The present invention relates to new peptides, to an _¬ t ⁱ bodies directed again ^β t said peptidβB and to d _i agnost _i c processes for determining the presence of mature nerve growth factors and precursors thereof.

Nerve growth factor, NGF, is a 118-amino acid prote _i n that acts as a trophic factor for many sensory and sympathe _¬ tic neurons in the peripheral nervous system and for choli- nergic neurons in the brain. Also other non-cholinergic neu _¬ rons in the central nervous ^β ystem may be sensitive to NGF. It is therefore of interest to find diagnostic means to de _¬ termine the presence and the level of NGF and precursors thereof to enable efficient treatment to alleviate different diseases of the nervous system, such as senile dementia of the Alzheimer type, Parkinson's disease, spinal cord injury, stroke, and developmental disturbances.

Accordingly, it is an object of the present invention to provide new peptides which can be efficiently used in such diagnostic procedures.

Another object of the invention is to provide antibo _¬ dies against such new peptides for use in determining NGF levels using for example enzyme or radioimmuno a _ββ ays to de _¬ termine said levels.

Yet another object of the invention is to provide new processes for determining the presence of mature NGF : s and/or precursors thereof .

For these and other purposes the present invention provides for a peptide having the formula:

Glu-Pro-Hi ^β -Ser-Glu-Ser-Asn-V l-Pro-Ala-Gly-His-Thr-Ile ,

Gly-Asp-Lys-Thr-Thr-Ala-Thr-Asp-I le-Lys-Gly-Lys-Glu

One preferred peptide has the following formula:

Glu-Pro-His-Ser-Glu-Ser-Asn-Val-Pro-Ala-Gly-His-Thr-Ilβ, said peptide corresponding to amino acid reβidues -103 to -90 in the human NGF precursor. The antibodies to this peptide

are effective in detecting the NGF precursor and can be used in immunological determination of the NGF precursor protein.

Another peptide according to the invention has the formula: Gly-Asp-Lys-Thr-Thr-Ala-Thr-Asp-Ile-Lyβ-Gly-Lys-Glu, and this peptide corresponds to the conserved region of the NGF prote¬ in corresponding to amino acid residues 23 to 35 of the matu¬ re NGF protein. Tests performed in connection with the re¬ search leading to the present invention demonstrate recogni- tion of the native NGF protein by antibodies raised to this peptide .

Relevant antibodies can be directed either against any of the peptides described above or both such peptides and native NGF: s . The antibodies may be either polyclonals or monoclonals.

The invention also provides for a process for determi¬ ning the presence of mature NGF or a precursor thereof in a biological test specimen. Said process is based on the inter¬ action between antibodies generated by immunization using a peptide according to the invention and said NGF or said pre¬ cursor. Such determination may be qualitative or quantitative depending on the type of diagnosis performed and the object of the assay.

The test specimen subject to such assay may be consti- tuted by mammalian brain tissue, such as brain tissue of hu¬ man origin.

In the present disclosure, when reference is made to living animal body or living mammal body said terms are in¬ tended to include also man. The present invention will be further illustrated be¬ low with reference to the appended drawings, wherein:

Fig. 1A shows homologous antibody binding to NGF pep¬ tides as a function of antiserum concentration,*

Fig. IB shows binding of affinity-purified peptide antibodies to their homologous peptides as a function of con¬ centration of the immunoglobulines CIg);

Fig. 1C shows antibody binding to different NGF-pepti- des shown as a function of antiserum concentration;

Fig. 2A shows peptide antibody binding to the β-NGF protein as a function of concentration of the affinity-puri¬ fied immunoglobulines ;

Fig. 2B shows comparison of binding to NGF peptide P3 in different antisera to β-NGF; and

Fig. 2C shows antibody binding to different synthetic NGF peptides by an antiserum to mouse β-NGF.

