EP0642665A4

EP0642665A4 - Method and assay involving improved specific binding reactivity of a polypeptide.

Info

Publication number: EP0642665A4
Application number: EP94912878A
Authority: EP
Inventors: Robert C Wohl; Gregory A Marchildon
Original assignee: Curative Technologies Inc
Current assignee: Curative Technologies Inc
Priority date: 1993-03-26
Filing date: 1994-03-25
Publication date: 1997-05-02
Also published as: EP0642665A1; AU6525494A; CA2136757A1; WO1994023297A1; JPH08504276A

Abstract

A method of improving the specific binding reactivity of a polypeptide which is capable of polypeptide complexing is disclosed. The method involves contacting the polypeptide with an anionic surfactant capable of decomplexing the polypeptide under suitable conditions resulting in at least some decomplexing of the polypeptide. The polypeptide is then contacted with a nonionic surfactant capable of complexing with the anionic surfactant under suitable conditions resulting in at least some of the decomplexed polypeptide remaining decomplexed. A method of assay of a sample containing a polypeptide which is capable of polypeptide complexing is also disclosed.

Description

METHOD AND ASSAY INVOLVING IMPROVED SPECIFIC BINDING REACTIVITY OF A POLYPEPTIDE

This is a continuation-in-part of U.S. patent application Ser. No. 08/037,596, filed March 26, 1993, now abandoned.

FIELD OF THE INVENTION The invention relates to a method of improving the specific binding reactivity of a polypeptide in a sample, and specific binding assays incorporating that method.

BACKGROUND OF THE INVENTION Many polypeptides are present in biological fluids and assay samples in a form in which a specific binding site or sites are inaccessible to binding partners specific to those binding sites. The inaccessibility of the binding sites is due to such phenomena as noncovalent binding between the polypeptide of interest and molecules with which it is complexed in precursor or latent form, or between the polypeptide of interest and other molecules in the sample. The noncovalently bound molecule or molecules block specific binding sites on the polypeptide of interest. The molecules with which the polypeptide of interest is complexed in the biological fluid or sample include other polypeptides, proteoglycans, glycoproteins, and lipoproteins. Some complexed polypeptides are formed in the fluid or sample, such as blood or serum, while other complexed polypeptides are secreted in complexed form from cells such as platelets. The noncovalent binding in these polypeptide complexes is very strong, and is usually reversible only by proteolytic digestion. This noncovalent binding hampers efforts to design specific binding assays that exploit the highly specific reaction between a specific binding site and its specific binding partner. Specific binding assays that target polypeptides which are present at least partially in complexed form are unreliable and produce inconsistent results due to the interference of the noncovalently bound polypeptide, proteoglycan, glycoprotein or lipoprotein with a specific binding site for a specific binding partner.

Illustrative of this problem is the assaying of transforming growth factor β (TGFβ). Transforming growth factor ^β is a ubiquitous molecule which regulates growth and differentiation in many cell types. TGFβ was first described as a factor which caused phenotypic transformation of rat fibroblasts, but is now known as a multifunctional regulator of cellular growth and differentiation; it is a potent growth inhibitor for many cells, such as epithelial, endothelial, and hematopoietic cells, and T and B lymphocytes, but induces proliferation in other cell types, primarily cells of mesenchymal origin. TGFβ has been found to stimulate, in vivo- fibrosis, angiogenesis, and wound healing. The mitogenic activity of TGFβ might be indirect, since it is now known that TGFβ induces production of platelet- derived growth factor isoforms, which are potent mitogens for cells of mesenchymal origin. The broad spectrum of known biological properties of TGFβ suggests that TGFβ might play important roles in such physiological processes as morphogenesis, wound healing, hemapoiesis, and immunoregulation, and such pathogenic processes as oncogenesis. At least three subtypes of TGFβ have been identified in mammals:

TGFβl- TGFβ2, and TGFB3 (the TGFβs). Other subtypes have been identified in other species. The mammalian subtypes are produced by a variety of cell types: in humans, platelets are a rich source of TGFβl, and TGFβl may also be produced in bone matrices, kidney, placenta, normal fibroblasts, endothelial cells, leukocytes, and some tumor cell lines. TGFB2 has been found in high concentrations in bone extracts, the pregnant uterus, and in the supernatant fluid of some cultured tumor cell lines. TGFβl, TGFβ2, and TGFB3 mRNAs have been found in a variety of adult human tissues. There is a high degree of primary sequence and structural homology among individual TGFβ subtypes of different species; the subtypes behave similarly in most assays, and exhibit a high degree of cross-species activity.

The TGFβs are secreted from their producer cells predominantly as a latent, high molecular weight complex form (LTGFβ). In the case of human TGFβl, for example, LTGFβ 1 is secreted as a complex of approximately 235,000 daltons. The TGFβ precursor contains a large N- terminal domain, and a highly conserved C-terminal domain of 112 amino acids that is present in active TGFβ. The mechanism of processing from the precursor TGFβ to active TGFβ is not well understood, but it appears that dimerization of precursor TGFβ commences with the formation of several disulphide bonds between C-terminal domains of precursor molecules, followed by proteolytic cleavage of the N-terminal domain. The cleaved, N-terminal domain (the latent peptide) remains noncovalently bound to the C-terminal dimer. The complex containing the latent peptide and the noncovalently associated C-terminal domain is referred to here as latent TGFβ (LTGFβ). This appears to be the main form in which TGFβ is secreted from cells. Active TGFβ is a dimer of two identical, 12,500 dalton, 112 amino acid chains linked by nine disulphide bonds. The 112 amino acid chains correspond to C-terminal domains of the TGFβ precursor. In the case of human LTGFβ 1, therefore, the latent peptide is approximately 210,000 daltons.

