- Email: [email protected]
Molecular Modeling of the Jak3 Kinase Domains and Structural Basis for Severe Combined Immunodeficiency
Clinical Immunology Vol. 96, No. 2, August, pp. 108 –118, 2000 doi:10.1006/clim.2000.4880, available online at http://www.idealibrary.com on
Molecular Modeling of the Jak3 Kinase Domains and Structural Basis for Severe Combined Immunodeficiency Mauno Vihinen,* Anna Villa,† Patrizia Mella,‡ R. Fabian Schumacher,‡ Gianfranco Savoldi,‡ John J. O’Shea,§ Fabio Candotti,‡ ,1 and Luigi D. Notarangelo‡ *Institute of Medical Technology, FIN-33014 University of Tampere, Finland, and Tampere University Hospital, FIN-33520 Tampere, Finland; †Istituto di Tecnologie Biomediche Avanzate, Consiglio Nazionale della Ricerche, 20090 Segrate (MI), Italy; ‡Istituto di Medicina Molecolare “Angelo Nocivelli,” Clinica Pediatrica, Universita` di Brescia, Spedali Civili, Brescia, Italy; and §Lymphocyte Cell Biology Section, Arthritis and Rheumatism Branch, NIAMSD, NIH, Bethesda, Maryland 20892-1820
Hereditary severe combined immunodeficiency (SCID) includes a heterogeneous group of diseases that profoundly affect both cellular and humoral immune responses and require treatment by bone marrow transplantation. Characterization of the cellular and molecular bases of SCID is essential to provide accurate genetic counseling and prenatal diagnosis, and it may offer the grounds for alternative forms of treatment. The Jak3 gene is mutated in most cases of autosomal recessive T ⴚB ⴙ SCID in humans. Jak3 belongs to the family of intracellular Janus tyrosine kinases. It is physically and functionally coupled to the common ␥ chain, ␥c, shared by several cytokine receptors. We have established the JAK3base registry for disease and mutation information. In order to study the structural consequences of the Jak3 mutations, the structure of the human Jak3 kinase and pseudokinase domains was modeled. Residues involved in ATP and Mg 2ⴙ binding were highly conserved in the kinase domain whereas the substrate binding region is somewhat different compared to other kinases. We have identified the first naturally occurring mutations disrupting the function of the human Jak3 kinase domain. The structural basis of all of the known Jak3 mutations reported so far is discussed based on the modeled structure. The model of the Jak3 protein also permits us to study Jak3 phosphorylation at the structural level and may thus serve in the design of novel immune suppressive drugs. © 2000 Academic Press Key Words: immunodeficiency; B cells; T cells; disease-causing mutations; structural basis of disease; structure–function relationships; JAK3base.
INTRODUCTION
Severe combined immunodeficiency (SCID) includes a group of hereditary diseases that severely affect both 1 Present address: Clinical Gene Therapy Branch, NHGRI, NIH, Bethesda, MD 20892.
1521-6616/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
cellular and humoral-based immune defenses (1). Unless treated with bone marrow transplantation, SCID patients die early in life due to severe infections. Characterization of the cellular and molecular bases of SCID is essential to provide accurate genetic counseling and prenatal diagnosis, and it may offer the basis for alternative forms of treatment, based on gene transfer. The most common form of SCID in humans is inherited as an X-linked trait (SCIDX1, OMIM 308380) and is due to mutations of the IL2RG gene (2), encoding for the common gamma chain (␥c), shared by cytokine receptors for interleukin- (IL-) 2, IL-4, IL-7, IL-9, and IL-15 (for a review, see Ref. 3). The immunological phenotype of SCIDX1 is characterized by the lack of circulating T (and NK) cells, with normal to increased numbers of B lymphocytes (T ⫺B ⫹ SCID) (1). The same phenotype, however, may also be observed in autosomal variants of SCID. We and others have shown that in most cases of autosomal recessive T ⫺B ⫹ SCID, the Jak3 gene, encoding for an intracellular Janus tyrosine kinase that is physically and functionally coupled to ␥c, is mutated (OMIM 600170) (4, 5). Jak3 is a member of the family of Janus kinases (Jaks), which in humans also includes Jak1, Jak2, and Tyk2 (6, 7). These proteins share seven conserved regions called Jak homology (JH) domains. One of the earliest events that follows binding of IL-2, IL-4, IL-7, IL-9, or IL-15 to the specific receptor is represented by phosphorylation of tyrosine residues of both Jak3 (associated with the ␥c) and Jak1 proteins (coupled to the major transducing subunit of the cytokine receptor) (6 – 8). Once activated, Jaks tyrosine-phosphorylate other proteins including cytokine receptor subunits, generating docking sites for the SH2 domains of STATs (signal transducers and activators of transcription). This leads to dimerization of the STATs and their translocation to the nucleus where they bind enhancer regions of DNA, thereby inducing the transcription of cytokine responsive genes (7).
108
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
The crucial role for intact ␥c-Jak3 expression and function is best illustrated by the occurrence of T ⫺B ⫹ SCID in humans with genetic defects of either of these genes and by similar observations in gene-targeted mice, although in mice the disruption of the IL2RG (9, 10) or of the Jak3 (11–13) genes also strongly affects B cell differentiation. Despite the essential role played by Jak3 in lymphoid development and function, little is known about the structural basis for its activity. The JH7 and JH6 domains of Jak3 have been shown to mediate interaction with the ␥c (14). Furthermore, Jak3 has several phosphorylation sites, and tyrosinephosphorylation at these residues may regulate the activity of the enzyme positively or negatively (15, 16). In the present paper, we report the three-dimensional structure of the human Jak3 kinase (JH1) and pseudokinase (JH2) domains. These models were used to assess the structural consequences of the Jak3 mutations thus far described in humans, including the first mutations in the kinase domain that we have recently reported (17). A disease and mutation database (JAK3base) has been developed, which includes information on the clinical and immunological phenotype of Jak3-deficient SCID patients and on the structural consequences of their respective Jak3 gene defects. This information could be of particular interest for understanding the structural basis of Jak3-deficient SCID. Furthermore, a better appreciation of the functional role played by distinct structural motifs in the Jak3 protein may help develop specific Jak3 inhibitors as a novel class of immune suppressive drugs. MATERIALS AND METHODS
Mutation Analysis Detection of mutations in the Jak3 gene in patients with T ⫺B ⫹ SCID was accomplished on genomic DNA by single-strand conformation polymorphism (SSCP) and DNA sequencing as described elsewhere (17). Molecular Modeling
109
The insertions and deletions were modeled by searching loops from a database, which contained either most of the PDB structures or an unbiased selection of PDB (23). The search was performed for fragments of the required length and end-point separation by using some residues at each end of the loop as anchor points. Fragments obtained were evaluated on three criteria: root mean square deviation from the anchor points, sequence similarity, and interference with the core region. The models were refined by energy minimization with the program Discover in stepwise manner using Amber force field. First, hydrogen atoms were relaxed, and then the side chains of the insertions and deletions were relaxed and the rest of the molecule was fixed. In the next step the borders of insertions and deletions and the C ␣ atoms of the conserved regions, and finally only the C ␣ atoms of the conserved regions, were harmonically constrained. JAK3base A program was developed to allow submission of clinical, immunological, and molecular data on Jak3deficient SCID patients to the JAK3base registry. The registry was built according to the guidelines adopted for the BTKbase (24), and its content has been made available through the Internet by using the World Wide Web. The database contains four main items: identification of the patient by means of a specific code related to the Jak3 mutation(s), reference either to published article(s) or to the submitting physician, mutation information, and data related to the clinical and immunological phenotype and to treatment. All of these items are organized as fields, which are easily accessible. Protection of the patients’ identity is assured in the registry. Patient data form an entry, which can be analyzed with the provided tools or with the sequence retrieval system (SRS) (25). RESULTS
Mutation Analysis in Jak3 Deficiency The kinase and pseudokinase domains of Jak3 were modeled based on the structure of the Src kinase at 1.7 Å resolution (18) (Protein Data Bank (PDB) (19) entry 1fmk). The Jak3 kinase domain was further modeled based on the Irk structure (20; PDB code 1ir3). The sequence alignment was performed with GCG (21) and MULTICOMP (22) program packages. The final alignment was performed using information from multiple sequence analysis and secondary structural information from the three-dimensional structures. The model was built with the program InsightII (Molecular Simulations, Inc., San Diego, CA). A side chain rotamer library was used to model amino acid substitutions.
In addition to previously published mutations, two novel Jak3 gene defects were identified in a 2-year-old female child (patient 13 in Table 1) with a mild phenotype, characterized by recurrent respiratory tract infections and persistent candidiasis, which did not affect her growth. Immunological investigations revealed reduced, but not absent, circulating T cells (CD3: 14.5%, CD4: 7.2%, CD8: 6.2%), an excess of B lymphocytes (CD19: 83.4%), absent NK cells (CD16: 0.7%), and lack of proliferative responses to phytohemagglutinin (0.8 vs 37 cpm ⫻ 10 ⫺3). Molecular analysis at highly polymorphic DNA loci ruled out the possibility that the
110
VIHINEN ET AL.
TABLE 1 Molecular and Structural Basis of Jak3-Deficient SCID Patient
Allele
Genomic mutation
Protein alteration
Domain
1 2
1,2 1 2 1,2 1 2 1,2 1 2 1,2 1,2 1 2 1,2 1 2 1,2 1 2
A394G C574G ? C1428T C1428T T2370C IVS9,⫺2a 3 g A1537G Del 1536–1582 C1839T C1862T G2259A ? IVS16,⫹2 t 3 c T2824C C3164A IVS21,⫹2 t 3 a G268C Del C 3167
Y100C P151R ? R445X R445X C759R Frameshift after codon E481 E481G Del aa 482–596 R582W Del aa 586–592 V722I ? Frameshift after codon K733 L910S Y1023X Not known A58P Frameshift after codon 1024
JH7 JH7 ? JH3 JH3 JH2 JH4 JH3 JH3 JH2 JH2 JH2 ? JH2-JH1 JH1 JH1 JH1 JH7 JH1
3 4 5 6 7 8 9 10 11 12 13
residual T cells could be of maternal origin. Use of the SSCP-based screening assay, followed by direct sequencing, revealed that the patient is a compound heterozygote for two mutations in exon 1 (G268C, resulting in A58P amino acid change) and in exon 21 (deletion of C3167, resulting in frameshift after codon 1024, with premature termination at codon 1037). The molecular data of all of the patients with Jak3 gene defects identified at our center are reported in Table 1. Structural Model of Jak3 Kinase and Pseudokinase Domains To address the structural features of Jak3 and Jak3deficient SCID, the three-dimensional structure of the kinase and pseudokinase domains were modeled based on Src (18). Further, the effect of phosphorylation was studied by modeling the Jak3 kinase domain based on insulin receptor kinase (Irk) in its closed, active conformation (20). The sequence alignment was based on multiple sequence analysis, hallmark residues, and locations of secondary structural elements (Fig. 1). The kinase domain of Jak3 has six insertions and one deletion compared to the Src kinase domain. One of the insertions is in the activation loop and two are located in the kinase insert region; both of these functionally important regions are highly variable with respect to length and amino acid sequence. These three loops were not modeled and coordinates are missing also for the activation loop of the Src crystal structure (18). The sequence identity of the Jak3 kinase and pseudokinase domains with the Src kinase is 36 and 23%, respec-
Effect
Ref.
