Patent Application
20050244911
The invention relates to the fields of protein kinases, autoimmune disease, autoimmune gets, and protein structure.
The idea that common pathogenic events exist at least for some autoimmune disorders is suggested by the significant number of patients displaying more than one autoimmune disease, and also by the strong and common linkage that some of these diseases show to specific MHC haplotypes. The experimental observation that the autoantigen is the leading moiety in autoimmunity and that a limited number of self-components are autoantigenic, suggests that these self-components share biological features which are relevant for self/non-self recognition by the immune system. One possibility is that triggering events by altering these features result in abnormal proteolysis. In certain individuals expressing a particular MHC specificity, the resulting abnormal peptides could be recognized by non-tolerized T cells and trigger an immune response
Type IV collagen (also referred to herein as collagen IV) networks scaffold the basement membranes, the laminar extracellular matrix structures often found between the cells and connective tissue. Six different type IV collagen α chains (α1-α6) exist, and three chains associate through the C terminal non-collagenous (NC1) domain to form a collagen IV molecule. In basement membranes, two type IV collagen molecules interact through their NC1 regions, yielding a hexameric globular quaternary structure (“hexamer”). Six disulfide bonds stabilize the native structure of each individual NC1 domain, and bonds generated by disulfide exchange between collagen IV molecules stabilize the “hexamer”. Bacterial collagenase digestion of basement membrane degrades the collagenous portion of collagen IV and releases the “hexamer”. Upon dissociation, this globular structure yields the individual NC1 domains as single polypeptides (“monomer”) or disulfide-related oligomers (dimers and higher molecular weight aggregates).
Recent data indicates that the information required to form a collagen IV “hexamer” resides in the covalent structure of the “monomer” as the individual NC1 domains select their partners and form “hexamers” without the assistance of other cellular factors. However the structural features mediating “monomer” association and the mechanism regulating the intermolecular disulfide bridging is presently unknown.
The chain composition of the collagen IV network varies among basement membranes and different collagen IV networks have been shown to exist. In the kidney, the glomerular basement membrane (GBM) results from assembly of two connected but independent collagen IV networks, one containing α1-α2(IV) and the other made of α3-α4-α5(IV). GBM plays a major role in plasma ultrafiltration since genetic and acquired diseases altering its collagen IV network impair renal function. In Alport syndrome, mutations in any of the α3, -α4 or -α5(IV) genes result in disruption of the corresponding collagen IV network and nephritis, whereas in Goodpasture (GP) disease an autoimmune response against the α3(IV)NC1 (also referred to as the GP antigen) cause linear deposits of autoantibodies along alveolar and glomerular BM, causing a rapidly progressive glomerulonephritis and often lung hemorrhage.
In GP disease, immunologically privileged epitopes buried in the GBM hexamer are exposed by an unknown pathogenic mechanism that engages the immune system in the deleterious production of antibodies. The human condition of this disorder and the exclusive involvement of the α3(IV)NC1 domain among six highly related domains, supported early comparative studies to identify biological features relevant in autoimmune pathogenesis. Accordingly, the human α3(IV)NC1 domain undergoes unique phosphorylation at Ser9 by type A protein kinases (cPKA) and structural diversification by alternative exon splicing generating multiple related products (GPΔIII, GPΔIII/IV/V and GPΔV).
The data presented herein indicate that the human α3(IV)NC1 domain exists as multiple phosphorylation-dependent conformational isoforms (conformers) that are stabilized by disulfide bonds. Furthermore our data indicate that phosphorylation of Ser9 induces conformational diversification of the α3(IV)NC1 domain, whereas the alternative products contain divergent C terminal ends that specifically induce cPKA phosphorylation of Ser9 in the primary product, suggesting that in humans the levels of expression of alternatively spliced products by regulating Ser9 phosphorylation control the conformational diversification process of the α3(IV)NC1 domain. All of the above suggests that Ser9 phosphorylation, alternative exon splicing and pathogenesis are related phenomenon.
The data presented herein further identify GPBP and GPBPΔ26 as two alternatively spliced isoforms of a novel non-conventional protein kinase that binds to the N terminal region of the human α3(IV)NC1 and phosphorylates Ser9. GPBP is a more active variant whose expression is highly restricted to histological structures targeted by common autoimmune responses including human alveolar and glomerular basement membranes. Each GPBP isoform likely represents a different strategy to perform the same function as we have found that for a particular tissue individuals expressing higher levels of GPBP express very little GPBPΔ26 and vice versa. An augmented expression of GPBP with respect to GPBPΔ26 has been associated with several autoimmune conditions including GP patients, cutaneous lupus erythematosus, pemphigus, pemphigoid and lichen planus, suggesting that GPBP expression and autoimmune pathogenesis are related processes. Our data herein (Example 5) further indicate that phosphorylation activates the α3(IV)NC1 domain for aggregation, a process that is catalyzed at least in part by GPBP and which comprises conformational isomerization reactions and disulfide-bond exchange.
Furthermore we show here that in GP kidneys, a relative increased in the level of expression of GPΔIII and GPBP co-exist with assembled “aberrant” conformers of the α3(IV)NC1 domain that conduct the autoimmune response, suggesting this human disease represents the legitimate response of the immune system against misfolded autoantigen which results from a coordinated increase in the expression of GPBP and GPΔIII.
Finally, we disclose that myelin basic protein (MBP), a known human autoantigen implicated in multiple sclerosis, contains a structurally related site (Ser8) for cPKA and GPBP whose phosphorylation regulates conformation and is under the control of a related alternative splicing mechanism when cPKA is phosphorylating enzyme, suggesting that phosphorylation-dependent conformation is the biological condition that renders self-components potentially immunogenic.
Based on all of the above, there exists a need in the art for methods and reagents to identify drug candidates to modify GPBP activity to treat autoimmune disorders.
The present invention provides methods and reagents for identifying compounds to treat autoimmune diseases. In one aspect, the present invention provides methods for identifying compounds to treat an autoimmune condition, comprising identifying compounds that (a) reduce phosphorylation of a first target protein selected from the group consisting of GPBP, an α3 type IV collagen NC1 domain polypeptide comprising the amino acid sequence of SEQ ID NO:26, and a polypeptide comprising the amino acid sequence of SEQ ID NO:64 and (b) reduce formation of conformational isomers of a second target protein selected from the group consisting of an α3 type IV collagen NC1 domain polypeptide and myelin basic protein, wherein such compounds are candidates for treating an autoimmune condition. In a preferred embodiment, phosphorylation assays are conducted in vitro. In a further preferred embodiment, conformer formation assays are conducted in cultured cells. In another preferred embodiment, the method further comprises identifying compounds that reduce oligomerization of the second target protein. In a further embodiment, the autoimmune condition is selected from the group consisting of Goodpasture Syndrome, multiple sclerosis, systemic and cutaneous lupus erythematosus, pemphigus, pemphigoid and lichen planus.
In another aspect, the invention provides isolated type IV collagen α3 NC1 domain conformational isomers, wherein the isolated conformational isomer has an amino acid sequence identical to that of wild type α3 type IV collagen NC1 domain, wherein the conformational isomer is stabilized by disulfide bonds, wherein the isolated conformational isomer has a molecular weight in a non-reducing sodium dodecyl sulfate gel selected from the group consisting of 22 kD, 23, kD, 25 kD, 27 kD, and 28 kD, and wherein the conformational isomer has a molecular weight of 29 kDa in a reducing sodium dodecyl sulfate gel.
In a further embodiment, the invention provides isolated type IV collagen α3 NC1 domain nucleic acids encoding a polypeptide consisting of an amino acid sequence selected from the group consisting of SEQ ID NO:66 and SEQ ID NO:68, as well as the corresponding isolated polypeptides.
The abbreviations used herein are: BM, basement membrane; bp, base pair; DTT, dithiothreitol; DMEM, Dulbecco's modified Eagle's medium; EDTA, ethylenediamine tetraacetic acid; EGTA, ethylene glycol-bis(β-aminoethyl ether) N,N,N′,N′-tetraacetic acid; GBM, glomerular basement membrane; GP, Goodpasture; rGPΔIII, rGPΔIII/IV/V and rGPΔV, recombinant material representing the alternative forms of the Goodpasture antigen resulting from splicing out exon III, exon III, IV and V or exon V, respectively; GPBP and rGPBP, native and recombinant Goodpasture antigen binding protein; GPBPΔ26 and rGPBPΔ26, native and recombinant alternative form of the GPBP; GSH and GSSG, glutathione reduced and oxidized respectively; HLA, human lymphocyte antigens; HPLC, high performance liquid chromatography; Kb, thousand base pairs; kDa (or kD), thousand daltons; MBP, rMBP, native and recombinant 21 kDa myelin basic protein; MBPΔII and rMBPΔII, native and recombinant 18.5 kDa myelin basic protein that results from splicing out exon II; MBPΔV and MBPΔII/V, myelin basic protein alternative forms resulting from splicing out exon V and exons II and V, respectively; MHC, major histocompatibility complex; NC1, non-collagenous domain; PH, pleckstrin homology; PDI, protein disulfide isomerase; PKA, cPKA, cAMP-dependent protein kinase and catalytic subunit thereof; PMSF, phenylmethylsulfonyl fluoride; SDS-PAGE, sodium dodecylsulfate polyacrylamide gel electrophoresis; TBS, tris buffered saline.
Within this application, unless otherwise stated, the techniques utilized may be found in any of several well-known references such as: Molecular Cloning: A Laboratory Manual (Sambrook, et al., 1989, Cold Spring Harbor Laboratory Press), Gene Expression Technology (Methods in Enzymology, Vol. 185, edited by D. Goeddel, 1991. Academic Press, San Diego, Calif.), “Guide to Protein Purification” in Methods in Enzymology (M. P. Deutshcer, ed., (1990) Academic Press, Inc.); PCR Protocols: A Guide to Methods and Applications (Innis, et al. 1990. Academic Press, San Diego, Calif. ), Culture of Animal Cells: A Manual of Basic Technique, 2nd Ed (R. I. Freshney. 1987. Liss, Inc. New York, N.Y.), Gene Transfer and Expression Protocols, pp. 109-128, ed. E. J. Murray, The Humana Press Inc., Clifton, N.J.), and the Ambion 1998 Catalog (Ambion, Austin, Tex.).
As used herein, the term “GPBP” refers to Goodpasture binding protein, and includes both monomers and oligomers thereof. Human (SEQ ID NO:2), mouse (SEQ ID NO:4), and bovine GPBP sequences (SEQ ID NO:6) are provided herein.
As used herein, the term “GPBPΔ26” refers to Goodpasture binding protein deleted for the 26 amino acid sequence shown in SEQ ID NO:14, and includes both monomers and oligomers thereof. Human (SEQ ID NO:8), mouse (SEQ ID NO:10), and bovine GPBP sequences (SEQ ID NO:12) are provided herein.
As used herein the term “GPBPpep1” refers to the 26 amino acid peptide shown in SEQ ID NO:14, and includes both monomers and oligomers thereof.
As used herein, the term “GP antigen” refers to the α3 NC1 domain of type IV collagen.
As used herein, the terms “an α3 NC1 domain of type IV collagen” and “α3(IV)NC1” includes all conformational isomers thereof, and oligomers thereof, and also includes the α3(IV)NC1 mutants, α3(IV)NC1Asp9 (SEQ ID NO:66) and α3(IV)NC1Ala9 (SEQ ID NO: 68), conformational isomers thereof and oligomers thereof, described below.
As used herein, the term “α3(IV)NC1Ser9” means the wilt type α3 NC1 domain of type IV collagen.
As used herein, the term “protein kinase A” refers to the cAMP-dependent protein kinase.
As used herein, “MBP” refers to myelin basic protein.
As used herein, “test compound” refers to any substance that is tested for ability to produce the desired effects as discussed herein. It will be understood that such test compounds can be added to the various assays at a wide variety of concentrations in order to determine their effect on the results of the assay.
The inventor has discovered that GPBP, a non-conventional protein kinase that in vitro binds to and phosphorylates α3(IV)NC1, the autoantigen in Goodpasture disease, also possesses chaperone, chaperonine, and protein disulfide isomerase (PDI) activities. The present invention demonstrates that GPBP activity includes (1) aggregate disruption (typical chaperone activity); (2) folding catalysis into multiple conformations (atypical chaperonine activity, since typically chaperonines catalyzes only one conformation) and (3) intra and intermolecular disulfide-bond shuffling. The present invention has established the importance of these activities in the autoimmune process, as well as the general importance of autoantigen aberrant phosphorylation and conformational isomerization, which can be influenced by factors in addition to GPBP.
In one aspect, the present invention provides methods for identifying compounds to treat an autoimmune condition, comprising identifying compounds that (a) reduce phosphorylation of a first target protein selected from the group consisting of GPBP, an α3 type IV collagen NC1 domain polypeptide comprising the amino acid sequence of SEQ ID NO:26, and a polypeptide comprising the amino acid sequence of SEQ ID NO:64; and (b) reduce formation of conformational isomers of a second target protein selected from the group consisting of an α3 type IV collagen NC1 domain polypeptide and myelin basic protein, wherein such compounds are candidates for treating an autoimmune condition. Thus the first and second target proteins can different (for example, when GPBP is the first target and an α3 type IV collagen NC1 domain polypeptide is the second target protein; or when GPpep1 is the first target and an α3 type IV collagen NC1 domain polypeptide is the second target), or they can be identical.
The phosphorylation assays can be conducted in vitro on isolated targets, or can comprise analyzing the effects of the one or more test compounds on phosphorylation in cultured cells, although in vitro assays are preferred. A preferred method for identifying compounds that reduce in vitro phosphorylation of the target protein comprises:
One of skill in the art is capable of determining suitable phosphorylation conditions for conducting the phosphorylation assay, and thus the present method is not limited by the details of the particular phosphorylation conditions employed. A non-limiting example of such suitable conditions for assaying phosphorylation of the first target comprises the use of 0.5 μg to 5 μg of the first target protein, Hepes buffer pH 7.5, and 5 mM MgCl2, optionally including 1 mM DTT, depending on the first target protein.
