Field of invention:

Immuno assay methods which specifically measures the amount of biological active GLP-1 . The invention includes the following methods:

a) a sandwich immuno assay employing two different antibodies, directed against two different epitopes: one antibody directed against the C-terminai region and one antibody directed against the N-terminal region. The C-terminal antibody must be specific i.e must cross-react less than 20% (preferably less than 5%) with both MPGF and GLP-K7-32) and the N-terminal antibody must cross-react less than 20% (preferably less than 5 %) with GLP-K9-37);

b) a sandwich immuno assay employing two different antibodies, directed against two different epitopes: one antibody directed against the C-terminal region and one antibody directed against the N-terminal region. The N-terminal antibody must be specific i.e. must cross-react less than 20% (preferably less than 5%) with both MPGF and GLP-K9-37) and the C-terminal antibody must cross-react less than 20% (preferably less than 5 %) with GLP-K7-32),

or

c) MPGF and/or other fragments of MPGF lacking the biological

activity of GLP-1 can be removed before assay by an immuno assay. This can be done in several ways. MPGF can be removed by immunoabsorption with antibodies directed to an epitope in MPGF which is not present in biological active GLP-1 . MPGF can be removed by chromatographic e.g. by fractionation of a plasma sample on HPLC.

The antibody combination in the sandwich immuno assay must be specific for biological active GLP-1 or the inactive GLP-1 immunoreac- tivity must be removed before assay.

Background of the invention

Glucagon - like peptide - 1 , GLP - 1 , is a recently discovered gut hor¬ mone [1 ,2]. Two biologically active forms of GLP - 1 are formed by post -translational processing of the precursor peptide pro - glucagon: GLP - 1 (7 - 36)amide and GLP - 1 (7 - 37), corresponding to proglucagon- (78 - 107)arπide and proglucagon(78 - 108), respectively [3]. Both peptides are present in plasma. However, in humans GLP - 1 (7 - 36)- amide is by far the most abundant form [3]. The amino acid sequence of GLP - 1 (7-37) is [4]:

HAEGTFTSDVSSYLEGQAAKEFIAWLVKGRG

GLP - 1 (7 - 36)amide is secreted from the the L-cells in the distal ileum as a response to oral intake of e.g. carbohydrates [1 ,2,5]. The peptide has several biological functions, with its main target in the islets of Langerhans in the endocrine pancreas. In β - cells, it enhances the glucose -stimulated insulin secretion [1 ,2] and the biosynthesis of insulin [6]. In D - cells, it enhances the secretion of somatostatin [1 ,2] and in a — cells, glucagon secretion is inhibited, either directly or through an intraislet pathway involving somatostatin [1 ,2]. In the gastro - intestinal

tract it inhibits gastric emptying and secretion [7].

Pro-glucagon is expressed primarily in two tissues, the L-cells as described above and in the σ-cells in the islets of Langerhans in the endocrine pancreas. In the σ-cells pro-glucagon is processed differently than in the intestine, fragments containing GLP-1 formed in the σ-cells are: proglucagon(72-158) and proglucagon(72-107). Neither of these two fragments have any known biological activity [1].

Deletion studies have shown that the biologically active forms of GLP- 1 are GLP-K7-37), GLP- 1(7-36)amide, GLP-K7-35) and GLP-1- (7-34) [10]. If the N-terminal histidine in position 7 or the lysine in position 34 are removed, the biological activity decreases by at least three orders of magnitude [10].

GLP -1(7 -37) and GLP-1(7-36)amide have been shown to be degraded to GLP -1(9 -37) and GLP-1(9-36)amide, respectively, in vitro, when incubated in human plasma by the enzyme Dipeptidyl Peptidase IV (11,12). The half-life for this conversion was 19.9 ±6.6 and 20.4 ±1.4 minutes for GLP-1 (7-37) and GLP-1 (7-36)amide, respectively.

Recently, several studies have shown exogenous GLP- 1(7-36)amide to exhibit a potent blood glucose lowering effect in patients with non insulin dependent diabetes mellitus (NIDDM). In these studies GLP-K7- 36)amide was infused intravenously to give plasma concentrations of

100-200 pmol/L. This corresponds to 3-5 times the increase in plasma concentration of GLP- 1(7-36)amide seen following a meal [5,8,9]. These results hold promise for GLP-1 as future therapeutic drug in the management of NIDDM.

Known GLP-1 assays

Several assays to measure GLP-1 in plasma samples have been described, the assays can primarily be divided into 4 groups:

1. Radioimmunoassay employing one C-terminal directed antibody

[13,14]

2. Radioimmunoassay employing one mid-region directed antibody

[15, 16,17]

3. Radioimmuoassay employing one N-terminal directed antibody

[13, 17]

4. Sandwich ELISA employing two different antibodies, directed against two different epitopes: an antibody directed against the C-terminal region and an antibody directed against the N-terminal region. [18,19]

These assays have several drawbacks:

1 and 2 have severe cross-reactions with GLP-1 (9-36)amide/(9-37), which are found in plasma in up to 10 times the concentration of GLP- 1 (7-36)amide/(7-37).

