PERMEABILITY CHANGES

The present invention relates to new antisecretory factors having fluid transport and/or inflammatory reac¬ tions regulating properties as well as polynucleic regu¬ lating properties, and polynucleic acids coding therefor, and the use thereof.

All cells and tissues of the body are critically dependent on a constant and normal fluid environment in combination with an adequate blood supply. Derangement of one or both of these supporting systems may rapidly become fatal. Concerning fluid imbalance, two principally different systems exist:

A. edema, which is characterised by the abnormal accumu¬ lation of fluid in the intercellular tissue spaces or body cavities, or B. dehydration, which, in a strict sense, means loss of water only, but is in fact commonly used to describe the combined loss of water and ions.

The most common forms of either edema or dehydration are: diarrheas, inflammatory bowel diseases, brain edema, asthma, rhinitis, conjunctivitis, arthritis, glaucoma, various forms of pathological intracranial pressure (increase or decrease), pressure alteration in the middle ear such as Morbus Meniere, dermatitis, chemical or phy- sical derangement of the skin and skin adjacent glands such as mastitis, various forms of endocrine disorders, such as diabetes insipidus. Conn's syndrome, Cushing's syndrome and Morbus Addison, kidney diseases such as pyelonephritis and glomerulonephritis, metabolic diseases such as myxedema and acute intermittent porphyria, side effects during treatment with various drugs such as anti- diabetics, tricyclic antidepressants, cytostatics, barbi¬ turates, narcotics and narcotic analogues.

Diarrhea is caused by a change in the permeability in the gut for electrolytes and water. This disturbance is often caused by bacterial enterotoxins such as those produced by Escerichia coli, Camp lobacter jejuni, Vibrio cholerae, Shigella dysenterlae and Clostridium difficile. The disturbance could also be caused by intestinal in¬ flammation. Since the uptake of water is coupled to the uptake of electrolytes and nutrients, animals with fre¬ quent diarrhea suffers from malnutrition, resulting in retardation of the daily weight gain in the growing ani¬ mal. The body counteracts these reactions by neuro-hormo- nal mechanisms such as the release of somatostatin and opiate peptides from interneurons in the intestinal mucosa. These polypeptides are capable of reversing fluid secretion and diarrhea.

The recently described antisecretory factor (AF) has been partially purified from pig pituitary gland and shown to reverse pathological secretion induced by various enterotoxins. High levels of AF in sow milk pro- tect the suckling piglets against neonatal diarrhea. Antimicrobial drugs have been widely used in the treatment of diarrhea in both human and veterinarian medicine. They are also used as feed additives for pigs, calves and chicken. However, due to the rapid development of resistant bacteria in the gut, the use of antibiotics against enteritis is generally not accepted in human medicine and their use is also diminishing in veterina¬ rian medicine.

Other antidiarrheal drugs counteract the secretion in the intestinal mucosa. Since these drugs are directed against the host animal, it is unlikely that resistance against the drugs will develop. These types of drugs include nerve-active drugs like phenothiazines and thioxanthenes. Due to some serious side effects these types of drugs have not been accepted for treatment of diarrhea in most countries. Other drugs are derivatives of opiates like codeine and loperamide. Since these drugs

mainly acts by inhibiting intestinal mobility, they also inhibit the clearance of pathogenic bacteria from the gut and should definitely not be recommended against dysen¬ teric bacteria or parasites. Derivatives of somatostatin have been introduced recently, but have so far a limited use due to difficulties in the administration of the drugs and possible interactions with the endocrine regu¬ lation of growth.

The antisecretory factor (AF) has so far not been used directly for treatment of diarrhea or malnutrition due to the difficulties involved in obtaining a pure pre¬ paration of this protein. However, it has been possible to induce similar proteins in domestic animals which have been given a specific feed (SE Patent No. 9000028-2). Pigs given this feed obtained high levels of AF-like pro¬ teins and had a significant increase in the daily growth rate compared to matched controls. AF in rats challenged with toxin A from C. difficile protects not only against intestinal secretion but also against inflammation and bleeding in the gut.

