IL97625A - Method for the measurement of glycosylated protein level in a body fluid - Google Patents

Description

Ί-ayvy omn^nn πο~ι -mai? η 1 A ^υη η>κ^>τι >^Α Method for the measurement of glycosylated protein level in a body fluid YISSUM RESEARCH DEVELOPMENT COMPANY i?w npnnn nm>3i> man ΟΊΪΡ» OF THE HEBREW UNIVERSITY OF JERUSALEM 0>¾n'2 n>niyn DO » Oil > _t 1 Nil and HADASSAH MEDICAL ORGANIZATION Π07Π ji'O'K'TiD im-mon-i The inventors: Jack GROSS ΟΊ Α p*A Amirav GORDON Erol GERASI >σΤΑ i -)* C: 74009 FIELD OF INVENTION The present invention is in the field of glucose blood level measurements in general and concerns an _in vitro method for measuring the level of glycosylated proteins in body fluid samples in particular.

In the context of the following description and claims "body fluid sample" denotes any of whole blood, blood plasma, urine, lymphatic fluid or fluid fractions thereof containing the proteins present in the native fluid.

LIST OF PRIOR ART REFERENCES The following list of references is believed to be the most relevant prior art to the subject of the present invention: 1) "" Brownlee M., Vlassara H. and Cerami A. (1983). Non-enzymic glycosylation reduces the susceptibility of fibrin to degradation by plasmin. Diabetes 32:680-684. 2) Dolhofer R. and Wieland O.H. (1981). Improvement of the thiobarbituric acid assay for serum glycosylprotein determination. Clinica Chimica Acta, 112:197-204. 3) Garlick R.L. and Mazer J.S. (1983). The principal site of nonenzymatic glycosylation of human serum albumin _in vivo. J. Biol. Chem. 258:6142-6146. 4) Hodge J.E. (1955). The amadori rearrangement. Advances of Carbohydrate Chemistry 10:169-205. 5) Kennedy L., Mehl T.D., Riley W.J. and Merimee T.J. (1981). Nonenzymatical ly glycosylated proteins in diabetes mellitus: An index of short term glycemia. Diabetologia 21:94-98. 6) Kennedy L. and Baynes J.W. (1984). Non-enzymatic glyco- sylation and the chronic complications of diabetes: An over- view. Diabetologia 26:93-98. 7) Paisey R., Hopton M., and Hartog M. (1984). Correlation between skin glycosylation and glycemic control in human diabetes. Clinical Endocrinology 20:521-525. 8) Rodbard D. and Hutt D.M. (1974).. Statistical analysis of radioimmunoassays and immunoradiomet ic (labelled antibody) assays: a generalized, weighted, iterative, least-square method for logistic curve fitting (IAEA-SM-177/208). In Radioimmunoassay and related procedures in medicine. International Atomic Energy Agency, Vienna. pl89-194. 9) Shaklai N., Gar lick R.L. and Bunn H/F. (1984). Nonenzymatic glycosylation of human serum albumin alters its conformation and function. J. Biol. Chem. 259:3812-3817. 10) Shin Y.S., Stern C, Rucker A. and Endres W. (1984). Glycated hemoglobin and glycated albumin: Evaluation of different methods in diabetic control. J. Clin. Chem. Clin. Biochem. 22:47-51. 11) Trueb B., Holenstein C.G., Fischer, R-W.and Winterhalter .H. (1980). Nonenzymatic glycosylation of proteins, J. Biol. Chem. 255:6717-6720. 12) Vlassara H., Brownlee M. and Cerami A. (1982). Assessment of diabetic control by measurement of urinary glyco-peptides. Diabetologia 23:252-254. 13) Vlassara H., Brownlee M. and Cerami A. (1983). Excessive nonenzymatic glycosylation of peripheral and central nervous system myelin components in diabetic rats. Diabetes 32:670-674. 14) Vogt B.W., Schleicher E.D. and Wieland D.H. (1982). e-Amino-lysine-bound glucose in human tissues obtained at autopsy.

Increase in diabetes mellitus. Diabetes 31:1123-1127. 15) Witztum J.L., Mahoney E.M., Branks M.J., Fisher M., Elam R. and Steinberg D. (1982). Nonenzymatic glycosylation of low-density lipoprotein alters its biologic activity. Diabetes 31:283-291. 16) Immunological Methods, Levkowitz I. and Pernice D. editors (1979) p. 58-60. 17) Curtis K.L. and Witztum J.L. (1983). A novel method for generating region specific antibody to modified proteins. J. Clin. Invest. 72: 1427-1438.

