Genotype-Proteotype Linkage in the Wiskott-Aldrich Syndrome1
http://www.100md.com
免疫学杂志 2005年第14期
Abstract
Wiskott-Aldrich syndrome (WAS) is a platelet/immunodeficiency disease arising from mutations of WAS protein (WASP), a hemopoietic cytoskeletal protein. Clinical symptoms vary widely from mild (X-linked thrombocytopenia) to life threatening. In this study, we examined the molecular effects of individual mutations by quantifying WASP in peripheral lymphocytes of 44 patients and identifying the molecular variant (collectively called proteotype). Nonpredicted proteotypes were found for 14 genotypes. These include WASP-negative lymphocytes found for five missense genotypes and WASP-positive lymphocytes for two nonsense, five frameshift, and two splice site genotypes. Missense mutations in the Ena/VASP homology 1 (EVH1) domain lead to decreased/absent WASP but normal mRNA levels, indicating that proteolysis causes the protein deficit. Because several of the EVH1 missense mutations alter WIP binding sites, the findings suggest that abrogation of WIP binding induces proteolysis. Whereas platelets of most patients were previously shown to lack WASP, WASP-positive platelets were found for two atypical patients, both of whom have mutations outside the EVH1 domain. WASP variants with alternative splicing and intact C-terminal domains were characterized for eight nonsense and frameshift genotypes. One of these, a nonsense genotype in a mild patient, supports expression of WASP lacking half of the proline-rich region. With one notable exception, genotype and proteotype were linked, indicating that a genotype-proteotype registry could be assembled to aid in predicting disease course and planning therapy for newly diagnosed infants. Knowledge of the molecular effect of mutations would aid also in identifying disease-modifying genes.
Introduction
The Wiskott-Aldrich syndrome (WAS)4 is a X-linked platelet and immune deficiency disease identified in 1937 in three infant brothers who presented with bloody diarrhea, decreased resistance to infections, eczema, and thrombocytopenia (1). In subsequent decades, the disease was considered lethal in the first decade of life, either from hemorrhage or overwhelming infection (2). Currently, diagnosed patients show remarkable heterogeneity of clinical symptoms, ranging from isolated thrombocytopenia (X-linked thrombocytopenia; XLT) to severe disease including life-threatening pyogenic and opportunistic infections, autoimmune disease, and malignancy, particularly B cell lymphoma in addition to thrombocytopenia present in all patients (3, 4, 5). T cell immunity is principally affected and might reflect the extent of WAS protein (WASP) depletion in cells. Symptoms of immune dysfunction appear shortly after birth in some patients, but are delayed in other patients by 2, 3, or more years (3). This disparity in time of onset of immune dysfunction, the cause of which is unknown, complicates diagnosis and delays treatment. The only definitive cure, bone marrow transplantation, is most effective if done before age 5 years (6).
The defective gene WASP encodes a multidomain blood cell protein (7) that regulates cytoskeletal rearrangements in activated cells. The multiple domains integrate signals that induce the following: 1) WASP recruitment to activation sites (e.g., immune synapse); and 2) conformational activation by exposure of the C-terminal VCA (verprolin/cofilin/acidic) domains that catalyze nucleation of actin filaments (8, 9). WASP is found in the cytoplasm of resting cells in the inactive autoinhibited conformation in which the VCA domains are bound to the central GTPase binding domain (GBD) (10). The resting cell molecule is further stabilized by binding of WIP to the large N-terminal Ena/VASP homology 1 (EVH1) domain (11, 12).
More than 100 patient mutations have been described. Early reports associated missense mutations, which cluster in the EVH1 domain with XLT or mild immune disease (4, 13, 14) and frameshift, nonsense, and splice site mutations with severe disease (4, 14, 15, 16). However, this linkage was not seen in other studies (17, 18), and many exceptions are known. Nonetheless, the association of many cases of mild clinical course with missense mutations that are expected to support WASP expression strongly suggested the importance of WASP expression as a determinant of disease severity.
In this study, we explore the hypothesis that the patient’s underlying mutation determines whether WASP is expressed in his cells, the level of expression, and molecular structure of the protein. These characteristics, collectively called "proteotype," are not always known based on genotype alone. The hypothesis predicts that WASP proteotype, once determined for a patient with a particular genotype, will be found for other patients with the same genotype, even when clinical course differs. To test the hypothesis, we examined WASP in PBMC of 44 patients newly diagnosed with WAS/XLT using a standardized Western blot. Molecular structures were defined by RNA analysis. The findings are analyzed in conjunction with published data on previous patients.
Materials and Methods
Patients
Patients were diagnosed with WAS/XLT based on male sex, thrombocytopenia with small platelets, eczema, immunodeficiency of variable clinical severity, family history, and, in some cases, nonrandom X chromosome inactivation of the mother’s blood cells. Most patients were newly diagnosed and <5 years old at the time of diagnosis. Based on the presence or absence of eczema, recurrent infections, autoimmune and malignant diseases in addition to thrombocytopenia, the patients’ disease was graded as mild, moderate, or severe. If only thrombocytopenia with or without mild eczema were present, the disease was considered mild. Thrombocytopenia, severe eczema, or frequent infections were graded as moderate. Thrombocytopenia, severe eczema, and severe frequent infections and/or autoimmune or malignant disease were classified as severe.