A major object of the present invention is to pro¬ vide a new recombinant protein and homologues and frag¬ ments (peptides) thereof for use in normalising patholo¬ gical fluid transport. These proteins and peptides are collectively called antisecretory factors (AF). The use of AF also partly inhibits, or totally eliminates the development of inflammatory reactions of various aetiolo¬ gies. Reconstitution back to normal (fluid transport or inflammation) is obtained by the use of proteins or pep- tides. Further the AF proteins or peptides are effective¬ ly absorbed via various mucus membranes without loosing in potency (when compared to intravenous administration). Consequently, a multitude of treatment regimens exist, and a correctly administrated protein or peptide make it possible to rapidly reconstitute a deranged fluid (water and ion) balance, an inflammatory reaction, or both.

In summary, the recombinant AF (rAF) and the homo¬ logues and fragments thereof could be used for immunode- tection, as feed additive for growing animals and as antidiarrheal and drugs against diseases involving edema, dehydration and/or inflammation.

The objects of the present invention are the fol¬ lowing:

A recombinant protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof.

A fragment of the recombinant protein shown in SEQ ID No. 1, which fragment is chosen from the group com¬ prising a) amino acids nos. 35-42 b) amino acids nos. 35-46 c) amino acids nos. 36-51 d) amino acids nos. 36-80 e) amino acids nos. 1-80 of the amino acid sequence shown in SEQ ID No. 1. A peptide X1VCX2X3KX4R corresponding to the fragment comprising the amino acids no. 35-42 of the recombinant protein shown in SEQ ID No. 1, wherein X is I or none, X2 is H, R or K, X3 is S, L or another neutral amino acid and X4 is T or A. Antibodies against a recombinant protein having es¬ sentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof.

A protein binding to antibodies specific to a recom¬ binant protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments there¬ of.

A composition for normalising pathological fluid transport and/or inflammatory reactions comprising as an active principal an effective amount of the recombi- nant protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments there¬ of.

Use of a recombinant protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof for manufacturing a composition for normalising pathological fluid transport and/or inflamma- tory reactions.

Feed for normalising pathological fluid transport and/or inflammatory reactions in vertebrates, comprising as an active agent a recombinant protein having essen¬ tially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof, or an organism capable of producing such a protein or homologues or fragments thereof.

A process of normalising pathological fluid trans¬ port and/or inflammatory reactions in vertebrates, com- prising administering to the vertebrate an effective amount of a recombinant protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof, or an organism producing said pro¬ tein or homologues or fragments. Use of specific antibodies against a recombinant protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof, for detecting said protein or fragments in organisms.

Nucleic acids coding for a recombinant protein hav- ing essentially the sequence shown in SEQ ID No. 1, or homologues or fragments thereof.

Use of nucleic acids coding for a recombinant pro¬ tein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or fragments thereof, for producing corresponding proteins or homologues or frag¬ ments.

Use of probes or primers derived from nucleic acids coding for a recombinant protein having essentially the sequence shown in SEQ. ID No. 1, or homologues or frag- ments thereof, for detecting the presence of nucleic acids in organisms.

Vector comprising nucleic acids coding for a recom¬ binant protein having essentially the amino acid sequence shown in SEQ. ID No. 1, or homologues or fragments there¬ of. Host except human comprising a vector including nucleic acids coding for a recombinant protein having essentially the amino acid sequence shown in SEQ. ID No. 1, or homologues or fragments thereof.

A strain of an organism except human capable of producing a protein having essentially the amino acid sequence shown in SEQ ID No. 1, or homologues or frag¬ ments thereof.

As organisms capable of producing the recombinant protein use can be made of different types of organisms, such as recombinant bacteria and eucaryotic organisms, such as yeast, plants and vertebrates except humans.