2 and 3 have severe cross-reactions with C-terminally truncated pepti- des, which might be present in plasma after exogenous administration of GLP-1 [19].

1 , 2, 3 and 4 have severe cross-reactions with GLP-1 (1-36)amide (proglucagon(72-107)amide) as none to our knowledge yet have raised N-terminally directed antibodies which have no cross-reaction with GLP-

1 (1-36)amide.

2, 3 and 4 have severe cross-reactions with MPGF (proglucagon(72- 158)) as none to our knowledge yet has raised N-terminally directed antibodies which have no cross-reaction with GLP-1 (1 -36)amide or has successfully used a C-terminal antibody with no cross-reactions with C- terminally extended GLP-1 in a sandwich ELISA.

Summary of the invention

The object of the present invention is to provide a method of quantifying GLP-1 in biological fluids. Thus the method can be used in search for variability of pharmacokinetic parameters. Determining this variability during the pre-clinical and clinical development of new drugs for regi¬ stration purposes is a requirement. It also contributes to an individual optimization of the treatment during clinical research and after introduc¬ tion into the market. This strategy would clearly result in an improve¬ ment of the benefit-risk ratio.

MPGF and inactive fragments of GLP-1 (e.g. GLP-1 (9-36)amide) in plasma samples is a major cause of erroneous measurements of GLP-1 . The "perfect" assay for biological active GLP-1 would have no cross- reaction (less than 20% (preferably less than 5%)) with MPGF, GLP-1 - (9-37)/(9-36)amide and GLP-K7-32). By the present invention, either the antibody specificity or the pretreatment of the samples will provide a method of specifically quantifying biological active GLP-1 . Thus these methods will be superior to other known assays. Particularly as shown in figure 1 and 2 the problems with the present assays make them unfit for use in pharmacokinetic studies. Pretreatment of the samples to remove

MPGF can be done in several ways e.g. by immunoabsorption, chromatographic.

The invention will now be described by way of examples thereof:

EXAMPLE

Figure 1 and 2 shows a clinical study where 7 patients with Non Insulin

Dependent Diabetes Mellitus (NIDDM) received mixed meals at time 0, from time 20-50 minutes they received either an infusion of GLP-K7- 36)amide 5 ng/(kg-min) or saline, respectively.

Figure 1 shows the plasma concentrations measured by a radioimmuno¬ assay employing one C-terminal directed antibody [14]. The antibody used has been shown to cross-react (> 10%) with GLP-1 (1 -36)amide, GLP-1 (8-36)amide, GLP-1 (9-36)amide and less than < 1 % GLP-K7-37), GLP-K7-35), GLP-K7-34) and GLP-K7-33), and < < 0.1 % with Glucagon.

Figure 2 shows the plasma concentrations measured by a sandwich ELISA [19],the assay employs a monoclonal mouse antibody directed against the (26-33) region and a polyclonal rabbit antibody directed against the (7-14) region as catching and detecting antibody, respecti¬ vely. This antibody combination used has been shown to cross-react 0 10%) with MPGF, GLP-1 (1 -36)amide, GLP-K7-35) and GLP-K7-34), only to a minor extent ( < 10%) with GLP-K7-33), less than < 1 % with GLP-1 (8-36)amide and GLP-1 (9-36)amide and < < 0.1 % with GLP-K7- 32), GLP-K7-31 ), GLP-1 (10-36)amide, GLP-K1 1 -36)amide and

Glucagon.

When comparing these two figures two things can be deducted:

- the incremental area under the plasma concentration curve corresponding to the infusion measured by the RIA was 2-3 times higher compared to the ELISA. This might be due to cross-

reactions with a metabolite, probably GLP-1 (9-36)amide, in the RIA.

the increase in plasma concentrations following a meal measured by the ELISA was 4 times higher compared to the RIA. This might be due to cross-reactions with MPGF in the ELISA. We propose that MPGF might be co-secreted with glucagon in response to the meal.

MPGF and inactive fragments of GLP-1 (e.g. GLP-1 (9-36)amide) in plasma samples is a major cause of erroneous measurements of GLP-1. The assay according to the invention for biological active GLP-1 would have no cross-reaction (less than 20% (preferably less than 5%)) with MPGF, GLP-1 (9-37)/(9-36)amide and GLP-K7-32).