EXAMPLE 1

Synthesis of NGF Peptides

Hydrophilic areas suggested by the hydrophilicity va¬ lues of Hopp TP, Woods KR ⁽ 1981 ⁾ : Prediction of protein anti- genie determinants from amino acid sequences. Proc . Natl .Acad. Sci. (USA) 78:3824-3828 in the pro-NGF and the mature NGF were selected for peptide synthesis (Table I) . Peptides PI to P6 were synthesized by Cambridge Research Biochemicals Ltd., England, at an estimated purity of greater than 80% and checked by amino-acid analysis, HPLC and mass spectrometry. Peptides P7 and P8 were kindly synthesized by Drs. Tamas Bartfai and Janis Abens at the Arrhenius Laboratory, Depart¬ ment of Biochemistry at the University of Stockholm, Sweden. Similarities to any other known protein sequences than NGF for each of these peptides were checked in the GeneBank with¬ out any extensive homologies found except to NGF.

EXAMPLE 2

Conjugation and Production of Peptide Antisera All peptides were conjugated to keyhole limpet hemo- cyanin (KLH) before use in immunizations. LH (8-10 g) was in each case coupled with 3-10 mg of the peptide. Couplings were achieved with m-maleimidobenzoic acid N-hydroxy-succini- mide or N-succinimidyl 3- ⁽ 2-pyridyldithio)propionate to pro- duce N-terminally bound conjugates of P1-P5 and P7-P8 and a C-terminally bound conjugate of P6. After dialysis the pep-

tide-KLH conjugates were used to immunize rabbits (Table II) . Each rabbit received an initial dose of about 400 μg of the conjugate in Freund'β complete adjuvant intramuscularly as well as via multiple intradermal injections. After 1 month, the rabbits received four weekly booster injections of about 200μg of the conjugate in incomplete adjuvant before being sacrificed .

Purification of Antibodies Affinity columns were prepared for each of the eight peptides; 3 ml of ECH-Sepharoβe 4B (Pharmacia) was coupled with 2 mg peptide (coupling via a spacer arm to amino groups) using N-ethyl-N- ⁽ 3-dimethylaminopropyl ⁾ -carbodiimide hydro- chloride according to the manufacturer's instructions. Each gel was packed in a small column, and antiserum diluted 15- -fold in 0.1 M Tris-HCl (pH 8 ⁾ with 0.5 M NaCl was passed several times over the column. In most cases, the antiserum first had been taken over another column with an unrelated peptide to remove any possible nonspecifically binding compo- nents. After washing, the column was eluted with 4.5 M MgCl2 in 0.05 M NaAc—buffer, pH 5.0, and the peak of eluting anti¬ bodies localized by the absorption at 280 nm. The peak was collected, dialysed extensively, and stored frozen.

Other Sera Tested

For comparison, normal rabbit antibodies were purified on protein A-Sepharoβe (rabbit no. 3) . Antibodies to mouse β- NGF from rabbit (Ig 17) or sheep ⁽ Ig 28) were affinity-puri¬ fied on NGF coupled to CNBr-activated Sepharose 4B as descri- bed earlier (Ebendal T, Olson L, Seiger A (1983): The level of nerve growth factor (NGF ⁾ as a function of innervation. Exp Cell Res 148:311-317; Ebendal T, Olson L, Seiger A, Belew M (1984): Nerve growth factors in chick and rat tissue. In Black IB (ed) : "Cellular and Molecular Biology of Neuronal Development." New York: Plenum pp 231-242.) In addition, sera from another normal rabbit ⁽ no. 19) and from two rabbits im-

munized with mouse β-NGF ⁽ nos. 10 and 30) were examined in the present study. Finally, one rabbit (no. 46) receiving the first immunization with β-NGF ⁽ 200 μg) and then four boosts with P3- LH (200 μg each time) also was included.

EXAMPLE 4

Enzyme Immunoassay (ELISA)

Ninety-six-well immunoplates were coated with diffe¬ rent peptides, KLH or β-NGF ⁽ all at 1 μg/ml) in 0.05 M carbo- nate buffer ⁽ pH 9.6 ⁾ . The plates were blocked with IX bovine serum albumin ⁽ BSA ⁾ , and then dilutions of the peptide anti¬ sera or the affinity-purified antibodies were added to the wells and left at 4°C overnight. After extensive washing, bound antibodies were detected with biotinylated antirabbit antibodies ⁽ Vector Labs., CA ⁾ followed by streptavidin-conju- gated β-galactosidase (Bethesda Research Labs., MD) . Enzyme activity was examined in a fluorometric plate reader (Dyna- tech Microfluor ⁾ after the addition of methylumbelliferyl-β- -galactoside .