17 Truncation Truncation Sterical clash Truncation
17
Fold alteration Altered ligand interaction Fold alteration Sterical clash in core Truncation Structural alterations Truncation
Truncation
tively. The Jak3 kinase domain is 34% identical with the Irk catalytic domain. The catalytic site amino acids as well as some of the conserved ATP-binding residues are missing from the structure of the pseudokinase domain. Because of the high similarity, the Jak3 kinase domain was modeled based both on Irk and on Src structures, and the Jak3 JH2 domain was modeled based on Src (Fig. 2). Compared to the templates, the major differences in the Jak3 model are in the the exposed surface loops. The upper lobe is twisted in a different way in the two JH1 domain models. The models of the insertions and deletions are rather reliable, because the length of the changes does not exceed four residues except for one insertion in the upper lobe of the pseudokinase model. The final models have typical globular structures according to several structural tests (26). Both the kinase and the pseudokinase models have a five-stranded anti-parallel -sheet in the upper lobe and a mostly ␣-helical C-terminal lobe. The upper domain of both the kinase and the pseudokinase domains, responsible for ATP binding, is highly conserved. The glycine-rich loop and the other residues in the upper domain that establish contacts with ATP are conserved in the Jak3 JH1, but not in the JH2, domain. The main difference between Src and Jak3 was observed in the kinase insert region (Fig. 1), which shows remarakable variability among different kinases. The hallmark residues of protein kinases were conserved in the kinase domain, whereas in the pseudokinase more alterations were noticed (Fig. 1).
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
111
FIG. 1. Sequence alignment of the Jak3 kinase and pseudokinase domain with cAPK and Irk sequences. Residues involved in the catalytic site of Irk are indicated with # and in ATP and Mg 2⫹ binding with @. The amino acids of binding sites -1, 1, and 3 are are denoted by corresponding numbers. The secondary structures are shown below the sequence with ␣ for ␣-helices and with  for -strands. The numbers are for secondary structural elements. The sites of the SCID-causing mutations are shown below the sequence (# denotes the start of deletion and X nonsense mutation).
JAK3base All of the reported SCID-causing Jak3 mutations (4, 5, 17, 28 –30) were collected into a database called JAK3base. The reported mutations are scattered into several domains; for some of them, the functional consequences have been previously investigated (29 –32). In addition to mutations (Table 1), the registry also contains information about immunoglobulin levels, lymphocyte counts, age at diagnosis, symptoms, mRNA
and protein data, putative structural consequences of the mutations, and details on therapy including hematopoietic stem cell transplantation, when available. The registry is constructed according to the concepts used in BTKbase, XLA mutation registry (24), by using the MUTbase system (33). The JAK3base is freely accessible through the World Wide Web at http://www.uta.fi/imt/bioinfo/JAK3base.html. Researchers are encouraged to send their patient and mutation information to the registry.
112
VIHINEN ET AL.
The JAK3base also provides additional information, such as alterations in restriction enzyme patterns due to mutations and mutation descriptive statistics. DISCUSSION
Mutations in tyrosine kinases are responsible for several genetic diseases in humans; disease-causing mutations in the kinase domains are accessible through the KinMutBase (http://www.uta.fi/imt/ bioinfo/KinMutBase.html) (34). Three-dimensional structural models may provide new insights into the effects of specific mutations. The crystal structures of Irk (35) and fibroblast growth factor receptor kinase (36), and the models of Btk kinase domain (37– 42), have been used to interpret the structural consequences of the defects in the respective genes. Here, the mutations in the human Jak3 gene have been explained with the models of the three-dimensional structure of Jak3 kinase and pseudokinase regions (Fig. 3). Many of the patients have consanguineous parents and are therefore homozygous for the gene defect. We have recently identified the first Jak3-deficient patient (patient 11 in Table 1) with a missense mutation in the kinase domain (17). Mutation L910S at the beginning of ␣-helix D affects the protein core where the side chain is pointing. The corresponding residue in Irk is involved in recognition of the residue prior to phosphotyrosine (P-1) of the substrate (20). Introduction of the polar hydroxyl group might require neutralisation by, e.g., water molecules and lead to structural and functional alterations due to sterical reasons. The other mutant allele in the patient (Y1023X) deletes the major part of the kinase domain. Patient 12 in Table 1 is homozygous for a splice-site mutation in intron 21; this mutation results in exon skipping, leading to a frameshift mutation in the cDNA and to premature termination of the protein chain. Most recently, we have identified a third patient (patient 13 in Table 1) with a milder phenotype, who is a compound heterozygote for a missense mutation in the JH7 domain (A58P) and a single nucleotide deletion (del C 3167) in exon 21 of the Jak3 gene; the last mutation results in a frameshift in the kinase domain after codon 1024, with premature termination at codon 1037 and loss of C-terminal parts of the kinase domain leading to impaired kinase activity. A number of additional Jak3 mutations have been previously identified, but in most cases their structural importance and function have not been investigated in detail. With the generated structural models we were able to predict the putative consequences of all of the disease-causing alterations. Although the pseudokinase JH2 domain is catalytically inactive, it is obviously involved in regulating Jak3 function, as muta-
tions in the JH2 domain have been shown to alter the function of various Jak proteins (43, 44). Moreover, it has been shown that the JH2 domain of Jaks may directly interact with STATs (45). More recently, direct interaction between the JH1 (kinase) and JH2 (pseudokinase) domains of the Jak3 protein has been demonstrated (32); this interaction appears to inhibit Jak3 catalytic activity, presumably by interfering with substrate access to the kinase domain. Evaluation of the structural consequences of SCID-causing mutations in the JH2 domain supports this notion. In patient 6, the deletional 482–596 mutant is expressed in reduced amounts as a nonfunctional protein (29). Lack of the kinase domain upper lobe (Fig. 3) in this mutant affects the overall organization of the protein, because the C-terminus, even if correctly folded, is wrongly positioned relative to the rest of the molecule. The same patient carries on the other allele the E481G mutation in the linker region between JH3 and JH2 domains. This mutation might alter the Jak3 fold and possibly misplace domains in relation to each other. Patient 8 is homozygous for the inframe deletion of amino acids 586 –592; this mutant is expressed, but it is nonfunctional (29). The 586 –592 deletion affects the lower lobe structure of the pseudokinase domain, in its -sheet 7, determining structural abnormalities due to the lack of a substantial part of a secondary structure scaffolding. The C759R mutant (allele 2 of patient 4) is expressed almost at normal amounts, but it is strongly functionally defective (29). It is constitutively tyrosine-phosphorylated and cannot be further phosphorylated in response to IL-2, nor can it mediate STAT5 phosphorylation. Recent data indicate that the constitutive tyrosine phosphorylation of the C759R mutant results from endogenous Jak3 phosphotransferase activity (32). This mutation alters an invariant cysteine to arginine in the end of ␣-helix H of the pseudokinase domain (3). Introduction of a charged residue (arginine) into the protein core most likely alters protein folding and possibly the scaffolding of the whole Jak3 protein or of its kinase domain. Residue C759 is tightly packed and the longer side chain of arginine would not fit without structural alterations. However, the C759R JH2 mutant apparently retains the ability to interact with the JH1 domain and strongly inhibits its catalytic activity, as demonstrated by in vitro cotransfection experiments (32). On the other allele, patient 4 carries a nonsense mutation at codon 445 that results in lack of protein expression (29). Patient 5 is homozygous for a transition in the acceptor splice site of intron 9, resulting in frameshift from codon 481 and leading to truncation at residue 517 (29). Consequently, a major part of the C-terminus, including the whole kinase domain, is missing. Another splice site mutation in intron 16 (detected in
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
patient 10) causes a deletion of 151 bases in the transcript (4, 28) leading to protein truncation at residue 733 in the G helix of the pseudokinase domain. The complete kinase domain is missing. Patient 7 is homozygous for the C1839T nucleotide mutation and expresses two distinct Jak3 transcripts. One of them codes for a mutant protein with the R582W amino acid substitution. R582 is exposed on the surface of the pseudokinase domain, and may be involved in intra- or intermolecular interactions which are destroyed when the charged side chain is substituted by a bulky, aromatic tryptophan. The other transcript detected in patient 7 misses 213 bp, corresponding to the whole exons 12 and 13 (30). This transcript encodes for an internally shortened protein, in which the kinase domain is misplaced as related to the Nterminus of Jak3. Patient 9 has in one allele V722I mutation in the loop between ␣F and ␣G in the pseudokinase. The substitutions seems to cause a sterical clash in the protein core and lead to misfolded protein. Finally, patient 1 is homozygous for the Y100C mutation in the JH7 domain (4). This mutant, albeit expressed at reduced levels, is unable to interact with the ␥c protein, thus preventing cytokine-mediated signal transduction (31). Two additional JH7 mutations have been reported in Jak3-deficient patients (P151R in patient 2 and A58P in patient 13). Although these mutations fall outside the minimal region (amino acids 72– 130) of the Jak3 protein required for interaction with ␥c (31), it has been demonstrated that the complete JH7 and JH6 domains are needed for intact functional response (31), and it is therefore likely that these mutants affect the efficiency of signal transduction. The available structural models for kinase domains were further used to interpret experimental observations and to better understand functional aspects of the Jak3 protein. The ATP and Mg 2⫹ binding site is well conserved in all protein kinases. The structures have been determined with ATP or derivatives for cyclic AMP-dependent kinase (cAPK) (46, 47), Irk (19), Src (48), and CDK (49). cAPK is a protein serine/threonine kinase. The ATP and Mg 2⫹ binding sites are conserved among cAPK and Irk. The ATP binding site is well conserved also in the Jak3 kinase domain (Figs. 1 and 3). The binding site is also structurally similar in Src family members, except for the catalytic loop residues which are different in the Src family compared to all other kinases. In the Irk complex with peptide substrate and the ATP analog, AMP-PNP, fewer residues are involved in binding, but they are structurally similar to those in Src and Jak3 kinases (Fig. 1). Of the eight ATP and Mg 2⫹ binding residues of Irk, five are invariant in the Jak3 kinase, but only three are conserved in the pseudokinase domain.
113
In most kinases, the substrate is bound in a cleft between the two lobes. The cAPK structure has been determined with and without inhibitor and substrate (46, 50 –58), and the Irk structure has been modeled as a complex with peptide ligand (20). Tyrosine residues are clearly differently bound to PTKs than are serine or threonine to cAPK, reflecting differences in the size of the side chains as previously discussed in the case of the Bruton’s tyrosine kinase kinase domain model (37). For Irk, substrate residues at positions ⫹1 and ⫹3 (with respect to the tyrosine that is the target of the phosphorylation event) are bound to hydrophobic pockets. Several of the substrate binding residues are conserved in the Jak3 kinase (Fig. 1). In comparison to Irk, three of the residues in the pocket for the P⫹3 residue are invariant or similar in the Jak3 kinase. Two of three site P-1 forming residues are arginines also in the Jak3 kinase. The substrate binding residues (Fig. 3) form a long extended surface, which facilitates numerous interactions with the target(s) (Fig. 4). The differences in the surfaces reflect different substrate specificities. The three residues interacting with tyrosine in the Irk active site (20) are invariant in the Jak3 kinase (Figs. 3 and 4) and are represented by D948, R952, and P990. It is likely that these residues have the same function as the corresponding amino acids in Irk. It is also likely that the pseudokinase cannot bind ATP and lacks catalytic activity, consistent with the observation that a Jak3 mutant devoid of the C-terminal kinase domain cannot mediate IL-2-induced tyrosine phosphorylation of target proteins and may actually act as a dominant negative molecule (59). Many proteins are regulated by phosphorylation. Phosphorylation in the activation loop of the kinase domain causes a conformational change (20, 48, 52, 60). Jak3 has two consecutive tyrosines, which were shown to regulate the activity of the enzyme in humans (15), whereas a weaker effect was observed in the mouse (16). Y980 is essential for full catalytic activity of human Jak3, while mutation of Y981 increases catalytic activity (15). In cAPK, the phosphorylated serine in the activation loop binds to conserved residues and forms one wall of the substrate binding region (46). Three tyrosines in the activation loop of Irk can be phosphorylated (Fig. 3). The tris-prosphorylated Irk permits unrestricted access to ATP and to substrate binding sites (20). PTKs are tightly regulated and their maximal enzymatic activity is not necessarily required. This may apply to Jak3 and possibly also to the other Jaks. The phosphorylated Y981 is predicted to point outward from the protein core and it could form an inter- or intramolecular docking site, which would block entry to the catalytic site when occupied. To test the feasibility of the possible new regulatory step, the Src SH2
114
VIHINEN ET AL.