In a further preferred embodiment, the first target protein is GPBP, and the assay comprises analyzing the effect(s) of the one or more test compounds on GPBP autophosphorylation. In such an embodiment, an exemplary amount of GPBP for use in the assay is between 50 to 200 ng. In an alternative embodiment, the first target protein is selected from the group consisting of an α3 type IV collagen NC1 domain polypeptide comprising the amino acid sequence of SEQ ID NO:26, and an MBP polypeptide comprising the amino acid sequence of SEQ ID NO:64, and the assay is conducted in the presence of GPBP to test for transphosphorylation of the first target protein by the protein kinase. In this embodiment, the first target protein can comprise a full length α3 type IV collagen NC1 domain polypeptide (including α3(IV)NC1Asp9 SEQ ID NO:66 or α3(IV)NC1Ala9 SEQ ID NO:68), full length MBP, or any fragments thereof containing the recited sequence.
For in vitro phosphorylation assays, detection of phosphorylation can be accomplished by any number of means, including but not limited to using 32P labeled ATP and carrying out autoradiography of a Western blot of the resulting protein products on a reducing or non-reducing gel, or by scintillation counting after a step to separate incorporated from unincorporated label.
Analysis of in vitro phosphorylation may further include identifying the effect of the one or more test compounds on phosphorylation of individual conformational isomers of the first target protein, when the first target protein is selected from the group consisting of an α3 type IV collagen NC1 domain polypeptide and MBP. Such identification can be accomplished, for example, by carrying out SDS-PAGE on the reaction products of the phosphorylation reaction, followed by Western blotting, autoradiography and immunodetection of the target protein.
Analysis of in vitro phosphorylation may further include identifying the effect of the one or more test compounds on Ser9 phosphorylation of the α3 type IV collagen NC1 domain. Such identification can be accomplished, for example, by comparing the immunoreactive patterns of antibodies specifically reacting with the N terminus of the α3(IV)NC1 (including but not limited to anti-GPpep1, MabM3/1 and MabP1/2, disclosed herein) and antibodies specifically reacting with Ser(P), such as those commercially available from Sigma Chemical Co. (St. Louis, Mo.). Alternatively, V8 protease digestion and anti-GPpep1 precipitation followed by reducing SDS-PAGE on the precipitated products and autoradiography can be used.
The data presented herein demonstrate that phosphorylation at Ser9 exerts a positive control over conformational isomerization of α3(IV)NC1, and efficiently changes the cohort of α3(IV)NC1 conformers produced by a cell. These findings indicate that Ser9 is, at least in part, the structural feature that renders the α3(IV)NC1 domain immunogenic, and suggest that, during pathogenesis, a phosphorylation event lead the formation of conformers for which the immune system has not established a tolerance. Thus, determining the effect of test compounds on phosphorylation of the Ser9 residue of α3 type IV collagen NC1 domain may be important in identifying especially useful candidate compounds for treating autoimmune disorders.
Alternatively, the effects of test compounds on phosphorylation of the first target protein can be analyzed in cultured cells. Such a method involves contacting cells that express a first target protein selected from the group consisting of an α3 type IV collagen NC1 domain polypeptide and MBP, under conditions to promote phosphorylation, detecting phosphorylation of the first target protein; and identifying test compounds that reduce phosphorylation of the first target protein relative to phosphorylation of the first target protein in the absence of the one or more test compounds. Appropriate cells for use are eukaryotic cells that express the appropriate first target protein. Methods of detecting phosphorylation are as described above.
As used herein, the phrase “reduce/reducing phosphorylation” means to lessen the phosphorylation of the target protein relative to phosphorylation of the target protein in the absence of the one or more test compounds. Such “reducing” does not require elimination of phosphorylation, and includes any detectable reduction in phosphorylation. Thus, a test compound that inhibits phosphorylation of the first target by, for example, as little as 10-20% would be considered a test compound that reduced phosphorylation. Such a compound may, for example, affect phosphorylation of Ser9, which is shown to exert a powerful control on conforinational diversification, and thus to be a strong candidate for an inhibitor of autoimmunity. Alternatively, a test compound may inhibit phosphorylation of a first target protein, such as an α3 type IV collagen NC1 domain polypeptide comprising the amino acid sequence of SEQ ID NO:26 by 90%, but have little inhibitory effect on conformational isomerization of the second target protein, because reduction affects phosphorylation at sites other than Ser9. By performing assays both for phosphorylation inhibition of the first target protein, and conformer inhibition of the second target protein, it is possible to identify those compounds with the best potential for use as therapeutics for autoimmune disorders.
Similarly, inhibition of conformational isomerization of the second target protein can be carried out in vitro using isolated components, or can be carried out in cultured cells, although the use of cultured cells is preferred. In a preferred embodiment using cultured cells, identifying compounds that reduce formation of conformational isomers of the second target protein comprises:
Appropriate cells for use are eukaryotic cells that express the appropriate second target protein. In a preferred embodiment, cell lines stably transfected to express the second target protein are used.
In this embodiment, detection of conformational isomers of, for example, the α3 type IV collagen NC1 domain polypeptide, and the effects of the test compounds thereon, generally involve immunodetection using Western blots of non-reducing SDS-PAGE gels containing the α3 type IV collagen NC1 domain polypeptide from the cells. The α3 type IV collagen NC1 domain polypeptide can be purified via standard techniques (such as using cells transfected with a recombinant second target protein that is linked to an epitope tag or other tag to facilitate purification), or cell extracts can be analyzed. In a most preferred embodiment, stable cell lines (such as those disclosed herein) expressing recombinant α3(IV)NC1 are used, which secrete the protein into the medium in a monomeric form, permitting running of serum-free media samples on SDS-PAGE gels and subsequent Western blot analysis and immunodetection. Preferably, immunodetection is carried out using, in parallel, an antibody that detects a native conformation of α3 type IV collagen NC1 domain polypeptide (including but not limited to Mab3 disclosed herein), and an antibody that detects all α3 type IV collagen NC1 domain polypeptide conformational isomers (including but not limited to Mab175 disclosed herein). Alternatively, serum free media or otherwise isolated proteins could be used to coat ELISA plates, followed by similar immunodetection using antibodies that selectively bind to native conformers and either aberrant conformers or all conformers, respectively, and analysis using plate readers.
In a preferred embodiment of an in vitro assay for inhibitors of conformational isomerization of the second target protein, the method comprises
As used herein, the phrase “reduce/reducing conformational isomerization” means to lessen the formation of conformers of the target protein relative to conformer production under control conditions. Such “reducing” does not require elimination of conformer formation, and includes any detectable reduction in conformer formation. Furthermore, such “reduction in conformer formation” may entail a reduction in only one, or fewer than all conformational isomers; one can envision that such a reduction in production of specific conformers may be accompanied by an increase in the formation of other conformers. For example, we show in the examples to follow that, for the α3 type IV collagen NC1 domain polypeptide, a 27 kD conformer is the primary product, from which the remaining conformers derive. Thus, in a further preferred embodiment, the method comprises identifying those compounds that do not alter the formation of the 27-kDa conformer, but reduce formation of one or more of the other conformers. A preferred method for monitoring this inhibition of specific conformers is to use Mab3 antibody (described below), which only reacts with the 27-kDa conformer, in parallel with Mab175, which is equally reactive with all α3 type IV collagen NC1 domain conformers.
In a further preferred embodiment of the assays to identify inhibitors of conformational isomerization of the second target protein, the second target protein is an α3 type IV collagen NC1 domain polypeptide, and analysis of test compound effect on conformer formation of each of wild type α3(IV)NC1 and α3(IV)NC1Asp9 (SEQ ID NO:66) is carried out in parallel. α3(IV)NC1Asp9 is modified to replace Ser9 with Asp9, an amino acid residue that mimics an always phosphorylated residue, which is used herein as an example of an aberrant phosphorylation of α3(IV)NC1, that leads to the production of aberrant conformers, as demonstrated in the Examples to follow. In example 4, we show that α3(IV)NC1Asp9 expressing cells produce a larger number of conformers than cells expressing α3(IV)NC1Ser9. Furthermore α3(IV)NC1Asp9 cells expresses a 27-kDa conformer that reacts more strongly with Mab3, as well as Goodpasture patient autoantibodies, than the 27-kDa conformer produced by α3(IV)NC1Ser9 expressing cells. It is most preferred to identify compounds that abolish these differences in conformer production between α3(IV)NC1Asp9 and α3(IV)NC1Ser9, because this will indicate that the compound inhibits the production of an aberrant 27-kDa conformer from α3(IV)NC1Asp9, while maintaining appropriate conformer production for α3(IV)NC1Ser9.
In a further preferred embodiment, identifying compounds for treating an autoimmune disorder further comprises identifying compounds that reduce oligomerization of the second target protein. While not being limited by a specific mechanism, the inventor proposes that the ideal drug candidate for treating autoimmune disorders would inhibit the kinase and chaperonine activity of GPBP, but would not inhibit its chaperone (ie: aggregate-disrupting) activity, in order to minimize the possibility that inhibition of GPBP activity would lead to increased random aggregate formation. Even more preferably, the ideal drug candidate would, in fact, enhance the chaperone activity of GPBP, to minimize secondary effects derived from undesirable aggregation of conformers.
Both in vitro assays and assays utilizing cultured cells can be used fore identifying compounds that reduce oligomerization of the second target protein, although in vitro methods are preferred. One embodiment of an in vitro assay comprises:
In a preferred embodiment, the second target protein is an α3(IV)NC1 domain polypeptide. Any appropriate redox system can be used, such as DTT/Mn2+ (exemplified in Material and Methods of Example 5 below), or with GSH/GSSG (glutathione reduced and oxidized respectively) at 1.0 mM/0.2 mM at pH 8.0 in a similar buffer.
One of skill in the art will be able to determine appropriate conditions for promoting GPBP-induced oligomerization of the second target protein, and thus the method is not limited to specific details of the conditions. A non-limiting example of such conditions is provided in Example 5 below.
Detection of oligomers, and the effect of test compounds thereon, is preferably carried out by Western blotting of a non-reducing SDS-PAGE gel of the isolated recombinant α3 type IV collagen NC1 domain polypeptides after incubation, and probing with antibodies that recognize the α3 type IV collagen NC1 domain polypeptides. Preferably, immunodetection is carried out using, in parallel, an antibody that detects a native conformation of α3 type IV collagen NC1 domain polypeptide (including but not limited to Mab3 disclosed herein), and an antibody that detects all α3 type IV collagen NC1 domain polypeptide conformational isomers (including but not limited to Mab175 disclosed herein).
In a preferred embodiment of the oligomerization assay using cultured cells, cells that express type IV collagen are contacted with the one or more test compounds, and the extracellular matrix produced by the cells is collagenase digested and analyzed for α3(IV)NC1 oligomers by Western blot analysis as described herein.
As used herein the phrase “reduce/reducing GPBP induced disulfide-mediated oligomerization of the α3 type IV collagen NC1 domain polypeptide” means to decrease the amount of GPBP induced disulfide-mediated oligomers of the α3 type IV collagen NC1 domain polypeptide relative to oligomerization under control conditions. Such “reducing” does not require elimination of oligomer formation, and includes any detectable reduction in oligomer formation, including reduction in only a single species of oligomer in the presence of increased in other species of oligomers.
In another aspect, the present invention provides isolated nucleic acids that encode α3(IV)NC1(Asp9) (SEQ ID NO:66) and α3(IV)NC1(Ala9) (SEQ ID NO:68). The production and use of these mutant α3(IV)NC1 domains are described below. The nucleic acid sequences are useful, for example, for the production of the respective encoded polypeptide.
An used herein, an “isolated nucleic acid sequence” refers to a nucleic acid sequence that is free of gene sequences which naturally flank the nucleic acid in the genomic DNA of the organism from which the nucleic acid is derived (i.e., genetic sequences that are located adjacent to the gene for the isolated nucleic molecule in the genomic DNA of the organism from which the nucleic acid is derived). An “isolated” nucleic acid sequence according to the present invention may, however, be linked to other nucleotide sequences that do not normally flank the recited sequence, such as a heterologous promoter sequence. It is not necessary for the isolated nucleic acid sequence to be free of other cellular material to be considered “isolated”, as a nucleic acid sequence according to the invention may be part of an expression vector that is used to transfect host cells
In another aspect, the present invention provides recombinant expression vectors comprising nucleic acid sequences that encode α3NC1(Asp9) (SEQ ID NO:66) or α3NC1(Ala9) (SEQ ID NO:68). In one embodiment, the vectors comprise nucleic acid sequences consisting of the sequences shown in SEQ ID NO:65 or SEQ ID NO:67.
“Recombinant expression vector” includes vectors that operatively link a nucleic acid coding region or gene to any promoter capable of effecting expression of the gene product. The promoter sequence used to drive expression of the disclosed nucleic acid sequences in a mammalian system may be constitutive (driven by any of a variety of promoters, including but not limited to, CMV, SV40, RSV, actin, EF) or inducible (driven by any of a number of inducible promoters including, but not limited to, tetracycline, ecdysone, steroid-responsive). The construction of expression vectors for use in transfecting prokaryotic cells is also well known in the art, and thus can be accomplished via standard techniques. (See, for example, Sambrook, Fritsch, and Maniatis, in: Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory Press, 1989; Gene Transfer and Expression Protocols, pp. 109-128, ed. E. J. Murray, The Humana Press Inc., Clifton, N.J.), and the Ambion 1998 Catalog (Ambion, Austin, Tex.)
The expression vector must be replicable in the host organism either as an episome or by integration into host chromosomal DNA. In a preferred embodiment, the expression vector comprises a plasmid. However, the invention is intended to include other expression vectors that serve equivalent functions, such as viral vectors.
The expression vector may encode additional sequences that are operably linked to the nucleic acid encoding that encode α3(IV)NC1(Asp9) (SEQ ID NO:66) and α3(IV)NC1(Ala9) SEQ ID NO:68). Such additional sequences can encode, for example, amino acid sequences useful for promoting purification of the protein, such as epitope tags and transport signals. Examples of such epitope tags include, but are not limited to FLAG (Sigma Chemical, St. Louis, Mo.), myc (9E10) (Invitrogen, Carlsbad, Calif.), 6-His (Invitrogen; Novagen, Madison, Wis.), and HA (Boehringer Manheim Biochemicals). Examples of such transport signals include, but are not limited to, export signals, secretory signals, nuclear localization signals, and plasma membrane localization signals. Other examples of additional sequences include, but are not limited to, polyadenylation signals to effect proper polyadenylation of the transcript, and termination signals.