Bioassay of NGF-Inhibitory Activity

NGF activity was demonstrated in an assay with sympa¬ thetic chick embryo ganglia explanted to a collagen matrix (Ebendal T, Olson L, Seiger A, Belew M (1984) : Nerve growth factors in chick-and rat tissue. In Black IB (ed) : "Cellular and Molecular Biology of Neuronal Development." New York: Plenum pp 231-242 ⁾ . The influence of increasingly higher an¬ tiserum concentrations on NGF-induced fibre outgrowth was determined using this assay.

EXAMPLE 5 Immunohietochemistry

The NGF peptide antibodies also were characterized im unohistochemically in the salivary glands of adult male mice that had been kept in isolation at least 24 hr prior to sacrifice to ensure a resting state of the salivary glands.

The salivary glands were fixed by perfusion with 4% parafor- maldehyde and further fixed by immersion in the same fixative for another 2 hr (Olson L, Ayer-LeLievre C, Ebendal T, Seiger A (1987): Nerve growth factor-like immunoreactivities in ro- dent salivary glands and testis. Cell Tieeue Res 248:275-

-286) . Cryostat section ⁽ 5-14 μm) were incubated (4°C, over¬ night) with the affinity-purified peptide antibodies at con¬ centrations 1-20 μg/ml in phosphate-buffered saline (PBS) with 0.3X Triton X-100. After washing, bound antibodies bound were visualized by incubation for 1 hr at room temperature with fluorescein-isothiocyanate-labelled antirabbit or anti- sheep antibodies. After further rinsing, sections were cover- slipped in a mixture of glycerol and PBS (9:1), containing 0. IX p-phenylene diamine as an anti ding agent (Johnson GD, de C Nogueira-Araujo GM (1981): A simple method of reducing the fading of immunofluorescence during microscopy. J. Immun¬ ol.Methods 43:349-350; Platt JL, Michael AF (1983); Retarda¬ tion of fading and enhancement of intensity of immunofluore¬ scence by p-phenylenediamine . J Histoche Bytochem 31:840- -842) . Sections were evaluated using epifluorescence micro¬ scopy (Nikon MicrophotFX ⁾ and the distribution and intensity of fluorescence estimated on a blind basis. Fluorescence in¬ tensities were βemiquantitatively estimated on a 0-5 scale.

RESULTS

Antisera to the eight synthetic NGF peptides were raised in rabbits (Table I) after their conjugation to KLH. All peptide conjugates resulted in antisera that in ELISA tests bound to the appropriate NGF peptide (Table II). Serum titres (fold dilutions) in the ELISA ranged from 1:10,000 to 1:1,000,000. Peptide P2 gave high background values when used to coat the immunoplates and had to be conjugated with bovine serum albumin (BSA) before coating in order to achieve ap¬ propriate estimations of the titres of the antisera to P2. The titres measured with KLH-coated wells exceeded the titres for peptide binding and were in the range 1:1,000,000 to

1 : 10,000,000.

All peptide antisera were subjected to affinity chro- matography on columns to which the peptides had been coupled. The recovery of immunoglobulinβ were in the range 50-200 μg per ml of antiserum ⁽ Table II ⁾ . Recoveries corresponded only to some extent to the relative titres of the different anti¬ sera .

Binding to the appropriate NGF peptides by some of the peptide antisera as a function of antiserum dilution is shown in Figure 1A. The antisera bound to coats of the re¬ spective peptides at much higher rates than did normal rabbit serum at corresponding concentrations. Some peptide antisera performed better than others, although strict comparisons are not possible, as different peptides may adsorb differentially to the microwells.