Platelet sizing
Platelet size was determined on platelet-rich plasma obtained by gravity sedimentation of blood cells from aliquots of acid-citrate dextran-anticoagulated blood. Using an Elzone Particle Analyzer Model 112 (Particle Data), the size of the platelets was measured relative to standardized latex spheres (19).
DNA isolation, amplification, and sequencing
Blood samples were obtained with informed consent under protocols approved by the Institutional Review Board of the CBR Institute for Biomedical Research. Blood of patients and paired normal individuals was collected in acid-citrate-dextrose (National Institutes of Health formula A) and processed immediately or after overnight shipment at ambient temperature. PBMC were isolated as described previously (20). Aliquots of whole blood were frozen as a source of DNA, which was isolated using QIAamp reagents (Qiagen). DNA for two patients was isolated from cryopreserved PBMC; DNA for another patient was isolated from a detergent-solubilized lysate of PBMC. WASP exons were amplified and sequenced as described above (21) by the Jeffrey Modell National Center for Primary Immunodeficiency Diagnosis at the CBR Institute for Biochemical Research.
Western blot analysis
Lysates of PBMC at 15 x 106/ml were prepared in buffer containing SDS and the protease inhibitors diisopropylfluorophosphate (2 mM) and leupeptin (25 μg/ml) (Sigma-Aldrich) as described previously (20). Lysates of 3 x 106 cells were electrophoresed on 8% acrylamide gels (14 cm x 10 cm x 1.5 mm) under reducing conditions, and the proteins were transferred to polyvinylidene difluoride membrane (Millipore) for staining with 100 ng/ml affinity-purified rabbit Ab to the WASP C-terminal 15-mer peptide (W485) as described previously (20). The reactive bands were detected by 125I-labeled secondary Ab and quantified relative to parallel dilutions of normal PBMC with the Phosphorimager Storm 860 Imager and ImageQuant v.1.1 Program (Molecular Dynamics). Except where noted, quantitation is based on duplicate analysis of two independent cell isolates. Linearity of the analysis of the responses was demonstrated previously for the range 2–100% (15).
Truncated proteins were detected using a combination of B9 and D1 mAb to different epitopes within WASP aa 1–250 (Santa Cruz Biotechnology). A total of 3 x 106 cells in SDS buffer was electrophoresed as described above, or 5 x 105 cells were electrophoresed on 10% Tris-glycine gels (Invitrogen Life Technologies). After overnight transfer, the polyvinylidene difluoride membrane was blocked with 2% goat or rabbit serum for 30 min at room temperature and incubated with B9/D1 mAb mixture for 2 h at room temperature followed by 1 h with rabbit anti-mouse 125I-labeled Ab and detected as described above. Alternatively, immunoreactive bands were detected by ECL using SuperSignal West Pico kit (Pierce).
RNA analysis
RNA was isolated from 107 PBMC using TRIzol reagent (Invitrogen Life Technologies), and RNA (2–8 μl) was reverse transcribed with 200 U of SuperScript II Reverse Transcriptase (Invitrogen Life Technologies), 1 μg of OligodT (Invitrogen Life Technologies), and 0.5 mM dNTP (Applied Biosystems) for 50 min at 45°C. cDNA (1 μl in 25 μl) was amplified with 1.25 U of Platinum Taq High Fidelity polymerase, 0.8 mM dNTP, 2 mM MgSO4, and 0.25 μM primers. GAPDH (BD Clontech) and WASP exons 1 to 4 (5'-ggcccaatgggaggaagg-3' and 5'-tagctggcgtctgtctcc-3') were amplified at 95°C x 2 min, and cycles of 95°C x 30 s, 57°C x 30 s, 72°C x 1 min, and extension for 10 min at 72°C. Limiting linear ranges were established by amplifying one to three serial dilutions of normal cDNA and analyzing products by densitometry (22). To compare patient and normal cDNA, GAPDH was amplified for 25 and 35 cycles and WASP for 32 and 40 cycles; dilutions of normal cDNA were amplified in parallel to verify linearity.
For qualitative analysis of alternatively spliced WASP, cDNA was amplified from exons 1 to 10 (5'-gcctcgccggagaagacaag-3' and 5'-gcaatccccaaaggtacagg-3') or exons 9 to 12 (5'-acgacttcattgaggaccag-3' and 5'-tgagtgtgaggaccaggcag-3') under the above conditions, except that 1x PCR Enhancer (Invitrogen Life Technologies) was included in the latter reaction.
Cloning and sequencing of RT-PCR products
PCR products were gel purified using Qiagen Gel Purification kit (Qiagen) and cloned with Perfectly Blunt reagents (Novagen). QIAamp miniprep plasmid reagents (Qiagen) were used to purify plasmids. Sequencing was done at Davis Sequencing using T7 and SP6 primers.