In a further aspect, the present invention provides host cells that have been transfected with the recombinant expression vectors disclosed herein, wherein the host cells can be either prokaryotic or eukaryotic. The cells can be transiently or stably transfected. Such transfection of expression vectors into prokaryotic and eukaryotic cells can be accomplished via any technique known in the art, including but not limited to standard bacterial transformations, calcium phosphate co-precipitation, electroporation, or liposome mediated-, DEAE dextran mediated-, polycationic mediated-, or viral mediated transfection. (See, for example, Molecular Cloning: A Laboratory Manual (Sambrook, et al., 1989, Cold Spring Harbor Laboratory Press; Culture of Animal Cells: A Manual of Basic Technique, 2nd Ed. (R. I. Freshney. 1987. Liss, Inc. New York, N.Y.),
In a still further aspect, the present invention provides isolated polypeptides selected from the group consisting of α3(IV)NC1Asp9 (SEQ ID NO:66) and α3(IV)NC1Ala9 (SEQ ID NO:68). These polypeptides represent mutant α3(IV)NC1, which have been substitute at the Ser9 residue to mimic an always phosphorylated position 9 (Asp9), or an always un-phosphorylated position 9 (Ala9). As described herein, such α3(IV)NC1 mimics can be used, for example, in carrying out the drug discovery assays of the invention, as described above.
As used herein, “α3(IV)NC1Asp9” and “α3(IV)NC1Ala9” include all conformational isomers, as well as oligomers thereof.
The protein may comprise additional sequences useful for promoting purification of the protein, such as epitope tags and transport signals. Examples of such epitope tags include, but are not limited to FLAG (Sigma Chemical, St. Louis, Mo.), myc (9E10) (Invitrogen, Carlsbad, Calif.), 6-His (Invitrogen; Novagen, Madison, Wis.), and HA (13oehringer Manheim Biochemicals). Examples of such transport signals include, but are not limited to, export signals, secretory signals, nuclear localization signals, and plasma membrane localization signals.
The experiments described below disclose the isolation of type IV collagen α3 NC1 domain conformational isomers (“conformers”). Thus, in a further embodiment, the present invention provides an isolated type IV collagen α3 NC1 domain conformational isomer, wherein the isolated conformational isomer has an amino acid sequence identical to that of wild type α3 type IV collagen NC1 domain (SEQ ID NO:69), wherein the conformational isomer is stabilized by disulfide bonds, wherein the isolated conformational isomer has a molecular weight in a non-reducing sodium dodecyl sulfate gel selected from the group consisting of 22 kD, 23, kD, 25 kD, 27 kD, and 28 kD, and wherein the conformational isomer has a molecular weight of 29 kDa in a reducing sodium dodecyl sulfate gel.
Isolation of the conformers can be accomplished by separation of the conformers on a non-reducing SDS-PAGE gel, cutting out of the relevant bands from the gel, and isolating the conformer away from the gel components. Alternatively, such conformers can be isolated by HPLC methods, such as those described in Example 4, below.
The invention further comprises an isolated, aberrant conformational isomer of α3(IV)NC1Asp9, wherein the isomer has the amino acid sequence of SEQ ID NO:66, wherein the conformational isomer is stabilized by disulfide bonds, wherein the isolated conformational isomer has a molecular weight in a non-reducing sodium dodecyl sulfate gel selected from the group consisting of 25 kD and 27 kD, and wherein the conformational isomer has a molecular weight of 29 kDa in a reducing sodium dodecyl sulfate gel.
As used herein, the term “isolated” means that the conformer is separated from its cellular environment, and purified away from any gel matrix, such as polyacrylamide. Such “isolated” conformers are substantially separated from other conformers, such that a particular “isolated conformer” constitutes at least 70% of the type IV collagen α3 NC1 domain polypeptide present in the isolated sample, more preferably 80%, even more preferably 90%, and even more preferably more than 95%. Such “isolated” conformers can be suspended in any appropriate buffer or pharmaceutical composition, and are useful, for example, for preparing antibodies to specific conformers, and for use in the drug discovery assays of the invention.
The present invention may be better understood with reference to the accompanying examples that are intended for purposes of illustration only and should not be construed to limit the scope of the invention, as defined by the claims appended hereto.
Here we report the cloning and characterization of a novel type of serine/threonine kinase that specifically binds to and phosphorylates the unique N-terminal region of the human GP antigen.
Synthetic polymers-Peptides. GPpep1, KGKRGDSGSPATWTTRGFVFT (SEQ ID NO:26), representing residues 3-23 of the human GP antigen and GPpep1Ala9, KGKRGDAGSPATWTTRGFVFT (SEQ ID NO:27), a mutant Ser9 to Ala9 thereof, were synthesized by MedProbe and CHIRON. FLAG peptide, was from Sigma
Oligonucleotides. The following as well as several other GPBP-specific oligonucleotides were synthesized by Genosys and GIBCO BRL:
Isolation and characterization of cDNA clones encoding human GPBP-Several human λ-gt11 cDNA expression libraries (eye, fetal and adult lung, kidney and HeLa S3, from CLONTECH) were probed for cDNAs encoding proteins interacting with GPpep1. Nitrocellulose filters (Millipore) prepared following standard immunoscreening procedures were blocked and incubated with 1-10 nmoles per ml of GPpep1 at 37° C. Specifically bound GPpep1 was detected using M3/1A monoclonal antibodies (7). A single clone was identified in the HeLa-derived library (HeLa1). Specificity of fusion protein binding was confirmed by similar binding to recombinant eukaryotic human GP antigen. The EcoRI cDNA insert of HeLa1 (0.5-kb) was used to further screen the same library and to isolate overlapping cDNAs. The largest cDNA (2.4-kb) containing the entire cDNA of HeLa1 (n4′) was fully sequenced.
Northern and Southern blots-Pre-made Northern and Southern blots (CLONTECH) were probed with HeLa1 cDNA following manufacturer instructions.
Plasmid construction, expression and purification of recombinant proteins-GPBP-derived material. The original λ-gt11 HeLa1 clone was expressed as a lysogen in E. Coli Y1089 (8). The corresponding β-galactosidase-derived fusion protein containing the N-terminal 150 residues of GPBP was purified from the cell lysate using an APTG-agarose column (Boehringer). The EcoRI 2.4-kb fragment of n4′ was subcloned in Bluescribe M13+ vector (Stratagene) (BS-n4′), amplified and used for subsequent cloning. A DNA fragment containing (from 5′ to 3′), an EcoRI restriction site, a standard Kozak consensus for translation initiation, a region coding for a tag peptide sequence (FLAG, DYKDDDDK (SEQ ID NO:32)), and the sequence coding for the first eleven residues of GPBP including the predicted Met; and a Ban II restriction site, was obtained by hybridizing ON-GPBP-54m and ON-GPBP-55c, and extending with modified T7 DNA polymerase (Amersham). The resulting DNA product was digested with EcoRI and BanII, and ligated with the BanII/EcoRI cDNA fragment of BS-n4′ in the EcoRI site of pIL-D2 (Invitrogen) to produce pHIL-FLAG-n4′.
This plasmid was used to obtain MutS transformants of the GS115 strain of Pichia pastoris and to express FLAG-tagged recombinant GPBP (rGPBP) either by conventional liquid culture or by fermentation procedures (Pichia Expression Kit, Invitrogen). The cell lysates were loaded onto an anti-FLAG M2 column (Sigma), the unbound material washed out with Tris buffered saline (TBS, 50 mM Tris-HCl, pH 7.4, 150 mM NaCl) or salt-supplemented TBS (up to 2M NaCl), and the recombinant material eluted with FLAG peptide.
For expression in cultured human kidney-derived 293 cells (ATCC 1573-CRL), the 2.4- or 2.0-kb EcoRI cDNA insert of either BS-n4′ or pHIL-FLAG-n4′ was subcloned in pcDNA3 (Invitrogen) to produce pc-n4′ and pc-FLAG-n4′ respectively. When used for transient expression, 18 hours after transfection the cells were lysed with 3.5-4 μl/cm2 of chilled lysis buffer (1% Nonidet P-40 or Triton-X100, 5 mM EDTA and 1 mM PMSF in TBS) with or without 0.1% SDS, depending on whether the lysate was to be used for SDS-PAGE or FLAG-purification, respectively. For FLAG purification, the lysate of four to six 175 cm2 culture dishes was diluted up to 50 ml with lysis buffer and purified as above.
For stable expression, the cells were similarly transfected with pc-n4′ and selected for three weeks with 800 μg/ml of G418. For bacterial recombinant expression, the 2.0-kb EcoRI cDNA fragment of pHIL-FLAG-n4′ was cloned in-frame downstream of the glutathione S-transferase (GST)-encoding cDNA of pGEX-5×−1 (Pharmacia). The resulting construct was used to express GST-GPBP fusion protein in DH5α cells (9).
GP antigen-derived material. Human recombinant GP antigen (rGP) was produced in 293 cells using the pRc/CMV-BM40 expression vector containing the α3-specific cDNA between ON-HNC-B-N-14m and ON-HNC-B-N-16c. The expression vector is a pRc/CMV (Invitrogen)-derived vector provided by Billy G. Hudson (Kansas University Medical Center) that contains cDNA encoding an initiation Met, a BM40 signal peptide followed by a tag peptide sequence (FLAG), and a polylinker cloning site. To obtain α3-specific cDNA, a polymerase chain reaction was performed using the oligonucleotides above and a plasrnid containing the previously reported α3(IV) cDNA sequence (3) as template (clone C2). For stable expression of rGP, 293 cells were transfected with the resulting construct (fα3VLC) and selected with 400 μg/ml of G418. The harvested rGP was purified using an anti-FLAG M2 column.
All the constructs were verified by restriction mapping and nucleotide sequencing.
Cell culture and DNA transfection-Human 293 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum. Transfections were performed using the calcium phosphate precipitation method of the Profection Mammalian Transfection Systems (Promega). Stably transfected cells were selected by their resistance to G418. Foci of surviving cells were isolated, cloned and amplified.
Antibody production-Polyclonal antibodies against the N-terninal region of GPBP. Cells expressing HeLa1 λ-g11 as a lysogen were lysed by sonication in the presence of Laemmli sample buffer and subjected to electrophoresis in a 7.5% acrylamide preparative gel. The gel was stained with Coomassie blue and the band containing the fusion protein of interest excised and used for rabbit immunization (10). The anti-serum was tested for reactivity using APTG-affinity purified antigen. To obtain affinity-purified antibodies, the anti-serum was diluted 1:5 with TBS and loaded onto a Sepharose 4B column containing covalently bound affinity purified antigen. The bound material was eluted and, unless otherwise indicated, used in the immunochemical studies.
Monoclonal antibodies against GPBP. Monoclonal antibodies were produced essentially as previously reported (7) using GST-GPBP. The supernatants of individual clones were analyzed for antibodies against rGPBP.
In vitro phosphorylation assays-About 200 ng of rGPBP were incubated overnight at 30° C. in 25 mM β-glycerolphosphate (pH 7.0), 0.5 mM EDTA, 0.5 mM EGTA, 8 mM MgCl2, 5 mM MnCl2, 1 mM DTT and 0.132 μM γ-32P-ATP, in the presence or absence of 0.5-1 μg of protein substrates or 10 nmoles of synthetic peptides, in a total volume of 50 μl.
In vivo phosphorylation assays-Individual wells of a 24-well dish were seeded with normal or with stably pc-n4′ transfected 293 cells. When the cells were grown to the desired density, a number of wells of the normal 293 cells were transfected with pc-FLAG-n4′. After 12 hours, the culture medium was removed, 20 μCi/well of H332PO4 in 100 μl of phosphate-free DMEM added, and incubation continued for 4 hours. The cells were lysed with 300 μl/well of TBS containing 1% Triton X-100, 2 mM EDTA, 1 mM PMSF, 50 mM NaF and 0.2 mM vanadate, and extracted with specific antibodies and Protein A-Sepharose. When anti-GPBP serum was used, the lysate was pre-cleared using pre-immune serum and Protein A-Sepharose.
In vitro dephosphorylation of rGPBP-About 1 μg of rGPBP was dephosphorylated in 100 μl of 10 mM Tris-acetate (pH 7.5), 10 mM magnesium acetate and 50 mM potassium acetate with 0.85 U of calf intestine alkaline phosphatase (Pharmacia) for 30 min at 30° C.
Renaturation assays-In-blot renaturation assays were performed using 1-5 μg of rGPBP as previously described (11).
Nucleotide sequence analysis-cDNA sequence analyses were performed by the dideoxy chain termination method using [α]35S-dATP, modified T7 DNA polymerase (Amersham) and universal or GPBP-specific primers (8-10).
32P-Phosphoamino acid analysis-Immunopurified rGPBP or HPLC gel-filtration fractions thereof containing the material of interest were phosphorylated, hydrolyzed and analyzed in one dimensional (4) or two dimensional thin layer chromatography (12). When performing two dimensional analysis, the buffer for the first dimension was formic acid:acetic acid:water (1:3.1:35.9) (pH 1.9) and the buffer for the second dimension was acetic acid:pyridine:water (2:0.2:37.8) (pH 3.5). Amino acids were revealed with ninhydrin, and 32P-phosphoamino acids by autoradiography.
Physical methods and immunochemical techniques-SDS-PAGE and Western-blotting were performed as in (4). Immunohistochemistry studies were done on human multi-tissue control slides (Biomeda, Biogenex) using the ABC peroxidase method (13).
Computer analysis-Homology searches were carried out against the GenBank and SwissProt databases with the BLAST 2.0 (14) at the NCBI server, and against the TIGR Human Gene Index database for expressed sequence tags, using the Institute for Genomic Research server. The search for functional patterns and profiles was performed against the PROSITE database using the ProfileScan program at the Swiss Institute of Bioinformatics (15). Prediction of coiled-coil structures was done at the Swiss Institute for Experimental Cancer Research using the program Coils (16) with both 21 and 28 residue windows.
Molecular cloning of GPBP-To search for proteins specifically interacting with the divergent N-terminal region of the human GP antigen, a 21-residue peptide (GPpep1; SEQ ID NO:26)), encompassing this region and flanking sequences, and specific monoclonal antibodies against it were combined to screen several human cDNA expression libraries. More than 5×106 phages were screened to identify a single HeLa-derived recombinant encoding a fusion protein specifically interacting with GPpep1 without disturbing antibody binding.