Because of TGFβ's importance as a multifunctional regulator of cellular growth and differentiation, efforts have been made to provide accurate and reproducible assays, including specific binding assays, for TGFβ. The fact that TGFβ is secreted predominantly in the latent complex described above presents a problem in the development of an accurate and reproducible specific binding assay for TGFβ in a sample wherein TGFβ is present at least partially in the latent, complexed form. Antibodies, or other binding partners specific to LTGFβ, are not readily available, and specific binding sites on active TGFβ, for which specific binding partners are available, are inaccessible due to noncovalent binding of the active, 25,000 dalton dimer to the latent peptide in the secreted, complexed form, or by noncovalent binding to other peptides in the sample, such as alpha-2 macroglobulin. Methods described in the literature to activate LTGFβ for specific binding assays focus principally on acidification of a sample containing LTGFβ to decomplex TGFβ from LTGFβ followed by neutralization prior to an immunoassay. The acidification/neutralization treatment is designed to decomplex LTGFβ by disrupting the noncovalent binding between active TGFβ and the latent peptide or other peptides in the sample, separating the active TGFβ from the latent peptide or other peptides, and thereby make epitope recognition sites on the active TGFβ accessible to specific binding partners in an immunoassay. Assays described in the literature employing this treatment report concentrations of 50-70 ng of TGFβ per billion platelets. Purification experiments for TGFβl, however, indicate that there is much more TGFβl in human platelets than is being measured in immunoassays reported in the literature using acidification/neutralization as a method of decomplexation. Semi-quantitative estimates of active TGFβl in thrombin activated platelet releasate (TPR) as derived from Western blots similarly indicate that much more TGFβl is present in human platelets than is reported in the literature. The likely reason for the disparity between reported levels of TGFβl from immunoassays using acidification/neutralization and levels observed in purification experiments is that the acidification/neutralization might not completely decomplex active TGFβ from the latent peptide, the decomplexed TGFβ might reassociate with the latent peptide or other peptides in the sample and lose immunoreactivity, or active TGFβ might interact with vessel surfaces or settle out of the sample solution. Also illustrative of this problem is the assaying of platelet factor 4

(PF4). PF4 is a multifunctional regulatory protein released from platelets. One of PF4's important functions is to bind heparin and heparin-like molecules on cell surfaces and endothelial surfaces. PF4 is also known to be a chemoattractant for monocytes and is a possible anti-cancer agent. PF4 is released from platelets as a 350,000 dalton complex containing eight tetramers of PF4 and two proteoglycan molecules of 59,000 daltons. The monomer form of PF4 is biologically functional and immunoreactive.

Because of the complexed form in which PF4 is released from platelets and found in biological fluids, it is very difficult to precisely quantify PF4 in a sample by specific binding assay, for example, of human serum or a platelet releasate preparation. At least some of the immunoreactive binding sites on the PF4 monomers are inaccessible to binding partners specific to those sites because of noncovalent complexing. Commercially available assay kits for PF4 produce variable and inconsistent results, principally because there is no means in such kits for making all specific binding sites available to a specific binding partner. Literature reports show a concentration of approximately 12 ± 5 μg per 10^ platelets in radioimmunoassays run on totally disrupted platelet supernatants. Several other polypeptides are known to form or be secreted as noncovalent complexes in biological fluids, making specific binding assays for such polypeptides unreliable and inconsistent. α2-macroglobulin is a large polypeptide found in serum that forms noncovalent complexes with proteases (and thus serves as a protease inhibitor) and growth factors such as TGFβs and platelet derived growth factor (PDGF). Plasminogen activator inhibitor 1 (PAT-1) is a polypeptide protease inhibitor found in blood which forms a noncovalent complex with vitronectin, an adhesion molecule found in blood. Osteonectin, a grown factor-like molecule found in blood, is known to form a noncovalent complex with PDGF. Each of these noncovalent complexes is characterized by strong noncovalent bonding between the polypeptide of interest and another molecule which may be a polypeptide, a proteoglycan, glycoprotein, or lipoprotein. In the complexed form, specific binding assays of the complexed polypeptide of interest are unreliable and produce inconsistent results due to the inaccessibility of specific binding sites in the complexed form. Such specific binding assays provide no mechanism for disrupting the noncovalent binding in the complex, retaining the disruption and keeping specific binding sites accessible for specific binding partners. There is thus a need, as exemplified by problems associated with specific binding assays for TGFβ, PF4 and the other complexed polypeptides discussed above, for an efficacious method of improving the specific binding reactivity of a polypeptide having specific binding sites inaccessible to specific binding partners. Such a method would make specific binding sites on the polypeptide accessible to binding partners used in specific binding assays by effecting complete decomplexation of the polypeptide, and the loss of specific binding activity due to reassociation of noncovalent complexes or interactions between the polypeptide and container surfaces would be m-inimized or eliminated. The desirable method would retain at least some of the polypeptide of interest in a decomplexed form in order to prepare the polypeptide for a specific binding assay, and would be incorporated into specific binding assays for the polypeptide. SUMMARY OF THE INVENTION The present invention provides a method of improving the specific binding reactivity of a polypeptide which is capable of polypeptide complexing. The polypeptide is contacted with an anionic surfactant capable of decomplexing the polypeptide under suitable conditions which result in at least some decomplexing of the polypeptide. The resultant decomplexed polypeptide is contacted with a nonionic surfactant capable of complexing with the anionic surfactant under suitable conditions resulting in at least some of the decomplexed polypeptide remaining decomplexed. The polypeptide may be any polypeptide capable of forming polypeptide complexes, including a transforming growth factor β (TGFβ), platelet factor 4, an c^-macroglobulin bound protease, an o^-macroglobulin bound growth factor, vitronectin bound plasminogen activator-inhibitor 1, and osteonectin bound platelet derived growth factor (PDGF). In a preferred embodiment, the polypeptide is a TGFβ. In another preferred embodiment, the polypeptide is platelet factor 4.

