Activity of imatinib in systemic mastocytosis with chronic basophilic leukemia and a PRKG2-PDGFRB fusion
Idoya Lahortiga, Cem Akin, Jan Cools, Todd M. Wilson, Nicole Mentens, Diane C. Arthur, Irina Maric, Pierre Noel, Can Kocabas, Peter Marynen, Lawrence S. Lessin, Iwona Wlodarska, Jamie Robyn, Dean D. Metcalfe

Author Affiliations

  1. Idoya Lahortiga1,2,
  2. Cem Akin3,
  3. Jan Cools1,2,
  4. Todd M. Wilson4,
  5. Nicole Mentens1,2,
  6. Diane C. Arthur5,
  7. Irina Maric6,
  8. Pierre Noel1,
  9. Can Kocabas4,
  10. Peter Marynen2,
  11. Lawrence S. Lessin7,
  12. Iwona Wlodarska2,
  13. Jamie Robyn4 and
  14. Dean D. Metcalfe4
  1. 1Human Genome Laboratory, Department of Molecular and Developmental Genetics, VIB, Leuven, Belgium
  2. 2Center for Human Genetics, University of Leuven, Leuven, Belgium
  3. 3Allergy and Clinical Immunology, University of Michigan, Ann Arbor, MI, USA
  4. 4Laboratory of Allergic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA
  5. 5Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA
  6. 6Department of Laboratory Medicine, National Institutes of Health, USA and
  7. 7Washington Cancer Institute, Washington Hospital Center, Washington, DC, USA
  1. Correspondence: Dean D. Metcalfe, Laboratory of Allergic Diseases, NIAID, NIH, DHHS, Building 10, Room 11C205, 10 Center Drive - MSC1881, Bethesda, MD 20892-1881 USA. E-mail: dmetcalfe{at}niaid.nih.gov
View Abstract

Abstract

Background Translocations involving region 5q31-32 (PDGFRB) have been reported in a variety of myeloproliferative diseases and are often associated with significant peripheral eosinophilia. We report an unusual case of a patient presenting with peripheral basophilia and systemic mastocytosis in whom cytogenetic analysis revealed a t(4;5)(q21.1;q31.3).

Design and Methods We used molecular analyses to determine the role of PDGFRB in this case. The patient was treated with imatinib.

Results Fluorescence in situ hybridization (FISH) documented a breakpoint in PDGFRB. In agreement with this, the patient responded very well to imatinib with resolution of clinical symptoms, basophilia, and mast cell disease. Molecular analyses revealed that PDGFRB, encoding an imatinib-sensitive tyrosine kinase, was fused to PRKG2. The fusion gene incorporates the first two exons of PRKG2 fused to the truncated exon 12 of PDGFRB, resulting in the disruption of its juxtamembrane domain. Functional studies confirmed that the activity and transforming properties of PRKG2-PDGFRβ were dependent on the disruption of the auto-inhibitory juxtamembrane domain.

Conclusions Our results identify a second case of the PRKG2-PDGFRB fusion and confirm the unusual PDGFRB breakpoint associated with this fusion. This work also illustrates the use of imatinib for the treatment of specific cases of systemic mastocytosis.

Introduction

Systemic mastocytosis (SM) is an uncommon myeloproliferative disease with variable clinical manifestations. Activating mutations of KIT (most commonly D816V) are found in more than 90% of cases of SM and are believed to be involved in disease pathogenesis.1 In about 10–35% of cases, SM is associated with a second hematologic neoplasm. This subtype of SM is referred to as SM with an associated hematologic non-mast cell lineage disease (SM-AHNMD). In order to establish the diagnosis, WHO criteria for both SM and the AHNMD must be met.