Using the cDNA insert of the original clone (HeLa1), we isolated a 2.4-kb cDNA (n4) that contains 408-bp of 5′-untranslated sequence, an open reading frame (ORF) of 1872-bp encoding 624 residues, and 109-bp of 3′-untranslated sequence (
Protein data bank searches revealed homologies almost exclusively within the approximately 100 residues at the N-terminal region harboring the PH domain. The PH domain of the oxysterol-binding protein is the most similar, with an overall identity of 33.5% and a similarity of 65.2% with GPBP. In addition, the Caenorhabditis elegans cosmid F25H2 (accession number Q93569) contains a hypothetical ORF that displays an overall identity of 26.5% and a similarity of 61% throughout the entire protein sequence, indicating that similar proteins are present in lower invertebrates. Several human expressed sequence tags (accession numbers AA287878, AA287561, AA307431, AA331618, AA040134, AA158618, AA040087, AA122226, AA158617, AA121104, AA412432, AA412433, AA282679 and N27578) possess a high degree of nucleotide identity (above 98%) with the corresponding stretches of the GPBP cDNA, suggesting that they represent human GPBP. Interestingly, the AA287878 EST shows a gap of 67 nucleotides within the sequence corresponding to the GPBP 5′-untranslated region, suggesting that the GPBP pre-mRNA is alternatively spliced in human tissues (not shown).
The distribution and expression of the GPBP gene in human tissues was first assessed by Northern blot analysis (
Southern blot studies analysis of genomic DNA from different species indicated that homologous genes exist throughout phylogeny (
Experimental determination of the translation start site-To experimentally confirm the predicted ORF, eukaryotic expression vectors containing either the 2.4-kb of cDNA of n4′, or only the predicted ORF tagged with a FLAG sequence (
Expression and characterization of yeast rGPBP-Yeast expression and FLAG-based affinity-purification were combined to produce rGPBP (
The rGPBP is a serine/threonine kinase that phosphorylates the N-terminal region of the human GP antigen-Although GPBP does not contain the conserved structural regions required to define the classic catalytic domain for a protein kinase, the recent identification and characterization of novel non-conventional protein kinases (19-27) encouraged the investigation of its phosphorylating activity. Addition of [γ32P]ATP to rGPBP (either from yeast or 293 cells (not shown)) in the presence of Mn2+ and Mg2+ resulted in the incorporation of 32P as PSer and PThr in the major and related products recognized by both anti-FLAG and specific antibodies (
Although the purification system provides high quality material, the presence of contaminants with a protein kinase activity could not be ruled out. The existence of contaminants was also suggested by the presence of a FLAG-containing 40 kDa polypeptide, which displayed no reactivity with specific antibodies nor incorporation of 32P in the phosphorylation assays (
Immunohistochemical localization of the novel kinase-To investigate GPBP expression in human tissues we performed immunohistochemical studies using specific polyclonal (
Our data show that GPBP is a novel, non-conventional serine/threonine kinase. We also present evidence that GPBP discriminates between human and bovine GP antigens, and targets the phosphorylatable region of human GP antigen in vitro. Several lines of evidence indicate that the 89 kDa polypeptide is the only kinase in the affinity purified rGPBP. First, we found no differences in auto- or trans-phosphorylation among rGPBP samples purified in the presence of 150 mM, 0.5 M, 1 M or 2 M salt (not shown), suggesting that rGPBP does not carry intimately bound kinases. Second, there is no FAG-containing, yeast-derived kinase in our samples, since material purified using GPBP-specific antibodies shows no differences in phosphorylation (not shown). Third, a deletion mutant (GPBPΔ26; see below) displays reduced auto- and trans-phosphorylation activities (not shown), demonstrating that the 89 kD polypeptide is the only portion of the rGPBP with the ability to carry out phosphate transfer.
Although GPBP is not homologous to other non-conventional kinases, they share some structural features including an N-termina α-helix coiled-coil (26, 27), serine-rich motifs (24), high phosphoamino acids content (27), bipartite nuclear localization signal (27), and the absence of a typical nucleotide or ATP binding motif (24, 27).
Immunohistochemistry studies show that GPBP is a cytosolic polypeptide also found in the nucleus, associated with the plasma membrane and likely at the extracellular matrix associated with the basement membrane, indicating that it contains the structural requirements to reach all these destinations. The nuclear localization signal and the PH domain confer to it the potential to reach the nucleus and the cell membrane, respectively (17, 29, 30). Although GPBP does not contain the structural requirements to be exported, the 5′-end untranslated region of its mRNA includes an upstream ORF of 130 residues with an in-frame stop codon at the beginning (
Alternatively, a splicing phenomenon could generate transcripts with additional unidentified exon(s) that would provide the structural requirements for exportation. The multiple cellular localization, the high content in PTyr, and the lack of tyrosine kinase activity in vitro, suggest that GPBP is itself the target of specific tyrosine kinase(s) and therefore likely involved in specific signaling cascade(s).
As discussed above, specific serine phosphorylation, as well as pre-mRNA alternative splicing, are associated with the biology of several autoantigens, including the GP antigen, acetylcholine receptor and myelin basic protein (MBP) (4). The latter is suspected to be the major antigen in multiple sclerosis (MS), another exclusively human autoimmune disease in which the immune system targets the white matter of the central nervous system. GP disease and MS are human disorders that display a strong association with the same HLA class II haplotype (HLA DRB1*1501)(32, 33). This, along with the recent report of death by GP disease of a MS patient carrying this HLA specificity (34), supports the existence of common pathogenic events in these human disorders.
Phosphorylation of specific serines has been shown to change intracellular proteolysis (35-40). Conceivably, alterations in protein phosphorylation can affect processing and peptide presentation, and thus mediate autoimmunity. GP antigen-derived peptide presentation by the HLA-DR15 depends more on processing than on preferences of relatively indiscriminate DR15 molecules (41), suggesting that if processing is influenced by abnormal phosphorylation, the resulting peptides would likely be presented by this HLA. Our more recent data indicate that in both the GP and MBP systems, the production of alternative splicing products serves to regulate the phosphorylation of specific and structurally homologous PKA sites, suggesting that this or a closely related kinase is the in vivo phosphorylating enzyme. Alterations in the degree of antigen phosphorylation, caused either by an imbalance in alternative products, or by the action of an intruding kinase that deregulates phosphorylation of the same motifs, could lead to an autoimmune response in predisposed individuals. rGPBP phosphorylates the human GP antigen at a major PKA phosphorylation site in an apparently unregulated fashion, since the presence of specific alternative products of the GP antigen did not affect phosphorylation of the primary antigen by GPBP (not shown).
Although GPBP is ubiquitously expressed, in certain organs and tissues it shows a preference for cells and tissue structures that are target of common autoimmune responses: the Langerhans cells (type I diabetes); the white matter of the central nervous system (multiple sclerosis); the biliary ducts (primary biliary cirrhosis); the cortical cells of the adrenal gland (Addison disease); striated muscle cells (myasthenia gravis); spermatogonium (male infertility); Purkinje cells of the cerebellum (paraneoplasic cerebellar degeneration syndrome); and intestinal epithelial cells (pernicious anemia, autoimmune gastritis and enteritis). All the above observations point to this novel kinase as an attractive candidate to be considered when envisioning a model for human autoimmune disease.
Here we report the existence of two isoforms of GPBP that are generated by alternative splicing of a 78-base pair (bp) long exon that encodes a 26-residue serine-rich motif. Both isoforms, GPBP and GPBPΔ26, exist as high molecular aggregates that result from polypeptide self-aggregation. The presence of the 26-residue peptide in the polypeptide chain results in a molecular species that self-interacts more efficiently and forms aggregates with higher specific activity. Finally, we present evidences supporting the observation that GPBP is implicated in human autoimmune pathogenesis.
Synthetic Polymers:
Peptides. GPpep1, KGKRGDSGSPATWTWTTRGFVFT (SEQ ID NO:26), is described in Example 1. GPBPpep1, PYSRSSSMSSIDLVSASDDVHRFSSQ (SEQ ID NO:14), representing residues 371-396 of GPBP was synthesized by Genosys.
Oligonucleotides. The following oligonucleotides were synthesized by Life Technologies, Inc., 5′ to 3′: ON-GPBP-11 m, G CGG GAC TCA GCG GCC GGA TTT TCT (SEQ ID NO:34); ON-GPBP-15m, AC AGC TGG CAG AAG AGA C (SEQ ID NO:35); ON-GPBP-20c, C ATG GGT AGC TTT TAA AG (SEQ ID NO;36); ON-GPBP-22m, TA GAA GAA CAG TCA CAG AGT GAA AAG G (SEQ ID NO;37); ON-GPBP-53c, GAATTC GAA CAA AAT AGG CTT TC (SEQ ID NO:38); ON-GPBP-56m, CCC TAT AGT CGC TCT TC (SEQ ID NO:39); ON-GPBP-57c, CTG GGA GCT GAA TCT GT (SEQ ID NO:40); ON-GPBP-62c, GTG GTT CTG CAC CAT CTC TTC AAC (SEQ ID NO:41); ON-GPBP-Δ26, CA CAT AGA TTT GTC CAA AAG GTT GAA GAG ATG GTG CAG AAC (SEQ ID NO:42).
Reverse transcriptase and polymerase chain rection (RT-PCR). Total RNA was prepared from different control and GP tissues as described in (15). Five micrograms of total RNA was retrotranscribed using Ready-To-Go You-Prime First-Strand beads (Amersham Pharmacia Biotech) and 40 pmol of ON-GPBP-53c. The corresponding cDNA was subjected to PCR using the pairs of primers ON-GPBP-11m/ON-GPBP-53c or ON-GPBP-15/ON-GPBP-62c. The identity of the products obtained with 15m-62c was further confirmed by Alu I restriction. To specifically amplify GPBP transcripts, PCR was performed using primers ON-GPBP-15m/ON-GPBP-57c.
Northern hybridization studies Pre-made human multiple-tissue and tumor cell-line Northern Blots (CLONTECH) were probed with a cDNA containing the 78-bp exon present only in GPBP or with a cDNA representing both isoforms. The corresponding cDNAs were obtained by PCR using the pair of primers ON-GPBP-56m and ON-GPBP-57c using GPBP as a template, or with primers ON-GPBP-22m and ON-GPBP-20c, using GPBPΔ26 as a template. The resulting products were random-labeled and hybridized following the manufacturers' instructions.
Plasmid construction, expression and purification of recombinant proteins. The plasmid pHIL-FLAG-n4′, used for recombinant expression of FLAG-tagged GPBP in Pichia pastoris has been described elsewhere (4). The sequence coding for the 78-bp exon was deleted by site-directed mutagenesis using ON-GPBP-Δ26 to generate the plasmid pHIL-FLAG-n4′Δ26. Expression and affinity-purification of recombinant GPBP and GPBPΔ26 was done as in (4).
Gel-filtration HPLC. Samples of 250 μl were injected into a gel filtration PE-TSK-G4000SW HPLC column equilibrated with 50 mM Tris-HC1 pH 7.5, 150 mM NaCl. The material was eluted from the column at 0.5 ml/min, monitored at 220 nm and minute fractions collected.
In vitro phosphorylation assays. The auto-, trans-phosphorylation and in-blot renaturation studies were performed as in Example 1.
Antibodies and immunochemical techniques. Polyclonal antibodies were raised by in chicken against a synthetic peptide (GPBPpep1) representing the sequence coded by the 78-bp exon (Genosys). Egg yolks were diluted 1:10 in water, the pH adjusted to 5.0. After 6 hours at 4° C., the solution was clarified by centrifugation (25 min at 10000×g at 4° C.) and the antibodies precipitated by adding 20% (w/v) of sodium sulfate at 20.000×g, 20′. The pellets were dissolved in PBS (1 ml per yolk) and used for immunohistochernical studies. The production of antibodies against GPBP/GPBPΔ26 or against α3(IV)NC1 domain are discussed above (see also 4,13).
Sedimentation velocity. Determination of sedimentation velocities were performed in an Optima XL-A analytical ultracentrifuge (Beckman Instruments Inc.), equipped with a VIS-UV scanner, using a Ti60 rotor and double sector cells of Epon-charcoal of 12 mm optical path-length. Samples of ca. 400 μl were centrifuged at 30,000 rpm at 20° C. and radial scans at 220 nm were taken every 5 min. The sedimentation coefficients were obtained from the rate of movement of the solute boundary using the program XLAVEL (supplied by Beckman).
Sedimentation equilibrium. Sedimentation equilibrium experiments were done as described above for velocity experiments with samples of 70 μl, and centrifuged at 8,000 rpm. The experimental concentration gradients at equilibrium were analyzed using the program EQASSOC (Beckman) to determine the corresponding weight average molecular mass. A partial specific volumes of 0.711 cm3/g for GPBP and 0.729 cm3/g for GPBPΔ26 were calculated from the corresponding amino acid compositions.
Physical methods and immunochemical techniques. SDS-PAGE and Western blotting were performed under reducing conditions as previously described (3). Immunohistochemistry studies were done on formalin fixed paraffin embedded tissues using the ABC peroxidase method (4) or on frozen human biopsies fixed with cold acetone using standard procedures for indirect immunofluorescence.
Two hybrid studies. Self-interaction studies were carried out in Saccharomyces cerevisiae (HF7c) using pGBT9 and pGAD424 (CLONTECH) to generate GAL4 binding and activation domain-fusion proteins, respectively. Interaction was assessed following the manufacture's recommendations. β-galactosidase activity was assayed with X-GAL (0.75 mg/ml) for in situ and with ortho-nitrophenyl β-D galactopyranoside (0.64 mg/ml) for the in-solution determinations.
Identification of two spliced GPBP variants. To characterize the GPBP species in normal human tissues, we coupled reverse transcription to a polymerase chain reaction (RT-PCR) on total RNA from different tissues, using specific oligonucleotides that flank the full open reading frame of GPBP. A single cDNA fragment displaying lower size than expected was obtained from skeletal muscle-derived RNA (
To investigate whether the 78-bp represent an exon skipped transcript during pre-mRNA processing, we used this cDNA fragment to probe a human-derived genomic library and we isolated a˜14-Kb clone. By combining Southern blot hybridization and PCR, the genomic clone was characterized and a contiguous DNA fragment of 12482-bp was fully sequenced (SEQ ID 25). The sequence contained (from 5′ to 3′), 767-bp of intron sequence, a 93-bp exon, an 818-bp intron, the 78-bp exon sequence of interest, a 9650-bp intron, a 96-bp exon and a 980-bp intron sequence (
The relative expression of GPBP in human-derived specimens was assessed by Northern blot analysis, using either the 78-bp exon or a 260-bp cDNA representing the flanking sequence of 78-bp (103-bp 5′ and 157-bp 3′) present in both GPBP and GPBPΔ26 (
All the above indicates that GPBP is expressed at low levels in normal human tissues, and that the initial lack of detection by RT-PCR of GPBP can be attributed to a preferential amplification of the more abundant GPBPΔ26. Indeed, the cDNA of GPBP could be amplified from human tissues (skeletal muscle, lung, kidney, skin and adrenal gland) when the specific RT-PCR amplifications were done using 78-bp exon-specific oligonucleotides (not shown). This also suggests that GPBPΔ26 mRNA is the major transcript detected in Northern blot studies when using the cDNA probe representing both GPBP and GPBPΔ26.
