|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, 15 December 2006, Vol. 108, No. 13, pp. 3967-3975. Prepublished online as a Blood First Edition Paper on August 22, 2006; DOI 10.1182/blood-2006-04-020610.
PLENARY PAPERS Arsenic trioxide: a promising novel therapeutic agent for lymphoproliferative and autoimmune syndromes in MRL/lpr miceFrom the Centre National de la Recherche Scientifique (CNRS) Unite Propre de Recherche (UPR) 9045, Villejuif, France; Université Paris-Sud XI, Orsay, France; and CNRS/Institut Gustave-Roussy, Unité Mixte de Recherche (UMR) 8121, Villejuif, France.
MRL/lpr mice develop a human lupuslike syndrome and, as in autoimmune lymphoproliferative syndrome (ALPS), massive lymphoproliferation due to inactivation of Fas-mediated apoptosis. Presently, no effective therapy exists for ALPS, and long term, therapies for lupus are hazardous. We show herein that arsenic trioxide (As2O3) is able to achieve quasi-total regression of antibody- and cell-mediated manifestations in MRL/lpr mice. As2O3 activated caspases and eliminated the activated T lymphocytes responsible for lymphoproliferation and skin, lung, and kidney lesions, leading to significantly prolonged survival rates. This treatment also markedly reduced anti-DNA autoantibody, rheumatoid factor, IL-18, IFN-  , nitric oxide metabolite, TNF- , Fas ligand, and IL-10 levels and immune-complex deposits in glomeruli. As2O3 restored cellular reduced glutathione levels, thereby limiting the toxic effect of nitric oxide, which is overproduced in MRL/lpr mice. Furthermore, As2O3 protected young animals against developing the syndrome and induced almost total disease disappearance in older affected mice, thereby demonstrating that it is a novel promising therapeutic agent for autoimmune diseases.
The pathogenesis of systemic lupus erythematosus (SLE) is multifactorial and polygenic. The apoptosis gene Fas is a candidate contributory gene in both human SLE and murine models of it. In humans, FAS mutations result in a familial autoimmune lymphoproliferative syndrome (ALPS), characterized by nonmalignant accumulation of an unusual population of double-negative (DN) B220+CD4–CD8– / + T lymphocytes and development of autoimmune diseases.1 Overexpression of Th2 cytokine IL-10 is associated with the manifestations of ALPS.2 In mice with the MRL genetic background, the autosomal recessive gene lpr (Tnfrsf6lpr) is responsible for a syndrome characterized by the progressive accumulation of DN T-cell population in peripheral lymphoid tissues. The lpr mutation engenders much less expression of the Fas death receptor, which results in the accumulation, in lymph nodes (LNs) and spleens, of numerous activated DN T lymphocytes that are not normally regulated by the Fas-mediated mechanism controlling apoptosis of mature T cells.3,4 Therefore, MRL/lpr T and B cells massively overexpress Fas ligand (FasL), which renders them able to kill Fas+ cells in vitro5-8 and in vivo.9-11 However, lpr mutation cannot account for the entire autoimmune syndrome of MRL/lpr mice and other genes of pathologic importance in the MRL background have been mapped.12,13 MRL/lpr mice spontaneously develop a lupuslike syndrome that can affect many organs. The murine cutaneous lesions resemble human discoid lupus erythematosus. Mononuclear cells infiltrate joints and lachrymal and salivary glands, resulting in rheumatoid arthritis and Sjögren syndrome, respectively.14 As in patients with SLE, MRL/lpr mice have neurologic manifestations caused by cellular infiltrations into the central nervous system.15 MRL/lpr mice develop hypergammaglobulinemia and high levels of autoantibodies, including anti-DNA antibodies, associated with immune-complex–mediated glomerulonephritis and vasculitis.16,17 Glomerulonephritis, interstitial nephritis, and vasculitis begin by 8 weeks of age, and MRL/lpr females die at about 17 weeks of age and males at 22 weeks.18
Because of the resemblance between the murine and human diseases, MRL/lpr mice have been used extensively to attempt to determine SLE etiology and to evaluate therapies. Indeed, MRL/lpr mice provide an attractive model because their syndrome is spontaneous, predictable, and rapid; it exhibits the characteristic multifaceted tissue destruction and sexual dimorphism, and its severity varies with the individual. Th1 cytokines, including IFN-