BACKGROUND OF THE INVENTION The determination of blood glucose levels is of great importance in medical diagnosis. For example, determination of blood glucose is used in the diagnosis of diabetes, and once diabetes has been diagnosed, blood glucose levels must be determined periodically for follow-up of the treatment, i.e. to determine the efficacy of glycemic control.

In individual humans the blood glucose level is not constant and may fluctuate significantly in the course of the day. Therefore, measuring the level of glucose in a single blood sample can be very misleading and not reflect the patient's real condition. Consequently, in order to gain a clear picture of a patient's condition, it is necessary to determine the patient's blood glucose level repeatedly and it is common practice to calculate the average blood glucose from such repeated measurements carried out several times a day over a period of about a week. Such repeated measurements, in addition to being unpleasant to the patient, are very onerous on the clinical laboratory personnel. In view of these shortcomings such procedures are applied in only relatively few cases.

A technically simpler and economically more attractive method for the determination of blood glucose levels, is by measuring the level of glucose in the urine. This method is widely used, and is based on a partial correlation between blood and urine glucose levels. This correlation is, however, limited in the sense that glucose can be detected in the urine only when the blood glucose reaches high, pathological levels. Consequently urine glucose measurements are reliable only for the diagnosis of severe cases of diabetes while being unreliable in cases of light diabetes. The reliablity of such methods in many other cases such as determining the efficacy of glycemic control, is also limited.

- Due to these limitations it is difficult to obtain a clear picture of an ambulatory patient's average blood glucose level. In many cases false negative results are obtained whereby patients having diabetes are classified as healthy and/or the determination of the efficacy of glycemic control is incorrect. In many cases these shortcomings lead to an inappropriate treatment, and this in turn may cause an increased morbidity among diabetes patients.

It is generally the object of the present invention to overcome the limitations imposed by the present state of the art and provide an improved diagnostic method for ambulatory and clinical application, by which the average blood glucose level may be inferred reliably from a single body fluid sample. Before explaining how this object is achieved the relevant prior art to the present invention will now briefly be discussed.

PRIOR ART It is known that proteins undergo a nonenzymatic glyco-sylation in the presence of glucose, in vitro (ref. 4) as well as in vivo (refs. 5, 6). This glycosylation may occur at any free amino group of the protein but occurs predominantly on lysine residues. This glycosylation of lysine residues is a two-step process, in the first of which-a labile Schiff base with a free amine group of the lysine is formed, and in the second step such labile Schiff base undergoes a stabilizing rearrangement, known as the "Amadori reaction", to form a 1-deoxy-fructosyllysine moiety (hereinafter "DFl"). The DF1 form is in equilibrium with the hemiketal ring form of this molecule. In vivo, most of the glycosylated proteins are in the form of DFl or its hemiketal and the labile Schiff base constitute only about 25%.

It has been shown that a correlation exists between the level of glucose in the blood and the level of glycosylated proteins. For example, the level of glycosylated proteins in the blood serum of diabetics was found to be more than double that in normal subjects and when glycemic control was improved a reduction in the level of blood serum glycosylated proteins occurred (ref. 5). Hemoglobin, an intracellular protein of the red blood cells, is known to become glycosylated in the presence of glucose and is used as a tool for the follow up of glycemic control. Several commercial methods are available for the determination of the degree of its glycosylation. They are all based on a chemico-physical separation of the glycosylated Hgb-Alc subunit and its colorimetric determination. However, since the life span of a red blood cell which contains the Hgb molecules is about 120 days, the glycosylation level of Hgb reflects in the average blood glucose levels over 120 days and therefore the above method is of limited value, for the determination of the efficacy of glycemic control (ref. 10). It was indeed shown that if the blood glucose level decreases in consequence of treatment, then although there is a decrease in the overall level of glycosylated proteins within several days, the decrease in glycosylated Hgb is negligible (ref. 5).

The life span of human serum albumin (HSA) in the blood is about 20 days, and thus its degree of glycosylation will closely reflect the mean blood glucose level over the recent several days preceding the test (ref. 5). Since glycosylated HSA constitutes about 90% of the blood's glycosylated proteins, measuring the total level of glycosylated proteins in the blood will be a reliable indication of a recent blood glucose level. In addition also the urinary level of glycosylated proteins is highly correlated with the mean blood glucose level during 8 to 9 days preceding the test (ref. 12).

It follows from the above that the level of glycosylated proteins in body fluids may serve as a very good indicator to the average blood glucose level over a short period of time preceding the test. Thus a reliable method by which glycosylated protein level is measured from a single blood sample would be very useful for screening the population for diabetics as well as for determining the efficacy of glycemic control. However, prior to the present invention no such method had been in clinical use. A certain cumbersome immunological method for the measurement of glycosylated protein was proposed in the literature (ref.17) but, to the best of the inventors' knowledge was never reduced to practice which may possibly be due to the complexity of that method.