The anionic surfactant may be sodium dodecyl sulfate, triethanolammomum lauryl sulfate, diethanolammonium lauryl sulfate, sodium cetyl sulfate, dioctyl sodium sulfosuccinate, discolium monococoamide sulfosuccinate, sodium isostearyl-2-lactylate. sodium cetearyl sulfate, and sodium cocoyl isethionate. In preferred embodiments of the present invention, the anionic surfactant is sodium dodecyl sulfate, also known as sodium lauryl sulfate, or SDS.

The nonionic surfactant may be an organic aliphatic or aromatic ethylene oxide adduct formed by the reaction of ethylene oxide with aliphatic or aromatic hydroxy substituted compounds. In preferred embodiments of the present invention, the nonionic surfactant is Tween-20. The specific binding reactivity which is improved by the method of the present invention may be immunoreactivity or the specific binding reactivity of a purified receptor or a purified receptor fragment.

The present invention further provides a method of assay of a sample containing a polypeptide which is capable of polypeptide complexing. The polypeptide is contacted with an anionic surfactant which is capable of decomplexing the polypeptide under suitable conditions which result in at least some decomplexing of the polypeptide. The resultant decomplexed polypeptide is contacted with a nonionic surfactant which is capable of complexing with the anionic surfactant under suitable conditions resulting in at least some of the decomplexed polypeptide remaining decomplexed. After the step of contacting the decomplexed polypeptide with the nonionic surfactant, the polypeptide is contacted with a specific binding partner under conditions suitable for binding of the polypeptide and the specific binding partner. The specific binding partner may be an antibody, and antibody fragment, a purified receptor, a purified receptor fragment, or a synthetic peptide. The extent of binding between the specific binding partner and the polypeptide is measured.

The method of assay may further include the step of correlating the extent of binding between the specific binding partner and the polypeptide with the presence or amount of the polypeptide in the sample.

The polypeptide may be a transforming growth factor β (TGFβ), platelet factor 4, an o^-macroglobulin bound protease, an o^-macroglobulin bound growth factor, vitronectin bound plasminogen activator-inhibitor 1, and osteonectin bound platelet derived growth factor (PDGF). In a preferred embodiment, the polypeptide is a TGFβ. In another preferred embodiment, the polypeptide is platelet factor 4. The anionic surfactant may be sodium dodecyl sulfate, triethanolammomum lauryl sulfate, diethanolammonium lauryl sulfate, sodium cetyl sulfate, dioctyl sodium sulfosuccinate, discolium monococoamide sulfosuccinate, sodium isostearyl-2-lactylate, sodium cetearyl sulfate, and sodium cocoyl isethionate. In preferred embodiments of the present invention, the anionic surfactant is sodium dodecyl sulfate, also known as sodium lauryl sulfate, or SDS.

The nonionic surfactant may be an organic aliphatic or aromatic ethylene oxide adduct formed by the reaction of ethylene oxide with aliphatic or aromatic hydroxy substituted compounds. In preferred embodiments of the present invention, the nonionic surfactant is Tween-20. The binding between the specific binding partner and the polypeptide may be immunoreactive binding or the specific binding reactivity of a purified receptor or a purified receptor fragment. BRIEF DESCRIPTION OF THE FIGURES Figure 1 is a standard curve for an immunoassay for human TGFβl. Figure 2 shows the loss of immunoreactivity of active TGFβl in Thrombin induced platelet releasate, over time in acidified/neutralized and acidified/heated/neutralized samples, and the retention of immunoreactivity in sodium dodecyl sulfate (SDS)/nonionic ethoxylate treated samples.

Figure 3 is a standard curve for an immunoassay for human platelet factor 4.

DETAILED DESCRIPTION OF THE INVENTION

The present invention provides a method of improving the specific binding reactivity of a polypeptide. The polypeptide can be any polypeptide which is capable of polypeptide complexing. The term polypeptide complexing encompasses and refers to the formation of noncovalent bonds between a polypeptide and another molecule or molecules. The other molecule or molecules may be a polypeptide, a proteoglycan, a glycoprotein, or a lipoprotein. The polypeptide is characterized in that specific binding sites on the polypeptide are inaccessible to binding partners specific to the specific binding sites. Examples of such complexes include latent TGFβ (polypeptide-polypeptide complex), platelet factor 4 (polypeptide-proteoglycan complex), α₂- macroglobulin bound proteases and o^-macroglobulin bound to growth factors such as TGFβ and platelet derived growth factor (polypeptide- polypeptide complexes), vitronectin bound plasminogen activator-inhibitor 1 (polypeptide-polypeptide complex), and osteonectin bound platelet derived growth factor (polypeptide-polypeptide complex). An example of a polypeptide-lipoprotein complex is plasminogen receptor-lipoprotein a.

Each of the above recited complexes is characterized by noncovalent bonding between the molecules of the complex. The noncovalent bonding in the various complexes is of the same type and the method of the invention is applicable to any polypeptide complex containing such noncovalent bonds and to improving the specific binding reactivity of any polypeptide molecule of interest which is capable of forming such noncovalent bonds. In accordance with the method of the present invention, the polypeptide which is capable of polypeptide complexing is contacted with an amount of an anionic surface active agent (surfactant) capable of decomplexing such polypeptide under suitable conditions resulting in at least some decomplexing of the polypeptide. This frees up the specific binding sites on the polypeptide in the sample that are normally inaccessible to specific binding partners for the specific binding sites because of noncovalent bonds, such as ionic or hydrogen bonds, between the polypeptide and other molecules. These other molecules may be polypeptides which are associated with the polypeptide of interest in a latent or precursor form of the polypeptide of interest, or may be other polypeptides, proteoglycans, glycoproteins, or lipoproteins such as those discussed above. Such molecules are associated with the polypeptide such that the specific binding sites on the polypeptide are rendered inaccessible to binding partners specific to the binding sites. The anionic surfactant is a salt possessing a negatively charged moiety capable of interacting with positively or negatively charged regions on the polypeptide. The hydrophobic portion of the surfactant provides compatibility with portions of the polypeptide. By this interaction the anionic surfactant disrupts and interferes with noncovalent associations, such as ionic bonds between negatively charged regions of one polypeptide and positively charged regions of another, or associations through weaker bonding forces, between the polypeptide and any other polypeptides in the sample. The binding of the anionic surfactant to the polypeptide creates a surface having a hydrophilic, negative charge which repels other molecules with similar surface charge. The result of contacting the polypeptide in the sample with the anionic surfactant is to separate polypeptides and other macromolecules in the sample from each other and to put individual polypeptides, including the polypeptide to be assayed, into solution in separate, dissociated form. Treatment with the anionic surfactant under optimized conditions such as those discussed below will effect a complete dissociation of the noncovalent complexes.