Results
WASP was analyzed in PBMC of 44 patients in 41 families, most of whom were infants and young children newly diagnosed with WAS. We used Ab W485 directed against a C-terminal epitope to detect normal WASP and variants that have an in-frame C terminus. The C-terminal region contains the functional VCA domains that initiate polymerization of new actin filaments in activated cells (reviewed in Ref. 23). Using this approach, we detected WASP in cells of a subgroup of the patients with missense mutations and a smaller subgroup of the patients with nonsense, frameshift, and splice site mutations (Fig. 1, Table I).
Discussion
To define the effects of mutations for a panel of WAS/XLT patients, we addressed three questions: whether WASP is expressed in the patient’s cells, the level of WASP expressed, and the molecular structure of the variant protein. These characteristics, collectively called proteotype, constitute the effect of the mutation at the molecular level. The overall finding is the nearly universal link of WASP genotype and proteotype. With few exceptions, patients sharing a mutation either did or did not express the same WASP protein in their PBMC and at levels that differed only within a small range. For example, for seven patients with the splice site mutation IVS 6 (+5) g>a, WASP levels in PBMC were clustered within the range 8–18% (mean, 12.7 ± 1.6%) (this study and Ref. 20) and thus readily distinguishable from most other splice site mutations, which were WASP-negative. Secondly, although the majority of patient mutations had proteotypes that could have been predicted from genotype, a substantial minority had proteotypes that were not obvious a priori. These include five missense genotypes associated with WASP-negative PBMC and two nonsense, five frameshift, and two splice site genotypes associated with WASP-positive PBMC. This significant number of nonpredicted proteotypes indicates that attempts to link genotype and clinical phenotype require knowledge of the actual molecular effect of the mutation.
The single exception to the genotype-proteotype linkage is patient W36 with WASP-negative PBMC and IVS6 (+5) g>a associated with WASP expression for seven previous patients. Whereas the seven previous patients ranged in age from 5 to 50 years, W36 is an infant of 2 years. The absence of WASP in W36 PBMC may be explained by recent findings for one of the adult patients. This study showed that WASP expression is confined to a subset of the patient’s memory T cells; his naive T cells are WASP-negative (M. I. Lutskiy et al., manuscript in preparation). For infants with WAS, although their lymphocyte numbers differ from normal, memory T cells are largely absent as in normal infants due to predominance of naive T cell populations (38). Further study of the mechanism of discordant WASP expression is needed particularly for the relatively frequent IVS 6 (+5) g>a genotype. Although these patients present with mild or moderate disease, their clinical phenotype is not really benign because the incidence of B cell lymphoma is high for these genotypes and for patients with WASP-negative proteotypes (39).
The presence of normal levels of WASP mRNA in PBMC of patients with missense mutations in the EVH1 domain implicates proteolysis as the cause of decreased/absent WASP in these cells. The genotypes with decreased/absent WASP include seven mutations that alter surface residues implicated in WIP binding, strongly suggesting that WIP binding is important for stabilization or survival of WASP. Whereas WIP is well known for its role in targeting WASP to activation sites (11, 35), additional evidence indicates that WIP functions to protect WASP in resting cells. For example, the recombinant EVH1 domain of N-WASP could not be expressed in the absence of WIP (12). Also, in resting lymphoid cells, 95% of WASP molecules are found in a 1:1 complex with WIP (11).
Current experiments did not address whether degradation of EVH1 missense WASP in PBMC was cotranslational or posttranslational or both, but the finding that cells of two patients had low- level WASP on one occasion and no WASP on another suggests that at least part of WASP degradation is posttranslational. Posttranslational proteolysis of WASP with EVH1 mutations is consistent with higher WASP levels in cells with more active protein synthesis, as found for EBV-transformed patient cell lines compared with peripheral lymphocytes (20), and is consistent also with complete degradation of WASP in platelets because platelets lack protein synthesis (20).
Despite the presence of WASP in megakaryocytes of patients with EVH1 mutations (n = 2) (40), the platelets of these patients are WASP-negative (n = 7) (Ref. 20 and this study), strongly suggesting that WASP absence is due to proteolysis as outlined above. WASP-negative platelets were also found as expected for patients with nonsense, frameshift, splice site mutations that lead to WASP-negative PBMC (n = 5) and discordant patients (n = 6) because WASP is unlikely to be synthesized in megakaryocytes with these genotypes. As a result of these two mechanisms, most WAS patients will have WASP-negative platelets, explaining the largely consistent and severe platelet defect in this disease. The finding that two exceptional patients, W21 with alternatively spliced WASP and W14 with an exon 11 missense mutation, have WASP-positive platelets suggests that their deviation from the common fate of platelets occurred because their altered WASP proteins have a normal EVH1 domain. A previous study described two additional patients with WASP-positive platelets, one of whom also has a normal EVH1 domain (exon 11 and I481N) (41). The single anomaly is a patient with the EVH1 mutation P58R who has WASP-positive platelets (41), suggesting that the mutation P58R may have only a modest negative effect on WIP binding.