SM-AHNMD has been described in a number of case series and while the AHNMD may be of myeloid or lymphoid origin, most often the associated neoplasm is myeloid.2 The molecular pathogenesis of SM-AHNMD appears to be variable. In one series of patients with SM-AHNMD, KIT point mutations were detected in 16 out of 20 cases,3 while in other cases, SM-AHNMD has been reported in association with a variety of cytogenetic abnormalities including t(8;21)(q22;q22), del 20(q12), and trisomy 9.2,4

Other myeloproliferative diseases are also frequently characterized by the presence of mutations or rearrangements of protein tyrosine kinase genes such as JAK2, FGFR1, PDGFRA and PDGFRB.5,6 A variety of chromosomal translocations involving PDGFRB have been identified and reveal that PDGFRB can fuse to many different partner genes.7 The most common fusion is the ETV6-PDGFRB (TEL-PDGFRB) fusion, for which it has been well documented that homooligomerization of ETV6-PDGFRβ is responsible for its constitutive kinase activation.8 PDGFRβ, as well as PDGFRα, can also be activated by disruption of the inhibitory juxtamembrane region, as recently demonstrated in the case of the FIP1L1-PDGFRα fusion found in patients with chronic eosinophilic leukemia.9

Walz and colleagues have recently described three novel PDGFRB fusions in myeloproliferative disorders, including a case with a t(4;5;5)(q23;q31;q33) resulting in the fusion of PRKG2 with PDGFRB.10 We here describe a patient with a novel clinical presentation of SM with associated chronic basophilic leukemia characterized by leukocytosis, dramatic peripheral basophilia, atypical bone marrow mast cells, and a t(4;5)(q21.1;q31.3). Reverse transcriptase-polymerase chain reaction (RT-PCR) showed the presence of a PRKG2-PDGFRB fusion transcript, with an unusual breakpoint located within exon 12 of PDGFRB.

Design and Methods

Immunophenotypic analysis of bone marrow mast cells by multiparameter flow cytometry

Bone marrow aspirates were washed and the samples were stained using directly conjugated antibodies against CD2, CD25, CD35, CD45 and CD117 (Becton Dickinson, San José, CA, USA). Red blood cells were lysed using FACS Lysing Buffer (Becton Dickinson). After washing, the cells were resuspended in phosphate-buffered saline 1% formaldehyde. Three-color flow cytometry was performed using a FACScan flow cytometer (Becton Dickinson). CELLQuest software was used for acquisition of data and analysis.

Cytogenetics

Cytogenetic studies were performed using standard methods. Metaphase chromosomes were obtained from 24- and 48-hour unstimulated cultures of buffy coat cells from fresh heparinized bone marrow aspirates. Chromosomes were G-banded using Wright’s stain. A minimum of 20 metaphase cells were fully analyzed at the 400–550 band level of resolution at all time points; 30 additional cells were scored for chromosome number and chromosomes #4 and #5 at diagnosis. Multiple digital images were captured and karyotyped using a CytoVision™ system (Applied Imaging Corp., San José, CA, USA). Karyotypes were designated according to the ISCN (2005).11

Fluorescence in situ hybridization (FISH) studies

FISH was performed according to standard procedures. The applied probes included four cosmid clones flanking the PDGFRB gene (c9-4, c12-a, c4-1, c4-6) and three bacterial artificial chromosome (BAC) clones flanking the PRKG2 gene (RP11-660B13, RP11-22A1 and RP11-29L8). BAC clones were selected from the Human Gene Project map (www.ensembl.org) and obtained from the Roswell Park Cancer Institute libraries (www.chori.org/BACPAC). The PDGFRB probes were kindly provided by Dr. M. Dixon (University of Manchester, Manchester, UK). Cosmid and BAC clones were directly labeled with SpectrumGreen- and SpectrumOrange-dUTP (Vysis, Downers Grove, USA) using random primer reactions (Invitrogen, Carlsbad, USA). FISH experiments were evaluated using an Axioplan 2 fluorescence microscope equipped with a charge-coupled device Axiophot 2 camera (Carl Zeiss Microscopy, Jena, Germany) and a MetaSystems Isis imaging system (MetaSystems, Altlussheim, Germany). Five abnormal metaphases and 200 interphase cells were evaluated in each FISH study.

RNA isolation

Bone marrow or peripheral blood mononuclear cells were obtained by Ficoll-Hypaque gradient centrifugation and cells were lysed in Trizol (Invitrogen). Total RNA was extracted from 1x106 to 1x107 cells using the RNAEasy RNA isolation kit (Qiagen, Valencia, CA USA).