Similar results were obtained with the affinity-puri¬ fied peptide antibodies (Fig. IB). The signal to KLH-coated immunoplates was at the same time reduced 50- to 500-fold (data not shown) . Thus, it was concluded that the applied affinity purification scheme is uβeful to enhance the speci¬ fic signal in, e.g., immunohistochemical studies.

The specificity of the peptide antibodies was further examined by ELISA, testing an antiserum raised to one peptide against coats of a range of the synthetic NGF peptides. An example is shown in Figure 1C, from which it is obvious that the homologous combination resulted in the best binding. How¬ ever, peptides other than the one used for immunization were recognized to some extent (Fig. 1C) .

These findings show that antibodies specific to the different synthetic NGF peptides resulted from immunizing the rabbits and that the antibodies could be efficiently purified by affinity chromatography . ELISA then was used to test whe¬ ther or not any of the peptide antibodies recognized their corresponding peptide sequence in the native β-NGF protein. For comparison, rabbit antibodies (no. 17) to mouse β-NGF was included in the ELISA of NGF-coated immunoplates (Fig. 2A) .

Of the. peptide antibodies, only Ig 37 and 38, both against peptide P3 , bound strongly to the mature β-NGF protein. The signal intensities obtained with these two antibodies were close to that of the affinity-purified, anti-NGF antibody (Fig. 2A) . The other peptide antibodies bound NGF at least two orders of magnitude less efficiently, some not above the background levels obtained with normal rabbit immunoglobulin alone. Low levels of binding to native NGF also were found using the precursor-specific peptide antibodies. The findings suggest that peptide P3 is unique in generating antibodies that recognize the mature, native NGF protein.

Whether or not P3 represents an epitope of β-NGF that is found in polyclonal antisera raised to β-NGF also was in¬ vestigated. Three antisera (AS 10, 17 and 30) were examined in a P3-ELISA, and all recognized the synthetic peptide well above background levels, although they differed in their re¬ lative ability to bind to P3 (Fig. 2B) . The three NGFantisera bound P3 about one order of magnitude less efficiently than the anti-P3 antiserum AS 38 itself. An antiserum (AS 46) pro- duced by immunization with β-NGF, followed by boosts with

P3-KLH, gave the same high binding to P3 coats as the homolo¬ gous P3 antiserum (Fig. 2B) .

In order to check if any of the other peptides would be recognized by NGF antisera, AS 10, 17 and 30 were applied to the entire range of NGF peptides in an ELISA. The peptide P3 consistently resulted in high binding of antibodies (Fig. 2C) . AS 17 also bound to P4, but this was less marked with AS 10 and 30. Antiserum AS 46 (the combined NGF/P3 antiserum) also recognized only P3 and none of the other peptides. The possible interference with NGF activity by the peptide antibodies were studied in a bioassay using explanted sympathetic chick ganglia (Ebendal T, Olson L, Seiger A, Be¬ lew M (1984): Nerve growth factors in chick and rat tissue. In Black IB (ed) : "Cellular and Molecular Biology of Neuronal Development." New York: Plenum pp 231-242). NGF-induced fibre outgrowths (stimulated by mouse β-NGF) were not blocked by

any of the peptide antisera, not even when present at high concentrations (3X) in the assays ⁽ data not shown) . Affinity- purified antibodies to peptide P3 also failed to block the outgrowth response (tested at concentrations of up to 8 μg of Ig 38 per ml) . This property of the peptide antibodies is reminiscent of the inability of monoclonal antibodies against NGF to block β-NGF-induced fibre responses in the ganglion assay. Furthermore, peptide P3 ⁽ at 100 ng/ml) could not sub¬ stitute for NGF in neurite-outgrowth stimulation, nor did peptide P3 added in vast molar excess interfere with the bio¬ logical activity of NGF. The NGF antiserum AS 46, at a thou¬ sand-fold dilution, was, however, blocking the sympathetic fibre outgrowth evoked by β-NGF.