A suitable anionic surfactant can be any of the following: sodium dodecyl sulfate, triethanolammonium lauryl sulfate, diethanolammonium lauryl sulfate, sodium cetyl sulfate, dioctyl sodium sulfosuccinate, disodium monococoamide sulfosuccinate, sodium isostearyl-2-lactylate, sodium cetearyl sulfate, sodium cocoyl isethionate, and the like. The most preferred anionic surfactant for use in the method of the present invention is sodium dodecyl sulfate, also known as sodium lauryl sulfate, or SDS. The sample is adjusted to a concentration of the anionic surfactant sufficient to react with and decomplex the polypeptide in the sample. For example, using a freshly made solution of the anionic surfactant, a sample can be incubated for a time sufficient to achieve decomplexing of the polypeptide in the sample. When SDS is used as the anionic surfactant, the concentration of SDS in the sample could range from 0.05% to 2.5%

(weight/volume), and the sample could be incubated for one hour at room temperature.

The anionic surfactant operates to decomplex the polypeptide in the sample by completely disrupting noncovalent bonding between the polypeptide and the molecule or molecules with which it is complexed; the polypeptide, however, is not yet prepared for a specific binding assay. Interactions between the anionic surfactant and the decomplexed polypeptide and/or between the anionic surfactant and a binding partner specific to the specific binding site or sites on the polypeptide, would interfere with or prevent the specific binding reaction between the specific binding site or sites on the polypeptide and the binding partner that is essential to a specific binding assay. For example, after decomplexing of a polypeptide with SDS, a specific binding assay would be ineffectual; the reaction between specific binding sites on the polypeptide and binding partners specific to those binding sites would be substantially inhibited by interactions between SDS and the specific binding sites and SDS and the binding partner.

In accordance with the invention, therefore, the decomplexed polypeptide in the sample is contacted with an amount of a nonionic surfactant which is capable of substantially complexing with the anionic surfactant while retaining the polypeptide as substantially decomplexed.

Nonionic surfactants contain molecular portions which are hydrophobic and other molecular portions which are hydrophilic. In the method of the present invention the nonionic surfactant is capable of complexing with the anionic surfactant through, for example, hydrogen bonding with the negatively charged moiety on the anionic surfactant, to displace or remove the anionic surfactant from its interaction with the polypeptide in the sample. The addition of the nonionic surfactant to a solution containing a polypeptide which has been decomplexed with an anionic surfactant effects, in preferred embodiments of the invention, a dilution of the anionic surfactant and the formation of mixed micelles of the anionic and nonionic surfactants. The formation of the mixed micelles depletes the polypeptide of bound anionic surfactant, which exposes the previously inaccessible binding sites and allows the polypeptide to bind to a specific binding partner in an assay. The dilution of the anionic surfactant takes the concentration of the anionic surfactant down to levels which no longer interfere with binding of the polypeptide to a specific binding partner. At the same time, there are sufficient concentrations of both the anionic and nonionic surfactants to keep the polypeptide in a soluble and measurable state. In the solution which has been treated with the anionic and nonionic surfactants, the complex containing the polypeptide of interest does not re-assemble because of the action of the surfactants and because the reformation of the complex is an energy-requiring step and few, if any, complexes will re-assemble. Thus, at equilibrium, it is unlikely that the complexes will re-assemble.

Nonionic surfactants suitable for use in the method of the present invention include the organic aliphatic and aromatic ethylene oxide adducts formed by the reaction of ethylene oxide with aliphatic or aromatic hydroxy substituted compounds, preferably alcohols and phenols. These surfactants may be the ethylene oxide adducts of polyols esterified with Cg-C₂o fatty acids, ethylene oxide adducts of C₈-C₂o aliphatic alcohols, ethylene oxide adducts of phenol and alkyl substituted phenols, and the like. The polyol adducts may be such compounds as glycerine, ethylene glycol, or sugars such as glucose, pentaerythritol, sorbitol, and the like, that are esterified with C₈-C2o fatty acids, such as hexanoic acid, 2-ethylhexanoic acid, lauric acid, oleic acid, and the like. Other nonionics include the Tritons® or Tergitols® (sold by Union Carbide Chemical and Plastics, Inc., Danbury, C ). Those are ethoxylated long chain C₈-C₂o alcohols, phenols and nonyl phenols. Preferred nonionic surfactants are the Tweens® (polyoxyethylene polyol mono fatty acid carboxylates, the Tritons and the Tergitols). Particularly preferred is Tween-20, a polyoxyethylene sorbitan monolaurate.

The nonionic surfactant is preferably added to the sample in substantial excess of the anionic surfactant. For example, in a method of decomplexing a polypeptide in a sample, where SDS is used as an anionic surfactant and adjusted to a concentration of 0.05% SDS (weight/volume), and Tween-20 is used as the nonionic surfactant, the concentration of Tween-20 in the sample would be adjusted to 0.8% Tween-20 (volume/volume) . The specific binding activity which is improved by the method of the present invention may be immunoreactivity or the specific binding activity of a purified receptor or purified receptor fragment. Immunoreactivity, as referred to here, encompasses the binding activity between an epitope recognition site on a molecule and a binding partner specific to that epitope recognition site. Such binding partner can be a monoclonal antibody, a polyclonal antibody, a Fab fragment of a monoclonal or polyclonal antibody, or any binding partner or fragment of such binding partner made, synthetically or through inoculation of a host organism, to bind specifically to the epitope recognition site. The term receptor binding activity encompasses specific binding relationships between purified receptor polypeptides or fragments thereof and binding partners displaying specificity for binding sites on the purified receptor or purified receptor fragment.