WASP variants with alternative splicing and intact C-terminal domains were characterized for mutations (n = 5) in the exon 10 upstream region. These patients tend to have moderate rather than the anticipated severe disease. Most notable was patient W21, diagnosed at age 5 years, who has mild disease, suggesting that his variant protein, which lacks half of the proline-rich region (63 aa), retains some degree of normal function. Although we studied only one patient with this mutation, genotype-proteotype linkage is likely because an unrelated family with the same "nonsense" mutation includes three siblings who also have mild disease (scores of 1, 1, and 2) (14). Of note, the less dramatic ameliorating effect of the alternatively spliced WASP variant that lacks 52 aa in the same region may be due to lower expression levels. The impact of the W20 alternatively spliced WASP protein, which lacks a small region at the extreme end of the GBD, is uncertain. On the one hand, the patient’s disease is severe; in contrast, he has survived for 50 years.
Knowledge of the molecular effect of mutations provides a logical basis for correlating genotype and clinical phenotype. The value of this information was enhanced recently by two large studies that correlated WASP proteotype and clinical outcome (42, 43). A comprehensive analysis of clinical data collected longitudinally for a panel of 50 Japanese patients found that several important markers of clinical phenotype correlated with presence or absence of WASP (43). WASP negativity was associated with a 4-fold increase of susceptibility to infections (bacterial, viral, and fungal), with increased frequency of severe eczema, intestinal hemorrhage, death from intracranial bleeding and incidence of malignancies and with lower overall survival rate. The findings led to the conclusion that WASP protein expression is a useful tool for predicting long-term prognosis and that hemopoietic stem cell transplantation should be considered, especially for WASP-negative patients, while the patients are young to improve prognosis (43). The patient panel studied here includes five mutations recently identified as hotspots: T45M, R86H, IVS 6 (+5) g >1, R211X, and IVS 8 (+1) g>a (43), and thus the genotype-proteotype correlations for these five genotypes alone provide molecular information for an estimated 22% of patients.
The current findings of genotype-proteotype linkages provide a molecular basis for predicting and understanding the extremely heterogeneous clinical phenotype of this disease. Because the diagnosis of WAS encompasses highly heterogeneous clinical phenotypes, the findings for 37 genotypes suggest that a genotype-proteotype registry would provide information that would aid in predicting clinical course and planning therapy for newly diagnosed infants. Genotype-proteotype information would aid also in efforts to identify modifier genes that additionally influence disease severity.
Acknowledgments
We thank the physicians who allowed us to study their patients and provided material that was invaluable for this investigation. We thank Drs. Anna Shcherbina for providing data on T cells of discordant patients, Uta Francke for valuable advice, and Raif Geha for critical review of the manuscript. We also thank Jessica Cooley for Ab purification and assistance with blood cell isolations. We are grateful to the patients and their families and to the normal blood donors for their cooperation.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by grants from the National Institutes of Health (HL59561 and AI39574) and the Jeffrey Modell Foundation.
2 Fred S. Rosen died May 21, 2005.
3 Address correspondence and reprint requests to Dr. Eileen Remold-O’Donnell, CBR Institute for Biomedical Research, 800 Huntington Avenue, Boston, MA 02115. E-mail address: remold@cbr.med.harvard.edu
4 Abbreviations used in this paper: WAS, Wiskott-Aldrich syndrome; XLT, X-linked thrombocytopenia; WASP, WAS protein; VCA, verprolin/cofilin/acidic; GBD, GTPase binding domain; EVH1, Ena/VASP homology 1.
Received for publication January 18, 2005. Accepted for publication April 29, 2005.
References
Wiskott, A.. 1937. Familiarer, angeobren Morbus Werlhofi?. Monatschrift Kinderheil 68: 212-216.
Perry, G. S., 3rd, B. D. Spector, L. M. Schuman, J. S. Mandel, V. E. Anderson, R. B. McHugh, M. R. Hanson, S. M. Fahlstrom, W. Krivit, J. H. Kersey. 1980. The Wiskott-Aldrich syndrome in the United States and Canada (1892–1979). J. Pediatr. 97: 72-78.
Sullivan, K. E., C. A. Mullen, R. M. Blaese, J. A. Winkelstein. 1994. A multiinstitutional survey of the Wiskott-Aldrich syndrome. J. Pediatr. 125: 876-885.
Zhu, Q., C. Watanabe, T. Liu, D. Hollenbaugh, R. M. Blaese, S. B. Kanner, A. Aruffo, H. D. Ochs. 1997. Wiskott-Aldrich syndrome/X-linked thrombocytopenia: WASP gene mutations, protein expression, and phenotype. Blood 90: 2680-2689.
Ochs, H. D., F. S. Rosen. 1999. The Wiskott-Aldrich syndrome. H. D. Ochs, 3rd, and C. I. E. Smith, 3rd, and J. M. Puck, 3rd, eds. Primary Immunodeficiency Diseases 292-305 Oxford University Press, New York. .
Filipovich, A. H., J. V. Stone, S. C. Tomany, M. Ireland, C. Kollman, C. J. Pelz, J. T. Casper, M. J. Cowan, J. R. Edwards, A. Fasth, et al 2001. Impact of donor type on outcome of bone marrow transplantation for Wiskott-Aldrich syndrome: collaborative study of the International Bone Marrow Transplant Registry and the National Marrow Donor Program. Blood 97: 1598-1603.
Derry, J. M., H. D. Ochs, U. Francke. 1994. Isolation of a novel gene mutated in Wiskott-Aldrich syndrome. Cell 78: 635-644.(Maxim I. Lutskiy, Fred S.)