It is the obj ect of the present invention to provide a relatively easy to perform immunological method by which the level of glycosylated proteins is determined in a body fluid sample.

GENERAL DESCRIPTION OF THE INVENTION An immunological method for measuring the level of glycosylated proteins has the advantage of being both sensitive and relatively easy to perform, but the main obstacle for such a method is, that DF1 is a very poor immunogen. However, this moiety may be reduced with an appropriate reducing agent such as sodium borohydrtde (NaBH^ to a glucitol lysine (hereinafter "Gl") moiety which is a strong immunogen. The present invention is based on the realisation that this may be exploited to advantage for making available an immunological method for the determination of glycosylated proteins , and thus, broadly speaking the method according to the present invention involves reducing the DF1 moieties to Gl moieties, and then detecting the amount of Gl moieties on the basis of antibodies specific thereto. However, one big hurdle had yet to be overcome for rendering such a method practical , namely that during reduction the medium containing the glycosylated proteins must be free of unbound glucose since the reducing agent also induces the binding of free glucose to free amino groups of the proteins which if it were to happen would give false results. Removing small molecules such as glucose from a medium containing macromolecules such as proteins is usually accomplished in the art by dialysis, which is however much too complicated and time consuming for a routine laboratory test. 3 It was found in accordance with the present invention that the need for dialysis may be avoided by adsorbing the glycosylated proteins on a solid carrier before the reduction stage, washing off the glucose with an aqueous solution and then subjecting the glucose-free adso bate to reduction.

Thus, the invention provides a method for measuring the level of glycosylated proteins in a body fluid sample (as hereinbefore defined) comprising: (i) Obtaining from a subject a body fluid sample (as herein defined) ; ( ii ) incubating said sample or an aqueous dilution thereof with a solid carrier capable of adsorbing the proteins contained therein ; (iii) washing the resulting adsorbate withr an aqueous solution to remove unadsorbed proteins and free glucose; (iv) incubating said adsorbate with an aqueous solution of a reducing agent for a sufficient time for any DF1 moieties of glycosylated proteins to be reduced to Gl moieties; (v) washing the so-reduced adsorbate with an aqueous solution to remove the reducing agent; (vi) incubating the reduced adsorbate with an aqueous solution of antibodies specific for Gl moieties for a time sufficient for binding of said antibodies to the Gl moieties; (vii ) washing the so-treated adsorbate with an aqueous solution to remove excess antibodies; and (viii) measuring the amount of bound antibodies.

The measurement of the amount of bound antibodies may be effected by any suitable known method, and from such measurement the amount of DF1 moieties of glycosylated proteins may readily be deduced.

In practising the method according to the present invention it may be desirable to compare the results with those obtained with known standards. Such standards may, for example, be body fluid into which a known amount of Gl-proteins, such as Gl-albumin, has been added. Many such standards can be prepared and tested in accordance with the above method, and the results obtained with these standards may be compiled into a standards' curve with which the result obtained from an unknown sample tested in accordance with the present invention, may be compared. In addition, in order to correct interassay variations, caused for example by minor variations in reagents used at different times, the results obtained by the above method should preferably be normalised in accordance with results obtained from a control sample run in parallel. Such a control sample is preferably pooled normal plasma (hereinafter "PNP"), which is a mixed plasma from many blood donors obtained from a blood bank.

In many cases, it is desirable to store body fluid samples over prolonged periods of time. However, since body fluids contain glucose, glycosylation continues _in vitro, and it was found in accordance with the present invention, that this in vitro glycosylation continues even if the sample is frozen. If this glycosylation is not avoided, a higher content of glycosylated protein level will be obtained in a stored body fluid, which will bring about an erroneous estimation of the average blood glucose level. It was found in accordance with the present invention that this _in vitro glycosylation can be avoided by acidifying the solution before storage. Thus, in accordance with the present invention the body fluid sample is preferably acidified shortly after it is taken from a subject.

The nature of the solid carrier which may be used for adsorbing the proteins is not critical , examples being various plastic materials, glass , latex, half solid gels and various polymers. A suitable reducing agent is, for example, NaBH^ The antibodies which are used in accordance with the present invention have to be specific against reduced glycosylated proteins , i.e. should bind only to Gl moieties. An obv ious way by which one may try to obtain such antibodies is by utilising the monoclonal antibodies technique. However since an antibody against a specific antigenic determinant is sought, isolating a clone producing such an antibody may prove rather difficult.