Arsenic trioxide (As2O3) has shown substantial efficacy in the treatment of patients with newly diagnosed or relapsed acute promyelocytic leukemia (APL).36,37 Side effects are more likely to occur with intravenous than with oral administration of As2O3.38 As2O3 induces remissions in patients with APL in part through degradation of the aberrant promyelocytic leukemia-retinoic acid receptor
Mice and As2O3 treatment Wild-type MRL/MpJ (MRL+/+), MRL/MpJ-Tnfrsf6lpr/J mutant (MRL/lpr), and C57BL/6J mice originally from the Jackson Laboratory (Bar Harbor, ME) were maintained in our animal facilities. All experiments were performed in accordance with institutional Animal Research Committee guidelines. A stock solution was prepared by dissolving As2O3 powder (Sigma, St Louis, MO) in 1 M NaOH, then further diluting it in PBS; mice were injected intraperitoneally daily with 2.5, 5, 7.5, 10, or 15 µg/g body weight; controls received a daily intraperitoneal injection of PBS (volume weight-determined). Phenotyping
Spleen- and LN-cell suspensions were phenotyped by flow cytometry using either FITC-, phycoerythrin-, or biotin-conjugated monoclonal antibody (mAb): rat anti–Thy-1.2, anti-CD4, anti-CD8, anti-B220 (all from PharMingen, San Diego, CA), and rat IgG2a mAb as the isotype control ((Serotec France, Cergy Saint-Christophe, France). Use of mAb to mouse Fc Western blotting Lymphoid-cell pellets were directly lysed in Laemmli sample buffer and boiled for 5 minutes. Protein samples (30 µg each) were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by Western blotting using rabbit polyclonal anti-FasL (Santa Cruz Biotechnology, Santa Cruz, CA) or antiactin antibodies (Sigma). Histopathology and immunohistochemistry Kidney, liver, skin, and lung samples were fixed overnight in GlyoFix (Shandon Lipshaw, Pittsburgh, PA), then dehydrated, and paraffin-embedded; 4-µm thick sections were cut. After hematoxylin, eosin, and saffron staining, sections were examined. For immunohistochemical labeling, kidney samples were frozen in OCT compound, and immune complexes deposited in glomeruli were detected by incubating cryosections with horseradish peroxidase-conjugated goat anti–mouse IgG (1:200, Vector Laboratories, Burlingame, CA). Sections were then counterstained with hemalum. The photomicrographs were obtained using an Axiophot microscope (Carl Zeiss, Oberkochen, Germany) and a PCO sensicam 12 bit cooled imaging camera with SensiControl Utility software (PCO, Klenheim, Germany). Images were processed with Adobe Photoshop (Adobe Systems, San Jose, CA). ELISA detection of anti-DNA autoantibodies, rheumatoid factor, and cytokines To detect serum IgG (IgG1, IgG2a, IgG2b, IgG3) anti-DNA autoantibodies, 96-well plates were coated with 10 µg/mL calf thymus DNA (Sigma). For detection of IgM and IgG rheumatoid factor (RF), plates were coated with rabbit IgG (1 µg/mL; Sigma), and for detection of IgG3 RF, plates were coated with mouse IgG2a (1 µg/mL; Sigma).43 After blocking with 1% BSA, serial serum dilutions (starting at 1:400) were added in triplicate for 2 hours at room temperature, then washed; bound IgG anti-DNA was detected with biotin-labeled rat anti–mouse IgG, streptavidin-alkaline phosphatase, and pNPP (Sigma). Bound RF was detected with alkaline phosphatase-labeled goat anti–mouse IgM, IgG (both Sigma), or IgG3 (Caltag Laboratories, Burlingame, CA) and pNPP. OD was determined at 405 nm.
Serum levels of IFN- Determination of NO production Serum levels of nitrite were measured using the NO quantification kit (Active Motif, Rixensart, Belgium), according to the manufacturer's instructions. Apoptosis analysis and GSH detection Activated caspase levels in spleen and LN cells were measured using the CaspaTag kit (Chemicon, Temecula, CA), according to the manufacturer's instructions. GSH levels were detected in spleen and LN cells using 5 µM CellTracker probe CMFDA, following the manufacturer's instructions (Molecular Probes, Eugene, OR) and flow cytometry. Statistical analysis Data are reported as the mean ± SE. Comparisons between experimental and control groups were made by Student t test. Statistical difference was accepted at P below .05.