FIG. 2. Left, superimposition of the Jak3 kinase domain (magenta) with the Irk kinase (yellow) (left), and right, the Jak3 pseudokinase (blue) with the Src kinase domain (red).
FIG. 3. Left, structure of the Jak3 kinase domain modeled based on Irk. The ATP analog is in red and the Mg 2⫹ ions are in green. The residues involved in ATP binding are in magenta, catalytic residues are in yellow, substrate binding residues in the P-1 site are in green, the P⫹1 site is in gray, the P⫹3 site is in dark blue, and tyrosines Y980 and Y981 are in orange. Right, location of the disease-causing mutations in the Jak3 kinase domain. L910 (top) and Y1023 (below) are in white. Residues corresponding to the mutations in the JH2 domain are indicated as follows: 520 –596 are in magenta except for 586 –592 in yellow and 565 in green. R582 is in orange, C759 in dark blue, 733 in magenta, and 722 in red. The images form a stereo pair.
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
115
FIG. 4. Surface of the Jak3 kinase domain and Irk and cAPK kinase domains, from left to right. Inhibitor peptides are drawn with sticks in Irk and cAPK strcutures. The coloring is as in Fig. 3, left, except for phosphorylation sites in orange and inhibitor peptides in magenta.
domain with high-affinity peptide (PDB entry 1sps, 61) was docked without any major clashes into the Jak3 JH1 domain. Src is most likely not a Jak3 ligand, and some other SH2 domains are likely to fit even better. Another, although less likely, explanation for the inhibitory regulatory role of Y981 could be that the phosphate groups of both Y980 and Y981 might independently link the activation loop to the rest of the structure. Of these two alternative forms, the one involving binding and phosphorylation of Y980 might have remarkably higher activity; according to this hypothesis, the Y981F mutation would increase kinase activity by allowing phosphorylated Y980 to always bind to the activation loop. Additional examples of different regulatory activities of adjacent tyrosine residues have been reported in other kinases. The mutation of residues Y1162 and Y1163 in the activation loop increases the basal exogenous kinase activity of Irk (62), whereas the Y1158F mutant could still be activated. In Zap-70, Lck phosphorylates Y493 and Zap-70 can then autophosphorylate at Y492 and other tyrosines (63– 65). Mutation at Y493 decreases Zap-70 activity (63– 65). In contrast, phosphorylated Y492 inhibits kinase activity, and mutations at Y492 increase the basal kinase activity. The unphosphorylated activation loop of Hck (60) and c-Src (48) inactivates the kinase by forming a short 3 10- or ␣-helix, respectively, and by binding to the cleft between the lobes. Because Jak3 has no sequence similarity in this region to Src family, it is not possible to predict whether it has a similar inactivation mechanism.
In summary, the SCID-causing Jak3 mutations identified thus far have been interpreted in structural terms with the aid of the Jak3 JH1 and JH2 domain models. Residues involved in ATP and Mg 2⫹ binding were highly conserved whereas the substrate binding region is somewhat different compared to other kinases. Finally, the crucial role of Jak3 in lymphoid activation in vivo has been substantiated by studies in tumors. In particular, it has been shown that Jak3 is expressed in human B cell malignancies (66) and that proliferation of adult T cell leukemia/lymphoma cells is associated with the constitutive activation of the Jak/ STAT pathway (67, 68). Since Jak3 is preferentially expressed in hematopoietic tissues, it represents an ideal target for novel immune suppressive and chemotherapeutic drugs. Also from this point of view, the structural model of the Jak3 kinase and pseudokinase domains presented here may help identify the Jak3 structural motifs that are crucial for Jak3 function and thus offer the basis for a solid pharmacodesign. ACKNOWLEDGMENTS This work was supported by the Sigrid Juselius Foundation, National Technology Agency of Finland, Medical Research Fund of Tampere University Hospital, EU Biomed 2 concerted action PL 963007, Telethon Grant E668, C.N.R. (Progetto Finalizzato Biotecnologie), and N.A.T.O. Grant CRG.CRG 973041. REFERENCES 1. WHO Scientific Group, Primary immunodeficiency diseases. Clin. Exp. Immunol. 109 (Suppl. 1), 1–28, 1997.
116
VIHINEN ET AL.
2. Puck, J. M., de Saint Basile, G., Schwarz, K., Fugmann, S., and Fischer, R. E., IL2RGbase: A database of c-chain defects causing human X-SCID. Immunol. Today 17, 507–511, 1996. 3. Candotti, F., O’Shea, J. J., and Villa, A., Severe combined immune deficiencies due to defects of the common chain-JAK3 signaling pathway. Springer Semin. Immunopathol. 19, 401– 415, 1998. 4. Macchi, P., Villa, A., Giliani, S., Sacco, M. G., Frattini, A., Porta, F., Ugazio, A. G., Johnston, J. A., Candotti, F., O’Shea, J. J., Vezzoni, P., and Notarangelo, L., Mutations of Jak-3 gene in patients with autsomal severe combined immune deficiency (SCID). Nature 377, 65– 68, 1995. 5. Russell, S. M., Tayebi, N., Nakajima, H., Riedy, M. C., Roberts, J. L., Aman, M. J., Migone, T.-S., Noguchi, M., Markert, M. L., Buckley, R. H., O’Shea, J. J., and Leonard, W. J., Mutation of Jak3 in a patient with SCID: Essential role of Jak3 in lymphoid development. Science 270, 797– 800, 1995. 6. Ihle, J. N., Cytokine receptor signalling. Nature 377, 591–594, 1995. 7. O’Shea, J. J., Jaks, STATs, cytokine signal transduction, and immunoregulation: Are we there yet? Immunity 7, 1–11, 1997. 8. Taniguchi, T., Cytokine signaling through nonreceptor protein tyrosine kinases. Science 268, 251–255, 1995. 9. Cao, X., Shores, E. W., Hu-Li, J., Ansver, M. R., Kelsall, B. L., Russell, S. M., Drago, J., Noguchi, M., Grinberg, A., Bloom, E. T., Paul, W. E., Katz, S. I., Love, P. E., and Leonard, W. J., Defective lymphoid development in mice lacking expression of the common cytokine receptor gamma chain. Immunity 2, 223–238, 1995. 10. DiSanto, J. P., Rieux-Laucat, F., Dautry-Varsat, A., Fischer, A., and de Saint Basile, G., Defective human interleukin 2 receptor gamma chain in an atypical X chromosome-linked severe combined immunodeficiency with peripheral T cell. Proc. Natl. Acad. Sci. USA 91, 9466 –9470, 1994. 11. Nosaka, T., van Deursen, J. M. A., Tripp, R. A., Therfelder, W. E., Witthuhn, B. A., McMickle, A. P., Doherty, P. C., Grosveld, G. C., and Ihle, J. N., Defective lymphoid development in mice lacking Jak3. Science 270, 800 – 802, 1995. 12. Park, S. Y., Saijo, K., Takahashi, T., Osawa, M., Arase, H., Hirayama, N., Miyake, K., Nakauchi, H., Shirasawa, T., and Saito, T., Developmental defects of lymphoid cells in Jak3 kinase-deficient mice. Immunity 3, 771–782, 1995. 13. Thomis, D. C., Gurniak, C. B., Tivol, E., Sharpe, A. H., and Berg, L. J., Defects in B lymphocyte maturation and T lymphocyte activation in mice lacking Jak3. Science 270, 794 –797, 1995. 14. Chen, M., Cheng, A., Chen, Y. Q., Hymel, A., Hanson, E. P., Kimmel, L., Minami, Y., Taniguchi, T., Changelian, P. S., and O’Shea, J. J., The amino terminus of JAK3 is necessary and sufficient for binding to the common ␥ chain and confers the ability to transmit interleukin 2-mediated signals. Proc. Natl. Acad. Sci. USA 94, 6910 – 6915, 1997. 15. Zhou, Y.-J., Hanson, E. P., Chen, Y.-Q., Magnuson, K., Chen, M., Swann, P. G., Wange, R. L., Chagelian, P. S., and O’Shea, J. J., Distinct tyrosine phosphorylation sites in Jak3 kinase domain positively and negatively regulate its enzymatic activity. Proc. Natl. Acad. Sci. USA 94, 13850 –13855, 1997. 16. Liu, K. D., Gaffen, S. L., Goldsmith, M. A., and Greene, W. C., Janus kinases in interleukin-2 mediated signaling: Jak1 and Jak3 are differently regulated by tyrosine phosphorylation. Curr. Biol. 7, 817– 826, 1997. 17. Schumacher, R. F., Mella, P., Badolato, R., Fiorini, M., Savoldi, G., Giliani, S., Villa, A., Candotti, F., Tampalini, A., O’Shea, J. J., and Notarangelo, L. D., Complete genomic organization of the human JAK3 gene and mutation analysis in severe combined
18.
19.
20.
21.
22.
23.
24.
25. 26.
27.
28.
29.
30.
31.
32.
immunodeficiency by single-strand conformation polymorphism. Hum. Genet. 106, 73–79, 2000. Xu, W., Harrison, S. C., and Eck, M. J., Three-dimensional structure of the tyrosine kinase c-Src. Nature 385, 595– 602, 1997. Abola, E. E., Sussman, J. L., Prilusky, J., and Manning, N. O., Protein Data Bank archives of three-dimensional macromolecular structures. Methods Enzymol. 277, 556 –571, 1997. Hubbard, S. R., Crystal structure of the activated insulin receptor tyrosine kinase in complex with peptide substrate and ATP analog. EMBO J. 16, 5572–5581, 1997. Devereux, J., Haeberli, P., and Smithies, O., A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12, 387–395, 1984. Vihinen, M., Euranto, A., Luostarinen, P., and Nevalainen, O., MULTICOMP, a program package for multiple sequence comparison. Comput. Appl. Biosci. 8, 35–38, 1992. Boberg, J., Salakoski, T., and Vihinen, M., Representative selection of proteins based on nuclear families. Prot. Eng. 8, 501–503, 1995. Vihinen, M., Brandau, O., Brande´n, L., Kwan, S.-P., Lappalainen, I., Lester, T., Noordzij, J. G., Ochs, H. D., Ollila, J., Pienaar, S. M., Riikonen, P., Saha, B. K., and Smith, C. I. E., BTKbase, mutation database for X-linked agammaglobulinemia (XLA). Nucleic Acids Res. 26, 242–247, 1998. Etzold, T., and Argos, P., SRS—An indexing and retrieval tool for flat file data libraries. Comput. Appl. Biosci. 9, 49 –57, 1993. Rodriguez, R., Chinea, G., Lopez, N., Pons, T., and Vriend, G., Homology modeling, model and software evaluation: Three related resources. Bioinformatics 14, 523–528, 1998. Vihinen, M., Lehva¨slaiho, H., and Cotton, R. G. H., Immunodeficiency mutation databases. In “Primary Immunodeficiency Diseases. A Molecular and Genetic Approach” (H. D. Ochs, C. I. E. Smith, and J. Puck, Eds.), pp. 443– 447, Oxford Univ. Press, Oxford, 1998. Villa, A., Sironi, M., Macchi, P., Matteucci, C., Notarangelo, L., Vezzoni, P., and Mantovani, A., Monocyte function in a severe combined immunodeficient patient with a donor splice site mutation in the Jak3 gene. Blood 88, 817– 823, 1996. Candotti, F., Oaker, S. A., Johnston, J. A., Giliani, S., Schumacher, R. F., Mella, P., Fiorini, M., Ugazio, A. G., Badolato, R., Notaragelo, L. D., Bozzi, F., Macchi, P., Strina, D., Vezzoni, P., Blaese, R. M., O’Shea, J. J., and Villa, A., Structural and functional basis of Jak3-deficient severe combined immunodeficiency. Blood 90, 3996 – 4003, 1997. Bozzi, F., Lefranc, G., Villa, A., Badolato, R., Schumacher, R. F., Khalil, G., Loislet, J., Bresciani, S., O’Shea, J. J., Vezzoni, P., Notarangelo, L. D., and Candotti, F., Molecular and biochemical characterization of Jak3 deficiency in a patient with severe combined immunodeficiency over 20 years after bone marrow transplantation: Implications for treatment. Br. J. Haematol. 102, 1363–1366, 1998. Cacalano, N. A., Migone, T.-S., Bazan, F., Hanson, E. P., Chen, M., Candotti, F., O’Shea, J. J., and Johnston, J. A., Autosomal SCID caused by a point mutation in the N-terminus of Jak3: Mapping of the Jak3-receptor interaction domain. EMBO J. 18, 1549 –1558, 1999. Chen, M., Cheng, A., Candotti, F., Zhou, Y.-J., Hymel, A., Fasth, A., Notarangelo, L. D., and O’Shea, J. J., Complex effects of naturally occurring mutations in the JAK3 pseudokinase domain: Evidence for interactions between the kinase and pseudokinase domains. Mol. Cell. Biol. 20, 947–956, 2000.