Wiskott-Aldrich syndrome (WAS) is a platelet/immunodeficiency disease arising from mutations of WAS protein (WASP), a hemopoietic cytoskeletal protein. Clinical symptoms vary widely from mild (X-linked thrombocytopenia) to life threatening. In this study, we examined the molecular effects of individual mutations by quantifying WASP in peripheral lymphocytes of 44 patients and identifying the molecular variant (collectively called proteotype). Nonpredicted proteotypes were found for 14 genotypes. These include WASP-negative lymphocytes found for five missense genotypes and WASP-positive lymphocytes for two nonsense, five frameshift, and two splice site genotypes. Missense mutations in the Ena/VASP homology 1 (EVH1) domain lead to decreased/absent WASP but normal mRNA levels, indicating that proteolysis causes the protein deficit. Because several of the EVH1 missense mutations alter WIP binding sites, the findings suggest that abrogation of WIP binding induces proteolysis. Whereas platelets of most patients were previously shown to lack WASP, WASP-positive platelets were found for two atypical patients, both of whom have mutations outside the EVH1 domain. WASP variants with alternative splicing and intact C-terminal domains were characterized for eight nonsense and frameshift genotypes. One of these, a nonsense genotype in a mild patient, supports expression of WASP lacking half of the proline-rich region. With one notable exception, genotype and proteotype were linked, indicating that a genotype-proteotype registry could be assembled to aid in predicting disease course and planning therapy for newly diagnosed infants. Knowledge of the molecular effect of mutations would aid also in identifying disease-modifying genes.
Introduction
The Wiskott-Aldrich syndrome (WAS)4 is a X-linked platelet and immune deficiency disease identified in 1937 in three infant brothers who presented with bloody diarrhea, decreased resistance to infections, eczema, and thrombocytopenia (1). In subsequent decades, the disease was considered lethal in the first decade of life, either from hemorrhage or overwhelming infection (2). Currently, diagnosed patients show remarkable heterogeneity of clinical symptoms, ranging from isolated thrombocytopenia (X-linked thrombocytopenia; XLT) to severe disease including life-threatening pyogenic and opportunistic infections, autoimmune disease, and malignancy, particularly B cell lymphoma in addition to thrombocytopenia present in all patients (3, 4, 5). T cell immunity is principally affected and might reflect the extent of WAS protein (WASP) depletion in cells. Symptoms of immune dysfunction appear shortly after birth in some patients, but are delayed in other patients by 2, 3, or more years (3). This disparity in time of onset of immune dysfunction, the cause of which is unknown, complicates diagnosis and delays treatment. The only definitive cure, bone marrow transplantation, is most effective if done before age 5 years (6).
The defective gene WASP encodes a multidomain blood cell protein (7) that regulates cytoskeletal rearrangements in activated cells. The multiple domains integrate signals that induce the following: 1) WASP recruitment to activation sites (e.g., immune synapse); and 2) conformational activation by exposure of the C-terminal VCA (verprolin/cofilin/acidic) domains that catalyze nucleation of actin filaments (8, 9). WASP is found in the cytoplasm of resting cells in the inactive autoinhibited conformation in which the VCA domains are bound to the central GTPase binding domain (GBD) (10). The resting cell molecule is further stabilized by binding of WIP to the large N-terminal Ena/VASP homology 1 (EVH1) domain (11, 12).
More than 100 patient mutations have been described. Early reports associated missense mutations, which cluster in the EVH1 domain with XLT or mild immune disease (4, 13, 14) and frameshift, nonsense, and splice site mutations with severe disease (4, 14, 15, 16). However, this linkage was not seen in other studies (17, 18), and many exceptions are known. Nonetheless, the association of many cases of mild clinical course with missense mutations that are expected to support WASP expression strongly suggested the importance of WASP expression as a determinant of disease severity.
In this study, we explore the hypothesis that the patient’s underlying mutation determines whether WASP is expressed in his cells, the level of expression, and molecular structure of the protein. These characteristics, collectively called "proteotype," are not always known based on genotype alone. The hypothesis predicts that WASP proteotype, once determined for a patient with a particular genotype, will be found for other patients with the same genotype, even when clinical course differs. To test the hypothesis, we examined WASP in PBMC of 44 patients newly diagnosed with WAS/XLT using a standardized Western blot. Molecular structures were defined by RNA analysis. The findings are analyzed in conjunction with published data on previous patients.
Materials and Methods
Patients
Patients were diagnosed with WAS/XLT based on male sex, thrombocytopenia with small platelets, eczema, immunodeficiency of variable clinical severity, family history, and, in some cases, nonrandom X chromosome inactivation of the mother’s blood cells. Most patients were newly diagnosed and <5 years old at the time of diagnosis. Based on the presence or absence of eczema, recurrent infections, autoimmune and malignant diseases in addition to thrombocytopenia, the patients’ disease was graded as mild, moderate, or severe. If only thrombocytopenia with or without mild eczema were present, the disease was considered mild. Thrombocytopenia, severe eczema, or frequent infections were graded as moderate. Thrombocytopenia, severe eczema, and severe frequent infections and/or autoimmune or malignant disease were classified as severe.