It should be recognized that the precise conditions for the decomplexing of a polypeptide in a sample in accordance with the present invention might vary depending upon the nature of the sample to be tested and the concentration or suspected concentration of the polypeptide of interest and other proteins in the sample. Each sample will require experimental determination of the amount of the anionic surfactant sufficient to decomplex the polypeptide in the sample. Once the requisite concentration of the anionic surfactant is determined, the amount of nonionic surfactant to be used can be adjusted such that the concentration of the nonionic surfactant in the sample is in appropriate excess of the concentration of the anionic surfactant. The method of decomplexing a polypeptide of this invention has the advantage over methods of decomplexing in the prior art, such as acidification-neutralization, of preventing the loss of specific binding activity through noncovalent reassociation between the polypeptide and other polypeptides in the sample or between the polypeptide and container surfaces present in a specific binding assay.

In the specific binding assay of the present invention, a sample containing or suspected of containing a polypeptide having specific binding sites inaccessible to binding partners specific to the specific binding sites is decomplexed in accordance with the invention as described above.

The decomplexed polypeptide is then contacted with a specific binding partner under conditions suitable for the binding of the decomplexed polypeptide and the specific binding partner. As discussed above, the specific binding partner can be a monoclonal or polyclonal antibody or fragment thereof, such as a Fab fragment or other fragment containing a region having specific binding reactivity with the specific binding site on the polypeptide, a purified receptor, a purified receptor fragment, or a synthetic peptide.

After an incubation in which the substantially decomplexed polypeptide binds with its specific binding partner, the extent of binding between the polypeptide and the specific binding partner is measured. Such measurement can be performed in accordance with known specific binding assay methods. Such methods include homogeneous and heterogeneous immunoassays using a variety of methodologies to signal and measure the extent of binding between the binding partner and the polypeptide. Such methodologies include enzyme-linked immunosorbent assay (ELISA), sandwich enzyme-linked immunosorbent assay (SELISA), direct or sandwich immunoassays using chemiluminescence, fluorescence, enzyme- substrate reactions, or radioactivity as a signaling means. SELISAs for TGFβ and platelet factor 4, incorporating the method of the present invention, are set forth below in the examples.

In a preferred embodiment of the specific binding assay of the present invention, the extent of binding between the binding partner and the polypeptide is correlated with the presence or amount of the polypeptide in the sample. The correlation can be effected by the use of a standard curve, where samples containing known concentrations of the polypeptide are assayed in the specific binding assay, and the results of the specific binding assay run on a sample containing unknown amounts of the polypeptide are compared with the standard curve. The preparation of standard curves for specific binding assays for TGFβl and platelet factor 4 in accordance with the present invention are described in Examples I and IV.

In a preferred embodiment of the specific binding assay, the polypeptide is TGFβ. As described above, TGFβ is present in a sample in predominantly latent form. In this embodiment, latent TGFβ is decomplexed as described above to prepare a sample containing or suspected of containing TGFβ for a specific binding assay. In another preferred embodiment the polypeptide is platelet factor 4 which, as described above, is secreted from platelets as a large complex containing eight platelet factor 4 tetramers and two proteoglycans. In this embodiment, complexed platelet factor 4 is decomplexed as described above to prepare the polypeptide for a specific binding assay. The platelet factor 4 may also be decomplexed with an anionic surfactant and then coated onto a plate or solid phase, and assayed (by immunoassay) in the presence of the anionic and nonionic surfactants, as described below in the examples.

In preferred embodiments of the specific binding assay, the anionic surfactant is SDS, as described above, and the nonionic surfactant is Tween-20. The specific binding activity which is improved by the present invention is preferably immunoreactivity or the specific binding activity of a purified receptor or a purified receptor fragment, as described above.

The present invention contemplates that the method of improving the specific binding reactivity of a polypeptide will have wide applicability to the preparation of complexed polypeptides in samples for a wide variety of specific binding assays. It should thus be recognized that any known or available specific binding assay can be performed on a sample containing a polypeptide has been decomplexed in accordance with the present invention.

The following examples demonstrate the protocol for SELISAs for human TGFβl and human platelet factor 4, the preparation of standard curves for such assays, and the superiority of the surfactant decomplexing method of the present invention over the acidification/neutralization protocol for decomplexing latent TGFβ described in the literature.

While the invention is exemplified by reference below to specific embodiments, examples and applications, the principles involved are susceptible of numerous other applications which will be apparent to persons skilled in the art. The numerous other applications of the principles of this invention can be practiced without departing from the scope of the invention as delimited in the appended claims. Aspects of the invention are illustrated by the following examples.

Example I: Preparation of a Standard Curve Using Sandwich Enzyme-Liked Immunosorbent Assay of Pure Human TGFβl

A standard curve for a human TGFβl assay was generated using the following protocol.

TGFβl samples were prepared for assay as follows: Pure, recombinant active human TGFβl (CRI cat. no. 40039-0004), in a working solution of 10 ng/ml in a standard buffer containing 0.10 mg/ml human serum albumin (NYBC ASA 25%), 0.05% (wt/v) SDS (electrophoresis grade, Bio-Rad cat. no. 161-0302), and 0.8% Tween-20 (v/v) (Sigma cat. no. P-1379) was used to prepare samples for assay at concentrations of 0.00, 0.100, 0.200, 0.300, 0.400, 0.500, 0.750, 1.000, 2.000, 2.500, 3.000, 4.000, 5.000, 6.000, 7.500, and 10.00 ng/ml. The standard curve was generated using samples containing known concentrations of pure, active human TGFβl in the presence of SDS and Tween-20; the TGFβl thus was not decomplexed from its latent form for these assays. The object of this example is to demonstrate that a specific binding assay for a polypeptide, specifically an immunoassay, can be accurate and reproducible when run on samples containing the decomplexing surfactants, and that a precise standard curve can be generated in such an assay.