Recombinant expression and functional characterization of GPBPΔ26. To investigate whether the absence of the 26-residue serine-rich motif would affect the biochemical properties of GPBP, we expressed and purified both isoforms (rGPBP and rGPBPΔ26), and assessed their auto- and trans-phosphorylation activities (
rGPBP and rGPBP-26 exist as very active high molecular weight aggregates. Gel filtration analysis of affinity-purified rGPBP or rGPBPΔ26 yielded two chromatographic peaks (I and II), both displaying higher MW than expected for the individual molecular species, as determined by SDS-PAGE studies (89 kDa and 84 kDa, respectively ) (
GPBP and GPBPΔ26 self-interact in a yeast two-hybrid system. To assess the physiological relevance of the self-aggregation, and to determine the role of the 26-residue motif, we performed comparative studies using a two-hybrid interaction system in yeast. In this type of study, the polypeptides whose interaction is under study are expressed as a part of a fusion protein containing either the activation or the binding domains of the transcriptional factor GAL4. An effective interaction between the two fusion proteins through the polypeptide under study would result in the reconstitution of the transcriptional activator and the subsequent expression of the two reporter genes, Lac Z and His3, allowing colony color detection and growth in a His-defective medium, respectively. We estimated the intensity of interactions by the growth-rate in histidine-defective medium, in the presence of different concentrations of a competitive inhibitor of the His3 gene product (3-AT), and a quantitative colorimetric liquid β-galactosidase assay. A representative experiment is presented in
GPBP is highly expressed in human but not in bovine and murine glomerulus and alveolus. We have shown that GPBP/GPBPΔ26 is preferentially expressed in human cells and tissues that are commonly targeted in naturally occurring autoimmune responses. To specifically investigate the expression of GPBP, we raised polyclonal antibodies against a synthetic peptide representing the 26-residue motif characteristic of this kinase isoform, and used it for immunohistochemical studies on frozen or formalin fixed paraffin embedded human tissues (
To further assess this hypothesis, we investigated the presence of GPBP and GPBPΔ26 in the glomerulus of two mammals that naturally do not undergo GP disease compared to human (
GPBP is highly expressed in several autoimmune conditions. We analyzed several tissues from different GP patients by specific RT-PCR to assess GPBP/GPBPΔ26 mRNA levels. As in control kidneys, the major expressed isoform in GP kidneys was GPBPΔ26. However, in the muscle of one of the patients, GPBP was preferentially expressed, whereas GPBPΔ26 was the only isoform detected in control muscle samples (
To investigate the expression of GPBP and GPBPΔ26 in autoimnune pathogenesis, we studied cutaneous autoimmune processes and compared them with control samples representing normal skin or non-autoimmune dermatitis (
Alternative pre-mRNA splicing is a fundamental mechanism for differential gene expression that has been reported to regulate the tissue distribution, intracellular localization, and function of different protein kinases (8-11). In this regard, and closely resembling GPBP, B-Raf exists as multiple spliced variants, in which the presence of specific exons renders more interactive, efficient and oncogenic kinases (12).
Although it is evident that rGPBPΔ26 still bears the uncharacterized catalytic domain of this novel kinase, both auto- and trans-phosphorylating activities are greatly reduced when compared to rGPBP. Gel filtration and two hybrid experiments provide some insights into the mechanisms that underlie such a reduced phosphate transfer activity. About 1-2% of rGPBP is organized in very high molecular weight aggregates that display about one third of the phosphorylating activity of rGPBP, indicating that high molecular aggregation renders more efficient quaternary structures. Recombinant GPBPΔ26, with virtually no peak I material, consistently displayed a reduced kinase activity. However, aggregation does not seem to be the only mechanism by which the 26-residues increases specific activity, since the rGPBPΔ26 material present in peak II also shows a reduced phosphorylating activity when compared to homologous fractions of rGPBP. One possibility is that rGPBP-derived aggregates display higher specific activities because of quaternary structure strengthening caused by the insertion of the 26-residue motif. The oligomers are kept together mainly by very strong non-covalent bonds, since the bulk of the material appears as a single polypeptide in non-reducing SDS-PAGE, and the presence of either 8 M urea or 6 M guanidine had little effect on chromatographic gel filtration profiles (not shown). How the 26-residue motif renders a more strengthened and active structure remains to be clarified. Conformational changes induced by the presence of an exon encoded motif that alter the activation status of the kinase have been proposed for the linker domain of the Src protein (24) and exons 8b and 10 of B-Raf (12). Alternatively, the 26-residue motif may provide the structural requirements such as residues whose phosphorylation may be necessary for full activation of GPBP.
We have reported (13) that the primary structure of the GP antigen (α3(IV)NC1) is the target of a complex folding process yielding multiple conformers. Isolated conformers are non-minimum energy structures specifically activated by phosphorylation for supramolecular aggregation and likely quaternary structure formation. In GP patients, the α3(IV)NC1 shows conformational alterations and a reduced ability to mediate the disulfide stabilization of the collagen IV network. The GP antibodies, in turn, demonstrate stronger affinity towards the patient α3(IV)NC1 conformers, indicating that conformationally altered material caused the autoimmune response. Therefore, it seems that in GP disease an early alteration in the conforming process of the α3(IV)NC1 could generate altered conformers for which the immune system is not tolerant, thus mediating the autoimmune response.
Other evidence (Raya et al., unpublished results) indicates that phosphorylation is the signal that drives the folding of the α3(IV)NC1 into non-minimum energy ends. In this scenario, three features of the human α3(IV)NC1 system are of special pathogenic relevance when compared to the corresponding antigen systems from species that, like bovine or murine, do not undergo spontaneous GP disease. First, the N-terninus of the human α3(IV)NC1 contains a motif that is phosphorylatable by PKA and also by GPBP (see above, and also 2-4). Second, the human gene generates multiples alternative products by alternative exon splicing (14,15). Exon skipping generates alternative products with divergent C-terminal ends that up-regulate the in vitro PKA phosphorylation of the primary α3(IV)NC1 product (See below Example 3). Third, the human GPBP is expressed associated with glomerular and alveolar basement membranes, the two main targets in GP disease. The phosphorylation-dependent conforming process is also a feature of non-pathogenic NC1 domains (13), suggesting that the phosphorylatable N-terminus, the alternative splicing diversification, and the expression of GPBP at the glomerular and alveolar basement membranes, are all exclusively human features that place the conformation process of α3(IV)NC1 in a vulnerable condition. The four independent GP kidneys studied expressed higher levels of GP antigen alternative products (15; Bernal and Saus, unpublished results), and an augmented expression of GPBP were found in a GP patient (see above). Both increased levels of alternative GP antigen products and GPBP are expected to have consequences in the phosphorylation-dependent conformational process of the α3(IV)NC1, and therefore with pathogenic potential.
GPBP is highly expressed in skin targeted by natural autoimmune responses. In the epidermis, GPBP is associated with cell surface blebs characteristic of the apoptosis-mediated differentiation process that keratinocytes undergo during maturation from basale to corneum strata (22, 23). Keratinocytes from SLE patients show a remarkably heightened sensitivity to UV-induced apoptosis (6, 18, 20), and augmented and premature apoptosis of keratinocytes has been reported to exist in SLE and dermatomyositis (7). Consistently, we found apoptotic bodies expanding from basal to peripheral strata of the epidermis in several skin autoimmune conditions including discoid lupus (not shown), SLE and lichen planus. Autoantigens, and modified versions thereof are clustered in the cell surface blebs of apoptotic keratinocytes (6,18,20). Apoptotic surface blebs present autoantigens (21), and likely release modified versions to the circulation (16-20). It has been suggested that the release of modified autoantigens from apoptotic bodies could be the immunizing event that mediates systemic autoimmune responses mediating SLE and scleroderma (18,19).
Our evidence indicates that both GPBP and GPBPΔ26 are able to act in vitro as protein kinases, with GPBP being a more active isoform than GPBPΔ26. Furthermore, recombinant material representing GPBP or GPBPΔ26 purified from yeast or from human 293 cells contained an associated proteolytic activity that specifically degrades the α3(IV)NC1 domain (unpublished results). The proteolytic activity operates on α3(IV)NC1 produced in an eukaryotic expression system, but not on recombinant material produced in bacteria (unpublished results), indicating that α3(IV)NC1 processing has some conformational or post-translational requirements not present in prokaryotic recombinant material. Finally, it has been reported that several autoantigens undergo phosphorylation and degradation in apoptotic keratinocytes (20). While not being limited to an exact mechanism, we propose, in light of all of the above data, that the machinery assembling GPBP at the apoptotic blebs likely performs a complex modification of the autoantigens that includes phosphorylation, conformational changes and degradation. Accordingly, recombinant protein representing autoantigens in SLE (P1 ribosomal phosphoprotein and Sm-D1 small nuclear ribonucleoproteins) and in dermatomyositis (hystidil-tRNA synthetase) were in vitro substrates of GPBP (unpublished results).
The down-regulation in cancer cell lines of GPBP, suggest that the cell machinery harboring GPBP/GPBPΔ26 is likely involved in signaling pathways inducing programmed cell death. The corresponding apoptotic pathway could be up regulated during autoimmune pathogenesis to cause an altered antigen presentation in individuals carrying specific MHC haplotypes; and down regulated during cell transformation to prevent autoimmune attack to the transformed cells during tumor growth.
In GP disease, the immune system attack is mediated by autoantibodies against the non-collagenous C-terminal domain (NC1) of the α3 chain of collagen IV (the GP antigen) (1). The N-terminus of the human α3(IV)NC1 contains a highly divergent and hydrophilic region with a unique structural motif, KRGDS9 (SEQ ID NO:63) that harbors a cell adhesion signal as an integral part of a functional phosphorylation site for type A protein kinases (2,3). Furthermore, the gene region encoding the human GP antigen characteristically generates multiple mRNAs by alternative exon splicing (4,5). The alternative products diverge in the C-terminal ends and all but one share the N-terminal KRGDS9 (SEQ ID NO:63) (4,5).
Multiple sclerosis (MS) is an exclusive human neurological disease characterized by the presence of inflamatory demyelization plaques at the central nervous system. (6). Several evidences indicate that this disease is caused by an autoimmune attack mediated by cytotoxic T cells towards specific components of the white matter including the myelin basic protein (MBP) (7, 8). In humans, the MBP gene generates four products (MBP, MBPΔII, MBPΔV and MBPΔII/V) that result from alternative exon splicing during pre-mRNA processing (9). Among these, MBPΔII is the more abundant form in the mature central nervous system, while MBP form containing all the exons is virtually absent (9).
Several biological similarities exist between the autoimme responses mediating GP disease and MS, namely: 1) both are human exclusive diseases and typically initiate after a viral flu-like disease; 2) a strong linkage exists to the same haplotype of the HLA-DR region of the class II MHC; 3) several products are generated by alternative splicing; and 4) the death of a MS patient by GP disease has recently been reported (10).
Synthetic polymers: GPΔIII derived peptide, QRAHGQDLDALFVKVLRSP (SEQ ID NO:43) and GPΔIII/IV/V derived peptide, QRAHGQDLESLFHQL (SEQ ID NO:44) were synthesized using either Boc- (MedProbe) or Fmoc- (Chiron, Lipotec) chemistry.
Plasmid Construction and Recombinant Expression.
GP derived material: The constructs representing the different GP-spliced forms were obtained by subcloning the cDNAs used elsewhere to express the corresponding recombinant proteins (5) into the BamHI site of a modified pET15b vector, in which the extraneous vector-derived amino-terminal sequence except for the initiation Met was eliminated. The extra sequence was removed by cutting the vector with NcoI and Bam HI, filling-in of the free ends with Klenow, and re-ligation. This resulted in the reformation of both restriction sites and placed the BamHI site immediately downstream of the codon for the amino-terminal Met.
The recombinant proteins representing GP or GPΔV (SEQ ID NO:46) were purified by precipitation (5). Bacterial pellets containing the recombinant proteins representing GPΔIII (SEQ ID NO:48) or GPΔIII/IV/V (SEQ ID NO:50) were dissolved by 8 M urea in 40 mM Tris-HCl pH 6.8 and sonication. After centrifugation at 40,000×g the supernatants were passed through a 0.22 μm filter and applied to resource Q column for FPLC. The effluent was acidified to pH 6 with HCl and applied to a resource S column previously equilibrated with 40 mM MES pH 6 for a second FPLC purification. The material in the resulting effluent was used for in vitro phosphorylation.
MBP-derived material: cDNA representing human MBPΔII (SEQ ID NO:51) was obtained by RT-PCR using total RNA from central nervous system. The cDNA representing human MBP was a generous gift from C. Campagnoni (UCLA). Both fragments were cloned into a modified version of pHIL-D2 (Invitrogen) containing a 6×His-coding sequence at the C-terminus to generate pHIL-MBPΔII-His and pHIL-MBP-His, respectively. These plasmids were used for recombinant expression in Pichia pastoris as described in (11). Recombinant proteins were, purified using immobilized metal affinity chromatography (TALON resin, CLONTECH) under denaturant conditions (8M urea) and eluted with 300 mM imidazole following manufacturers' instructions. The affinity-purified material was then renatured by dilution into 80 volumes of 50 mM Tris-HCl pH 8.0, 10 mM CHAPS, 400 mM NaCl, 2 mM DTT, and concentrated 50 times by ultrafiltration through a YM10-type membrane (AMICON). The Ser to Ala mutants were produced by site-directed mutagenesis over native sequence-containing constructs using transformer mutagenesis kit from CLONTECH and the resulting proteins were similarly produced.
Phosphorylation studies. Phosphorylation studies were essentially done as described above (see also 3 and 11). In some experiments, the substrates were in-blot renatured and then, phosphorylated for 30 min at room temperature by overlaying 100 μl of phosphorylation buffer containing 0.5 μg of rGPBP. Digestion with V8 endopeptidase and immunoprecipitation were performed as described in (3).