Screening for KIT and interleukin-3 receptor β chain mutations

To screen for D816V KIT mutation by restriction fragment length polymorphism (RFLP) analysis, approximately 1 μg of total RNA was reverse transcribed as described above. A 373-bp region of the c-kit gene was amplified by PCR using the following primers: primer1 (KITAsp816F 5'-AAA GGA GAT CTG TGA GAA TAG GCT C-3') and primer2 (KITAsp816R: 5'-AGC TCC CAA AGA AAA ATC CCA TAG G-3'). PCR products were digested with the restriction enzymes Hae III and Hinf I (New England Biolabs, USA) and electrophoresed on a 10% trisborate EDTA polyacrylamide gel (Invitrogen, San Diego, CA, USA) stained with ethidium bromide. Digestion of the wild-type PCR product with these enzymes yields fragments of 171 bp, 127 bp, 43 bp, and 31 bp. An A to T substitution in KIT nucleotide 2468 (within the 171-bp fragment) causing the D816V mutation introduces a new Hinf I recognition site resulting in fragments of 157 bp, 127 bp, 43 bp, 31 bp, and 14 bp. Full length coding sequences of KIT and IL3 receptor beta genes were analyzed by RT-PCR amplification using overlapping primers followed by direct sequencing of the amplified PCR fragments on a capillary sequencer. Total RNA was reverse transcribed to cDNA with Moloney murine leukemia virus reverse transcriptase without RNaseH activity and oligo (dT) primer (Invitrogen). cDNA was amplified by PCR using overlapping oligonucleotide primers. Resulting PCR products were sequenced directly using the Big Dye Terminator kit (PE Biosystems, Foster City, CA, USA) and analyzed on a capillary automated sequencer (ABI Prism 310 Genetic Analyzer; PE Biosystems).

Reverse transcription PCR for the PRKG2-PDGFRB fusion gene

RNA was reverse transcribed with random hexamers using standard techniques. The PRKG2-PDGFRB fusion gene was amplified using a nested PCR with primers for the first round PRKG2-F1: 5′-GAGCGGGAGTAC-CATTTGAA-3′ and PDGFRB-R1: 5′-TAAGCATCTT-GACGGCCACT-3′ and primers PRKG2-F2: 5'-GCCATTGCTGAACTCACAGA-3′ and PDGFRB-R2: 5'-CTCCACCACCTGCCCAAA-3′ for the second round of amplification.

PDGFRB and PDGFRA constructs

PRKG2-PDGFRB, PRKG2-JM-PDGFRB and myc-W-PDGFRA constructs were generated by PCR and were cloned into the MSCV-puromycin vector (Clontech).

Cell culture

Ba/F3 cells were cultured in RPMI-1640 medium supplemented with 10% FBS and 1 ng/mL mouse IL3. Ba/F3 cells expressing the different PRKG2-PDGFRB fusions were generated by retroviral transduction and subsequent puromycin selection. To test IL3-independent growth, stable Ba/F3 cell lines were washed three times in PBS and cultured in medium without IL3. The number of viable cells was counted with a Vi-cell XR cell viability analyzer (Beckman Coulter, Fullerton, CA, USA). Retroviral vectors were generated by cotransfection of MSCV vectors with a packaging construct, as described elsewhere.12

Western blotting

Total cell lysates were obtained by lysing cells in cold lysis buffer (PBS with 1 mM Na2EDTA, 1M NaF, 0.1% Triton X-100, 5 mM Na3VO4, 200 mM phenylarsine oxide, pH 7.2) (Calbiochem) and complete protease inhibitor tablets (Roche). Thirty micrograms of protein lysate were combined with SDS loading buffer plus DTT (Cell Signaling Technology, Beverly, MA, USA) before electrophoresis on 4–12% Bis-Tris gradient gels (Invitrogen) and transferred to nitrocellulose membranes. The following antibodies were used: anti-PDGFRβ (Santa Cruz Biotechnology), anti-phosphotyrosine (4g10; Upstate Biotechnology, Lake Placid, NY, USA), anti-ERK2 (Santa Cruz Biotechnology), anti-phospho ERK, and anti-mouse peroxidase and anti-rabbit peroxidase (Amersham Pharmacia Biotech).

