al. Acute-phase-specific heptapeptide epitope for diagnosis of parvovirus B19 infection. J Clin Microbiol. 1999;37:3952–3956. 11. Shields LE, Andrews RG. Gestational age changes in circulating CD34⫹ hematopoietic stem/progenitor cells in fetal cord blood. Am J Obstet Gynecol. 1998;178:931–937.
1087
12. Shields LE, Gaur LK, Gough M, Potter J, Sieverkropp A, Andrews RG. In utero hematopoietic stem cell transplantation in nonhuman primates: the role of T cells. Stem Cells. 2003;21:304–314. 13. Kurtzman GJ, Ozawa K, Cohen B, Hanson G, Oseas R, Young NS. Chronic bone marrow failure due to persistent B19 parvovirus infection. N Engl J Med. 1987;317:287–294. 14. McClain K, Estrov Z, Chen H, Mahoney DH Jr. Chronic neutropenia of childhood: frequent association with parvovirus infection and correlations with bone marrow culture studies. Br J Haematol. 1993;85:57–62. 15. Ozawa K, Kurtzman G, Young N. Productive infection by B19 parvovirus of human erythroid bone marrow cells in vitro. Blood. 1987;70: 384–391. 16. Potter CG, Potter AC, Hatton CS, et al. Variation of erythroid and myeloid precursors in the marrow and peripheral blood of volunteer subjects infected with human parvovirus (B19). J Clin Invest. 1987;79: 1486–1492. 17. Srivastava A, Bruno E, Briddell R, et al. Parvovirus B19-induced perturbation of human megakaryocytopoiesis in vitro. Blood. 1990; 76:1997–2004. 18. Weigel-Kelley KA, Yoder MC, Srivastava A. Recombinant human parvovirus B19 vectors: erythrocyte P antigen is necessary but not sufficient for successful transduction of human hematopoietic cells. J Virol. 2001;75:4110–4116. 19. Zadori Z, Szelei J, Lacoste MC, et al. A viral phospholipase A2 is required for parvovirus infectivity. Dev Cell. 2001;1:291–302. 20. Saikawa T, Anderson S, Momoeda M, Kajigaya S, Young NS. Neutralizing linear epitopes of B19 parvovirus cluster in the VP1 unique and VP1-VP2 junction regions. J Virol. 1993;67:3004–3009. 21. Bansal GP, Hatfield JA, Dunn FE, et al. Candidate recombinant vaccine for human B19 parvovirus. J Infect Dis. 1993;167:1034–1044. 22. Ballou WR, Reed JL, Noble W, Young NS, Koenig S. Safety and immunogenicity of a recombinant parvovirus B19 vaccine formulated with MF59C.1. J Infect Dis. 2003;187:675–678. 23. Takahashi T, Ozawa K, Takahashi K, Asano S, Takaku F. Susceptibility of human erythropoietic cells to B19 parvovirus in vitro increases with differentiation. Blood. 1990;75:603–610. 24. Mortimer PP, Humphries RK, Moore JG, Purcell RH, Young NS. A human parvovirus-like virus inhibits haematopoietic colony formation in vitro. Nature. 1983;302:426–429. 25. Shields LE, Lindton B, Andrews RG, Westgren M. Fetal hematopoietic stem cell transplantation: a challenge for the twenty-first century. J Hematother Stem Cell Res. 2002;11:617–631. 26. Spivak JL. Polycythemia vera: myths, mechanisms, and management. Blood. 2002;100:4272–4290. 27. Kaptan K, Beyan C, Cetin T, et al. Anemia following human parvovirus B19 infection in a patient with polycythemia vera. Am J Hematol. 2002;69:296–297. 28. O’Sullivan MG, Anderson DC, Fikes JD, et al. Identification of a novel simian parvovirus in cynomolgus monkeys with severe anemia. A paradigm of human B19 parvovirus infection. J Clin Invest. 1994;93: 1571–1576. 29. Green SW, Malkovska I, O’Sullivan MG, Brown KE. Rhesus and pigtailed macaque parvoviruses: identification of two new members of the erythrovirus genus in monkeys. Virology. 2000;269:105–112. 30. O’Sullivan MG, Anderson DK, Goodrich JA, et al. Experimental infection of cynomolgus monkeys with simian parvovirus. J Virol. 1997; 71:4517–4521. 31. Brown KE, Young NS. The simian parvoviruses. Rev Med Virol. 1997;7: 211–218.