Immunohistochemistry

The properties of the peptide antisera and affinity- purified antibodies also were examined using immunohistoche- ical staining of the male mouse submandibular gland. Various immunoglobulin preparations from the 19 rabbits .immunized with the eight different peptides ⁽ Table II) were therefore tested immunohistochemically. In several cases, preadsorption tests with appropriate peptides were carried out. In addi¬ tion, for peptides 4 and 5, the affinitypurified preparations were compared with the excluded fractions from the affinity- -purification columns. Finally, comparisons were made to the distribution of immunofluorescence in the salivary gland u- sing affinity-purified anti-NGF-antibodiee .

EXAMPLE 6 The following peptide:

Glu-Pro-His-Ser-Glu-Ser-Asn-Val-Pro-Ala-Gly-His-Thr-Ile. was synthesized by Cambridge Research Biochemicals Ltd. , Eng¬ land, at an estimated purity of >80X and checked by amino acid analysis, HPLC and mass spectrometry . The peptide corre- sponde to position -103 to -90 of the human NGF precursor.

This peptide was subjected to the same treatments and procedures as described in Examples 2 to 5 above giving the same useful results. It was also noted that although the pep¬ tide corresponds to a sequence of the human NGF precursor it resulted in good binding of purified peptide antibodies also to NGF precursor of rat origin in spite of differencieβ in protein compositions .

The invention gives the important contribution to the assay techniques in that the levels of both NGF precursor and mature NGF per se can be determined using the peptides of the present invention. This is important in obtaining a correct diagnosis on nerve disorders since the relation between level of precursor and level of mature NGF may be decisive for de¬ termining the kind of treatment to be applied. Thus, abnormal levels of mature NGF in a human being can be either caused by overproduction thereof or by reduced consumption thereof. Therefore, the level of precursor can be decisive in establi¬ shing the nature of the disorder making it possible to find the correct treatmen .

TABLE I

Synthetic NGF and Pro-NGF Peptides*

Peptide Position

No Species Sequence Start End PI Chicken ⁽ Tyr ⁾ -Glu-Thr-Lys-Cys- 54 69

Arg-Aβp-Pro-Arg-Pro-Val-

Ser-Ser-Gly-Cys-Arg-Gly

P2 Chicken Cys-Val-Leu-Ser-Arg-Lys- 109 118 Ser-Gly-Arg-Pro

P3 Rat(mouse, ⁽ Cys ⁾ -Gly-Asp-Lys-Thr-Thr- 23 35 chicken, Ala-Thr-Asp-Ile-Lys-Gly- human) Lys-Glu

P4 Rat Cys-Arg-Ala-Pro-Asn-Pro- 58 67

Val-Glu-Ser-Gly

P5 Rat(mouse) (Cys)-Glu-Pro-Tyr-Thr- -103 -90 Asp-Ser-Asn-Val-Pro-Glu-

Gly-Asp-Ser-Val

P6 Rat(mouββ, Ser-Pro-Arg-Val-Leu-Phe- -38 -28 human) Ser-Thr-Gln-Pro-Pro- ⁽ Cys )

P7 Rat ⁽ Lys ⁾ -Val-Leu-Sβr-Arg-Lys- 111 120

Ala-Ala-Arg-Arg-Gly

P8 Rat (Cys ⁾ -Val-Lys-Ala-Leu-Thr- 87 97 Thr-AsP-ASP-LVS-Gin-Ala

*Amino acid residues given within parentheses were added to the NGF sequence for conjugation purposes.

TABLE II

The NGF-Peptide Antisera Studied

Peptide Rabbit Recovery of No. AS No. ELISA titre (tig/ml)*

PI 32 200,000 60

33 100,000 90

35 10,000 15

44 1:1 000,000 70

48 200,000 120

P2 34 10,000 30

36 100,000 200

P3 37 400,000 40

38 1:1 ,000,000 80

P4 39 1:1 ,000,000 90

40 1:1 ,000,000 100

P5 41 1 ; 100,000 60

42 1 ; 10,000 30

P6 43 1 : 100,000 80

49 1 ; 100,000 100

P7 45 10,000 150

^• 50 1 ; 20,000 90

P8 51 1 : 200,000 220

52 . 100.000 40

*Values refer to recovery of specific peptide antibodies per ml antiserum by affinity chromatography.