It was found in accordance with the invention that conventionally purified polyclonal antibodies which are obtained from an animal immunized against a reduced glycosylated protein that is foreign to the human body, are as specif ic to the Gl moiety as monoclonal antibodies. Any foreign protein characterized in that the antibodies against it do not cross react with unaltered normal body fluid proteins is suitable for raising antibodies by the above method.

The antibodies which are used in accordance with this preferred embodiment of the present invention are novel and constitute one of the aspects of the present invention. Broadly they are prepared as follows: (a) A foreign protein, characterized in that the antibodies against it do not cross react with unaltered normal human body fluid proteins , such as Keyhole Limpet Hemocyanin (hereinafter "KLH"), is glycosylated; (b) the glycosylated protein obtained in (a) is reduced with a suitable reducing agent such as aBH^, to yield this said foreign protein with Gl moieties; (c) the reduced glycosylated protein obtained in (b) is injected into an animal whereby an immunogenic response is initiated by which antibodies against the reduced glycosylated foreign protein are produced; (d) the antibodies formed in (c) are isolated and, if desired, purified both by conventional methods.

The antibodies obtained in (d) are a mixture of polyclonal antibodies directed against a large variety of moieties on the protein used for the immunization. But, since the immunizing protein is foreign to the human body, none of these antibodies will bind to proteins which are present in normal body fluids. However, after the body fluid proteins are subjected to the action of a reducing agent and Gl moieties are formed thereon, this moiety will be "recognized" by part of the antibodies which will then bind thereto. Thus any unspecific binding of antibodies to body fluid proteins will be avoided, and the binding will therefore truely reflect the amount of Gl moieties.

In many cases it is not necessary to purify the Gl moiety-specific antibodies. There are, however, cases where pure anti-Gl polyclonal antibodies are required. In such cases the antibodies are purified to obtain a pure aqueous solution. Purification of the anti-Gl antibodies may be done by conventional methods, e.g. affinity purification of the crude solution of the antibodies by passing it through a column containing bound Gl-proteins such as Gl-HSA, and thereafter eluting the antibodies from the bound proteins by methods known per se.

The anti Gl antibodies may be labelled by conventional methods known per se.

In performing the method according to the invention, the amount of antibodies bound to the adsorbate may be determined either by a direct mode i.e. by using labelled anti-Gl antibodies, or by an indirect mode i.e. by reacting the bound anti-Gl antibodies with label led antibodies directed against the anti-Gl antibodies.

The antibodies may be label led by conventional methods, e.g. by radioactive substances, enzymes, fluorescent substances, and the like. In accordance with the direct mode radioactively label led anti-Gl antibodies may, for example, be used and the results may then be evaluated by the intmuno radiometric assay (IR A). In accordance with the indirect mode the label may, for example, be an enzyme and the results may then be evaluated using the ELISA method.

Preferably, the body fluid sampled in the performance of the method according to the present invention is blood. In cases where there exists a correlation betwen blood glucose levels and Gl-proteins found in the urine, a urine sample may also be used.

The degree of protein glycosylation in the blood depends on the concentration of glucose in the blood and the affinity between glucose and the proteins, which affinity is an individual proper ty that seems to vary f rom one subj ect to another. Consequently, the level of glycosylated proteins determined in accordance with the present invention reflects both the average glucose level in the blood and said affinity. In view of this affinity, different results are obtained from different subjects even though their blood glucose level may be the same. However, where the blood sugar is periodically determined in a single subject, the issue of affinity does not arise and any changes in the results truly reflect changes in the average blood glucose level.

In practising the method according to the present invention, it may be useful that all ingredients be provided in a form of a kit. Such a kit, which constitutes yet another aspect of the present invention, may comprise, for example, a solid carrier material, a reducing agent, anti-Gl antibodies either labelled for use in the direct mode or unlabel led for use in the indirect mode. In the latter case the kit will preferably also include labelled antibodies against the anti-Gl antibodies.

Preferably the kit according to the invention also includes instructions for use. Optionally the kit may also comprise a standards curve and/or table and control body fluid samples.

The method according to the invention is relatively simple to perform and does not require the use of complicated and costly instruments. The reagents needed for the purposes of the present invention are all commercially available, with the exception of the anti-Gl antibodies which have to be prepared in accordance with the invention. In view of all this, the invention makes it possible to upgrade substantially the capability of an existing laboratory to determine average glucose levels in body fluids in a reliable manner.

DESCRIPTION OF THE DRAWINGS In the following description reference will at times be made to the following drawings in which: Fig. 1 shows a graphical representation of changes in the glucose and in the Gl-HSA ("GSA") levels in a single patient over time; and Fig. 2 is a graphical representation showing the degree of glycosylation versus time after incubation with and without an acetate buffer.