Antibody production. Synthetic peptides representing the C-terminal divergent ends of GP□III or GP□III/IV/V comprised in SEQ ID NO:43 or SEQ ID NO:44 respectively were conjugated to a cytochrome C, BSA or ovoalbumine using a glutaraldehyde coupling standard procedure. The resulting protein conjugates were used for mouse immmunization to obtain polyclonal antibodies specific for GPΔIII and monoclonal antibodies specific for GPΔIII/IV/V (Mab153). To obtain monoclonal antibodie specific for GPΔV (Mab5A mouse were immunized using recombinant bacterial protein representing the corresponding alternative form comprising the SEQ ID NO:50. The production of monoclonal (M3/1, P1/2) or polyclonal (anti-GPpep1) antibodies against SEQ ID NO: 26 which represents the N-terminal region of the GP alternative forms have been previously described (3,5).
Boc-Based Peptide Synthesis.
Assembling The peptide was assembled by stepwise solid phase synthesis using a Boc-Benzyl strategy. The starting resin used was Boc-Pro-PAM resin (0.56 meq/g, batch R4108). The deprotection/coupling procedure used was: TFA (1×1 min) TFA (1×3 min) DCM (flow flash) Isopropylalcohol (1×30 sec) DMF (3×1 min) COUPLING/DMF (1×10 min) DMF (1×1 min) COUPLING/DMF (1×10 min) DMF (2×1 min) DCM (1×1 min). For each step 10 ml per gram of peptide-resin were used. The coupling of all amino acids (fivefold excess) was performed in DMF in the presence of BOP, Hobt and DIEA. For the synthesis the following side-chain protecting groups were used: benzyl for serine; 2 chlorobenzyloxycarbonyl for lysine; cyclohexyl for aspartic and glutamic acid; tosyl for histidine and arginine.
Cleavage. The peptide was cleaved from the resin and fully deprotected by a treatment with liquid Hydrogen Fluoride (HF): Ten milliliters of HF per gram of peptide resin were added and the mixture kept at 0° C. for 45 min in the presence of p-cresol as scavengers. After evaporation of the HF, the crude reaction mixture is washed with ether, dissolved in TFA, precipitated with ether and dried.
Purification. Stationary phase: Silica C18, 15 μm, 120 A; Mobile phase: solvent A: water 0.1% TFA and solvent B: acetonitrile /A, 60/40 (v/v); Gradient: linear from 20 to 60% B in 30 min; Flow rate: 40 ml/min; and detection was U.V (210 nm). Fractions with a purity higher than 80% were pooled and lyophilized. Control of purity and identity was performed by analytical HPLC and ES/MS. The final product had 88% purity and an experimental molecular weight of 2192.9.
Fmoc-Based Peptide Synthesis.
Assembling. The peptides were synthesized by stepwise linear solid phase on Pro-clorotrityl-resin (0.685 meq/g) with standard Fmoc/tBu chemistry. The deprotection/coupling procedure used was: Fmoc aa (0.66 g) HOBt (0.26 g) DIPCDI (0.28 ml) for 40 min following a control by Kaiser test. If the test was positive the time was extended until change to negative. Then DMF (31 min), piperidine/DMF 20% (11 min) piperidine/DMF 20% (15 min) and DMF (41 min). Side chain protectors were: Pmc (pentamethylcromane sulfonyl) for arginine, Bcc (tert-butoxycarbonyl) for lysine, tBu (tert-butyl) for aspartic acid and for serine and Trl (trityl) for histidine.
Cleavage. The peptide was cleaved and fully deprotected by treatment cleavage with TFA/water 90/10. Ten milliliters of TFA solution per gram of resin were added. Water acts as scavenger. After two hours, resin was filtered and the resulting solution was precipitated five times with cold diethylether. The final precipitated was dried.
Purification. Stationary phase: Kromasil C18 10 μm; Mobile phase: solvent A: water 0.1% TFA and solvent B: acetonitrile 0.1% TFA; Isocratic: 28% B; Flow rate: 55 ml/min; Detection: 220 nm. Fractions with the higher purity were pooled and lyophilized, and a second HPLC purification round performed. Control of purity and identity was performed by analytical HPLC and ES/MS. The final product had 97% purity and an experimental molecular weight of 2190.9.
Regulation of the phosphorylation of the human GP antigen by alternative splicing. We produced bacterial recombinant proteins representing the primary antigen (GP) or the individual alternative products GPΔV (SEQ ID NO:46), GPΔIII (SEQ ID NO:48) and GPΔIII/IV/V (SEQ ID NO:50), and we tested their ability to be phosphorylated by PKA (
Regulation of the phosphorylation of the MBP by alternative splicing. The MBP contains at its N terminal region two PKA phosphorylation sites (Ser8, Ser57) that are structurally similar to the N terminus site (Ser9) present in GP antigen products (
In similar experiments assessing GPBP phosphorylation of the recombinant MBP proteins, GPBP preferentially phosphorylated MBP, while little phosphorylation of MBPΔII was observed (
From all these data we concluded that in the MBP system, alternative splicing regulates the phosphorylation of specific serines by either PKA or GPBP.
Synthetic peptides representing the C terminal region of GPΔII influence GPBP phosphorylation. To assess the effect of the C terminal region of GPΔIII on GPBP activity, peptides representing this region were synthesized using two different chemistries (Boc or Fmoc), and separately added to a phosphorylation mixture containing GPBP (
We show (here and in the following examples) that the α3(IV)NC1 domain undergoes a complex structural diversification by two different mechanism: 1) alternative splicing (4,5) and 2) conformational isomerization of the primary product. Both mechanisms generate products that are distinguished by PKA, indicating that PKA phosphorylation is a critical event in the biology of the α3(IV)NC1 domain. Phosphorylation guides at least in part the folding, but also the supramolecular assembly of the α3(IV)NC1 domain in the collagen IV network (below). Altered conformers of the α3(IV)NC1 lead the autoimmune response mediating GP disease (See the following examples), suggesting that an alteration in antigen phosphorylation could be the primary event in the onset of the disease. Accordingly, we have found increased expression levels of GPΔIII in several GP kidneys (4 and Bernal and Saus, unpublished results), and an increased expression of GPBP has been detected in another Goodpasture patient (
An alternative splicing mechanism also regulates the PKA phosphorylation of specific serines in the MBP antigenic system. MBP is also a substrate for GPBP suggesting that GPBP may play a pathogenic role in multiple sclerosis, and other autoimmune responses.
All of the above data identify GPBP as a potential target for therapeutics in autoimmune disease. In
Here we show that the human α3(IV)NC1 domain exists as multiple phosphorylation-dependent conformational isoforms (conformers) that are stabilized by disulfide bonds. We present evidence supporting that phosphorylation of Ser9 can lead to the formation of α3(IV)NC1 conformers for which tolerance has not been established.
Production of native and recombinant NC1 material. Human collagen IV NC1 “hexamer” and “monomers” were prepared from renal cortex as previously described (21). The “monomers” were further analyzed by reverse-phase HPLC using a C18 column from Vydac and a 30-48% acetonitrile gradient developed during 36 min in the presence of 0.1% TFA. The most hydrophobic fractions containing α3(IV)NC1 domain with no detectable traces of other chains, as assessed by enzyme-linked immunosorbent assay (EISA) and individual α(IV) chain specific antibodies, were pooled and concentrated (27-kDa). The more hydrophilic fractions, containing both α3 material and the other a chains, were re-analyzed by reverse-phase HPLC using a C4 column from Vydac and a 24-44% isopropanol gradient developed during 36 min in the presence of 0.2% TFA. Fractions containing mainly α3, but also α4 and α5 chains, were pooled and concentrated (22-25-kDa).
Recombinant FLAG-tagged α1(IV)NC1-α6(IV)NC1 (fa1-fa6) were prepared as reported in Ref. 22. A site-directed mutagenesis approach (Clontech) and the fα3 construct were used to obtain fα3Ala9 and fα3Asp9. The constructs were assessed by nucleotide sequencing, and used to generate stably transfected human kidney 293 (ATCC # CRL-1573) cell lines as described in Ref. 23. Individual clones secreting similar levels of protein to the culture media, as estimated by Western blot analysis, were further selected and used for comparative studies. For these purposes, the individual cell lines were grown in Dubelcco's modified Eagle's medium supplemented with 10% fetal calf serum. When the culture reached ˜80% confluence, the serum-containing media was removed and cells were brought to quiescence in serum-free medium supplemented with Ham's F-12 nutrient mixture. After 24 hours, the media were changed, and the media of an additional period of 24 hours were separately collected, centrifuged to remove cell debris and analyzed. by Western-blot using α3(IV)NC1 specific antibodies.
Physical, chemical and immunochemical methods. When indicated, SDS-electrophoresis was performed on a fusible acrylamide (National Diagnostics) following manufacturer instructions. After electrophoresis, the gel region between 21- and 30-kDa was split into eight horizontal slices of similar height Each of these was further split in two, separately melted in the presence of reducing or non-reducing Laemmli sample buffer, and re-analyzed in SDS-PAGE for immunoblot purposes.
Otherwise indicated SDS-PAGE studies were carried out in the absence of a reducing agent and the immunoblots were performed following standard procedures using PVDF membranes (Millipore) and 27.5% methanol in the transfer buffer.
Reduction/Oxidation studies. In a standard assay, ˜1 μg of recombinant human α3(IV)NC1 (fα3) in 25 mM β-glycerol phosphate (pH 7.0), 0.5 mM EDTA, 0.5 mM EGTA, 8 mM MgCl2 was incubated with or without 2 units of calf intestine alkaline phosphatase (Pharmacia). After 1 hour at 30° C., 5 mM MnCl2 and 1 mM DTT were added (redox conditions) and incubation continued until the DTT was fully oxidized ([DTT]<50 nM). To monitor the reaction, aliquots were taken at several times and DTT measured as described in Ref. 24. When the reaction was completed, the remaining material was analyzed by immunoblot. Phosphatase-treated materials were subjected to phosphorylation with the catalytic subunit of PKA to assess dephosphorylation effectiveness.
Phosphorylation, V8 protease digestion and immunoprecipitation assays. Phosphorylation with the catalytic subunit of the cAMP-dependent protein kinase (Promega), digestion with V8 protease (Sigma), and immunoprecipitation with anti-GPpep1 antibodies was performed essentially as previously described (17).
Antibodies. We have described the production and characterization of Mab3 antibodies (previously called Mab17), which recognize a conformational disulfide-dependent epitope in the α3(IV)NC1 (25). The epitope of Mab3 implicates residues 29-44 and more critically the two Ser and a Pro therein, and residues 139-153 (15,16). We have previously reported (17,20) the production of the antibodies specific for the N-terinus of the human α3(IV)NC1 domain (anti-GPpep1, MabM3/1 and MabP1/2). MabP1/2 epitope implicates Ser9, as substitution of this residue by Ala or Asp effectively abolishes antibody binding to the corresponding α3(IV)NC1 mutants. The remaining α3(IV)NC1-specific monoclonal antibodies, Mab175 and Mab189, were raised against bacterial randomly folded human recombinant α3(IV)NC1 (20). For these purposes, the α3(IV)NC1 was analyzed by SDS-PAGE under reducing conditions, stained with Coomassie blue, and the polyacrylamide band containing the material of interest excised and used for mice immunization following standard procedures. The two monoclonal antibodies showed similar binding to reduced α3(IV)NC1 material in Western blot studies (not shown) and recognize linear epitopes that involve residues 103-117 of the α3(IV)NC1 domain (15). However, whereas Mab175 reactivity does not vary significantly with antigen reduction or conformation (15), the binding of Mab189 to the α3(IV)NC1 varies among conformers (see
Individual sera from fifty GP patients, six healthy blood donors, or three autoimmune patients containing either rheumathoid factor, p-ANCA or ANCA autoantibodies, were used at 1:10 dilution in the immunoblot studies. Tissue-bound antibodies were acid-extracted as described in Ref. 27 from a control and from a GP kidney and used in a 1:2 or 1:5 dilutions for immunoblot purposes.
The GP antibodies recognize multiple α3(IV)NC1 conformers. The reactivity of the GP antibodies towards human “monomers” was assessed using 50 individual patient sera. The reactivity greatly varied among patients, resulting in multiple reactive patterns (
To assess this the GP antibodies bound to the GBM of a patient kidney, and therefore with the highest affinity, were eluted and assayed for reactivity towards the recombinant proteins (
Identification of multiple conformers of the human α3(IV)NC1. The structural diversification of the α3(IV)NC1 domain detected with the GP antibodies was confirmed by identifying multiple α3(IV)NC1 molecular species in human “hexamer” using monoclonal antibodies (Mab) (
Differential phosphorylation of the α3(IV)NC1 conformers by PKA. We have shown that human α3(IV)NC1 undergoes phosphorylation by type A protein kinases (17). To assess the susceptibility of the different α3(IV)NC1 conformers to phosphorylation, purified α3(IV)NC1 from human renal cortex, mainly consisting of the 27-kDa conformer, was incubated with the catalytic subunit of the cAMP-dependent protein kinase in the presence of [γ32P] ATP (
This was independently confirmed by demonstrating that an α3(IV)NC1 fraction enriched in the 22-25-kDa species showed higher susceptibility to phosphorylation than the fraction which is enriched in the 27-kDa conformer (
To assess the physiological significance of these findings, we determined the presence of phosphoserine [Ser(P)] in the different human α3(IV)NC1 polypeptides by comparing the immunoreactive patterns of antibodies specifically reacting with the N terminus of the α3IVNC1 (MabP1/2) and antibodies specifically reacting with Ser(P) (
Phosphorylation regulates the conformation of the α3(IV)NC1 domain. The role of phosphorylation regulating the conformation of the α(IV)NC1 domain was further investigated by assessing the ability of dephosphorylated domain to maintain its native structure. Untreated or alkaline phosphatase-treated human recombinant α3(IV)NC1 domain was allowed to rearrange its disulfide bonds in the presence of a DTT-metal-based redox system until DTT was fully oxidized. The material was then analyzed by SDS-PAGE and blotted either with Mab3, a monoclonal antibody binding to a native disulfide-dependent epitope present in the 27-kDa conformer (
During DTT consumption, most of the untreated material forms disulfide-bond high molecular weight aggregates, which do not enter into the running gel, and only a limited amount of material remains monomeric. Phosphatase treatment efficiently inhibited disulfide-based aggregation, and most of the material remains in a monomeric form. The untreated material that remained in a monomeric form maintained both apparent molecular weight (27-kDa) and the relative reactivity with the two antibodies of the starting material, whereas monomeric phosphatase-treated material contained multiple molecular species between 22 and 29 kD, which were poorly reactive with Mab3. All the molecular species, however, displayed the same apparent mobility (29 kDa) under reducing conditions, thus confirming that they represented different disulfide-based conformers.