Results

Clinicopathological features and response to imatinib

The patient is a 61-year old Caucasian male with a history of familial polycystic kidney disease, who received a cadaveric renal transplant in 1991. In 1997 he developed generalized pruritus with an erythematous, reticulated-appearing rash. In 2003 he developed myalgia, abdominal pain, and diarrhea; and was noted to have splenomegaly on examination. A complete blood count showed a white blood cell count of 26,300/mm3, with 26% basophils, 4% eosinophils, and a platelet count of 118,000/mm3. Total serum tryptase concentration was 200 ng/mL. His bone marrow was hypercellular with 10% basophils and 2% mast cells. There were some atypical megakaryocytes but normal myeloid and erythroid maturation. Focal interstitial and peri-vascular mast cell aggregates with 11–18% spindle-shaped mast cells were observed (Figure 1A–C). Reticulin staining revealed grades 2–3/4 fibrosis. Flow cytometric analysis of mast cells revealed aberrant surface expression of CD25 and CD35 but not CD2.

RT-PCR followed by RFLP analysis was negative for the KIT D816V mutation, and sequencing of the entire KIT cDNA revealed no mutations. The bcr-abl translocation was not found. Because of the involvement of the IL-3 receptor common β chain in both basophil and eosinophil growth and survival, the entire coding sequence of this gene was also analyzed and no mutations were found. This patient was thought to have SM with chronic basophilic leukemia. Cytogenetics showed a translocation, t(4;5)(q21.1;q31.3), in 47 (94%) of 50 bone marrow metaphases analyzed (Figure 2).

Due to the lack of a codon 816 KIT mutation, imatinib mesylate treatment was initiated empirically at a dose of 200 mg daily. The pruritus, myalgia and abdominal complaints improved dramatically within 1–2 weeks after starting imatinib. Basophilia and elevated serum tryptase levels normalized over the next 3–4 weeks (Figure 3A–B). A repeat bone marrow analysis 1 month after initiating imatinib showed only scattered mast cells, mostly round in shape (Figure 1D–E). The dose of imatinib was increased to 300 mg/day. Cytogenetic analysis 4 months after the dose increase showed persistence of the t(4;5)(q21.1;q31.3) in ten (50%) of 20 metaphase cells analyzed. Cytogenetic remission was documented 1 year after diagnosis. The patient remains in clinical and cytogenetic remission 3 years after imatinib was initiated.

Figure 1.

Pre-treatment and post-treatment peripheral blood smears and bone marrow biopsies. (A) Pre-treatment peripheral blood smear showing marked basophilia with partially degranulated basophils. (B,C) Pre-treatment bone marrow biopsy stained with hematoxylin and eosin (B) and tryptase mmunostaining ( C) showing a hypercellular marrow with mast cell aggregates and spindle-shaped mast cells. (D) Post-treatment peripheral blood smear showing disappearance of circulating basophils. (E,F) Post-treatment bone marrow biopsy stained with hematoxylin and eosin (E) and tryptase immunostaining (F) showing a mildly hypocellular marrow with scattered tryptase-positive cells, mostly round in shape.

Molecular analyses of the t(4;5)(q21.1;q31.3)

A variety of chromosomal translocations involving the PDGFRB gene in chromosome band 5q31-32 have been reported.7 Using FISH, we investigated whether PDGFRB was also involved in this case. FISH analysis using probes flanking the PDGFRB gene documented a breakpoint within PDGFRB in this case, with the 5’ part of PDGFRB translocated to chromosome 4 (Figure 4).

Walz and colleagues recently identified three novel PDGFRB fusion partners in myeloproliferative diseases, including a case with a complex t(4;5;5)(q23;q31;q33). This case harbored a PRKG2-PDGFRB fusion with an unusual breakpoint within exon 12 of PDGFRB.10 Since our case showed similar breakpoints on chromosome 4, we investigated whether our case also harbored the PRKG2-PDGFRB fusion. RT-PCR confirmed an in-frame fusion between exon 2 of PRKG2, a 5 bp insert derived from intron 2 of PRKG2, and a truncated version (last 33 bp) of exon 12 of PDGFRB (Figure 5A).

This was similar to the fusion described by Walz et al., in which there was a fusion between exon 5 of PRKG2, a 17 bp insert derived from intron 5 of PRKG2, and a truncated version (last 41 bp) of exon 12 of PDGFRB. In both fusions, the region encoding the catalytic domain of PDGFRβ was retained, with PRKG2 contributing a coiled-coil domain that may potentially lead to dimerization and constitutive activation of the fusion protein.

Figure 2.

Cytogenetic analysis of the patient. Composite of chromosomes 4 and 5 from two G-banded metaphase cells (A, B) showing the t(4;5)(q21.1;q31.3).