As2O3 prevents and reduces lymphoproliferation in MRL/lpr mice Accumulation of DN B220+CD4–CD8– T lymphocytes in MRL/lpr LN and spleen starts at 2 months and this lymphoproliferation peaks at 4 to 5 months. In experiments starting at 2 months of age, MRL/lpr and MRL+/+ mice were treated for 2 months with As2O3 daily (2.5-10 µg/g body weight) or PBS. With 2.5 µg/g As2O3 the lymphoproliferation was significantly reduced (P < .05) when compared to PBS-treated MRL/lpr mice. However, with 5, 7.5, or 10 µg/g, no lymphoproliferation was seen in the spleens and LNs of treated MRL/lpr mice (Figure 1 and data not shown). Because 5 µg/g/d was the lowest dose of As2O3 efficient to block the lymphoproliferation, this dose was used for subsequent experiments. To assess As2O3 activity on nonlymphoid organs, MRL+/+ and MRL/lpr mice received 5 µg/g As2O3 daily for 2 months and then their spleens, mesenteric LNs, axillary LNs, livers, hearts, lungs, and kidneys were weighed (Table 1). In MRL/lpr mice, mean spleen weight increased 6-fold and their mean axillary and mesenteric LN weights rose 45- and 48-fold, respectively, versus MRL+/+. Mean MRL/lpr liver and lung weights were also higher than MRL+/+ values, at about 1.5- and 1.7-fold, respectively. Under As2O3, the abnormally high MRL/lpr lymphoid organ, liver, and lung weights were sharply reduced, whereas kidney and heart weights remained unaffected (Table 1). In As2O3-treated MRL+/+ mice, spleen- and LN-cell numbers as well as liver, lung, kidney, and heart weights were unchanged (Table 1).
To determine whether As2O3 could also be curative, 4-month-old MRL/lpr mice, with severe lymphoproliferation, were given daily As2O3 (5 µg/g) injections for 2 months; lymphoproliferation in their spleens and LNs regressed under therapy (Figure 2A). Importantly, As2O3 had no effect on lymphoid organs from MRL+/+ mice (Figure 2A). Flow cytometry quantification of spleen-cell subpopulations from MRL/lpr mice treated with PBS or As2O3 using mAb specific to CD3, CD4, CD8, CD19, and B220 cell-surface antigens yielded (mean ± SE), for 4-month-old PBS-treated MRL/lpr mice, 7.2 ± 0.7 x 108 splenocytes (n = 8), among which 77% ± 1.5% and 13% ± 0.5% were CD3+ T and CD19+ B lymphocytes, respectively. DN represented 75% ± 7% of the CD3+ T cells, whereas CD4+ and CD8+ subsets accounted for 13% ± 2% and 12% ± 1.5%, respectively. The number of splenocytes in As2O3-treated MRL/lpr mice was reduced about 15-fold to 0.49 ± 0.6 x 108 (n = 10), among which 51% ± 2% and 49% ± 3% were CD3+ T and CD19+ B lymphocytes, respectively. Abnormal DN T lymphocytes represented 23% ± 1.5% of CD3+ T lymphocytes, with 50% ± 4% and 25% ± 2% normal CD4+ and CD8+ subpopulations, respectively. Similarly, mesenteric LN-cell counts were 10-fold lower. These data demonstrate the ability of As2O3 to specifically eliminate abnormal DN T lymphocytes. To completely exclude that As2O3 acts as a cytotoxic agent by causing generalized immunosuppression, we determined bone marrow, thymus, blood, spleen, and LN-cell numbers in autoimmune MRL/lpr and nonautoimmune C57BL/6J mice treated with As2O3 (5 µg/g/d) for 2 months. In As2O3-treated C57BL/6J mice, bone marrow, thymus, red blood, spleen, and LN-cell numbers were unchanged. White blood cell numbers from these mice were slightly, but not significantly, decreased (Table 2). In As2O3-treated MRL/lpr mice, the number of white blood cells is normalized and the lymphoproliferation in their spleens and LNs regressed, whereas bone marrow, thymus, and red blood cell numbers remained unchanged (Table 2). Taken together our results suggest that As2O3 has a selective effect on DN T lymphocytes, which accumulate in autoimmune MRL/lpr mice.