Therefore, it appears that upon dephosphorylation, the 27-kDa conformer was unable to keep its native conformation, recognized by Mab3, antibodies, but adopted multiple conformations (22-29 kDa conformers) during DTT consumption, and that disulfide-based aggregation of the α3(IV)NC1 is a specific phenomenon which requires phosphorylation and native conformation to occur.
The Ser9 phosphorylation promotes conformational diversification of the (α3(IV)NC1 domain. Phosphorylation at Ser9 is a biological hallmark of the human α3(IV)NC1 when compared to other NC1 domains. To assess the implication of Ser9 phosphorylation on the formation of multiple conformers of the α3(IV)NC1 domain, cell lines expressing α3(IV)NC1 (fαSer9) or mutants thereof in which Ser9 have been replaced by Ala (fα3Ala9) (SEQ ID NO:68) or Asp (fα3Asp9) (SEQ ID NO:66) were generated. Although the two mutants are non-phosphorylatable at this site the Asp-based mutant is expected to mimic the Ser(P) derivative, because the acidic lateral chain Asp mimics Ser(P), whereas the Ala mutant is expected to represent the non phosphorylated counterpart, since, chemically, Ser is hydroxy-Alanine. The recombinant materials produced were separately collected and analyzed using Mab175 or Mab3 antibodies (
Disulfide bond distribution represents the folding state of domains that are resident at the extracellular compartment (29). We have presented physical, chemical, immunochemical, biochemical and cell biological data supporting the existence of multiple disulfide bond-stabilized conformers of the α3(IV)NC1 domain in basement membrane collagen. The evidence presented in this example indicates that phosphorylation plays a critical role in the production of these multiple conformers, and suggest that differential phosphorylation is at least part of the strategy for cellular production of conformers. Differential phosphorylation of a single unique native structure could occur prior or during chain association, yielding multiple structures, each one stabilized by individual disulfide-bond distributions. Individual molecular species would have enciphered in their covalent structure the assembly partner and the final conformation that would be acquired once assembled and stabilized into a “hexamer”. The multiple conformers produced by the cells expressing the phosphorylated version of the α3(IV)NC1 domain at Ser9 (fα3Asp9) sharply contrasts with the limited structural diversification of the material representing the non-phosphorylated counterpart (fα3Ala9). The molecular mechanism by which Ser9 (P) promotes the assembly of the α3(IV)NC1 domain in multiple conformers is presently unknown. However, the presence of a cell adhesion motif as an integral part of the sequence that conforms the PKA recognition site (KRGDS9) (SEQ ID NO:63) suggest that Ser9 phosphorylation promotes cell attachment of the α3(IV)NC1 and induce conformational diversification through an integrin-mediated mechanism.
The consequences on conformation derived from the presence of Asp9 are unlikely to represent a physiological phenomenon, since the Mab3 reactive conformers of 25- and 27-kDa present in fα3Asp9 are not produced by the cells expressing the native sequence (fα3Ser9). More likely, the phenomenon represents the aberrant consequences of a permanently phosphorylated Ser9 intruding in the phosphorylation-dependent conforming process. These findings, in addition to further implicating phosphorylation in conformation, reveal that a breakage in the homeostatic phosphorylation of Ser9 can promote the formation of conformers for which the immune system has not established a tolerance and thus trigger the immune response mediating GP disease. Overall, our studies establish the phosphorylation-dependent nature of the α3(IV)NC1 folding system and point to Ser9 phosphorylation as the biological feature that renders the human system vulnerable for autoimmune pathogenesis.
Here we show that the isolated α3(IV)NC1 conformers show a state of activation that depends on phosphorylation and which is required for “hexamer” assembly. GPBP exerts a complex catalysis over isolated α3(IV)NC1 conformers, which comprises conformational isomerization and specific intermolecular disulfide bond formation, suggesting that GPBP is a novel type of molecular enzyme that assists “hexamer” formation in vivo.
Production of native and recombinant material. Human collagen IV NC1 “hexamer” and “monomers” were prepared from renal cortex as described in Example 4. Bovine testis α3(IV)NC1 “monomer” was prepared as described in Zashai et al. (1997). To produce prokaryotic human recombinant α3(IV)NC1, the cDNA used elsewhere to express the corresponding recombinant protein (Penadés et al, 1995) was subcloned into the BamHI site of a modified version of pET-15b vector (Novagen), in which the vector-derived N-terminal sequence except for the initiation Met was eliminated. The recombinant α3(IV)NC1 was purified by precipitation as described in Penadés et al. (1995) and the final pellet was dissolved in 8M urea.
Recombinant FLAG-tagged α3(IV)NC1 (fα3) was prepared as previously reported in Sado et al. (1998).
Recombinant GPBP and GPBPΔ26 (rGPBP and rGPBPΔ26) were prepared as described in Raya et al. (1999).
Physical, chemical and immunochemical methods. Immunoblot studies were performed as described in Example 4. For far-Western, after protein transfer the membrane was blocked with non-fat milk, incubated with 30 ng/μl of fα3 or recombinant GPBP and the bound recombinant material detected with α-FlAG or Mab14, respectively.
Steady-state fluorescence measurements were carried out at 25° C. on a Perkin-Elmer LS-50 spectrofluorimeter in Tris-buffered saline. The spectra were corrected by comparison to a quinine sulfate standard. The buffer was used as baseline in all the experiments and subtracted.
Unless indicated, SDS-PAGE studies were performed in the absence of a reducing agent
DTT oxidation and oligomerization studies. In a standard assay, “monomer” or “hexamer” were reduced for 4 h with 2 mM DTT in 10 mM Tris pH 7.5 at 30° C. The mixtures were brought to 25 mM β-glycerol phosphate (pH 7.0), 0.5 mM EDTA, 0.5 mM EGTA, 8 mM MgCl2, 5 mM MnCl2 and 1 mM DTT (oligomerization buffer) in a final volume of 25-50 μl and incubation continued until the DTT was fully oxidized ([DTT]<50 nM). To monitor the reaction, aliquots of 2-5 μl were taken at several times and DTT measured as described in Riddles et al. (1983). In some experiments, when the reaction was completed, the remaining material was analyzed by immunoblot. For some purposes, “monomers” were first dephosphorylated with 2 units of calf intestine alkaline phosphatase (Pharmacia) in oligomerization buffer without MnCl2 and DTT. After 1h at 30° C., these components were added to reach oligomerization conditions and mixtures were monitored and analyzed as above. For some purposes alkaline phosphatase-treated fα3 were brought to the oligomerization conditions (DTT/Mn2+) in the presence of Tris-buffered saline and the process monitored by fluorescence emission spectra. The untreated materials used in these assays were carried in parallel in the absence of alkaline phosphatase. Phosphatase-treated materials were subjected to phosphorylation with cAMP-dependent protein kinase as previously described (Revert et al, 1995) to assess dephosphorylation effectiveness. For other purposes when the material was brought to oligomerization conditions equivalent amounts of bovine serum albumin (BSA), rGPBP or rGPBPAΔ26 were added and mixtures were similarly monitored and analyzed.
Antibodies. The production of monoclonal antibodies against GPBP (Mab 14) was described in Raya et al., (1999), for the other antibodies see details in Example 4.
Phosphorylation promotes the supramolecular aggregation of the α3(IV)NC1 domain. At the endoplasmic reticulum, ATP is required to maintain the non-assembled monomers in a metastable conformation that is critical for physiological oligomerization (Braakman et al., 1992). Consequently, ATP could be used to phosphorylate and to place the α3(IV)NC1 domain into a metastable condition required for “hexamer” formation. Upon dissociation, the “hexamer” yields the different α3(IV)NC1 conformers as individual polypeptides (“monomer”) but also as disulfide-based oligomers (Fessler and Fessler, 1982; Weber et al., 1984; Butkowski et al, 1985; Siebold et al., 1988; Reddy et al., 1993), which, in turn, represent disassembled and partially assembled α3(IV) chains, respectively. Conceivably, the transition from the “hexameric” (assembled) to “monomeric” (disassembled) condition could return the individual α3(IV)NC1 species to a non-minimum energy condition that still may promote disulfide-based aggregation in vitro.
To explore this idea, we first dissociated human “hexamer” by SDS-PAGE and performed specific far-Westem studies to assess “monomer-monomer” interactions. For these purposes, we used human recombinant FLAG-tagged α3(IV)NC1 domain (fα3) to probe in-blot renatured human “monomers” after SDS-PAGE, and FLAG-specific antibodies to detect fα3 binding (
The ability to form disulfide-based aggregates of the isolated “monomers”, in comparison with assembled counterparts present in the “hexamer”, was first investigated by assessing spontaneous disulfide-based aggregation of disassembled (27-kDa and 22-25-kDa), unassembled (fα3), or assembled (hexamer) human α3(IV)-monomers in the presence of a DTT-metal-based redox system (
Differences in DTT oxidation rates could be attributed to the different capacity for disulfide-based aggregation displayed by each individual “monomeric” sample. This was confirmed by assessing the ability of each disassembled “monomeric” sample (27-kDa, 22-25-kDa) to disulfide-aggregate with recombinant fα3, which displayed the lowest DTT oxidation rate and contained an engineered recognition site (FLAG) that allowed specific antibody detection (
The role of phosphorylation mediating disulfide-based aggregation was further investigated by assessing fα3 aggregation of 22-25-kDa conformers in the presence or absence of alkaline phosphatase (
GPBP catalyzes disulfide-based aggregation of the α3(IV)NC1 domain through specific conformational isomerization reactions. We have shown that GPBP is expressed associated with glomerular basement membranes, the main target of the GP autoantibodies, and that GPBP binds to recombinant material representing the human α3(IV)NC1 domain (see above). GPBP binding to human native NC1 material was tested over in-blot renatured human “monomers” after SDS-PAGE (
To investigate the role of GPBP in the supramolecular assembly of the α3(IV)NC1 domain, we assessed disulfide-mediated oligomerization of samples mainly consisting of the 27-kDa conformer in the presence of GPBP, or GPBPΔ26 (
As shown above, in the absence of GPBP or GPBPΔ26, DTT consumption resulted in a reduction of monomeric material mainly due to disulfide-dependent molecular aggregation as the reactivity of Mab175, an α3(IV)NC1-specific antibody which reactivity does not vary significantly upon antigen reduction (Borza et al, 2000), towards monomeric molecular species largely increased upon sample reduction. Essentially the same results were obtained when blotting the samples that contained GPBPΔ26. In contrast, when GPBP was present in the reaction mixture during DTT consumption, the resulting material displayed different reactive patterns in the Western-blot studies. Thus, Mab3 reacted with a previously unidentified polypeptide of approximately 28-kDa, in addition to the 27-kDa conformer, indicating that during DTT consumption GPBP catalyzed specific conformational isomerization reactions over the 27-kDa conformer that still maintained the native disulfide bonds arrangement required for Mab3 recognition. Accordingly, after DTT consumption, GPBPΔ26 samples contained a relatively greater abundance of 27-kDa conformer than samples containing GPBP, suggesting that this conformer was the substrate, whereas the 28-kDa polypeptide was the product in the conformational isomerization reaction catalyzed by GPBP. Western-blot analysis using Mab175 antibodies revealed that, in the samples containing GPBP, most of the α3(IV)NC1 material existed as molecular species displaying Mr from 22 to 29 kDa all of which yielded a single molecular species of 29 kDa upon reduction, indicating that GPBP impaired random monomer disulfide-aggregation and catalyzed multiple conformational isomerizations other than the 27- to 28-kDa monitored by Mab3. The catalysis performed by GPBP was ATP independent, required the presence of the DTT-metal-based redox system (not shown), and could be observed with both human recombinant (not shown) or bovine native (shown) α3(IV)NC1 materials.
The presence of α3(IV)NC1 Mab3-reactive material organized in high molecular weight oligomers was also investigated (
However, the most evident effect of GPBP over the α3(IV)NC1 material was to stabilize the different conformers in a monomeric form and to impair random disulfide-aggregation, suggesting that GPBP, and to a minor extent GPBPΔ26, are acting in the in vitro assays as molecular chaperones. Accordingly, GPBP and, to a lesser extent GPBPΔ26 disrupted disulfide-based high molecular weight aggregates characteristic of recombinant material representing human α3(IV)NC1 produced in bacteria which do not enter into the running gel of an SDS-PAGE analysis, and promoted the formation of lower molecular weight disulfide-based oligomers which reacted with Mab 3 (
Finally, we assessed the involvement of phosphate groups present in the α3(IV)NC1 in the overall process catalyzed by GPBP by comparing its action over alkaline phosphatase-treated or untreated fα3 (
Although it is widely accepted that the NC 1 domain of individual chains plays a leading role in collagen formation (Fessler and Fessler, 1982; Ries et al., 1995; Boutaud et al., 2000), the precise mechanism mediating chain selection and assembly is unknown. As indicated herein, the individual NC1 domains are generated as phosphorylation-dependent metastable conformations that become stable once assembled in the “hexamer”.
The mechanism by which α3(IV)NC1 conformers are generated remains to be established. However, the reduced ability of phosphatase-treated material to maintain the native structure and the high phosphoserine content of the non-conventional α3(IV)NC1 conformers, suggest that phosphorylation plays a critical role in the production of multiple non-minimum energy structures.
Phosphorylation also mediates at least in part the molecular recognition and DTT consumption in the oligomerization assays. The latter reveal the existence of a high turnover in the intermolecular disulfide bonds that likely reflects the search for the proper partner, but also suggests the existence of a machinery with the potential to assist disulfide-based cross-linking of the NC1 domain in vivo. We show here that GPBP catalyzes disulfide-based aggregation of the α3(IV)NC1 domain through a process that comprises specific conformational isomerization reactions in vitro, suggesting that GPBP catalyzes at least in part the intermolecular cross-linkage of the “hexamer” in vivo.
The information required to form a collagen IV “hexamer” resides in the covalent structure of the “monomer,” as the individual NC1 domains select their partners to form “hexameric” structures without the assistance of other cellular factors (Boutaud et al., 2000). This suggests that GPBP catalysis is occurring, at least in part, after chain association and during disulfide stabilization of the collagen IV network, a process that occur necessarily outside of the cell (Fessler and Fessler, 1982). Consistently, GPBP is abundantly expressed associated with GBM (Raya et al, 2000), and recent data using confocal microscopy demonstrate that α3(IV)NC1 and GPBP co-localize at the human GBM (Burgués and Saus, unpublished observations).