DESCRIPTION OF SOME SPECIFIC EMBODIME TS In the following description some examples of specific embodiments are given in order to illustrate the present invention. However, many other variations including modifications of these examples are all possible within the general teachings of the present invention, as will be readily understood by those skilled in the art.

EXAMPLE 1: Preparation of anti-Gl antibodies KLH, which is a protein obtained from a mollusk, was used as the foreign protein. (a) Glycosylation and reduction: KLH was dissolved to a final concentration of 0.4 mg/ml/ in a phosphate buffer (hereinafter "PB") solution (O.075M at PH 7.65), which contains also glucose at a concentration of 0.080M and NaCNBH3 (cyanoborohydride, SIGMA catalogue No. S-8628, serving as a reducing agent), at a concentration of 0.2M, and the mixture was incubated for seven days at room temperature. After the incubation the solution was dialysed against PB to remove glucose and aCNBH3f and a purified Gl-KLH solution was obtained. (b) Preparation of antibodies; Four parts of the Gl-KLH solution obtained as above were mixed with six parts of complete Freund's adjuvant. 0.5 ml of the resulting suspension was injected into each of the two gluteal muscles of the same rabbit. After two weeks the rabbit was reinjected with a similar suspension prepared however with an incomplete Freund's adjuvant, and this was repeated again after an additional period of two weeks. Ten days after the last injection the animal was bled and the antibodies were separated from the whole blood by a conventional manner.

The solution of antibodies obtained was thereafter subjected to affinity purification through columns containing Gl-HSA prepared in a similar manner as described hereinbefore in connection with the preparation of Gl-KLH. The affinity purification was performed in a manner similar to that described by Levkowitz and Pernice (Ref. 16), and was performed as described hereinbelow. (c) Preparation of the columns for the affinity purification 4 ml of 20% Gl-HSA solution was dialysed against 4 volumes of 500 ml of 0.2M sodium acetate buffer at pH 5. Thereafter the protein solution was transferred into a 50 ml polyethylene tube and while mixing, 0.2 ml of a 2.5% aqueous solution of glutaraldehyde per each ml of Gl-HSA solution was added dropwise. The solution was incubated overnight in a refrigerator and this was followed by 3 hours incubation at room temperature and as a result a gel was formed. The supernatant was removed from the gel, 5 ml 0.2M PB at pH 7.2 was added and using an homogenizer the gel was then homogenized for a few seconds. After homogenization the solution was centrifuged at 5000 rpm for 10 minutes at 4°C, the supernatant was removed and the resulting pellet was resuspended in 5 ml of the same buffer. The resulting suspension was mixed with an equal volume of a hydrated sephadex G-25 fine (manufactured by Pharmacia, Sweden) also suspended in the same buffer.

Each 1 ml of the resulting suspension was transferred into an 8 ml polyethylene column pretreated with gelatin (which pretreatment consisted of incubating each column for 30 min. with a 1.5% PB (pH 7.2) solution of gelatin followed by washing with pure PB). The columns were then washed 4 times with 7 ml of PB (pH 7.2) and this was followed by an incubation for 30 minutes in an eluting buffer (0.2M of glycine-HCl at PH 2.8). Finally the columns were washed again 4 times with 7 ml of PB (pH 7.2) whereupon they were ready for use and could be stored filled with PB over long periods of time in a refrigerator.

For quality control the void volume of one of the columns was determined by using labelled Gl-HSA, which void volume ranged from 0.5 to 1.0 ml. (d) Affinity-purification of the antibodies: The antibodies obtained in (b) above were precipitated in an aqueous solution of 45% ammonium sulfate. The ammonium sulfate was removed and the antibodies were resuspended in PB to a final concentration of 20-40 mg/ml. A fraction of the suspension was removed, the. antibodies in this fraction iodinated with 125j and the antibodies thus labelled were added into the antibody suspension for recovery measurements. 100 ul of the resulting, mixed antibody solution was loaded into a column and washed-in by the addition of 100 ul of PB pH 7.2, and was then incubated for 1 hour at room temperature. After incubation the column was washed with several void volumes of PB pH 7.2 and a radioactivity count was performed on the effluent. Washing was continued until the count stabilised.

The column was loaded and washed repeatedly until it column was saturated with antibodies, which was established by determining when the radioactivity in the column reached an asym-ptot. This was then followed by washing with at least 10 more void volumes of PB pH 7.2 after which the column was ready for elution.

One void volume of the elution buffer was loaded into the column and incubated therein at room temperature for 30 minutes. The column was then eluted with the elution buffer in volumes equal to the void volume and each void volume fraction was collected in a polyethylene test-tube containing 1 part of a neutralizing solution (2M tris-HCl buffer, pH 8.6, containing 0.03. NaNH3) per 10 parts of elution fractions.