As2O3 suppresses skin lesions in MRL/lpr mice Human lupus exhibits several clinical phenotypes. Discoid lupus affects the skin, causing rashes and lesions, usually on the face and upper torso, and the cutaneous lesions of MRL/lpr mice resemble it. MRL/lpr mice develop necrotic skin lesions on their ears, hair loss, and scab formation, typically on the upper back (Figure 2B). Light microscopy examination of the skin lesions revealed hyperkeratosis, acanthosis, liquefaction, vasodilatation, and mononuclear-cell infiltrates in the dermis (Figure 2B left panel). Unexpectedly, injecting As2O3 (5 µg/g/d, for 1-2 months) into MRL/lpr mice led to the quasi-total disappearance of cervical cutaneous lesions and hair grew back (Figure 2B). The quasi-total regression of skin lesions in As2O3-treated mice was confirmed by light microscopy examination of skin sections (Figure 2B right panel). Thus, in arsenic-treated animals, dermal lymphoid infiltration was dramatically attenuated, and hyperkeratosis, acanthosis, hypergranulosis, liquefaction, and dermal vasodilatation disappeared. As2O3 returns FasL levels to normal in MRL/lpr mice FasL is synthesized by lymphoid cells in membrane-associated and soluble forms.44,45 Both forms were studied in MRL+/+ and MRL/lpr mice treated with PBS or As2O3. Cell-surface–anchored FasL was analyzed by Western blotting on lymphoid-cell lysates, whereas soluble FasL in mouse sera was quantified by ELISA. As expected, in Fas-deficient MRL/lpr, but not MRL+/+ mice, both FasL forms were overexpressed (Figure 3A-B). As2O3 treatment of MRL/lpr mice sharply decreased soluble and membrane-associated FasL levels to approach those found in MRL+/+ mice (Figure 3A-B). Administering As2O3 to MRL+/+ mice did not affect FasL levels, suggesting that FasL-overexpressing T lymphocytes are the main target of this agent.
IFN- , IL-18, TNF-, and IL-10 syntheses decline in As2O3-treated MRL/lpr mice
Because IFN- Like patients with ALPS2 and human SLE patients,35 untreated MRL/lpr mice produce large amounts of IL-10 when compared to MRL+/+ mice (Figure 5A). As expected, PBS-treated MRL/lpr mice had very high serum IL-10 levels, but As2O3 treatment significantly reduced IL-10 concentrations, in a dose-dependent manner (Figure 5B). However, even at the highest dose of As2O3 tested (10 µg/g), serum from MRL/lpr mice contained significantly higher concentrations of IL-10 than those from MRL+/+ mice. The DN T cells are primary producers of IL-10 in ALPS patients2; we therefore quantified these cells in MRL/lpr mice treated with PBS or As2O3 by flow cytometry using mAbs specific to CD3, CD4, CD8, CD19, and B220 molecules. In LNs from PBS-treated MRL/lpr mice, DN T cells represented 66.8% ± 4.35% of CD3+ T cells. In MRL/lpr mice treated with 2.5, 5, 7.5, or 10 µg/g As2O3, DN T cells represented 48% ± 5.40%, 19.33% ± 3.80%, 15% ± 2.34%, or 8.51% ± 2.75% of CD3+ T lymphocytes, respectively. Importantly, the reduction of serum IL-10 paralleled the decrease in DN T-cell numbers. Concentrations of nitrites decline in As2O3-treated MRL/lpr mice Measuring NO synthase (NOS) activity by monitoring the accumulation of nitrite, a stable oxidative end product of NO, is a standard assay for NOS activity. MRL/lpr mice overproduce NO as autoimmune disease progresses.29 Therefore, serum from MRL/lpr mice contained significantly higher concentrations of nitrite than those from MRL+/+ mice (Figure 4D). Although As2O3 did not modify nitrite levels in serum from MRL+/+ mice, it sharply decreased nitrite levels in serum from MRL/lpr mice to approach those found in MRL+/+ mice (Figure 4D).