We have developed three strategies to overcome this problem:

1. An sandwich immuno assay employing two different antibodies, directed against two different epitopes: one antibody directed against the C-terminal region and one antibody directed against the N-terminal region. The C-terminal antibody must be specific i.e must cross-react less than 20% (preferably less than 5%) with both MPGF and GLP-K7-32) and the N-terminal antibody must cross-react less than 20% (preferably less than 5%) with GLP-K9-37).

or

2. An sandwich immuno assay employing two different antibodies, directed against two different epitopes: one antibody directed against the C-terminal region and one antibody directed against the N-terminal region. The N-terminal antibody must be specific

i.e. must cross-react less than 20% (preferably less than 5%) with both MPGF and GLP-K9-37) and the C-terminal antibody must cross-react less than 20% (preferably less than 5%) with GLP-K7-32).

or

3. MPGF and/or other biological inactive fragments of MPGF can be removed before assay by an immuno assay. This can be done in several ways.

MPGF can be removed by immunoabsorption with antibodies directed to an epitope in MPGF which is not present in biological active GLP-1 . Subsequently measurement of the amount of biological active GLP-1 can be done by using a sandwich immuno assay method employing two different antibodies, directed against two different epitopes: an antibody directed against the C-terminal region and an antibody directed against the N-terminal region. The N-terminal antibody must cross-react less than 20% (preferably less than 5%) with GLP-K9-37) and the C-terminal antibody must cross-react less than 20% (preferably less than 5%) with GLP-K7-32).

MPGF and/or other fragments of MPGF lacking the GLP-1 biologi- cal activity can be removed by chromatographic e.g. by fractionation of a plasma sample on HPLC. Subsequently mea¬ surement of the amount of biological active GLP-1 can be done by using an immuno assay employing at least one antibody directed against an epitope in GLP-K7-37) or GLP-1 (7-36)amide.

References

[1 ] C. ørskov, Diabetologica. 35, 701 -71 1 (1992).

[2] H.-C- Fehmann and J.F. Habener, TEM. 3, 158-163 (1992).

[3] C. ørskov, A. Wettergren and J.J. Hoist, Diabetes. 42, 658-661 (1993).

[4] S. Mojsov, Int. J. Peptide Protein Res. 40, 333-343 (1992).

[5] R.M. Elliott, L.M. Morgan, J.A. Tredger, S. Deacon, J. Wright and V. Marks, Journal of Endocrinology. 138, 159-166 (1993).

[6] H.-C. Fehmann J.F. and Habener, Endocrinology. 130, 159-166

(1992).

[7] A. Wettergren, B. Schjoldager, P.E. Mortensen, J. Myhre and J.J.

Hoist, Digestive Diseases and Sciences. 38, 665-673 (1993).

[8] M.A. Nauck, N. Kleine, C. ørskov, J.J. Hoist, B. Willms, and W.

Creutzfeldt, Diabetologia. 36, 741-744 (1993).

[9] M.D. Gutniak, C. ørskov, J.J. Hoist, B. Ahren, and S. Efendic, The New England Journal of Medicine. 326, 1316-1322 ( 1992) .

[10] D. Gefel, G.K. Hendrick, S. Mojsov, J. Habener, and G. Weir, Endocrinology. 126, 2164-2168 (1990).

[1 1 ] D.I. Buckley and P. Lundquist, Regulatory Peptides. 40(2)

(1993).

[12] C. Deacon, A.H. Johnsen, J.J. and Hoist. Journal of Clinical Endocrinology and Metabolism. 1995. 80: 952-57.

[13] H. Takahashi, H. Manaka, T. Suda, N. Fukase, M. Tominaga, H. Sasaki, K. Kawai and S. Ohashi. Biomedical Research. 11 (2), 99-

108. (1990)

[14] C. ørskov, L. Rabenhøj, A. Wettergren, H. Kofod and J.J. Hoist. Diabetes. 43, 535-539. 1994.

[15] N. Yanaihara, T. Mochizuki, M. Hoshino, M. Suzuki, T. Nagashi- ma, J. Ishikawa, K. sato, N. Takatsuka and C. Yanaihara. Gastrointestinal Endocrinology: Receptors and Post-Receptor Mechanisms. Sec. Galveston symp. 359-370. 1990.

[16] H. Takahashi, H. Manaka, T. Suda, N. Fukase, K. Takahashi, M. Tominaga and H. Sasaki. Biomedical Research. 9(Λ ), 90 abs P- 12. (1988).

[17] C.F. Deacon, M. A. Nauck, M.T. Nielsen, L. Pridal, B. Helms and

J.J. Hoist. Submitted to Diabetes. 1995.

[18] L.O. Uttenthal, M. Ghiglione, J. Van Delft, O.G. Hermida, T. Fontela and C. Koch. Digestion. 54(6). 1993.

[19] L. Pridal, S.H. Ingwersen, F.S. Larsen, J.J. Hoist, K. Adelhorst and O. Kirk. Journal of Pharmaceutical and Biomedical Analysis. In press. 1995.