The yield of the purification was calculated from the recovered CPM and rechecked by protein determination by an OD 280nm measurement. The purified anti-Gl antibody solution was frozen in small aliquots and kept in cold storage until use.

A portion of the antibodies obtained was labelled with ^■^1. Alternatively enzymatic labelling, e.g. with horseradish peroxidase or alkaline phosphatase, or fluorescent labelling and the like, may also be performed. All these labelling methods are conventional and known per se.

EXAMPLE 2 Determination of glycosylated protein in human plasma in accordance with the invention, using the IRMA procedure: 1. A plasma sample was mixed with an equal volume of 0.2M NaAc buffer having a pH of 4.85, to yield a solution having a final pH of about 5. This solution which will hereinafter be referred to as "acidified plasma", may be stored for prolonged periods of time without the occurence of _in vitro glycosylation. Prior to testing for glycosylated protein content the acidified plasma was diluted with 500 volumes of PB buffer (0.075M, pH 7.65, containing also 0.03M of Na-azide, serving as a preservative). 2. 0.5 ml of the diluted acidified plasma was added into a NUNC STAR (trade name) tube (manufactured by Nunc, Kumstrup, Denmark) and incubated therein overnight at room temperature. This brought about the adsorption of the plasma proteins on the walls of the tubes. 3. The unadsorbed proteins and free glucose were removed from the tubes by washing three times with distilled water. 4. In order to reduce the DF1 moieties to Gl moieties, 1.5 ml of a 0.075M NaBH^ solution was added and the mixture was incubated at room temperature for 1 hour. 5. The aBt-4 was removed from the tubes by washing three times with distilled water. 6. 125χ Labelled anti-Gl antibodies prepared in accordance with Example 1 were added into the tubes and incubated therein for 48 hours at room temperature. During the incubation period the tubes were subjected to a first radioactivity count to yield a total count with which the count obtained after removal of the unbound antibody (see 7 and 8 below), was later compared. 7. Unbound antibodies were removed from the tubes by washing three times with distilled water. 8. A further radioactivity count was performed.

The results were compared to known standards, based on acidified P P containing known amounts of Gl-HSA added thereto, e.g. in concentrations of 1 to 10,000 ng/ml. These standards were established by means of a test protocol similar to the one described above with the omission, however, of the reducing steps (4) and (5) so that the DF1 moieties on the proteins present in the normal plasma pool would not be reduced to Gl and thereby distort the standard measurement. From the results, a Standard's curve and table were compiled which served as auxiliaries in the evaluation of results obtained with unknown samples.

For the correction of interassay variations in the performance of the test method according to the invention as described above, a control test was run in parallal. The control sample used for the control test was an acidified plasma obtained from a pool of normal blood bank donors, and in each test the result was normalized on the basis of the result of the control.

EXAMPLE 3 Determination of glycosylated proteins in human plasma in accordance with the invention using the ELISA procedure 1. The acidified plasma was prepared and diluted in the same manner as in clause 1 of Example 2. 2. 200 μΐ of the diluted plasma was introduced into wells of a 96 ELISA well plate and incubated for 15 minutes. 3. The free glucose and the unadsorbed proteins were removed by washing with distilled water containing 0.05% tween. 4. - In order to reduce the DF1 moieties to Gl moieties, 400 lof a 0.225M of NaBH4 was added into each well and incubated for 15 minutes. 5. The reducing agent was removed by washing with distilled water containing 0.05% tween. 6. The wells were thereafter postcoated in order to block unadsorbed site, by the addition of 400 al of a 5% bovine serum albumin (BSA) and incubation for 15 minutes. 7. The unadsorbed BSA was removed by washing with distilled water containing 0.05% tween. 8. 125 μΐ of the unbound an i-Gl antibodies prepared in accon with Example 1 and diluted 500 fold were added into each well and incubated therein for 30 minutes. 9. Unbound antibodies were removed by washing with distilled water containing 0.05% tween. 10. 200 μΐ of a diluted solution (1:1500) of anti-rabbit IgG antibodies conjugated to alkaline phosphatase was added into each well and incubated therein for 90 minutes. 11. Unbound conjugates were removed by washing with distilled water containing 0.05% tween. 12. 200 μΐ of a solution containing p-nitrophenyl phosphate, which is a substrate for the alkaline phosphotase, was added and incubated for 60 minutes for the development of a colour reaction. The colour product was fixed with 50 }il of 1M aOH. 13. The results were measured by an ELISA optical reader.

The results were compared with a Standards' curve or table prepared in a similar manner as in Example 2. In addition, similarly as in Example 2 a control test was run in parallel.