As2O3 prevents inflammatory infiltrates in lungs and kidneys and inhibits immune-complex deposition in kidneys of MRL/lpr mice SLE, usually more severe than discoid lupus, can attack any organ. Pleurisy is a common pulmonary manifestation of SLE and, more rarely, patients develop acute lupus pneumonitis. Similarly, the MRL/lpr disease is characterized by massive accumulation of leukocytes in the lungs,47 with widespread perivascular and peribronchiolar mononuclear-cell infiltrates (Figure 6A). In contrast, in As2O3-treated MRL/lpr mice, the pulmonary architecture was normal, with mononuclear cells rarely seen around vessels and never around the airways (Figure 6A). Lupus nephritis or glomerulonephritis is another severe manifestation of human SLE. The kidney disease developing in MRL/lpr mice is fatal and caused by lymphoid cells infiltrating into glomeruli, interstitium, and perivascular compartments (Figure 6B). In contrast, MRL/lpr mice treated with As2O3 for 2 months had normal kidney structure, with a few mononuclear cells around glomeruli (Figure 6B). As in humans, MRL/lpr lupus nephritis is caused by the deposition of immune complexes (Figure 6C) in kidney parenchyma. Treating 2-month-old MRL/lpr mice with As2O3 for 2 months completely prevented immune-complex deposits in their glomeruli (Figure 6C).
As2O3 inhibits autoantibody and RF production in MRL/lpr mice Anti-DNA antibodies and RF, commonly found in patients with SLE, are thought to play important pathogenic roles in lupus nephritis.17,48,49 IgG1, IgG2a, IgG2b, and DNA-reactive IgG3, and IgM, IgG, and IgG3 RF were quantified by ELISA in sera from MRL+/+ and MRL/lpr mice treated with PBS or As2O3. IgG1 and IgG3 anti-DNA and IgG and IgG3 RF levels were markedly lower in As2O3-treated mice, and IgG2a and IgG2b anti-DNA and IgM RF were moderately lower (Figure 7).
As2O3 significantly improves survival of MRL/lpr mice As2O3 significantly prolonged survival (P < .001) of male and female MRL/lpr mice with established disease (Figure 8 and data not shown). Indeed, 18 weeks after starting treatment, all As2O3-treated female mice were alive, whereas all PBS-treated MRL/lpr mice had died. After 45 weeks of treatment, As2O3-treated MRL/lpr mice were still alive with no sign of lymphoproliferation. Finally, 75 weeks after starting treatment, all As2O3-treated MRL/lpr had died. Likewise, most of the MRL+/+ mice died by 77 weeks. Therefore, the survival curves of As2O3-treated MRL/lpr are similar to those of MRL+/+ mice.
As2O3 activates caspases and normalizes glutathione levels in MRL/lpr mice Because As2O3 was shown to induce apoptosis of various cell types,42 we determined the levels of activated caspase-2, -8, and -9 in lymphocyte populations from mice treated with As2O3 for 1 up to 90 days. Surprisingly, as early as day 4 of treatment, about 50% of MRL/lpr B220+ DN T lymphocytes contained activated caspases, as opposed to only a few MRL/lpr B cells, and MRL/lpr B220–CD4+ and B220–CD8+ T cells (Figure 9A). No activated caspases were detected in B lymphocytes or the B220–CD4+ and B220–CD8+ T-cell subpopulations from MRL+/+ mice treated with As2O3 for 4 days or longer (data not shown). Because NO is overproduced in the diseased organs of MRL/lpr mice,29 we measured levels of GSH, a critical antioxidant in cellular defenses against free radical-induced damage. Notably, GSH concentrations were 4 to 5 times lower in MRL/lpr spleen and LN-cell populations than in those from MRL+/+ mice (Figure 9B; data not shown). Although MRL+/+ GSH levels were not significantly modified by As2O3, intracellular GSH levels in MRL/lpr spleen and LN-cell populations had increased significantly (P < .001) 3- to 5-fold as of day 4 of treatment to reach the levels found in MRL+/+ mice (Figure 9B and data not shown).
Our data presented herein demonstrate that As2O3 is a novel therapeutic agent to delete autoreactive lymphocytes and block the progression of autoimmune diseases. Arsenic therapy strongly limited or even abrogated, depending on the mouse, the following autoimmune manifestations of MRL/lpr disease: cytokine and autoantibody production, lymphoid hyperplasia, skin lesions, and mononuclear-cell infiltration into lungs and kidneys, immune-complex deposition in glomeruli, and early mortality. Remarkably, As2O3 administration achieved the quasi-total disappearance of the disease, even when it was initiated at an advanced stage, and its prevention when initiated at the predisease stage. This model of disease reversal makes As2O3 application in humans with lupus or other autoimmune diseases highly promising. In MRL/lpr mice, DN T-lymphocyte accumulation in LNs and spleen starts at 2 months and peaks at 4 to 5 months. In MRL/lpr mice treated with As2O3 more than 50% of DN T lymphocytes contained activated caspases-2, -8, and -9, and the number of DN T cells was sharply diminished, suggesting that these cells were eliminated by apoptosis. In contrast, in As2O3-treated MRL+/+ mice only a few T cells contained activated caspases, and the number of T cells was unchanged. Therefore, we showed herein that As2O3 specifically activated caspases in the abnormal FasL-overexpressing DN T cells, thereby maintaining normal FasL levels.