MOLECULAR MODELING OF Jak3 KINASE DOMAINS 33. Riikonen, P., and Vihinen, M., MUTbase program package for maintenance and analysis of mutation databases on World Wide Web. Bioinformatics 15, 852– 859, 1999. 34. Stenberg, K. A. E., Riikonen, P. T., Vihinen, M., KinMutBase—A database of human disease-causing protein kinase mutations. Nucleic Acids Res. 27, 362–364, 1999. 35. Hubbard, S. R., Wei, L., Ellis, L., and Hendrickson, W. A., Crystal structure of the tyrosine kinase domain of the human insulin receptor. Nature 372, 746 –754, 1994. 36. Mohammadi, M., Schlessinger, J., and Hubbard, S. R., Structure of the FGF receptor tyrosine kinase domain reveals a novel autoinhibitory mechanism. Cell 86, 577–587, 1996. 37. Vihinen, M., Vetrie, D., Maniar, H. S., Ochs, H. D., Zhu, Q., Vorˇechovsky´, I., Webster, A. D. A., Notarangelo, L. D., Nilsson, L., Sowadski, J. M., and Smith, C. I. E., Structural basis for chromosome X-linked agammaglobulinemia: A tyrosine kinase disease. Proc. Natl. Acad. Sci. USA 91, 12803–12807, 1994. 38. Vihinen, M., Zvelebil, J. J. M., Zhu, Q., Brooimans, R. A., Ochs, H. D., Zegers, B. J. M., Nilsson, L., Waterfield, M. D., and Smith, C. I. E., Structural basis for pleckstrin homology domain mutations in X-linked agammaglobulinemia. Biochemistry 34, 1475– 1481, 1995. 39. Jin, H., Webster, A. D. B., Vihinen, M., Sideras, P., Vorˇechovsky´, I., Hammarstro¨m, L., Bernatowska-Matuszkiewicz, E., Smith, C. I. E., Bobrow, M., and Vetrie, D., Identification of Btk mutations in 20 unrelated patients with X-linked agammaglobulinemia (XLA) Hum. Mol. Genet. 4, 693–700, 1995. 40. Vorˇechovsky´, I., Vihinen, M., de Saint Basile, G., Honsova´, S., Hammarstro¨m, L., Mu¨ller, S., Nilsson, L., Fischer, A., and Smith, C. I. E., Mutation analysis and molecular modelling of Bruton’s tyrosine kinase in patients with X-linked agammaglobulinemia. Hum. Mol. Genet. 4, 51–58, 1995. 41. Saha, B. K., Curtis, S. K., Vogler, L. B., and Vihinen, M., Molecular and structural characterization of five novel mutations in the Bruton’s tyrosine kinase gene from patients with X-linked agammaglobulinemia. Molec. Med. 3, 477– 485, 1997. 42. Holinski-Feder, E., Weiss, M., Brandau, O., Jedele, K. B., Nore, B., Ba¨ckesjo¨, C.-M., Vihinen, M., Go¨tz, G., Hubbard, S. R., Belohradsky, B. H., Smith, C. I. E., and Meindl, A., Mutation screening of the btk gene in 56 families with X-linked agammaglobulinemia (XLA)— 47 unique mutations without correlation to clinical course. Pediatrics 101, 276 –284, 1998. 43. Luo, H., Rose, P., Barber, D., Hanratty, W. P., Lee, S., Roberts, T. M., D’Andrea, A. D., and Dearolf, C. R., Mutation in the Jak kinase JH2 domain hyperactivate Drosophila and mammalian Jak-STAT pathways. Mol. Cell. Biol. 17, 1562–1571, 1997. 44. Velazquez, L., Mogensen, K. E., Barbieri, G., Fellous, M., Uze, G., and Pellegrini, S., Distinct domains of the protein tyrosine kinase tyk2 required for binding of interferon-alpha/beta and for signal transduction. J. Biol. Chem. 270, 3327–3334, 1995. 45. Fujitani, Y., Hibi, M., Fukada, T., Takahashi-Tezuka, M., Yoshida, H., Yamaguchi, T., Sugiyama, K., Yamanaka, Y., Nakajima, K., and Hirano, T., An alternative pathway for STAT activation that is mediated by the direct interaction between Jak and STAT. Oncogene 14, 51–761, 1997. 46. Zheng, J., Knighton, D. R., Ten Eyck, L. F., Karlsson, R., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MgATP and peptide inhibitor. Biochemistry 32, 2154 –2161, 1993. 47. Bossemeyer, D., Engh, R. A., Kinzel, V., Ponstingl, H., and Huber, R., Phosphotransferase and substrate binding mechanism of the cAMP-dependent protein kinase catalytic subunit from porcine heart as deduced from the 2.0 Å structure of the
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
117
complex with Mn 2⫹ adenylyl imidodisphosphate and inhibitor peptide PKI(5-24). EMBO J. 12, 849 – 859, 1993. Xu, W., Doshi, A., Lei, M., Eck, M. J., and Harrison, S. C., Crystal structures of c-Src reveal features of its autoinhibitory mechanism. Mol. Cell 3, 629 – 638, 1999. De Bondt, H. L., Rosenblatt, J., Jancarik, J., Jones, H. D., Morgan, D. O., and Kim, S. H., Crystal structure of cyclin-dependent kinase 2. Nature 363, 595– 602, 1993. Knighton, D. R., Zheng, J., Ten Eyck, L. F., Ashford, V. A., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Crystal structure of the catalytic subunit of cyclic adenosine monophosphatedependent protein kinase. Science 253, 407– 414, 1991. Knighton, D. R., Zheng, J., Ten Eyck, L. F., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Structure of a peptide inhibitor bound to the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science 253, 414 – 420, 1991. Knighton, D. R., Bell, S. M., Zheng, J., Ten Eyck, L. F., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., 2.0 Å refined crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with a peptide inhibitor and detergent. Acta Crystallogr. D49, 357–361, 1993. Karlsson, R., Zheng, J., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Structure of the mammalian catalytic subunit of cAMP-dependent protein kinase and an inhibitor peptide displays an open conformation. Acta Crystallogr. D49, 381–388, 1993. Zheng, J., Knighton, D. R., Xuong, N.-H., Taylor, S. S., Sowadski, J. M., and Ten Eyck, L. F., Crystal structures of the myristoylated catalytic subunit of cAMP-dependent protein kinase reveal open and closed conformations. Prot. Sci. 2, 1559 –1573, 1993. Zheng, J., Trafny, E. A., Knighton, D. R., Xuong, N.-h., Taylor, S. S., Ten Eyck, L. F., and Sowadski, J. M., 2.2 Å refined crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MnATP and peptide inhibitor. Acta Crystallogr. D49, 362–365, 1993. Madhusudan, Trafny, E. A., Xuong, N.-h., Adams, J. A., Ten Eyck, L. F., Taylor, S. S., and Sowadski, J. M., cAMP-dependent protein kinase: Crystallographic insights into substrate recognition and phosphotransfer. Prot. Sci. 3, 176 –187, 1994. Narayama, N., Cox, S., Xuong, N.-h., Ten Eyck, L. F., and Taylor, S. S., A binary complex of the catalytic subunit of cAMP-dependent protein kinase and adenosine further defines conformational flexibility. Structure 5, 921–935, 1997. Prade, L., Engh, R. A., Girod, A., Kinzel, V., Huber, R., and Bossemeyer, D., Staurosporine -induced conformational changes of cAMP-dependent protein kinase catalytic subunit explain inhibitory potential. Structure 5, 1627–1637, 1997. Kawahara, A., Minami, Y., Miyazaki, T., Ihle, J. N., and Taniguchi, T., Critical role of the interleukin 2 (IL-2) receptor -chainassociated Jak3 in the IL-2-induced c-fos and c-myc, but not bcl-2, gene induction. Proc. Natl. Acad. Sci. USA 92, 8724 – 8728, 1995. Schindler, T., Sicheri, F., Pico, A., Gazit, A., Levitzki, A., and Kuriyan, J., Crystal structure of Hck in complex with a Src family-selective tyrosine kinase inhibitor. Mol. Cell 3, 639 – 648. Waksman, G., Shoelson, S. E., Pant, N., Cowburn, D., and Kuriyan, J., Binding of a high affinity phosphotyrosyl peptide to the Src SH2 domain: crystal structures of the coplexed and peptide-free forms. Cell 72, 779 –790, 1993. Zhang, B., Tavare, J. M., Ellis, L., and Roth, R. A., The regulatory role of known tyrosine autophosphorylation sites of the insulin receptor kinase domain. An assessment by replacement with neutral and negatively charged amino acids. J. Biol. Chem. 266, 990 –996, 1991.
118
VIHINEN ET AL.
63. Chan, A. C., Dalton, M., Johson, R., Kong, G.-h., Wang, T., Thoma, R., and Kurosaki, T., Activation of ZAP-70 kinase by phosphorylation of tyrosine 493 is required for lymphocyte antigen receptor function. EMBO J. 14, 2499 –2508, 1995. 64. Wange, R. L., Guitia´n, R., Isakov, N., Watts, J. D., Aebersold, R., and Samelson, L. E., Activating and inhibitory mutations in adjacent tyrosines in the kinase domain of ZAP-70. J. Biol. Chem. 270, 18730 –18733, 1995. 65. Kong, G., Dalton, M., Wardenburg, J. B., Straus, D., Kurosaki, T., and Chan, A. C., Distinct tyrosine phosphorylation sites in ZAP-70 mediate activation and negative regulation of antigen receptor function. Mol. Cell. Biol. 16, 5026 –5035, 1996. 66. Tortolani, P. J., Lal, B. K., Riva, A., Johnston, J. A., Chen, Y.-Q., Reaman, G. R., Beckwith, M., Longo, D., Ortaldo, J. R., Bhatia, Received June 8, 1999; accepted with revision April 11, 2000
K., McGrath, I., Kehrl, J., Tuscano, J., McVicar, D. W., and O’Shea, J. J., Regulation of JAK3 expression and activation in human B cells and B cell malignancies. J. Immunol. 155, 5220 – 5226, 1995. 67. Migone, T.-S., Lin, J.-X., Cereseto, A., Mulloy, J. C., O’Shea, J. J., Franchini, G., and Leonard, W. J., Constitutively activated JakSTAT pathway in T cells transformed with HTLV-I. Science 269, 79 – 81, 1995. 68. Takemoto, S., Mulloy, J. C., Cereseto, A., Migone, T.-S., Patel, B. K. R., Matsuoka, M., Yamaguchi, K., Takatsuki, K., Kamihira, S., White, J. D., Leonard, W. J., Waldmann, T., and Franchini, G., Proliferation of adult T cell leukemia/lymphoma cells is associated with the constitutive activation of JAK/STAT proteins. Proc. Natl. Acad. Sci. USA 94, 13897–13902, 1997.