The above procedure may be modified and simplified, e.g. by combining steps 6 to 10 so that after step 5, the wells will be incubated for say 60 mins. with 125 il of a solution containing 5% BSA, 500 folds diluted anti-Gl antibodies and 500 folds diluted anti-rabbit Igb antibodies conjugated to albumin phosphates. Thereafter the procedure continues as in steps 11 to 13.

EXAMPLE 4 Tests: In the fol lowing some tests are described all of which were performed in accordance with the IRMA method of Example 2.

Similar results were obtained by the ELISA method of Example 3, as specifically demonstrated in Test No. 5.

Al l the results given in the following tables are normalised to results obtained in parallel control runs with PNP.

Test No. 1 The level of Gl-HSA level was determined in normal subjects, in normal pregnant women and in diabetic patients.

The results are summarised in the following Table 1: Table 1 Group n* Meantsd Range p** Normals 45 - 0." 71±0 .' 15 0 . 43-1 . 18 Normal pregnant women 46 ' 0. 37± 0 . 18 0. 05-0 ..77 < . 001 Diabetics 59 5. 31 ± 3 . 72 _ 1. 37- 14. 82 '< . 001 . *n - number of subjects tested. **p - probability, compared to a normal group.

From the above results the following may be concluded: (i) The method in accordance with the present invention can clearly distinguish between normals and diabetics, with no overlap between the two groups. (ii) The method in accordance with the present invention is also sensitive enough to detect pregnancy hypoglycemia.

Test No. 2 In order to test repeatability of the results obtained by the method of the present invention, samples were tested twice at an interval of several weeks.

The results are summarised in the following Table 2: Table 2 Sample identification 1st assay 2nd assay D-5 14.31 12'.27 D-13 9.63 8.09 D-50 9".4.6 9.56 D-53 10.61 10.03 D-58 3.54 3.27 The above results show that the method in accordance with the present invention produces repeatable results.

Test No. 3 The following parameters were measured independently in blood serum samples: (i) Blood glucose levels - mean blood glucose for the same week determined by the conventional enzymatic methods for blood glucose . (ii) Gl-HSA levels - normalised values obtained by the method in accordance with the present invention. (iii) Fructosamine levels - the level of fructosyl lysine moieties, measured by the Roche dye test (using a kit manufactured by Hoffmann-La Roche, Basle, Switzerland) for fructosyl-albumin levels, in which tetrazolium blue interacts with DF1 moieties to form a coloured complex.

All samples were obtained from diabetics, and the results are summarised in Table 3 below: Table 3 Variables n r* P Blood glucose vs. Gl-HSA 79 0. 421 <.001 Blood glucose vs. fructosamine 52 -0. 078 ns ** Gl-HSA vs. fructosamine 52 0. 228 ns *r - correlation coefficient. **ns - not significant The results show that in diabetics there is a statistically significant correlation between blood glucose levels and Gl-HSA levels. In the same population, the fructosamine test which is sometimes used in the art shows no correlation with blood glucose levels.

Test No. 4 The same parameters of Test 3 were determined for three single patients (NI, PI and Bu) and the above correlations were determined. The results are shown in the following Table 4: Table 4 Patient Weeks of Correlations between Correlation (Inventors' code) observation Gl-HSA and glucose between fructosamine and glucose _r R r_ NI 6 0.83 <..05 0.42 ns PI 4 0.-99- <. or 0.75 ;. ns BU 7 0.87- <. or 0.40 ns The results show again the high degree of correlation between Gl-HSA weekly measurement and average glucose levels determined throughout the week. In contrast, there exists a lack of correlation between a weekly measurement of the fructosamine levels and blood glucose.

In addition, the above results show the higher correlation of blood glucose levels to Gl-HSA levels in a single patient, than the correlation of the same parameters in average population measurements. As explained above, this is due to the variability in the affinity of proteins from different individuals to glycosylation.

It may further be concluded from the above results that a single weekly Gl-HSA determination in accordance with the present invention, is a significant indicator of average weekly glycemia levels.

The data of patient BU is plotted in Fig. 1 which also demonstrates the high degree of correlation between blood glucose levels and the level of Gl-HSA.

Test No. 5 Blood glycosylated albumin levels were determined in accordance with the present invention by both the ELISA and the IRMA methods. The results are shown in the following Table 5: Table 5 Sample # ELISA IRMA ( inventor ' s designation) 417 .1.37 1.08 418 .53 .93 419 .1.91 1.92 420 1.52 1.79 421 2.19 1.90 422 1.95 Γ..92 424 4.33 4.63 425 2.33 2.50 426 2.55 2.47 428 2.80 2.64 429 1.98 2.16 431 3.53 3.75 433 2.22 2.28 434 2.41 2.33 435 2.20 2.50 436 2.16 1.70 437 .1.83 1.63 438 1.70 Γ.62 It can be seen from the above table that very similar results are obtained by both methods.