96-well plates were coated as follows: Corning microtiter plates with 1 x 8 strips in frames (cat. no. 24106-8) were coated overnight at 4°C with 0.1 ml murine monoclonal antibody IgG against human TGFβl (Genzyme cat. no. 1835-01) in PBS at 2.5 μg/ml (1:400 dilution of a 1 mg/ml stock solution of antibody). The wells were covered with cover sheets (Fisher cat. no. 14-245-20) for the microtiter frames.

After coating, wells were washed five times with 0.20 ml PBST (PBS plus 0.10%) (v/v) Tween-20, sterile filtered through 0.20 μm filter) at room temperature with 12 channel multichannel pipette (TiterTek) using ICN FLOW, Inc. pipette tips. Each well was blocked with 0.20 ml freshly prepared PBSTB (PBST plus 1.0 mg/ml BSA (immunological grade) sterile filtered through 0.20 μm filter) for one hour at 37°C in an incubator, with the microtiter plates covered with cover sheets. To apply samples, a sample template was prepared for the wells on each 96-well plate. 0.10 ml of sample was applied in each of three wells and incubated overnight at 37°C in an incubator, covered with the cover sheet.

The wells were then washed five times as described above, and 0.10 ml turkey polyclonal IgG antibody against human TGFβl (CRI cat. no. 40091) in PBSTB at 5.0 μg/ml was applied to each well. Wells, covered with the cover sheet, were incubated for two hours at 37°C in an incubator.

The wells were then washed five times as described above. Goat polyclonal IgG against turkey IgG coupled to alkaline phosphatase (CRI TGFβ kit cat. no. 001-30040.80 ug vial) in PBSTB at 0.5 μg/ml was then applied to the wells. 0.10 ml of the solution was applied to each well and the wells, covered with the cover sheet, were incubated at 37°C for one hour in the incubator.

The wells were then washed five times as described above. Substrate solution was prepared by mixing 1 ml diethanolamine buffer with 4 ml of deionized water and then add one PNPP tablet (PNPP substrate/buffer kit, Kierkegaard-Perry Labs, cat. no. 50-80-00). The substrate solution was sonicated for approximately thirty seconds until PNPP was dissolved and inverted three times to insure mixing. Enough substrate was prepared to have 15 ml substrate per 96-well plate.

0.10 ml substrate solution was applied to each well and the enzyme- substrate reaction was monitored at 405 nm in a Molecular Devices, Inc. model Vmax plate reader (using Softmax software and an appropriate computer). The Vmax for Kinetic analysis was set with 10 second data acquisition for a total of 30 minutes. The Automix and Autoprint was set to ON, with Standard Curve fit and with unknowns extrapolated by the

Quadratic function. (Note: When pipetting the substrate into the wells, make sure that the tips touch the side of the well above the bottom so that no air bubbles are generated; the air bubbles would give false readings. When evaluating results, values above 1 ng ml can be considered two places beyond the decimal point, while lower values are significant only to the first decimal point.)

An end-point reading was taken immediately after the first report.

The reactio was quenched with 0.10 EDTA solution (5.0% wt/v disodium- EDTA in deionized water, pH adjusted to 9.4 with 10 N NaOH), and a second end-point reading was taken.

Data for hextuplicate wells for each TGFβl concentration between

0.1 ng/ml and 10.0 ng/ml is shown in Table 1. A quadratic regression curve for this data is depicted in Figure 1.

TABLE 1

Standard (ng/ml) mOD/min (mean) Standard Deviation

0.000 1.014 0.041

0.100 2.155 0.106

0.200 2.626 0.134

0.300 3.818 0.226

0.400 4.955 0.340

0.500 4.946 0.353

0.750 7.599 0.313

1.000 8.695 0.446

2.000 13.51 0.49

2.500 15.31 1.71

3.000 19.27 1.15

4.000 20.92 1.22

5.000 21.23 1.87

6.000 26.95 1.29

7.500 29.03 1.77

10.00 32.09 1.41 The data presented in Table 1 and Figure 1 demonstrate the measurements with a TGFβl standard using a SELISA are accurate and sensitive to 0.1 ng/ml, in the presence of SDS and Tween. The correlation coefficient for this standard curve was 0.994. Assays run on the TGFβl standard preparations remained precise and sensitive over more than two months and thirty separate assays.

Example II: Comparison of Results Obtained from Acidification/Neutralization and SDS/TWEEN Treatment of Thrombin-Induced Platelet Releasate.

To compare results obtained from a TGFβl immunoassay of human TGFβl decomplexed by acidification/neutralization versus TGFβl decomplexed by the method of this invention, Thrombin induced platelet releasate (TPR) was assayed for TGFβl following each of these two treatments.

TPR was prepared as follows. After separation from human blood using well-known procedures, platelets were washed twice in HEPES buffered saline. The platelets were then counted and the concentration of platelets was adjusted to 1 x 10^ platelets/ml. The platelets were then activated with one unit of Thrombin per lO^/platelets at room temperature for ten minutes.