Molecular Modeling of the Jak3 Kinase Domains and Structural Basis for Severe Combined Immunodeficiency Mauno Vihinen,* Anna Villa,† Patrizia Mella,‡ R. Fabian Schumacher,‡ Gianfranco Savoldi,‡ John J. O’Shea,§ Fabio Candotti,‡ ,1 and Luigi D. Notarangelo‡ *Institute of Medical Technology, FIN-33014 University of Tampere, Finland, and Tampere University Hospital, FIN-33520 Tampere, Finland; †Istituto di Tecnologie Biomediche Avanzate, Consiglio Nazionale della Ricerche, 20090 Segrate (MI), Italy; ‡Istituto di Medicina Molecolare “Angelo Nocivelli,” Clinica Pediatrica, Universita` di Brescia, Spedali Civili, Brescia, Italy; and §Lymphocyte Cell Biology Section, Arthritis and Rheumatism Branch, NIAMSD, NIH, Bethesda, Maryland 20892-1820
Hereditary severe combined immunodeficiency (SCID) includes a heterogeneous group of diseases that profoundly affect both cellular and humoral immune responses and require treatment by bone marrow transplantation. Characterization of the cellular and molecular bases of SCID is essential to provide accurate genetic counseling and prenatal diagnosis, and it may offer the grounds for alternative forms of treatment. The Jak3 gene is mutated in most cases of autosomal recessive T ⴚB ⴙ SCID in humans. Jak3 belongs to the family of intracellular Janus tyrosine kinases. It is physically and functionally coupled to the common ␥ chain, ␥c, shared by several cytokine receptors. We have established the JAK3base registry for disease and mutation information. In order to study the structural consequences of the Jak3 mutations, the structure of the human Jak3 kinase and pseudokinase domains was modeled. Residues involved in ATP and Mg 2ⴙ binding were highly conserved in the kinase domain whereas the substrate binding region is somewhat different compared to other kinases. We have identified the first naturally occurring mutations disrupting the function of the human Jak3 kinase domain. The structural basis of all of the known Jak3 mutations reported so far is discussed based on the modeled structure. The model of the Jak3 protein also permits us to study Jak3 phosphorylation at the structural level and may thus serve in the design of novel immune suppressive drugs. © 2000 Academic Press Key Words: immunodeficiency; B cells; T cells; disease-causing mutations; structural basis of disease; structure–function relationships; JAK3base.
INTRODUCTION
Severe combined immunodeficiency (SCID) includes a group of hereditary diseases that severely affect both 1 Present address: Clinical Gene Therapy Branch, NHGRI, NIH, Bethesda, MD 20892.
1521-6616/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
cellular and humoral-based immune defenses (1). Unless treated with bone marrow transplantation, SCID patients die early in life due to severe infections. Characterization of the cellular and molecular bases of SCID is essential to provide accurate genetic counseling and prenatal diagnosis, and it may offer the basis for alternative forms of treatment, based on gene transfer. The most common form of SCID in humans is inherited as an X-linked trait (SCIDX1, OMIM 308380) and is due to mutations of the IL2RG gene (2), encoding for the common gamma chain (␥c), shared by cytokine receptors for interleukin- (IL-) 2, IL-4, IL-7, IL-9, and IL-15 (for a review, see Ref. 3). The immunological phenotype of SCIDX1 is characterized by the lack of circulating T (and NK) cells, with normal to increased numbers of B lymphocytes (T ⫺B ⫹ SCID) (1). The same phenotype, however, may also be observed in autosomal variants of SCID. We and others have shown that in most cases of autosomal recessive T ⫺B ⫹ SCID, the Jak3 gene, encoding for an intracellular Janus tyrosine kinase that is physically and functionally coupled to ␥c, is mutated (OMIM 600170) (4, 5). Jak3 is a member of the family of Janus kinases (Jaks), which in humans also includes Jak1, Jak2, and Tyk2 (6, 7). These proteins share seven conserved regions called Jak homology (JH) domains. One of the earliest events that follows binding of IL-2, IL-4, IL-7, IL-9, or IL-15 to the specific receptor is represented by phosphorylation of tyrosine residues of both Jak3 (associated with the ␥c) and Jak1 proteins (coupled to the major transducing subunit of the cytokine receptor) (6 – 8). Once activated, Jaks tyrosine-phosphorylate other proteins including cytokine receptor subunits, generating docking sites for the SH2 domains of STATs (signal transducers and activators of transcription). This leads to dimerization of the STATs and their translocation to the nucleus where they bind enhancer regions of DNA, thereby inducing the transcription of cytokine responsive genes (7).
108
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
The crucial role for intact ␥c-Jak3 expression and function is best illustrated by the occurrence of T ⫺B ⫹ SCID in humans with genetic defects of either of these genes and by similar observations in gene-targeted mice, although in mice the disruption of the IL2RG (9, 10) or of the Jak3 (11–13) genes also strongly affects B cell differentiation. Despite the essential role played by Jak3 in lymphoid development and function, little is known about the structural basis for its activity. The JH7 and JH6 domains of Jak3 have been shown to mediate interaction with the ␥c (14). Furthermore, Jak3 has several phosphorylation sites, and tyrosinephosphorylation at these residues may regulate the activity of the enzyme positively or negatively (15, 16). In the present paper, we report the three-dimensional structure of the human Jak3 kinase (JH1) and pseudokinase (JH2) domains. These models were used to assess the structural consequences of the Jak3 mutations thus far described in humans, including the first mutations in the kinase domain that we have recently reported (17). A disease and mutation database (JAK3base) has been developed, which includes information on the clinical and immunological phenotype of Jak3-deficient SCID patients and on the structural consequences of their respective Jak3 gene defects. This information could be of particular interest for understanding the structural basis of Jak3-deficient SCID. Furthermore, a better appreciation of the functional role played by distinct structural motifs in the Jak3 protein may help develop specific Jak3 inhibitors as a novel class of immune suppressive drugs. MATERIALS AND METHODS
Mutation Analysis Detection of mutations in the Jak3 gene in patients with T ⫺B ⫹ SCID was accomplished on genomic DNA by single-strand conformation polymorphism (SSCP) and DNA sequencing as described elsewhere (17). Molecular Modeling
109
The insertions and deletions were modeled by searching loops from a database, which contained either most of the PDB structures or an unbiased selection of PDB (23). The search was performed for fragments of the required length and end-point separation by using some residues at each end of the loop as anchor points. Fragments obtained were evaluated on three criteria: root mean square deviation from the anchor points, sequence similarity, and interference with the core region. The models were refined by energy minimization with the program Discover in stepwise manner using Amber force field. First, hydrogen atoms were relaxed, and then the side chains of the insertions and deletions were relaxed and the rest of the molecule was fixed. In the next step the borders of insertions and deletions and the C ␣ atoms of the conserved regions, and finally only the C ␣ atoms of the conserved regions, were harmonically constrained. JAK3base A program was developed to allow submission of clinical, immunological, and molecular data on Jak3deficient SCID patients to the JAK3base registry. The registry was built according to the guidelines adopted for the BTKbase (24), and its content has been made available through the Internet by using the World Wide Web. The database contains four main items: identification of the patient by means of a specific code related to the Jak3 mutation(s), reference either to published article(s) or to the submitting physician, mutation information, and data related to the clinical and immunological phenotype and to treatment. All of these items are organized as fields, which are easily accessible. Protection of the patients’ identity is assured in the registry. Patient data form an entry, which can be analyzed with the provided tools or with the sequence retrieval system (SRS) (25). RESULTS
Mutation Analysis in Jak3 Deficiency The kinase and pseudokinase domains of Jak3 were modeled based on the structure of the Src kinase at 1.7 Å resolution (18) (Protein Data Bank (PDB) (19) entry 1fmk). The Jak3 kinase domain was further modeled based on the Irk structure (20; PDB code 1ir3). The sequence alignment was performed with GCG (21) and MULTICOMP (22) program packages. The final alignment was performed using information from multiple sequence analysis and secondary structural information from the three-dimensional structures. The model was built with the program InsightII (Molecular Simulations, Inc., San Diego, CA). A side chain rotamer library was used to model amino acid substitutions.
In addition to previously published mutations, two novel Jak3 gene defects were identified in a 2-year-old female child (patient 13 in Table 1) with a mild phenotype, characterized by recurrent respiratory tract infections and persistent candidiasis, which did not affect her growth. Immunological investigations revealed reduced, but not absent, circulating T cells (CD3: 14.5%, CD4: 7.2%, CD8: 6.2%), an excess of B lymphocytes (CD19: 83.4%), absent NK cells (CD16: 0.7%), and lack of proliferative responses to phytohemagglutinin (0.8 vs 37 cpm ⫻ 10 ⫺3). Molecular analysis at highly polymorphic DNA loci ruled out the possibility that the
110
VIHINEN ET AL.
TABLE 1 Molecular and Structural Basis of Jak3-Deficient SCID Patient
Allele
Genomic mutation
Protein alteration
Domain
1 2
1,2 1 2 1,2 1 2 1,2 1 2 1,2 1,2 1 2 1,2 1 2 1,2 1 2
A394G C574G ? C1428T C1428T T2370C IVS9,⫺2a 3 g A1537G Del 1536–1582 C1839T C1862T G2259A ? IVS16,⫹2 t 3 c T2824C C3164A IVS21,⫹2 t 3 a G268C Del C 3167
Y100C P151R ? R445X R445X C759R Frameshift after codon E481 E481G Del aa 482–596 R582W Del aa 586–592 V722I ? Frameshift after codon K733 L910S Y1023X Not known A58P Frameshift after codon 1024
JH7 JH7 ? JH3 JH3 JH2 JH4 JH3 JH3 JH2 JH2 JH2 ? JH2-JH1 JH1 JH1 JH1 JH7 JH1
3 4 5 6 7 8 9 10 11 12 13
residual T cells could be of maternal origin. Use of the SSCP-based screening assay, followed by direct sequencing, revealed that the patient is a compound heterozygote for two mutations in exon 1 (G268C, resulting in A58P amino acid change) and in exon 21 (deletion of C3167, resulting in frameshift after codon 1024, with premature termination at codon 1037). The molecular data of all of the patients with Jak3 gene defects identified at our center are reported in Table 1. Structural Model of Jak3 Kinase and Pseudokinase Domains To address the structural features of Jak3 and Jak3deficient SCID, the three-dimensional structure of the kinase and pseudokinase domains were modeled based on Src (18). Further, the effect of phosphorylation was studied by modeling the Jak3 kinase domain based on insulin receptor kinase (Irk) in its closed, active conformation (20). The sequence alignment was based on multiple sequence analysis, hallmark residues, and locations of secondary structural elements (Fig. 1). The kinase domain of Jak3 has six insertions and one deletion compared to the Src kinase domain. One of the insertions is in the activation loop and two are located in the kinase insert region; both of these functionally important regions are highly variable with respect to length and amino acid sequence. These three loops were not modeled and coordinates are missing also for the activation loop of the Src crystal structure (18). The sequence identity of the Jak3 kinase and pseudokinase domains with the Src kinase is 36 and 23%, respec-
Effect
Ref.