Test No. 6 To a PNP sample an equal vol ume of either 0.2M NaAc solution having a pH of 4.84 or 0.2M PB hav ing a pH of 7.65 was added- Thereafter, 5 mg/ml of glucose was added into each sample and the samples were frozen.

At various times after the onset of incubation, samples were taken and tested by the IRMA method of Example 2. Resul ts are shown in the upper curve of Fig. 2. The level of Gl-HSA is seen to increase over time, which shows that in vitro glycosyla-tion occurs even in a frozen sample. However, no such glycosy- lation is seen for the acidified sample, i.e. the one mixed with the NaAc solution - lower curve of Fig. 2.

Claims (18)

1. A method for measuring the level of glycosylated proteins in a body fluid sample selected from the group consisting of whole blood, blood plasma, urine or fluid fraction thereof containing the proteins of the unfractioned body fluid, which comprises: (i) Obtaining from a subject a body fluid sample (as defined above ) ; ( ii) incubating said sample or an aqueous dilution thereof with a solid carrier capable of adsorbing the proteins contained therein; (iii) washing the resulting adsorbate with an aqueous solution to remove unadsorbed proteins and free glucose; (iv) incubating said adsorbate with an aqueous solution of a reducing agent for a sufficient amount of time for any DF1 moieties of glycosylated proteins to be reduced to Gl moieties; ( v) washing the so reduced adsorbate with an aqueous solution to remove the reducing agent; (vi) incubating the reduced adsorbate having the proteins adsorbed thereon with an aqueous solution of antibodies specific for Gl moieties for an amount of time sufficient for binding of said antibodies to the Gl moieties; (vii) washing the so treated adsorbate with an aqueous solution to remove excess antibodies; and (viii) measuring the amount of bound antibodies.

2. The method according to Claim 1 wherein said body fluid sample is acidified before use.

3. The method according to any one of the preceding claims wherein the reducing agent is NaBH^.

4. Antibodies specific for glucitollysine moieties on protein obtained by a) subjecting to glycosylation a foreign protein being a protein that the antibodies against it do not cross-react with unaltered normal body fluid proteins; b) reducing the so glycosylated foreign protein whereby glucitollysine moieties are formed thereon; c) injecting a solution containing the so-reduced glycosylated foreign protein into an animal whereby an immunogenic response is initiated by which antibodies against the reduced glycosylated foreign protein are produced; d) isolating said antibodies from the animal.

5. Antibodies obtained as in Claim 4 and further by purification after isolation by methods known per se.

6. Antibodies according to Claim 5 affinity purified by passing an aqueous solution of a crude antibody mixture through a column containing bound glucitollysine-proteins, and thereafter eluting the antibodies from the bound protein by methods known per se.

7. Antibodies according to any one of Claims 4 to 6 wherein said foreign protein used in the preparation is Keyhole Limpet hemocyanin.

8. Labelled antibodies obtained by labelling in a manner known per se antibodies obtained in accordance with any one of Claims 4 to 6.

9. A method according to any one of Claims 1 to 3 comprising using antibodies according to one of Claims 4 to 8.

10. A method according to any one of Claims 1 to 3 wherein the level of glucitol lysine moieties is determined directly by employing labelled anti-glucitollysine antibodies.

11. A method according to Claim 10 wherein said antibodies are radioactively labelled, and the results are determined by immunoradiometric assay.

12. A method according to any one of Claims 1 to 3 comprising determining the level of glucitollysine moieties is determined indirectly by contacting the adsorbate with anti-glucitol-lysine antibodies and then determining the amount of bound antibodies employing labelled antibodies directed against the anti-glucitollysine antibodies.

13. A method according to Claim 12 wherein enzyme labelled antibodies are used and results are obtained by the ELISA method.

14. A kit for carrying out the method of any one of Claims 1 to 3 and 9 to 13, comprising an anti-Gl antibody preparation, a solid support and a reducing agent.

15. A kit according to Claim 14 comprising also a control body fluid sample.

16. A kit according to Claim 14 and 15 comprising also a Standards' curve and/or table.

17. A kit according to any one of Claims 14 to 16 wherein the antibodies of said anti-Gl antibody preparation are labelled.

18. A kit according to any one of Claims 14 to 16 wherein the antibod ies of said anti-Gl antibody preparation are unlabel led, which kit also comprises a preparation of antibodies against the anti-Gl IC/rb/rb