Under As2O3, IgG1, IgG2a, IgG2b, and IgG3 anti-DNA autoantibodies as well as IgM, IgG, and IgG3 RF concentrations were much lower. Pathogenic autoantibodies in lupus mice generally belong to the IgG2a, IgG2b, and IgG3 subclasses, which are predominantly regulated by IFN-
IL-18 also participates in the progression of this autoimmune syndrome. This multifunctional cytokine is expressed by various cell types, including antigen-presenting cells (macrophages and dendritic cells [DCs]) and cells not primarily involved in immune responses, for example, keratinocytes, articular chondrocytes, synovial fibroblasts, and osteoblasts, and also within the adrenal cortex and pituitary gland.28 IL-18 induces IFN- Overexpression of Th2 cytokine, IL-10, is associated with the manifestations of ALPS2 and has a disease-promoting effect in humans and mice with lupus.33,35 IL-10 can be secreted by many cell types, with multiple biologic effects.59 For example, IL-10 can inhibit inflammatory responses as well as promote B-cell proliferation and antibody production. IL-10 contributed to the abnormal production of immunoglobulin and of autoantibodies in SLE.34 Administration of anti-IL-10 antibodies delayed onset of autoimmunity in NZB/W F1 mice.60 Moreover, anti-IL-10 mAb administration to SLE patients with active disease led to a reduction in disease activity.61 However, the precise role of IL-10 in the pathogenesis of lupus remains uncertain.62 IL-10 levels, which were abnormally high in MRL/lpr mice, decreased significantly in As2O3-treated MRL/lpr mice. Importantly, the reduction of serum IL-10 paralleled the decrease in DN T-cell numbers in peripheral lymphoid tissues. These observations suggest that like DN T cells from ALPS patients,2 the DN T cells are primary producers of IL-10 in MRL/lpr mice. As2O3-mediated regulation of IL-10 probably reduced serum immunoglobulins and autoantibodies, and renal immune-complex deposition. The prototypical autoimmune disease, SLE is currently treated with corticosteroids and cytotoxic or immunosuppressive drugs. These therapies prolong survival but are associated with severe side effects, particularly infections. The therapeutic intraperitoneal route of arsenic in our mice model of autoimmune disease is different from intravenous or oral administration used in patients with APL. It would be interesting to compare the 3 routes of AS2O3 administration in MRL/lpr mice.38 As2O3 significantly prolonged survival of MRL/lpr mice by preventing young mice from developing the syndrome and quasi-totally reversing established disease in older animals, thereby demonstrating its potential as a novel therapeutic agent for autoimmune diseases.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
We would like to thank Dr Jurgen Brosius (University of Muenster, Germany) for his critical reading of the manuscript and Magalis Berger and Elisabeth Connault for technical assistance. This work was supported by the Centre National de la Recherche Scientifique and grants from Ligue Nationale Contre le Cancer and Association for Research on Rheumatoid Arthritis (ARP).
Submitted April 28, 2006; accepted August 2, 2006.
Prepublished online as Blood First Edition Paper, August 22, 2006; DOI 10.1182/blood-2006-04-020610.
An Inside Blood analysis of this article appears at the front of this issue.
The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked "advertisement" in accordance with 18 USC section 1734.
Correspondence: Pierre Bobé or Mounira K. Chelbi-Alix, CNRS UPR 9045, 7 rue Guy Môquet, 94801, Villejuif Cedex France; e-mail: bobe{at}infobiogen.fr or mchelbi{at}vjf.cnrs.fr.
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
Related Article in Blood Online:
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|||||||
 |
 |
 |
 |
 |
 |
 |
 |
||
| Copyright © 2006 by American Society of Hematology Online ISSN: 1528-0020 | |||||||||
Top
Abstract
Introduction
) or PBS (
). Error bars indicate SE.