At the endoplasmic reticulum, differential phosphorylation of a single unique native structure could occur prior or during chain association, yielding multiple metastable structures each one stabilized by individual disulfide-bond distributions. Individual molecular species would have enciphered in their covalent structure the assembly partner and the final conformation that will be acquired once assembled and stabilized into a “hexamer”. In this model, GPBP could be the machinery assisting, deciphering and catalyzing the stabilization of the corresponding quaternary structures.
In the absence of ATP, GPBP catalyzed the formation of multiple conformers and specific oligomers of the α3(IV)NC1 domain, suggesting that the phosphorylated structure of this domain has enciphered multiple assembly programs that require GPBP assistance to be accomplished, and the kinase activity of GPBP could represent an auxiliary function required for specific in vivo folding-assembly reactions which are not occurring in the in vitro assays.
Humans have acquired an additional phosphorylation site for type A protein kinases at the N-terminal region of the α3(IV)NC1 domain (Ser9) ( Revert et al, 1995; Raya et al. 1999 and 2000), yielding a comparatively more phosphorylatable polypeptide (Revert et al., 1995; Raya et al., 1999) with a remarkable susceptibility to undergo autoimmune attack. Recent evidence indicates that phosphorylation of Ser9 (P) regulates at least in part the conformational diversification perhaps operating through an integrin recognition motif adjacent to it. Interestingly we have found that the recombinant counterparts for the α-1,-2,-4,-5 and -6(IV) chains also show a phosphorylation-dependent metastability in the in vitro oligomerization assays, and that human α1(IV)NC1 as well as bovine α3(IV)NC1 domains exist as multiple conformers (unpublished results). This indicates that the phosphorylation-dependent conformational diversification and “activation” for disulfide-aggregation are not a human α3(IV)NC1 exclusive conditions, and therefore cannot be considered the structural feature that renders this system vulnerable to pathogenesis. However, it is conceivable that vulnerability to pathogenesis of the human α3(IV)NC1 system comes from the potential intrusion in conformation of the human exclusive phosphorylation process at Ser9. Accordingly we have presented evidences supporting that a phosphorylation event involving Ser9 can lead to the formation of α3(IV)NC1 conformers for which the immune system has not established a tolerance and trigger an autoimmune attack, which therefore can be envisioned as a legitimate response of the immune system against a misfolded autoantigen
Here we present evidence suggesting that in GP patients an augmented expression of both GPBP and GPΔIII results in the assembly at the glomerular basement membrane of aberrant non-tolerized α3(IV)NC1 conformers that induce and conduct the autoimmune response. Our findings further support previous observations indicating that a phosphorylation event can lead the formation of α3(IV)NC1 conformers for which the immune system have not established a tolerance and therefore induce an immune response.
Synthetic oligonucleotides. The following oligonucleotides and other used for DNA sequencing were synthesized by Genosys, Life Technology Inc., Roche or Pharmacia:
Production of native and recombinant NC1 domain. These materials were prepared as described in the accompanying Examples.
RNA purification. Frozen human tissues were ground in the presence of liquid nitrogen and further disrupted with a Polytron-like device in the presence of either TRI-REAGENT™ (Sigma) and total RNA purified using manufacturer's recommendations, or with 4M guanidine thiocyanate 1% β-mercaptoethanol in 0.1 M Tris pH 7.5 and RNA purification carried out by standard CsCl gradient approach.
Reverse transcriptase coupled polymerase chain reaction studies (RT-PCR). To obtain the cDNA for the α3(IV)NC1 domain and for its alternatively spliced products, total RNA from each individual kidney (0.5 μg) was retro-transcribed using ON-B-HNC-1c. The corresponding single stranded cDNAs were subjected to PCR using ON-HNC-3m and ON-HNC-6c. The products were further identified by nucleotide sequence or restriction map.
The mRNA levels for all the COL4A3 and COL4A3BP products (GPt and GPBPt), GPΔIII, GPBP, GPBPΔ26, or glyceraldehyde 3-phosphate dehydrogenase (GAPDH) in each individual human kidney was estimated by measuring the corresponding cDNAs in the reverse transcription mixtures obtained as above using a random hexamer priming and 5 μg of total RNA. This was accomplished by quantitative PCR using a SDS 7700 Applied Biosystems apparatus and the following primers: ON-GP-F1 and ON-GP-R1; ON-hGPBPc-F1 and ON-hGPBPc-R1; ON-GP-3-F1 and ON-GP-3-R2; ON-GPBPe26-F1 and ON-GPBPe26-R1; ON-GPBP-26-F1 and ON-GPBP-26-R1; or, ON-GAPDH-F and ON-GAPDH-R, respectively. PCR reactions were done using 5 μl of 1:100 and 1:1000 dilutions of the reverse transcriptase except for GAPDH for which determinations the dilutions used were 1:1000 and 1:10000. Standard curves for each PCR were done using the same oligonucleotides and different amounts of individual plasmids containing the corresponding cDNAs.
Immunochemical studies. Immunoblot studies and in situ fα3 binding assays were performed as detailed in Example 5.
Antibodies. The production and specificity of the antibodies are detailed in the accompanying Examples 4 and 5. Tissue-bound antibodies were extracted from a control and from each of two GP kidneys from which NC1 hexamer was prepared for use.
GPΔIII is expressed at higher levels in GP kidneys. We have made the observation that the mRNA level for GPΔIII was augmented with respect to the primary product in a GP kidney and that this could have pathogenic significance (Bemal et al, 1993). This was investigated in additional patient and control kidneys using two different PCR approaches coupled to reverse transcription (
Identification of aberrant α3(IV)NC1 conformers in GP kidneys. Since GPΔIII positively regulates the phosphorylation of the primary α3(IV)NC1 product in vitro, and in this domain phosphorylation plays a critical role in conformation, we investigated the presence of disease associated α3(IV)NC1 conformers in GP kidneys. We have previously reported that there are not differences in the primary structure of patient α3(IV)NC1 that could account for its immunogenic condition, and therefore if there are structural differences between patient and control α3(IV)NC1 domains which account for the immunogenicity they must be post-translational (Bernal et al, 1993). Thus, after confirming by direct cDNA sequencing the fidelity of the primary structure of the α3(IV)NC1 domain in each individual patient kidney, we isolated the collagen IV NC1 domain (“hexamer”) from patient kidneys 2 and 3, and also from control kidneys and we assessed the binding of α3(IV)NC1-specific antibodies, which reactivity largely depends on antigen conformation (
The disulfide-bond cross-linkage of the NC1 domain is defective in GP kidneys. Since conformational differences are expected to be reflected in the quaternary structure (“hexamer”), the disulfide-based oligomeric subunits representing this structural level were analyzed in both patient and control “hexamers” (
The aberrant α3(IV)NC1 conformers conduct the immune response in GP disease. The conformational alterations present in the α3(IV)NC1 of GP patients, however, does not significantly reduce the gross amount of α3(IV) chain assembled into the collagen IV network since the reduced proportion of high molecular weight oligomers is compensated by a higher content in dimers (
The expression of GPBP is augmented in GP kidneys. We have shown that GPBP phosphorylates the N terminal region of the α3(IV)NC1 domain including Ser9 in vitro (Raya et al.,1999) and that Ser9 phosphorylation determines the cohort of conformers produced by the cell (Example 4). Furthermore, GPBP is expressed associated with alveolar and glomerular basement membranes and an augmented expression of GPBP has been associated with different autoimmune conditions including a GP patient (Raya et al, 2000). Consequently, to investigate the implication of GPBP in GP pathogenesis, we estimated by reverse transcriptase coupled to quantitative PCR, the transcriptional activity of COL4A3BP, the gene encoding GPBP and GPBPΔ26, in both patient and control kidneys (
The higher specificity of the pathogenic antibodies towards aberrant α3(IV)NC1 conformers present in disease-affected tissues indicates that this material is the antigen conducting the autoimmune response, and suggests that alterations in the tertiary structure of α3(IV)NC1 domain cause GP disease.
The data presented here and in the accompanying Examples support that phosphorylation activates the α3(IV)NC1 domain for disulfide bond-aggregation, a process that is catalyzed by GPBP, involves specific conformational isomerization reactions and which results in the assembly and stabilization of multiple conformers of this domain in the basement membrane. In the absence of ATP, GPBP catalyzes the formation of multiple conformers and specific oligomers of the α3(IV)NC1 domain in vitro Example 5), suggesting that the phosphorylated structure of this domain has enciphered multiple assembly programs which require GPBP assistance to be accomplished. Consistently, alkaline phosphatase-treated α3(IV)NC1 did not aggregate efficiently and this material was unable to follow a disulfide bond-aggregation program in the presence of GPBP (Example 5).
In vitro, PKA and GPBP phosphorylate the human α3(IV)NC1 domain at Ser9, a site that is also targeted by the endogenous phosphorylation process (Revert et al, 1995; Raya et al., 1999). The evidence indicates that the homeostasis of Ser9 phosphorylation is critical for physiological conformer production (Example 4). In addition to Ser9, the N-terminal region of the human α3(IV)NC1 contains additional phosphorylation sites not present in other species (Ser11 and Thr14,16,17), which are also targeted by the two kinases in vitro ( Raya et al, 1999; Revert et al, unpublished observations) suggesting that N-terminal phosphorylability is critical for pathogenesis.
In a yeast two hybrid system, the fly counterpart of GPBP interacts with the corresponding fly cPKA. (Carine Rossé and Jacques Camonis, personal communication) Bovine cPKA phosphorylates GPBP in vitro (not shown). Finally, type A protein kinases and GPBP have been found associated with cell plasma membrane and endothelial basement membranes, respectively (Revert et al., 1995; Raya et al., 2000). All these suggest that the two kinases can interact and form stable complexes in vivo and which operate during the molecular and supramolecular assembly of the collagen IV.
In addition to divergence at the N-terminal region of the α3(IV)NC1 domain (Quinones et al, 1992), humans have developed a unique alternative splicing mechanism to regulate phosphorylation of Ser9 by cPKA (herein), resulting in a comparatively more vulnerable polypeptide to undergo conformational alterations and an autoimmune attack.
The GP antibodies recognize a potent immunogenic region adjacent to the exclusive N-terminus that harbors also Mab3 epitope (Borza et al, 2000). The main epitope for the GP antibodies is maintained by disulfide bonds and depends on hydrophobic residues that require dissociation of the “hexamer” to be exposed (Netzer et al, 1999; Hellmark et al, 1999; Borza et al., 2000; David et al, 2001). Mab3 epitope is maintained by the same disulfide bonds but involves hydrophilic residues that are accessible in the “hexamer” (Saus et al, 1988; Johansson et al, 1991; Borza et al., 2000; David et al, 2001). Thus, during pathogenesis an aberrant N-terminal phosphorylation could result in conformers with a higher exposure of the hydrophobic residues, which because of the disulfide bonds would still maintain the reactivity with Mab3. Consistently, permanently phosphorylated versions of the α3(IV)NC1 domain at Ser9 show a relative higher specificity with Mab3 (Example 4) and with GP autoantibodies (not shown). Our data also indicate that a similar pathogenic mechanism is operating in every patient, therefore the resulting conformational alterations are expected to be highly similar among patients as no alterations in the primary structure of the patient α3(IV)NC1 have been found. This would account for the large cross-reaction among patient autoantibodies but also for the high affinity that tissue-bound autoantibodies from one patient display for the 27-kDa conformer of other patient in comparison with the affinity displayed towards control material (not shown).
COL4A3BP, the gene encoding GPBP and GPBPΔ26, and POLK the gene encoding for pol κ, a member of the UmuC/DinB superfamily of DNA polymerases which can extend aberrant replication forks are transcribed in a divergent mode from a bi-directional promoter (Granero et al, unpublished results). This promoter shows high sequence homology with a number of other bi-directional promoters including that transcribing COL4A3 and COL4A, the genes encoding the α3 and α4 chains of collagen IV. The homology between promoters transcribing otherwise unrelated structural genes reveals the existence of a convergent evolution phenomenon to coordinate their expression (Granero et al, unpublished results). Accordingly during pathogenesis we found a transcriptional induction of the two genes. Moreover, the signal(s) to coordinate the expression of these genes seems to reach the machinery regulating pre-mRNA processing, since GPΔIII and GPBP, which represent minor mRNA forms in each individual gene system, are the mRNA species more significantly increased.
Taking all these data together, it is plausible to think that during pathogenesis triggering events by increasing the expression of both GPΔIE and GPBP, cause an aberrant N-terminal phosphorylation generating activated α3(IV)NC1 structures with an altered disulfide bond-aggregation program. Subsequently, GPBP would catalyze its assembly into the collagen IV network resulting in the presence of altered conformers in the basement membrane. Finally, aberrant assembled α3(IV)NC1 conformers would induce and drive a T cell-dependent antibody-mediated immune response (
We have shown above in an in vitro system that during GPBP catalysis, and prior disulfide bond-aggregation of the α3(IV)NC1 domain, the 27-kDa conformer undergoes conformational isomerization to generate α28-kDa conformer similar to that found in Patient 2, suggesting that the Mab3-reactive 28-kDa conformer found in the “hexamer” of Patient 2 likely represents a trapped intermediate which derive from an aberrant 27-kDa conformer that is incapable to follow the correct disulfide bond-aggregation.
These and previous data which show that GPBP is abundantly expressed in structures that either are the target of common autoimmune responses or are undergoing an autoimmune attack (Raya et al, 1999 and 2000) reveal that GPBP plays a major role in human autoimmunity and suggest that the production of non-tolerized conformational versions of different autoantigens is operating in other autoimmune pathogenesis.
The molecular basis of the autoimmune responses has been elusive. The findings presented in this and the accompanying Examples lead to a new concept of the human autoimmune response, which is envisioned as a legitimate reaction of the immune system towards a non-physiologically folded but still assembled autoantigen.
The present invention is not limited by the aforementioned particular preferred embodiments. It will occur to those ordinarily skilled in the art that various modifications may be made to the disclosed preferred embodiments without diverting from the concept of the invention. All such modifications are intended to be within the scope of the present invention.
This application claims priority to U.S. Provisional Patent Application Ser. No. 60/265,249, filed Jan. 31, 2001
| Number | Date | Country | |
|---|---|---|---|
| 60265249 | Jan 2001 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10060607 | Jan 2002 | US |
| Child | 11090878 | Mar 2005 | US |