Latent TGFβ in TPR was decomplexed by acidification/neutralization as follows. The pH of TPR samples was adjusted to 2.5 with IN HC1. After a one hour incubation at room temperature, the samples were neutralized with 5 N NaOH. The acidification/neutralization-activated samples were then assayed using the sandwich enzyme-linked immunosorbent assay as described in Example I. Latent TGFβl in TPR was decomplexed in TPR with SDS/Tween treatment as follows. The sample was adjusted to 0.1% SDS (1: 100 dilution of freshly prepared 10% (wt/v) stock solution and incubated for one hour at room temperature. At the end of the incubation, the sample is diluted 1 : 100 into a buffer containing 0.10 mg/ml human serum albumin (NYBC HSA 25%), 0.05% (wt/v) SDS, and 0.8% Tween-20 (v/v). Four sandwich enzyme-linked immunosorbent assays for TGFβl as described in Example I were run on 5 samples of acidification/neutralization-activated TPR (10⁹ platelets/ml). A sample of SDS/Tween-activated TPR (10⁹ platelets/ml) was assayed eighty-four times in hextuplicate.

The results obtained were as follows:

TABLE 2 Method of _TGFβl (ng/ml) Standard Activation Average Deviation

ACIDIFICATION/

NEUTRALIZATION 111 14

SDS/TWEEN 220 12

The data in Table 2 show the superior results obtained from the SDS/Tween method of decomplexing a polypeptide for specific binding assay over the acidification/neutralization method described in the TGFβ literature. The results confirm what is suggested by purification experiments and semiquantitative measurements: that there is much more TGFβl in platelets than is indicated by immunoassays run on acidified/neutralized samples. The data in Table 2 demonstrates that the anionic/nonionic surfactant treatment of the present invention results in the detection of about 100% more TGFβ than is detected in the prior art acidification/neutralization method. The data in Table 2 demonstrate that the surfactant decomplexation method of the present invention produces more accurate results than techniques employed heretofore.

Example III: Demonstration of Rapid, Time- Dependent Loss of Immunoreactivity of TGFβ 1 in

Acidified/Neutralized Samples, and of Relative Time- Independence of TGFβl Activity in SDS/Tween- Activated Samples To demonstrate the rapid loss of immunoreactive TGFβl in an acidified/neutralized sample over time, and the retention of decomplexed, immunoreactive TGFβl over time in samples treated with SDS/Tween, the following experiment was performed. TGFβl in TPR samples was decomplexed by acidification/neutralization and SDS/Tween as described in Example II. In this experiment, some acidified samples were boiled for five minutes just prior to neutralization.

The sandwich enzyme-linked immunosorbent assay as described in Example I was performed at 2, 6 and 24 hours on TPR samples in which TGFβl was decomplexed by SDS/Tween, acidification/neutralization, or acidification heat/neutralization and incubated at 37° C for 2, 6 or 24 hours.

The results of this experiment are shown in Figure 2. The acidified/neutralized samples showed initial values for TGFβ in nanograms/ml lower than SDS/Tween-treated samples, followed by a rapid, precipitous loss of immunoreactivity. The addition of a heating step reduced, but did not eliminate, the loss of immunoreactivity of TGFβl in the samples.

There was no loss of immunoreactivity in the SDS/Tween-treated samples over the 24 hour period. The data from this experiment, coupled with the data of Example II, demonstrate that the SDS/Tween decomplexing method of this invention not only produces more accurate results by completely separating the active form of TGFβ from the latent peptide, but also provides a stable sample for assay wherein the TGFβl does not lose specific binding reactivity over time.

Example IV: Preparation of a Standard Curve Using Sandwich Enzyme-Linked Immunosorbent Assay of Pure Human Platelet Factor 4

A standard curve for a human platelet factor 4 assay was generated using the same protocol which is described above in Example I for a standard curve for TGFβ, with the following modifications:

Platelet factor 4 samples were prepared for assay as follows. Pure platelet factor 4 (Celsus lOOμg/vial) was dissolved in 2.0 ml of 0.05M HEPES/0.1M NaCl/0.004M Kcl/0.54% glucose buffer, pH 7.5, to a concentration of 50μg/ml. An aliquot was diluted in the same buffer 5-fold to 10 μg/ml. SDS is added to the sample from a 10% (w/v) stock solution to a final concentration of 0.13% and incubated at 40°C for one hour. The solution is centrifuged at 12000 rpm for one minute and serial dilutions were made to prepare samples of platelet factor 4 at 0.000, 1.000, 2.000, 3.000, 4.000, 5.000, 6.000, 7.000, 8.000, 9.000, and 10.00 ng/ml. 0.1 ml of each sample was applied to each of three wells in a 96-well plate (Corning microtiter, cat. no. 24106-8) and the plates were incubated at 4° for 16 hours.

After washing of the PF4 coated wells, polyclonal rabbit anti-platelet factor 4 antibody was applied to the wells at 5 μg/ml and incubated for 2 hours at 37°C.

After washing, 1 mg/ml of goat-anti-rabbit polyclonal alkaline phosphatase (Jackson Laboratories) conjugated IgG was added to the wells to a final dilution of 1 : 1000 and incubated for 1 hour at 37°C.

The washes and antibody binding conditions were identical to those described above in Example I. The washes and antibody binding, therefore, occurred in the presence of Tween after decomplexing of the platelet factor 4 with SDS prior to coating of the plates with the platelet factor 4 through the use of PBST and PBSTB.

The alkaline phosphatase assay was performed as described above in example I. The steps of blocking, washing and addition of antibodies were identical to those described in Example I. Data for quadruplicate wells for each PF4 concentration between

0.00 ng/ml and 10.00 ng/ml is shown in Table 3. A quadratic regression curve for this data is depicted in Figure 3.

TABLE 3

Standard PF4 (ng/ml) mOD/min (mean) Standard Deviation

0.000 0.775 0.195

1.000 2.554 0.193

2.000 4.603 0.321

3.000 6.272 0.202

4.000 8.586 0.232

5.000 9.195 0.248

6.000 9.318 0.231

7.000 10.93 0.305

8.000 12.11 0.452

9.000 12.48 1.424

10.10 14.17 0.822

The correlation coefficient for this standard curve was 0.993.

The data in Table 3 depicted in Figure 3 demonstrate that measurements with platelet factor 4 standards using a SELISA are accurate and sensitive to 0.1 ng/ml in the presence of SDS and Tween.