Interruption of the juxtamembrane domain of PDGFRβ is required for kinase activation and for the transforming properties of PRKG2-PDGFRβ

To test whether the interruption of the juxtamembrane was required for constitutive activation and the transforming properties of PRKG2-PDGFRβ, we generated two constructs. The first construct represented the PRKG2-PDGFRB fusion, as identified in the case described in this report with the juxtamembrane domain interrupted. A second construct was generated representing the PRKG2-JM-PDGFRB fusion, with the complete transmembrane and juxtamembrane regions present, as usually found in almost all PDGFRB fusions (Figure 5B).

Figure 3.

Decrease in basophil count and serum tryptase levels over time post-treatment. Decrease in total serum tryptase (A) and absolute basophil count (B) over time after initiation of imatinib (arrow)

Ba/F3 is a murine hematopoietic cell line that requires IL3 for proliferation and survival. It has been shown that Ba/F3 cells can be transformed by activated forms of PDGFRα and PDGFRβ.9

Here we used an activated form of PDGFRα (Myc-PDGFRα) as a positive control for Ba/F3 cell transformation. Expression of PRKG2-PDFGRβ also transformed Ba/F3 cells to IL3-independent growth. Western blot analysis confirmed autophosphorylation of PRKG2-PDGFRβ and activation of the downstream signaling proteins ERK1/2 (Figure 5C–D). In contrast, expression of PRKG2-JM-PDFGRβ did not result in transformation of the Ba/F3 cells or autophosphorylation of the fusion protein (Figure 5C–D).

Discussion

SM is a myeloproliferative disease with diverse clinical features which has been sub-classified by the WHO into four variants: indolent, aggressive, SM-AHNMD and mast cell leukemia. While the prognosis of indolent SM is usually good, such that most patients do not require cytoreductive therapy, patients with aggressive forms of the disease have a much worse prognosis and may survive for only a few years. The variant of mastocytosis known as SM-AHNMD occurs in 10–35% of patients and is believed to be associated with a less favorable prognosis. While peripheral blood eosinophilia is not uncommonly seen in association with SM-AHNMD, basophilia has been rarely described. The association of SM and chronic basophilic leukemia has not been reported.

Figure 4.

Analysis of the breakpoints on chromosomes 4q and 5q by FISH. Examples of FISH analysis showing the 5q33 and 4q21 breakpoints within the PDGFRB A ( ) and PRKG2 loci (B), respectively. The applied flanking assays consisted A) c9-4/c12-a (green) and c4- of ( 1/c4-6 (red) and (B) RP11-660B13 (red) and RP11-22A1/RP11-29L8 (green). The same metaphase could be used to analyze both the PDGFRB and PRKG2 loci.

Figure 5.

Molecular characterization and in vitro functional studies of the PRKG2-PDGFRB fusion. (A) Sequence of the identified PRKG2-PDGFRB fusion. PRKG2 sequences are shown in black and PDGFRB sequences in gray. The location of the breakpoints is represented with arrows. (B) Schematic representation of the proteins studied in this work. JM: juxtamembrane domain; TM: trans-membrane domain; m indicates the myc epitope tag. (C) Transformation assay of Ba/F3 cells in the absence of IL-3 reveals that expression of PRKG2-PDGFRβ, but not of PRKG2-JM-PDGFRβ transforms Ba/F3 cells to growth factor-independent growth. (D) Phosphorylation of the PRKG2-PDGFRβ, PRKG2-JM-PDGFRβ, and ERK1/2 proteins in Ba/F3 cells. Anti-ERK was detected as a control for loading.

Chronic basophilic leukemia is a rare disorder, which was first described in 1966.13 A detailed retrospective review at the Mayo Clinic revealed only four cases of chronic basophilic leukemia over a 28-year period.14 Symptoms related to basophil mediator release, including ulcers of the gastrointestinal tract, diarrhea, coagulopathy, pruritis and urticaria, were noted. An abnormal infiltrate of atypical mast cells was present in two of these cases. Cytogenetics were normal in two patients, one had trisomy 8 along with 11q- and the fourth had monosomy 7. Two of the four patients in this series died, one after progression to acute leukemia, and the second as a result of complications after bone marrow transplantation.