Platelet sizing
Platelet size was determined on platelet-rich plasma obtained by gravity sedimentation of blood cells from aliquots of acid-citrate dextran-anticoagulated blood. Using an Elzone Particle Analyzer Model 112 (Particle Data), the size of the platelets was measured relative to standardized latex spheres (19).
DNA isolation, amplification, and sequencing
Blood samples were obtained with informed consent under protocols approved by the Institutional Review Board of the CBR Institute for Biomedical Research. Blood of patients and paired normal individuals was collected in acid-citrate-dextrose (National Institutes of Health formula A) and processed immediately or after overnight shipment at ambient temperature. PBMC were isolated as described previously (20). Aliquots of whole blood were frozen as a source of DNA, which was isolated using QIAamp reagents (Qiagen). DNA for two patients was isolated from cryopreserved PBMC; DNA for another patient was isolated from a detergent-solubilized lysate of PBMC. WASP exons were amplified and sequenced as described above (21) by the Jeffrey Modell National Center for Primary Immunodeficiency Diagnosis at the CBR Institute for Biochemical Research.
Western blot analysis
Lysates of PBMC at 15 x 106/ml were prepared in buffer containing SDS and the protease inhibitors diisopropylfluorophosphate (2 mM) and leupeptin (25 μg/ml) (Sigma-Aldrich) as described previously (20). Lysates of 3 x 106 cells were electrophoresed on 8% acrylamide gels (14 cm x 10 cm x 1.5 mm) under reducing conditions, and the proteins were transferred to polyvinylidene difluoride membrane (Millipore) for staining with 100 ng/ml affinity-purified rabbit Ab to the WASP C-terminal 15-mer peptide (W485) as described previously (20). The reactive bands were detected by 125I-labeled secondary Ab and quantified relative to parallel dilutions of normal PBMC with the Phosphorimager Storm 860 Imager and ImageQuant v.1.1 Program (Molecular Dynamics). Except where noted, quantitation is based on duplicate analysis of two independent cell isolates. Linearity of the analysis of the responses was demonstrated previously for the range 2–100% (15).
Truncated proteins were detected using a combination of B9 and D1 mAb to different epitopes within WASP aa 1–250 (Santa Cruz Biotechnology). A total of 3 x 106 cells in SDS buffer was electrophoresed as described above, or 5 x 105 cells were electrophoresed on 10% Tris-glycine gels (Invitrogen Life Technologies). After overnight transfer, the polyvinylidene difluoride membrane was blocked with 2% goat or rabbit serum for 30 min at room temperature and incubated with B9/D1 mAb mixture for 2 h at room temperature followed by 1 h with rabbit anti-mouse 125I-labeled Ab and detected as described above. Alternatively, immunoreactive bands were detected by ECL using SuperSignal West Pico kit (Pierce).
RNA analysis
RNA was isolated from 107 PBMC using TRIzol reagent (Invitrogen Life Technologies), and RNA (2–8 μl) was reverse transcribed with 200 U of SuperScript II Reverse Transcriptase (Invitrogen Life Technologies), 1 μg of OligodT (Invitrogen Life Technologies), and 0.5 mM dNTP (Applied Biosystems) for 50 min at 45°C. cDNA (1 μl in 25 μl) was amplified with 1.25 U of Platinum Taq High Fidelity polymerase, 0.8 mM dNTP, 2 mM MgSO4, and 0.25 μM primers. GAPDH (BD Clontech) and WASP exons 1 to 4 (5'-ggcccaatgggaggaagg-3' and 5'-tagctggcgtctgtctcc-3') were amplified at 95°C x 2 min, and cycles of 95°C x 30 s, 57°C x 30 s, 72°C x 1 min, and extension for 10 min at 72°C. Limiting linear ranges were established by amplifying one to three serial dilutions of normal cDNA and analyzing products by densitometry (22). To compare patient and normal cDNA, GAPDH was amplified for 25 and 35 cycles and WASP for 32 and 40 cycles; dilutions of normal cDNA were amplified in parallel to verify linearity.
For qualitative analysis of alternatively spliced WASP, cDNA was amplified from exons 1 to 10 (5'-gcctcgccggagaagacaag-3' and 5'-gcaatccccaaaggtacagg-3') or exons 9 to 12 (5'-acgacttcattgaggaccag-3' and 5'-tgagtgtgaggaccaggcag-3') under the above conditions, except that 1x PCR Enhancer (Invitrogen Life Technologies) was included in the latter reaction.
Cloning and sequencing of RT-PCR products
PCR products were gel purified using Qiagen Gel Purification kit (Qiagen) and cloned with Perfectly Blunt reagents (Novagen). QIAamp miniprep plasmid reagents (Qiagen) were used to purify plasmids. Sequencing was done at Davis Sequencing using T7 and SP6 primers.
Results
WASP was analyzed in PBMC of 44 patients in 41 families, most of whom were infants and young children newly diagnosed with WAS. We used Ab W485 directed against a C-terminal epitope to detect normal WASP and variants that have an in-frame C terminus. The C-terminal region contains the functional VCA domains that initiate polymerization of new actin filaments in activated cells (reviewed in Ref. 23). Using this approach, we detected WASP in cells of a subgroup of the patients with missense mutations and a smaller subgroup of the patients with nonsense, frameshift, and splice site mutations (Fig. 1, Table I).