17 Truncation Truncation Sterical clash Truncation
17
Fold alteration Altered ligand interaction Fold alteration Sterical clash in core Truncation Structural alterations Truncation
Truncation
tively. The Jak3 kinase domain is 34% identical with the Irk catalytic domain. The catalytic site amino acids as well as some of the conserved ATP-binding residues are missing from the structure of the pseudokinase domain. Because of the high similarity, the Jak3 kinase domain was modeled based both on Irk and on Src structures, and the Jak3 JH2 domain was modeled based on Src (Fig. 2). Compared to the templates, the major differences in the Jak3 model are in the the exposed surface loops. The upper lobe is twisted in a different way in the two JH1 domain models. The models of the insertions and deletions are rather reliable, because the length of the changes does not exceed four residues except for one insertion in the upper lobe of the pseudokinase model. The final models have typical globular structures according to several structural tests (26). Both the kinase and the pseudokinase models have a five-stranded anti-parallel -sheet in the upper lobe and a mostly ␣-helical C-terminal lobe. The upper domain of both the kinase and the pseudokinase domains, responsible for ATP binding, is highly conserved. The glycine-rich loop and the other residues in the upper domain that establish contacts with ATP are conserved in the Jak3 JH1, but not in the JH2, domain. The main difference between Src and Jak3 was observed in the kinase insert region (Fig. 1), which shows remarakable variability among different kinases. The hallmark residues of protein kinases were conserved in the kinase domain, whereas in the pseudokinase more alterations were noticed (Fig. 1).
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
111
FIG. 1. Sequence alignment of the Jak3 kinase and pseudokinase domain with cAPK and Irk sequences. Residues involved in the catalytic site of Irk are indicated with # and in ATP and Mg 2⫹ binding with @. The amino acids of binding sites -1, 1, and 3 are are denoted by corresponding numbers. The secondary structures are shown below the sequence with ␣ for ␣-helices and with  for -strands. The numbers are for secondary structural elements. The sites of the SCID-causing mutations are shown below the sequence (# denotes the start of deletion and X nonsense mutation).
JAK3base All of the reported SCID-causing Jak3 mutations (4, 5, 17, 28 –30) were collected into a database called JAK3base. The reported mutations are scattered into several domains; for some of them, the functional consequences have been previously investigated (29 –32). In addition to mutations (Table 1), the registry also contains information about immunoglobulin levels, lymphocyte counts, age at diagnosis, symptoms, mRNA
and protein data, putative structural consequences of the mutations, and details on therapy including hematopoietic stem cell transplantation, when available. The registry is constructed according to the concepts used in BTKbase, XLA mutation registry (24), by using the MUTbase system (33). The JAK3base is freely accessible through the World Wide Web at http://www.uta.fi/imt/bioinfo/JAK3base.html. Researchers are encouraged to send their patient and mutation information to the registry.
112
VIHINEN ET AL.
The JAK3base also provides additional information, such as alterations in restriction enzyme patterns due to mutations and mutation descriptive statistics. DISCUSSION
Mutations in tyrosine kinases are responsible for several genetic diseases in humans; disease-causing mutations in the kinase domains are accessible through the KinMutBase (http://www.uta.fi/imt/ bioinfo/KinMutBase.html) (34). Three-dimensional structural models may provide new insights into the effects of specific mutations. The crystal structures of Irk (35) and fibroblast growth factor receptor kinase (36), and the models of Btk kinase domain (37– 42), have been used to interpret the structural consequences of the defects in the respective genes. Here, the mutations in the human Jak3 gene have been explained with the models of the three-dimensional structure of Jak3 kinase and pseudokinase regions (Fig. 3). Many of the patients have consanguineous parents and are therefore homozygous for the gene defect. We have recently identified the first Jak3-deficient patient (patient 11 in Table 1) with a missense mutation in the kinase domain (17). Mutation L910S at the beginning of ␣-helix D affects the protein core where the side chain is pointing. The corresponding residue in Irk is involved in recognition of the residue prior to phosphotyrosine (P-1) of the substrate (20). Introduction of the polar hydroxyl group might require neutralisation by, e.g., water molecules and lead to structural and functional alterations due to sterical reasons. The other mutant allele in the patient (Y1023X) deletes the major part of the kinase domain. Patient 12 in Table 1 is homozygous for a splice-site mutation in intron 21; this mutation results in exon skipping, leading to a frameshift mutation in the cDNA and to premature termination of the protein chain. Most recently, we have identified a third patient (patient 13 in Table 1) with a milder phenotype, who is a compound heterozygote for a missense mutation in the JH7 domain (A58P) and a single nucleotide deletion (del C 3167) in exon 21 of the Jak3 gene; the last mutation results in a frameshift in the kinase domain after codon 1024, with premature termination at codon 1037 and loss of C-terminal parts of the kinase domain leading to impaired kinase activity. A number of additional Jak3 mutations have been previously identified, but in most cases their structural importance and function have not been investigated in detail. With the generated structural models we were able to predict the putative consequences of all of the disease-causing alterations. Although the pseudokinase JH2 domain is catalytically inactive, it is obviously involved in regulating Jak3 function, as muta-
tions in the JH2 domain have been shown to alter the function of various Jak proteins (43, 44). Moreover, it has been shown that the JH2 domain of Jaks may directly interact with STATs (45). More recently, direct interaction between the JH1 (kinase) and JH2 (pseudokinase) domains of the Jak3 protein has been demonstrated (32); this interaction appears to inhibit Jak3 catalytic activity, presumably by interfering with substrate access to the kinase domain. Evaluation of the structural consequences of SCID-causing mutations in the JH2 domain supports this notion. In patient 6, the deletional 482–596 mutant is expressed in reduced amounts as a nonfunctional protein (29). Lack of the kinase domain upper lobe (Fig. 3) in this mutant affects the overall organization of the protein, because the C-terminus, even if correctly folded, is wrongly positioned relative to the rest of the molecule. The same patient carries on the other allele the E481G mutation in the linker region between JH3 and JH2 domains. This mutation might alter the Jak3 fold and possibly misplace domains in relation to each other. Patient 8 is homozygous for the inframe deletion of amino acids 586 –592; this mutant is expressed, but it is nonfunctional (29). The 586 –592 deletion affects the lower lobe structure of the pseudokinase domain, in its -sheet 7, determining structural abnormalities due to the lack of a substantial part of a secondary structure scaffolding. The C759R mutant (allele 2 of patient 4) is expressed almost at normal amounts, but it is strongly functionally defective (29). It is constitutively tyrosine-phosphorylated and cannot be further phosphorylated in response to IL-2, nor can it mediate STAT5 phosphorylation. Recent data indicate that the constitutive tyrosine phosphorylation of the C759R mutant results from endogenous Jak3 phosphotransferase activity (32). This mutation alters an invariant cysteine to arginine in the end of ␣-helix H of the pseudokinase domain (3). Introduction of a charged residue (arginine) into the protein core most likely alters protein folding and possibly the scaffolding of the whole Jak3 protein or of its kinase domain. Residue C759 is tightly packed and the longer side chain of arginine would not fit without structural alterations. However, the C759R JH2 mutant apparently retains the ability to interact with the JH1 domain and strongly inhibits its catalytic activity, as demonstrated by in vitro cotransfection experiments (32). On the other allele, patient 4 carries a nonsense mutation at codon 445 that results in lack of protein expression (29). Patient 5 is homozygous for a transition in the acceptor splice site of intron 9, resulting in frameshift from codon 481 and leading to truncation at residue 517 (29). Consequently, a major part of the C-terminus, including the whole kinase domain, is missing. Another splice site mutation in intron 16 (detected in
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
patient 10) causes a deletion of 151 bases in the transcript (4, 28) leading to protein truncation at residue 733 in the G helix of the pseudokinase domain. The complete kinase domain is missing. Patient 7 is homozygous for the C1839T nucleotide mutation and expresses two distinct Jak3 transcripts. One of them codes for a mutant protein with the R582W amino acid substitution. R582 is exposed on the surface of the pseudokinase domain, and may be involved in intra- or intermolecular interactions which are destroyed when the charged side chain is substituted by a bulky, aromatic tryptophan. The other transcript detected in patient 7 misses 213 bp, corresponding to the whole exons 12 and 13 (30). This transcript encodes for an internally shortened protein, in which the kinase domain is misplaced as related to the Nterminus of Jak3. Patient 9 has in one allele V722I mutation in the loop between ␣F and ␣G in the pseudokinase. The substitutions seems to cause a sterical clash in the protein core and lead to misfolded protein. Finally, patient 1 is homozygous for the Y100C mutation in the JH7 domain (4). This mutant, albeit expressed at reduced levels, is unable to interact with the ␥c protein, thus preventing cytokine-mediated signal transduction (31). Two additional JH7 mutations have been reported in Jak3-deficient patients (P151R in patient 2 and A58P in patient 13). Although these mutations fall outside the minimal region (amino acids 72– 130) of the Jak3 protein required for interaction with ␥c (31), it has been demonstrated that the complete JH7 and JH6 domains are needed for intact functional response (31), and it is therefore likely that these mutants affect the efficiency of signal transduction. The available structural models for kinase domains were further used to interpret experimental observations and to better understand functional aspects of the Jak3 protein. The ATP and Mg 2⫹ binding site is well conserved in all protein kinases. The structures have been determined with ATP or derivatives for cyclic AMP-dependent kinase (cAPK) (46, 47), Irk (19), Src (48), and CDK (49). cAPK is a protein serine/threonine kinase. The ATP and Mg 2⫹ binding sites are conserved among cAPK and Irk. The ATP binding site is well conserved also in the Jak3 kinase domain (Figs. 1 and 3). The binding site is also structurally similar in Src family members, except for the catalytic loop residues which are different in the Src family compared to all other kinases. In the Irk complex with peptide substrate and the ATP analog, AMP-PNP, fewer residues are involved in binding, but they are structurally similar to those in Src and Jak3 kinases (Fig. 1). Of the eight ATP and Mg 2⫹ binding residues of Irk, five are invariant in the Jak3 kinase, but only three are conserved in the pseudokinase domain.
113
In most kinases, the substrate is bound in a cleft between the two lobes. The cAPK structure has been determined with and without inhibitor and substrate (46, 50 –58), and the Irk structure has been modeled as a complex with peptide ligand (20). Tyrosine residues are clearly differently bound to PTKs than are serine or threonine to cAPK, reflecting differences in the size of the side chains as previously discussed in the case of the Bruton’s tyrosine kinase kinase domain model (37). For Irk, substrate residues at positions ⫹1 and ⫹3 (with respect to the tyrosine that is the target of the phosphorylation event) are bound to hydrophobic pockets. Several of the substrate binding residues are conserved in the Jak3 kinase (Fig. 1). In comparison to Irk, three of the residues in the pocket for the P⫹3 residue are invariant or similar in the Jak3 kinase. Two of three site P-1 forming residues are arginines also in the Jak3 kinase. The substrate binding residues (Fig. 3) form a long extended surface, which facilitates numerous interactions with the target(s) (Fig. 4). The differences in the surfaces reflect different substrate specificities. The three residues interacting with tyrosine in the Irk active site (20) are invariant in the Jak3 kinase (Figs. 3 and 4) and are represented by D948, R952, and P990. It is likely that these residues have the same function as the corresponding amino acids in Irk. It is also likely that the pseudokinase cannot bind ATP and lacks catalytic activity, consistent with the observation that a Jak3 mutant devoid of the C-terminal kinase domain cannot mediate IL-2-induced tyrosine phosphorylation of target proteins and may actually act as a dominant negative molecule (59). Many proteins are regulated by phosphorylation. Phosphorylation in the activation loop of the kinase domain causes a conformational change (20, 48, 52, 60). Jak3 has two consecutive tyrosines, which were shown to regulate the activity of the enzyme in humans (15), whereas a weaker effect was observed in the mouse (16). Y980 is essential for full catalytic activity of human Jak3, while mutation of Y981 increases catalytic activity (15). In cAPK, the phosphorylated serine in the activation loop binds to conserved residues and forms one wall of the substrate binding region (46). Three tyrosines in the activation loop of Irk can be phosphorylated (Fig. 3). The tris-prosphorylated Irk permits unrestricted access to ATP and to substrate binding sites (20). PTKs are tightly regulated and their maximal enzymatic activity is not necessarily required. This may apply to Jak3 and possibly also to the other Jaks. The phosphorylated Y981 is predicted to point outward from the protein core and it could form an inter- or intramolecular docking site, which would block entry to the catalytic site when occupied. To test the feasibility of the possible new regulatory step, the Src SH2
114
VIHINEN ET AL.