Example V: Immunoassays of Thrombin-Induced Platelet Releasate for Platelet Factor 4

Thrombin induced platelet releasate (TPR) was assayed for PF4 after decomplexing by the method of this invention as follows.

TPR was prepared as described above in Example II. PF4 in TPR was decomplexed and samples prepared as described above in Example IV. The platelet factor 4 in TPR was diluted either 1: 1000, 1:2000, or 1:4000 based on an expected concentration of platelet factor 4 of 10-20 μg/ml per 10⁹ platelets.

Enzyme-linked immunosorbent assays for PF4 as described above in Example IV were run on 4 samples of SDS/Tween treated TPR.

The results obtained were as follows: TABLE 4 Sample of TPR Platelet Factor 4/ 10⁹ Platelets

1 6.1 μg/ml

2 6.9 μg/ml

3 6.2 μg/ml

4 5.2 μg/ml

Standard deviations for the TPR samples assayed were approximately 6%.

The results obtained in the enzyme linked immunosorbent assay of TPR for the concentration of platelet factor 4 per billion platelets are in agreement with reports in the literature of approximately 12 ± 5 μg/10⁹ platelets in totally disrupted platelet supernatants. See B. Rucinski et al. Blood 61_, p. 1072 (1983). This result demonstrates the accuracy and reliability of the method of decomplexing of the present invention as applied to biological samples such as TPR (which is a relatively milder treatment than total platelet disruption) for the specific binding assay of complexed polypeptides in a biological sample.

Claims

We claim:

1. A method of improving the specific binding reactivity of a polypeptide which is capable of polypeptide complexing, said method comprising the steps of:

(a) contacting the polypeptide with an anionic surfactant capable of decomplexing such polypeptide under suitable conditions resulting in at least some decomplexing of the polypeptide; and

(b) contacting the polypeptide of step (a) with a nonionic surfactant capable of complexing with the anionic surfactant under suitable conditions resulting in at least some of the decomplexed polypeptide remaining decomplexed.

2. The method of claim 1 wherein the polypeptide is selected from the group consisting of a TGFβ, platelet factor 4, an α2-macroglobulin bound protease, an ot2-macroglobulin bound growth factor, vitronectin bound plasminogen activator-inhibitor 1, and osteonectin bound PDGF.

3. The method of claim 2 wherein the polypeptide is a TGFβ.

4. The method of claim 2 wherein the polypeptide is platelet factor 4.

5. The method of claim 1 wherein the anionic surfactant is selected from the group consisting of sodium dodecyl sulfate, triethanolammonium lauryl sulfate, diethanolammomum lauryl sulfate, sodium cetyl sulfate, dioctyl sodium sulfosuccinate, disodium monococoamide sulfosuccinate, sodium isostearyl-2-lactylate, sodium cetearyl sulfate, and sodium cocoyl isethionate.

6. The method of claim 5 wherein the anionic surfactant is sodium dodecyl sulfate. 7. The method of claim 1 wherein the nonionic surfactant is an organic aliphatic or aromatic ethylene oxide adduct formed by the reaction of ethylene oxide with aliphatic or aromatic hydroxy substituted compounds.

8. The method of claim 7 wherein the nonionic surfactant is Tween-20.

9. The method of claim 1 wherein the specific binding reactivity is immunoreactivity.

10. The method of claim 1 wherein the specific binding reactivity is the specific binding reactivity of a purified receptor or purified receptor fragment.

11. A method of assay of a sample containing a polypeptide, said polypeptide being capable of polypeptide complexing, said method comprising:

(a) contacting the polypeptide with an anionic surfactant capable of decomplexing such polypeptide under suitable conditions resulting in at least some decomplexing of the polypeptide;

(b) contacting the polypeptide resulting from step (a) with a nonionic surfactant capable of complexing with the anionic surfactant under suitable conditions resulting in at least some of the decomplexed polypeptide remaining decomplexed;

(c) contacting the polypeptide of step (b) with a specific binding partner under conditions suitable for binding of the polypeptide of step (b) and the specific binding partner where the specific binding partner is selected from the group consisting of an antibody, an antibody fragment, a purified receptor, a purified receptor fragment, and a synthetic peptide; and

(d) measuring the extent of binding between the specific binding partner and the polypeptide . 12. The method of claim 11 wherein the method further comprises:

(a) correlating the extent of binding between the specific binding partner and the polypeptide with the presence or amount of the polypeptide in the sample.

13. The method of claim 11 wherein the polypeptide is selected from the group consisting of a TGFβ, platelet factor 4, an α^-macroglobulin bound protease, α2-macroglobulin bound growth factor, vitronectin bound plasminogen activator-inhibitor 1, and osteonectin bound PDGF.

14. The method of claim 13 wherein the polypeptide is a TGFβ.

15. The method of claim 13 wherein the polypeptide is platelet factor 4.

16. The method of claim 11 wherein the anionic surfactant is selected from the group consisting of sodium dodecyl sulfate, triethanolammonium lauryl sulfate, diethanolammonium lauryl sulfate, sodium cetyl sulfate, dioctyl sodium sulfosuccinate, disodium monococoamide sulfosuccinate, sodium isostearyl-2-lactylate, sodium cetearyl sulfate, and sodium cocoyl isethionate.

17. The method of claim 16 wherein the anionic surfactant is sodium dodecyl sulfate.

18. The method of claim 11 wherein the nonionic surfactant is an organic aliphatic or aromatic ethylene oxide adduct formed by the reaction of ethylene oxide with aliphatic or aromatic hydroxy substituted compounds.

19. The method of claim 18 wherein the nonionic surfactant is Tween- 20. 20. The method of claim 11 wherein the binding between the specific binding partner and the polypeptide is immunoreactive binding.

21. The method of claim 11 wherein the binding between the specific binding partner and the polypeptide is the specific binding reactivity of a purified receptor or a purified receptor fragment.