Unfortunately, options for cytoreductive treatment in most cases of systemic mastocytosis are limited. The KIT D816V mutation found in most patients with SM appears to confer resistance to imatinib.15,16 Rare patients with novel KIT mutations have shown responses to imatinib, including a patient with a novel transmembrane mutation in KIT.17 A recent report described a patient with aggressive SM and eosinophilia with a fusion gene involving PDGFRB who showed a clinical response to imatinib.18 The pathogenetic role of PDGFRB or its 5’ fusion partner was not identified in this case. Earlier, published descriptions of responses to imatinib in patients with SM associated with eosinophilia19 prompted a series of clinical trials. The molecular basis of response was ultimately determined to be a novel fusion gene, FIP1L1-PDGFRA,20,21 and these patients appear to have a disease that is both clinically and molecularly distinct from KIT D816V-associated SM.22

Durable responses to imatinib have been described in rare patients with atypical myeloproliferative diseases. In most instances these imatinib-responsive diseases have been associated with fusion genes involving PDGFRB and a total of 15 PDGFRB fusion partners have been described to date: ETV6 (TEL, 12p13), TRIP11 (CEV14, 14q32), HIP1 (7q11), CCDC6 (H4, 10q21), RABEP1 (17p13), PDE4DIP (1q22), SPECCI (HCMOGT-1, 17p11), NIN (14q24), KIAA1509 (14q32), TP53BP1 (15q22), NDE1 (16p13), TPM3 (1q21), GIT2 (12q24), GPIAP1 (11p13) and PRKG2 (4q21).7 The breakpoints in most of these fusion genes are upstream of exon 10, joining the transmembrane domain of PDGFRβ to the 5’ partners, in each case containing putative oligomerization domains.

The case we describe here is unusual in many respects. This is the first demonstration of an auto-activating rearrangement of PDGFRβ in mast cell disease and the second report of a PDGFRB-associated fusion gene in SM. Furthermore, this is the second reported case of a fusion of PDGFRB with PRKG2, which encodes a type II, cyclic GMP-dependent protein kinase that seems to be involved in neuronal adaptation, smooth muscle relaxation and growth, intestinal water secretion, bone growth and renin secretion.10 In both cases with the PRKG2-PDGFRβ fusion, the juxtamembrane domain of PDGFRβ was interrupted. This is in contrast to all other known PDGFRβ fusions in which the entire juxtamembrane domain of PDGFRβ is present.9 In most characterized examples of fusion tyrosine kinases, oligomerization domains of the fusion partners have been shown to be necessary for both kinase activation and for transforming properties of the fusion proteins.6,8,23,24 Although the PRKG2 moiety contains a predicted coiled-coil domain, a recent study has shown that disruption of the juxtamembrane domain of PDGFRα or PDGFRβ alone is sufficient to cause constitutive activation of kinase activity irrespective of the presence of a fusion partner that mediates oligomerization.9 Indeed, the juxtamembrane domain has been shown to have an auto-inhibitory role in receptor tyrosine kinases.25 We show that in the context of the PRKG2-PDGFRβ fusion, the interruption of the juxtamembrane domain of PDGFRβ is required for the constitutive kinase activity and transforming properties.

Our study illustrates the importance of searching for PDGFRB rearrangements in patients with SM associated with myeloproliferative disorders, as this subgroup responds well to imatinib. This disease variant should be considered in patients without codon 816 KIT mutations and with cytogenetic abnormalities involving 5q31-32. The prevalence of this variant of SM is unclear at present, but this case illustrates the importance of cytogenetic analysis when evaluating patients with SM-AHNMD.

Footnotes

  • Authorship and Disclosures

    CA, TMW, PN, LSL, JR and DDM provided clinical care; IL, CA, JC, TMW, NM, DCA, IM, CK, IW and JR performed experiments and analyzed data and IL, CA, JC, IM, PN, PM, LSL, IW, JR and DDM contributed to writing the paper.

    The authors reported no potential conflicts of interest.

  • Funding: this research was supported in part by the Intramural Research Program of the NIH, NIAID and NCI Center for Cancer Research; a grant from the Foundation Against Cancer (a public interest foundation) (J.C.); a Concerted Research Action Grant from K.U. Leuven (J.C., I.W., P.M.), and the NIH/NIAID Division of Intramural Research.

  • Received June 12, 2007.
  • Accepted August 9, 2007.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
View Abstract