FIG. 2. Left, superimposition of the Jak3 kinase domain (magenta) with the Irk kinase (yellow) (left), and right, the Jak3 pseudokinase (blue) with the Src kinase domain (red).
FIG. 3. Left, structure of the Jak3 kinase domain modeled based on Irk. The ATP analog is in red and the Mg 2⫹ ions are in green. The residues involved in ATP binding are in magenta, catalytic residues are in yellow, substrate binding residues in the P-1 site are in green, the P⫹1 site is in gray, the P⫹3 site is in dark blue, and tyrosines Y980 and Y981 are in orange. Right, location of the disease-causing mutations in the Jak3 kinase domain. L910 (top) and Y1023 (below) are in white. Residues corresponding to the mutations in the JH2 domain are indicated as follows: 520 –596 are in magenta except for 586 –592 in yellow and 565 in green. R582 is in orange, C759 in dark blue, 733 in magenta, and 722 in red. The images form a stereo pair.
MOLECULAR MODELING OF Jak3 KINASE DOMAINS
115
FIG. 4. Surface of the Jak3 kinase domain and Irk and cAPK kinase domains, from left to right. Inhibitor peptides are drawn with sticks in Irk and cAPK strcutures. The coloring is as in Fig. 3, left, except for phosphorylation sites in orange and inhibitor peptides in magenta.
domain with high-affinity peptide (PDB entry 1sps, 61) was docked without any major clashes into the Jak3 JH1 domain. Src is most likely not a Jak3 ligand, and some other SH2 domains are likely to fit even better. Another, although less likely, explanation for the inhibitory regulatory role of Y981 could be that the phosphate groups of both Y980 and Y981 might independently link the activation loop to the rest of the structure. Of these two alternative forms, the one involving binding and phosphorylation of Y980 might have remarkably higher activity; according to this hypothesis, the Y981F mutation would increase kinase activity by allowing phosphorylated Y980 to always bind to the activation loop. Additional examples of different regulatory activities of adjacent tyrosine residues have been reported in other kinases. The mutation of residues Y1162 and Y1163 in the activation loop increases the basal exogenous kinase activity of Irk (62), whereas the Y1158F mutant could still be activated. In Zap-70, Lck phosphorylates Y493 and Zap-70 can then autophosphorylate at Y492 and other tyrosines (63– 65). Mutation at Y493 decreases Zap-70 activity (63– 65). In contrast, phosphorylated Y492 inhibits kinase activity, and mutations at Y492 increase the basal kinase activity. The unphosphorylated activation loop of Hck (60) and c-Src (48) inactivates the kinase by forming a short 3 10- or ␣-helix, respectively, and by binding to the cleft between the lobes. Because Jak3 has no sequence similarity in this region to Src family, it is not possible to predict whether it has a similar inactivation mechanism.
In summary, the SCID-causing Jak3 mutations identified thus far have been interpreted in structural terms with the aid of the Jak3 JH1 and JH2 domain models. Residues involved in ATP and Mg 2⫹ binding were highly conserved whereas the substrate binding region is somewhat different compared to other kinases. Finally, the crucial role of Jak3 in lymphoid activation in vivo has been substantiated by studies in tumors. In particular, it has been shown that Jak3 is expressed in human B cell malignancies (66) and that proliferation of adult T cell leukemia/lymphoma cells is associated with the constitutive activation of the Jak/ STAT pathway (67, 68). Since Jak3 is preferentially expressed in hematopoietic tissues, it represents an ideal target for novel immune suppressive and chemotherapeutic drugs. Also from this point of view, the structural model of the Jak3 kinase and pseudokinase domains presented here may help identify the Jak3 structural motifs that are crucial for Jak3 function and thus offer the basis for a solid pharmacodesign. ACKNOWLEDGMENTS This work was supported by the Sigrid Juselius Foundation, National Technology Agency of Finland, Medical Research Fund of Tampere University Hospital, EU Biomed 2 concerted action PL 963007, Telethon Grant E668, C.N.R. (Progetto Finalizzato Biotecnologie), and N.A.T.O. Grant CRG.CRG 973041. REFERENCES 1. WHO Scientific Group, Primary immunodeficiency diseases. Clin. Exp. Immunol. 109 (Suppl. 1), 1–28, 1997.
116
VIHINEN ET AL.
2. Puck, J. M., de Saint Basile, G., Schwarz, K., Fugmann, S., and Fischer, R. E., IL2RGbase: A database of c-chain defects causing human X-SCID. Immunol. Today 17, 507–511, 1996. 3. Candotti, F., O’Shea, J. J., and Villa, A., Severe combined immune deficiencies due to defects of the common chain-JAK3 signaling pathway. Springer Semin. Immunopathol. 19, 401– 415, 1998. 4. Macchi, P., Villa, A., Giliani, S., Sacco, M. G., Frattini, A., Porta, F., Ugazio, A. G., Johnston, J. A., Candotti, F., O’Shea, J. J., Vezzoni, P., and Notarangelo, L., Mutations of Jak-3 gene in patients with autsomal severe combined immune deficiency (SCID). Nature 377, 65– 68, 1995. 5. Russell, S. M., Tayebi, N., Nakajima, H., Riedy, M. C., Roberts, J. L., Aman, M. J., Migone, T.-S., Noguchi, M., Markert, M. L., Buckley, R. H., O’Shea, J. J., and Leonard, W. J., Mutation of Jak3 in a patient with SCID: Essential role of Jak3 in lymphoid development. Science 270, 797– 800, 1995. 6. Ihle, J. N., Cytokine receptor signalling. Nature 377, 591–594, 1995. 7. O’Shea, J. J., Jaks, STATs, cytokine signal transduction, and immunoregulation: Are we there yet? Immunity 7, 1–11, 1997. 8. Taniguchi, T., Cytokine signaling through nonreceptor protein tyrosine kinases. Science 268, 251–255, 1995. 9. Cao, X., Shores, E. W., Hu-Li, J., Ansver, M. R., Kelsall, B. L., Russell, S. M., Drago, J., Noguchi, M., Grinberg, A., Bloom, E. T., Paul, W. E., Katz, S. I., Love, P. E., and Leonard, W. J., Defective lymphoid development in mice lacking expression of the common cytokine receptor gamma chain. Immunity 2, 223–238, 1995. 10. DiSanto, J. P., Rieux-Laucat, F., Dautry-Varsat, A., Fischer, A., and de Saint Basile, G., Defective human interleukin 2 receptor gamma chain in an atypical X chromosome-linked severe combined immunodeficiency with peripheral T cell. Proc. Natl. Acad. Sci. USA 91, 9466 –9470, 1994. 11. Nosaka, T., van Deursen, J. M. A., Tripp, R. A., Therfelder, W. E., Witthuhn, B. A., McMickle, A. P., Doherty, P. C., Grosveld, G. C., and Ihle, J. N., Defective lymphoid development in mice lacking Jak3. Science 270, 800 – 802, 1995. 12. Park, S. Y., Saijo, K., Takahashi, T., Osawa, M., Arase, H., Hirayama, N., Miyake, K., Nakauchi, H., Shirasawa, T., and Saito, T., Developmental defects of lymphoid cells in Jak3 kinase-deficient mice. Immunity 3, 771–782, 1995. 13. Thomis, D. C., Gurniak, C. B., Tivol, E., Sharpe, A. H., and Berg, L. J., Defects in B lymphocyte maturation and T lymphocyte activation in mice lacking Jak3. Science 270, 794 –797, 1995. 14. Chen, M., Cheng, A., Chen, Y. Q., Hymel, A., Hanson, E. P., Kimmel, L., Minami, Y., Taniguchi, T., Changelian, P. S., and O’Shea, J. J., The amino terminus of JAK3 is necessary and sufficient for binding to the common ␥ chain and confers the ability to transmit interleukin 2-mediated signals. Proc. Natl. Acad. Sci. USA 94, 6910 – 6915, 1997. 15. Zhou, Y.-J., Hanson, E. P., Chen, Y.-Q., Magnuson, K., Chen, M., Swann, P. G., Wange, R. L., Chagelian, P. S., and O’Shea, J. J., Distinct tyrosine phosphorylation sites in Jak3 kinase domain positively and negatively regulate its enzymatic activity. Proc. Natl. Acad. Sci. USA 94, 13850 –13855, 1997. 16. Liu, K. D., Gaffen, S. L., Goldsmith, M. A., and Greene, W. C., Janus kinases in interleukin-2 mediated signaling: Jak1 and Jak3 are differently regulated by tyrosine phosphorylation. Curr. Biol. 7, 817– 826, 1997. 17. Schumacher, R. F., Mella, P., Badolato, R., Fiorini, M., Savoldi, G., Giliani, S., Villa, A., Candotti, F., Tampalini, A., O’Shea, J. J., and Notarangelo, L. D., Complete genomic organization of the human JAK3 gene and mutation analysis in severe combined
18.
19.
20.
21.
22.
23.
24.
25. 26.
27.
28.
29.
30.
31.
32.
immunodeficiency by single-strand conformation polymorphism. Hum. Genet. 106, 73–79, 2000. Xu, W., Harrison, S. C., and Eck, M. J., Three-dimensional structure of the tyrosine kinase c-Src. Nature 385, 595– 602, 1997. Abola, E. E., Sussman, J. L., Prilusky, J., and Manning, N. O., Protein Data Bank archives of three-dimensional macromolecular structures. Methods Enzymol. 277, 556 –571, 1997. Hubbard, S. R., Crystal structure of the activated insulin receptor tyrosine kinase in complex with peptide substrate and ATP analog. EMBO J. 16, 5572–5581, 1997. Devereux, J., Haeberli, P., and Smithies, O., A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12, 387–395, 1984. Vihinen, M., Euranto, A., Luostarinen, P., and Nevalainen, O., MULTICOMP, a program package for multiple sequence comparison. Comput. Appl. Biosci. 8, 35–38, 1992. Boberg, J., Salakoski, T., and Vihinen, M., Representative selection of proteins based on nuclear families. Prot. Eng. 8, 501–503, 1995. Vihinen, M., Brandau, O., Brande´n, L., Kwan, S.-P., Lappalainen, I., Lester, T., Noordzij, J. G., Ochs, H. D., Ollila, J., Pienaar, S. M., Riikonen, P., Saha, B. K., and Smith, C. I. E., BTKbase, mutation database for X-linked agammaglobulinemia (XLA). Nucleic Acids Res. 26, 242–247, 1998. Etzold, T., and Argos, P., SRS—An indexing and retrieval tool for flat file data libraries. Comput. Appl. Biosci. 9, 49 –57, 1993. Rodriguez, R., Chinea, G., Lopez, N., Pons, T., and Vriend, G., Homology modeling, model and software evaluation: Three related resources. Bioinformatics 14, 523–528, 1998. Vihinen, M., Lehva¨slaiho, H., and Cotton, R. G. H., Immunodeficiency mutation databases. In “Primary Immunodeficiency Diseases. A Molecular and Genetic Approach” (H. D. Ochs, C. I. E. Smith, and J. Puck, Eds.), pp. 443– 447, Oxford Univ. Press, Oxford, 1998. Villa, A., Sironi, M., Macchi, P., Matteucci, C., Notarangelo, L., Vezzoni, P., and Mantovani, A., Monocyte function in a severe combined immunodeficient patient with a donor splice site mutation in the Jak3 gene. Blood 88, 817– 823, 1996. Candotti, F., Oaker, S. A., Johnston, J. A., Giliani, S., Schumacher, R. F., Mella, P., Fiorini, M., Ugazio, A. G., Badolato, R., Notaragelo, L. D., Bozzi, F., Macchi, P., Strina, D., Vezzoni, P., Blaese, R. M., O’Shea, J. J., and Villa, A., Structural and functional basis of Jak3-deficient severe combined immunodeficiency. Blood 90, 3996 – 4003, 1997. Bozzi, F., Lefranc, G., Villa, A., Badolato, R., Schumacher, R. F., Khalil, G., Loislet, J., Bresciani, S., O’Shea, J. J., Vezzoni, P., Notarangelo, L. D., and Candotti, F., Molecular and biochemical characterization of Jak3 deficiency in a patient with severe combined immunodeficiency over 20 years after bone marrow transplantation: Implications for treatment. Br. J. Haematol. 102, 1363–1366, 1998. Cacalano, N. A., Migone, T.-S., Bazan, F., Hanson, E. P., Chen, M., Candotti, F., O’Shea, J. J., and Johnston, J. A., Autosomal SCID caused by a point mutation in the N-terminus of Jak3: Mapping of the Jak3-receptor interaction domain. EMBO J. 18, 1549 –1558, 1999. Chen, M., Cheng, A., Candotti, F., Zhou, Y.-J., Hymel, A., Fasth, A., Notarangelo, L. D., and O’Shea, J. J., Complex effects of naturally occurring mutations in the JAK3 pseudokinase domain: Evidence for interactions between the kinase and pseudokinase domains. Mol. Cell. Biol. 20, 947–956, 2000.