Discussion
To define the effects of mutations for a panel of WAS/XLT patients, we addressed three questions: whether WASP is expressed in the patient’s cells, the level of WASP expressed, and the molecular structure of the variant protein. These characteristics, collectively called proteotype, constitute the effect of the mutation at the molecular level. The overall finding is the nearly universal link of WASP genotype and proteotype. With few exceptions, patients sharing a mutation either did or did not express the same WASP protein in their PBMC and at levels that differed only within a small range. For example, for seven patients with the splice site mutation IVS 6 (+5) g>a, WASP levels in PBMC were clustered within the range 8–18% (mean, 12.7 ± 1.6%) (this study and Ref. 20) and thus readily distinguishable from most other splice site mutations, which were WASP-negative. Secondly, although the majority of patient mutations had proteotypes that could have been predicted from genotype, a substantial minority had proteotypes that were not obvious a priori. These include five missense genotypes associated with WASP-negative PBMC and two nonsense, five frameshift, and two splice site genotypes associated with WASP-positive PBMC. This significant number of nonpredicted proteotypes indicates that attempts to link genotype and clinical phenotype require knowledge of the actual molecular effect of the mutation.
The single exception to the genotype-proteotype linkage is patient W36 with WASP-negative PBMC and IVS6 (+5) g>a associated with WASP expression for seven previous patients. Whereas the seven previous patients ranged in age from 5 to 50 years, W36 is an infant of 2 years. The absence of WASP in W36 PBMC may be explained by recent findings for one of the adult patients. This study showed that WASP expression is confined to a subset of the patient’s memory T cells; his naive T cells are WASP-negative (M. I. Lutskiy et al., manuscript in preparation). For infants with WAS, although their lymphocyte numbers differ from normal, memory T cells are largely absent as in normal infants due to predominance of naive T cell populations (38). Further study of the mechanism of discordant WASP expression is needed particularly for the relatively frequent IVS 6 (+5) g>a genotype. Although these patients present with mild or moderate disease, their clinical phenotype is not really benign because the incidence of B cell lymphoma is high for these genotypes and for patients with WASP-negative proteotypes (39).
The presence of normal levels of WASP mRNA in PBMC of patients with missense mutations in the EVH1 domain implicates proteolysis as the cause of decreased/absent WASP in these cells. The genotypes with decreased/absent WASP include seven mutations that alter surface residues implicated in WIP binding, strongly suggesting that WIP binding is important for stabilization or survival of WASP. Whereas WIP is well known for its role in targeting WASP to activation sites (11, 35), additional evidence indicates that WIP functions to protect WASP in resting cells. For example, the recombinant EVH1 domain of N-WASP could not be expressed in the absence of WIP (12). Also, in resting lymphoid cells, 95% of WASP molecules are found in a 1:1 complex with WIP (11).
Current experiments did not address whether degradation of EVH1 missense WASP in PBMC was cotranslational or posttranslational or both, but the finding that cells of two patients had low- level WASP on one occasion and no WASP on another suggests that at least part of WASP degradation is posttranslational. Posttranslational proteolysis of WASP with EVH1 mutations is consistent with higher WASP levels in cells with more active protein synthesis, as found for EBV-transformed patient cell lines compared with peripheral lymphocytes (20), and is consistent also with complete degradation of WASP in platelets because platelets lack protein synthesis (20).
Despite the presence of WASP in megakaryocytes of patients with EVH1 mutations (n = 2) (40), the platelets of these patients are WASP-negative (n = 7) (Ref. 20 and this study), strongly suggesting that WASP absence is due to proteolysis as outlined above. WASP-negative platelets were also found as expected for patients with nonsense, frameshift, splice site mutations that lead to WASP-negative PBMC (n = 5) and discordant patients (n = 6) because WASP is unlikely to be synthesized in megakaryocytes with these genotypes. As a result of these two mechanisms, most WAS patients will have WASP-negative platelets, explaining the largely consistent and severe platelet defect in this disease. The finding that two exceptional patients, W21 with alternatively spliced WASP and W14 with an exon 11 missense mutation, have WASP-positive platelets suggests that their deviation from the common fate of platelets occurred because their altered WASP proteins have a normal EVH1 domain. A previous study described two additional patients with WASP-positive platelets, one of whom also has a normal EVH1 domain (exon 11 and I481N) (41). The single anomaly is a patient with the EVH1 mutation P58R who has WASP-positive platelets (41), suggesting that the mutation P58R may have only a modest negative effect on WIP binding.
WASP variants with alternative splicing and intact C-terminal domains were characterized for mutations (n = 5) in the exon 10 upstream region. These patients tend to have moderate rather than the anticipated severe disease. Most notable was patient W21, diagnosed at age 5 years, who has mild disease, suggesting that his variant protein, which lacks half of the proline-rich region (63 aa), retains some degree of normal function. Although we studied only one patient with this mutation, genotype-proteotype linkage is likely because an unrelated family with the same "nonsense" mutation includes three siblings who also have mild disease (scores of 1, 1, and 2) (14). Of note, the less dramatic ameliorating effect of the alternatively spliced WASP variant that lacks 52 aa in the same region may be due to lower expression levels. The impact of the W20 alternatively spliced WASP protein, which lacks a small region at the extreme end of the GBD, is uncertain. On the one hand, the patient’s disease is severe; in contrast, he has survived for 50 years.