MOLECULAR MODELING OF Jak3 KINASE DOMAINS 33. Riikonen, P., and Vihinen, M., MUTbase program package for maintenance and analysis of mutation databases on World Wide Web. Bioinformatics 15, 852– 859, 1999. 34. Stenberg, K. A. E., Riikonen, P. T., Vihinen, M., KinMutBase—A database of human disease-causing protein kinase mutations. Nucleic Acids Res. 27, 362–364, 1999. 35. Hubbard, S. R., Wei, L., Ellis, L., and Hendrickson, W. A., Crystal structure of the tyrosine kinase domain of the human insulin receptor. Nature 372, 746 –754, 1994. 36. Mohammadi, M., Schlessinger, J., and Hubbard, S. R., Structure of the FGF receptor tyrosine kinase domain reveals a novel autoinhibitory mechanism. Cell 86, 577–587, 1996. 37. Vihinen, M., Vetrie, D., Maniar, H. S., Ochs, H. D., Zhu, Q., Vorˇechovsky´, I., Webster, A. D. A., Notarangelo, L. D., Nilsson, L., Sowadski, J. M., and Smith, C. I. E., Structural basis for chromosome X-linked agammaglobulinemia: A tyrosine kinase disease. Proc. Natl. Acad. Sci. USA 91, 12803–12807, 1994. 38. Vihinen, M., Zvelebil, J. J. M., Zhu, Q., Brooimans, R. A., Ochs, H. D., Zegers, B. J. M., Nilsson, L., Waterfield, M. D., and Smith, C. I. E., Structural basis for pleckstrin homology domain mutations in X-linked agammaglobulinemia. Biochemistry 34, 1475– 1481, 1995. 39. Jin, H., Webster, A. D. B., Vihinen, M., Sideras, P., Vorˇechovsky´, I., Hammarstro¨m, L., Bernatowska-Matuszkiewicz, E., Smith, C. I. E., Bobrow, M., and Vetrie, D., Identification of Btk mutations in 20 unrelated patients with X-linked agammaglobulinemia (XLA) Hum. Mol. Genet. 4, 693–700, 1995. 40. Vorˇechovsky´, I., Vihinen, M., de Saint Basile, G., Honsova´, S., Hammarstro¨m, L., Mu¨ller, S., Nilsson, L., Fischer, A., and Smith, C. I. E., Mutation analysis and molecular modelling of Bruton’s tyrosine kinase in patients with X-linked agammaglobulinemia. Hum. Mol. Genet. 4, 51–58, 1995. 41. Saha, B. K., Curtis, S. K., Vogler, L. B., and Vihinen, M., Molecular and structural characterization of five novel mutations in the Bruton’s tyrosine kinase gene from patients with X-linked agammaglobulinemia. Molec. Med. 3, 477– 485, 1997. 42. Holinski-Feder, E., Weiss, M., Brandau, O., Jedele, K. B., Nore, B., Ba¨ckesjo¨, C.-M., Vihinen, M., Go¨tz, G., Hubbard, S. R., Belohradsky, B. H., Smith, C. I. E., and Meindl, A., Mutation screening of the btk gene in 56 families with X-linked agammaglobulinemia (XLA)— 47 unique mutations without correlation to clinical course. Pediatrics 101, 276 –284, 1998. 43. Luo, H., Rose, P., Barber, D., Hanratty, W. P., Lee, S., Roberts, T. M., D’Andrea, A. D., and Dearolf, C. R., Mutation in the Jak kinase JH2 domain hyperactivate Drosophila and mammalian Jak-STAT pathways. Mol. Cell. Biol. 17, 1562–1571, 1997. 44. Velazquez, L., Mogensen, K. E., Barbieri, G., Fellous, M., Uze, G., and Pellegrini, S., Distinct domains of the protein tyrosine kinase tyk2 required for binding of interferon-alpha/beta and for signal transduction. J. Biol. Chem. 270, 3327–3334, 1995. 45. Fujitani, Y., Hibi, M., Fukada, T., Takahashi-Tezuka, M., Yoshida, H., Yamaguchi, T., Sugiyama, K., Yamanaka, Y., Nakajima, K., and Hirano, T., An alternative pathway for STAT activation that is mediated by the direct interaction between Jak and STAT. Oncogene 14, 51–761, 1997. 46. Zheng, J., Knighton, D. R., Ten Eyck, L. F., Karlsson, R., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MgATP and peptide inhibitor. Biochemistry 32, 2154 –2161, 1993. 47. Bossemeyer, D., Engh, R. A., Kinzel, V., Ponstingl, H., and Huber, R., Phosphotransferase and substrate binding mechanism of the cAMP-dependent protein kinase catalytic subunit from porcine heart as deduced from the 2.0 Å structure of the
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
117
complex with Mn 2⫹ adenylyl imidodisphosphate and inhibitor peptide PKI(5-24). EMBO J. 12, 849 – 859, 1993. Xu, W., Doshi, A., Lei, M., Eck, M. J., and Harrison, S. C., Crystal structures of c-Src reveal features of its autoinhibitory mechanism. Mol. Cell 3, 629 – 638, 1999. De Bondt, H. L., Rosenblatt, J., Jancarik, J., Jones, H. D., Morgan, D. O., and Kim, S. H., Crystal structure of cyclin-dependent kinase 2. Nature 363, 595– 602, 1993. Knighton, D. R., Zheng, J., Ten Eyck, L. F., Ashford, V. A., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Crystal structure of the catalytic subunit of cyclic adenosine monophosphatedependent protein kinase. Science 253, 407– 414, 1991. Knighton, D. R., Zheng, J., Ten Eyck, L. F., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Structure of a peptide inhibitor bound to the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science 253, 414 – 420, 1991. Knighton, D. R., Bell, S. M., Zheng, J., Ten Eyck, L. F., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., 2.0 Å refined crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with a peptide inhibitor and detergent. Acta Crystallogr. D49, 357–361, 1993. Karlsson, R., Zheng, J., Xuong, N.-h., Taylor, S. S., and Sowadski, J. M., Structure of the mammalian catalytic subunit of cAMP-dependent protein kinase and an inhibitor peptide displays an open conformation. Acta Crystallogr. D49, 381–388, 1993. Zheng, J., Knighton, D. R., Xuong, N.-H., Taylor, S. S., Sowadski, J. M., and Ten Eyck, L. F., Crystal structures of the myristoylated catalytic subunit of cAMP-dependent protein kinase reveal open and closed conformations. Prot. Sci. 2, 1559 –1573, 1993. Zheng, J., Trafny, E. A., Knighton, D. R., Xuong, N.-h., Taylor, S. S., Ten Eyck, L. F., and Sowadski, J. M., 2.2 Å refined crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MnATP and peptide inhibitor. Acta Crystallogr. D49, 362–365, 1993. Madhusudan, Trafny, E. A., Xuong, N.-h., Adams, J. A., Ten Eyck, L. F., Taylor, S. S., and Sowadski, J. M., cAMP-dependent protein kinase: Crystallographic insights into substrate recognition and phosphotransfer. Prot. Sci. 3, 176 –187, 1994. Narayama, N., Cox, S., Xuong, N.-h., Ten Eyck, L. F., and Taylor, S. S., A binary complex of the catalytic subunit of cAMP-dependent protein kinase and adenosine further defines conformational flexibility. Structure 5, 921–935, 1997. Prade, L., Engh, R. A., Girod, A., Kinzel, V., Huber, R., and Bossemeyer, D., Staurosporine -induced conformational changes of cAMP-dependent protein kinase catalytic subunit explain inhibitory potential. Structure 5, 1627–1637, 1997. Kawahara, A., Minami, Y., Miyazaki, T., Ihle, J. N., and Taniguchi, T., Critical role of the interleukin 2 (IL-2) receptor -chainassociated Jak3 in the IL-2-induced c-fos and c-myc, but not bcl-2, gene induction. Proc. Natl. Acad. Sci. USA 92, 8724 – 8728, 1995. Schindler, T., Sicheri, F., Pico, A., Gazit, A., Levitzki, A., and Kuriyan, J., Crystal structure of Hck in complex with a Src family-selective tyrosine kinase inhibitor. Mol. Cell 3, 639 – 648. Waksman, G., Shoelson, S. E., Pant, N., Cowburn, D., and Kuriyan, J., Binding of a high affinity phosphotyrosyl peptide to the Src SH2 domain: crystal structures of the coplexed and peptide-free forms. Cell 72, 779 –790, 1993. Zhang, B., Tavare, J. M., Ellis, L., and Roth, R. A., The regulatory role of known tyrosine autophosphorylation sites of the insulin receptor kinase domain. An assessment by replacement with neutral and negatively charged amino acids. J. Biol. Chem. 266, 990 –996, 1991.
118
VIHINEN ET AL.
63. Chan, A. C., Dalton, M., Johson, R., Kong, G.-h., Wang, T., Thoma, R., and Kurosaki, T., Activation of ZAP-70 kinase by phosphorylation of tyrosine 493 is required for lymphocyte antigen receptor function. EMBO J. 14, 2499 –2508, 1995. 64. Wange, R. L., Guitia´n, R., Isakov, N., Watts, J. D., Aebersold, R., and Samelson, L. E., Activating and inhibitory mutations in adjacent tyrosines in the kinase domain of ZAP-70. J. Biol. Chem. 270, 18730 –18733, 1995. 65. Kong, G., Dalton, M., Wardenburg, J. B., Straus, D., Kurosaki, T., and Chan, A. C., Distinct tyrosine phosphorylation sites in ZAP-70 mediate activation and negative regulation of antigen receptor function. Mol. Cell. Biol. 16, 5026 –5035, 1996. 66. Tortolani, P. J., Lal, B. K., Riva, A., Johnston, J. A., Chen, Y.-Q., Reaman, G. R., Beckwith, M., Longo, D., Ortaldo, J. R., Bhatia, Received June 8, 1999; accepted with revision April 11, 2000
K., McGrath, I., Kehrl, J., Tuscano, J., McVicar, D. W., and O’Shea, J. J., Regulation of JAK3 expression and activation in human B cells and B cell malignancies. J. Immunol. 155, 5220 – 5226, 1995. 67. Migone, T.-S., Lin, J.-X., Cereseto, A., Mulloy, J. C., O’Shea, J. J., Franchini, G., and Leonard, W. J., Constitutively activated JakSTAT pathway in T cells transformed with HTLV-I. Science 269, 79 – 81, 1995. 68. Takemoto, S., Mulloy, J. C., Cereseto, A., Migone, T.-S., Patel, B. K. R., Matsuoka, M., Yamaguchi, K., Takatsuki, K., Kamihira, S., White, J. D., Leonard, W. J., Waldmann, T., and Franchini, G., Proliferation of adult T cell leukemia/lymphoma cells is associated with the constitutive activation of JAK/STAT proteins. Proc. Natl. Acad. Sci. USA 94, 13897–13902, 1997.