Knowledge of the molecular effect of mutations provides a logical basis for correlating genotype and clinical phenotype. The value of this information was enhanced recently by two large studies that correlated WASP proteotype and clinical outcome (42, 43). A comprehensive analysis of clinical data collected longitudinally for a panel of 50 Japanese patients found that several important markers of clinical phenotype correlated with presence or absence of WASP (43). WASP negativity was associated with a 4-fold increase of susceptibility to infections (bacterial, viral, and fungal), with increased frequency of severe eczema, intestinal hemorrhage, death from intracranial bleeding and incidence of malignancies and with lower overall survival rate. The findings led to the conclusion that WASP protein expression is a useful tool for predicting long-term prognosis and that hemopoietic stem cell transplantation should be considered, especially for WASP-negative patients, while the patients are young to improve prognosis (43). The patient panel studied here includes five mutations recently identified as hotspots: T45M, R86H, IVS 6 (+5) g >1, R211X, and IVS 8 (+1) g>a (43), and thus the genotype-proteotype correlations for these five genotypes alone provide molecular information for an estimated 22% of patients.
The current findings of genotype-proteotype linkages provide a molecular basis for predicting and understanding the extremely heterogeneous clinical phenotype of this disease. Because the diagnosis of WAS encompasses highly heterogeneous clinical phenotypes, the findings for 37 genotypes suggest that a genotype-proteotype registry would provide information that would aid in predicting clinical course and planning therapy for newly diagnosed infants. Genotype-proteotype information would aid also in efforts to identify modifier genes that additionally influence disease severity.
Acknowledgments
We thank the physicians who allowed us to study their patients and provided material that was invaluable for this investigation. We thank Drs. Anna Shcherbina for providing data on T cells of discordant patients, Uta Francke for valuable advice, and Raif Geha for critical review of the manuscript. We also thank Jessica Cooley for Ab purification and assistance with blood cell isolations. We are grateful to the patients and their families and to the normal blood donors for their cooperation.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by grants from the National Institutes of Health (HL59561 and AI39574) and the Jeffrey Modell Foundation.
2 Fred S. Rosen died May 21, 2005.
3 Address correspondence and reprint requests to Dr. Eileen Remold-O’Donnell, CBR Institute for Biomedical Research, 800 Huntington Avenue, Boston, MA 02115. E-mail address: remold@cbr.med.harvard.edu
4 Abbreviations used in this paper: WAS, Wiskott-Aldrich syndrome; XLT, X-linked thrombocytopenia; WASP, WAS protein; VCA, verprolin/cofilin/acidic; GBD, GTPase binding domain; EVH1, Ena/VASP homology 1.
Received for publication January 18, 2005. Accepted for publication April 29, 2005.
References
Wiskott, A.. 1937. Familiarer, angeobren Morbus Werlhofi?. Monatschrift Kinderheil 68: 212-216.
Perry, G. S., 3rd, B. D. Spector, L. M. Schuman, J. S. Mandel, V. E. Anderson, R. B. McHugh, M. R. Hanson, S. M. Fahlstrom, W. Krivit, J. H. Kersey. 1980. The Wiskott-Aldrich syndrome in the United States and Canada (1892–1979). J. Pediatr. 97: 72-78.
Sullivan, K. E., C. A. Mullen, R. M. Blaese, J. A. Winkelstein. 1994. A multiinstitutional survey of the Wiskott-Aldrich syndrome. J. Pediatr. 125: 876-885.
Zhu, Q., C. Watanabe, T. Liu, D. Hollenbaugh, R. M. Blaese, S. B. Kanner, A. Aruffo, H. D. Ochs. 1997. Wiskott-Aldrich syndrome/X-linked thrombocytopenia: WASP gene mutations, protein expression, and phenotype. Blood 90: 2680-2689.
Ochs, H. D., F. S. Rosen. 1999. The Wiskott-Aldrich syndrome. H. D. Ochs, 3rd, and C. I. E. Smith, 3rd, and J. M. Puck, 3rd, eds. Primary Immunodeficiency Diseases 292-305 Oxford University Press, New York. .
Filipovich, A. H., J. V. Stone, S. C. Tomany, M. Ireland, C. Kollman, C. J. Pelz, J. T. Casper, M. J. Cowan, J. R. Edwards, A. Fasth, et al 2001. Impact of donor type on outcome of bone marrow transplantation for Wiskott-Aldrich syndrome: collaborative study of the International Bone Marrow Transplant Registry and the National Marrow Donor Program. Blood 97: 1598-1603.
Derry, J. M., H. D. Ochs, U. Francke. 1994. Isolation of a novel gene mutated in Wiskott-Aldrich syndrome. Cell 78: 635-644.(Maxim I. Lutskiy, Fred S.)