Despite ten years of attempts to purify AF by con¬ ventional biochemical techniques, it has not been pos¬ sible to obtain AF in a homogeneous form. However, by means of a new procedure of preparing a semipure AF for immunisation and selecting antiserum by means of an immu- nohistochemical method a suitable antiserum was chosen. With this antiserum it has now been possible to clone recombinant human cDNA expressing AF in E. coll . The sequence of the new cDNA was determined and shown to be unique. By knowledge of this sequence, oligonucleotide probes were constructed which hybridise with human and porcine pituitary RNA. The size of this RNA, about 1400 basepairs, complies with the size of the sequenced cDNA comprising 1309 basepairs plus a poly(A)tail. A partial cDNA sequence from rat pituitary gland has been shown to be identical with that of the human cDNA reflecting a ubiquitous structure conserved in AF genes from different species. This resemblance makes it possible to use the same oligonucleotide probes to identify AF-coding RNA and DNA from different species.

It has furthermore been possible to express the rAF in a biological active form. The AF protein in form of a fusion protein with glutathione S-transferase was ex¬ pressed in large amounts in E. coli and purified to homo- geneity by affinity chromatography. After cleavage of the fusion protein with thrombin, the recombinant AF (rAF) was shown to be extremely potent, 44 ng (lO ^-1 ^ _m ol) giv¬ ing a half-maximal inhibition of cholera toxin-induced fluid secretion in rat intestine. By gene technique smaller fragments of rAF was produced. The activity was shown to reside in a small sequence consisting of 7 to 8 amino acids. This was confirmed by help of chemical solid phase synthesis by which technique an octapeptide was produced and shown to be almost as biological potent as rAF on molar basis. With help of site directed synthesis a variety of sequences within the active site was constructed and replacements of certain amino acids shown to be possible without abolishing the biological activity. The fluid secretion was measured by the intestinal loop model: a section (loop) of the small intestine is ligated by means of two satures; in the loop a certain amount of enterotoxin is injected. If antisecretory drugs are tested they are injected between one hour before and two hours after toxin challenge. The injection was made by three different routes; intravenously, intraintesti- nally and intranasally. The fluid is accumulating in the loop 5 h after toxin challenge. The secretion is calcu¬ lated from the weight of the accumulated fluid per cm intestine.

The sequence of the protein was determined both directly by amino acid sequencing and indirectly by deduction from the cDNA sequence.

Recombinant AF seems to exert very little toxic or systemic effects since no obvious toxic reactions were noted in rats given 100 fold higher doses than that caus¬ ing half-maximal inhibition. Since it is efficient when

injected in the small intestine it could be administrated perorally.

The recombinant AF inhibits secretion also when injected after toxin challenge in contrast to the prepa- rations of natural AF tested which seem to efficient only when injected before the toxin. Thus, rAF could be used both prophylactically and therapeutically.

Further, rAF and its peptide fragments were shown to inhibit cytotoxic reactions and inflammation in the gut caused by toxin A from Clostridiiππ difficile . By help of a dye permeability test rAF and its fragements were shown to reverse pathological permeability changes induced by cholera toxin not only in the intestinal mucusa but also in plexus choroideus which regulates the fluid pressure in the brain.

Antisera against rAF were produced in rabbits and used in enzyme-linked immuno assays (ELISA). This assay might be used to measure AF in body fluids or feed.

A method of purifying antibodies against AF (natural or recombinant) by means of affinity chromatography on columns with agarose coupled rAF is reported below.

The antibodies were also shown to be efficient for detection of AF in tissue sections by means of imrauno- histochemical techniques and for detection of AF in Western-blot.

The invention will now be described further by means of the following non-limiting Examples together with the accompanying drawings.

Example 1. Antibodies against AF produced for cloning of cDNA

Antisecretory factor was prepared from pig blood by means of affinity chromatography on agarose and isoelectric focusing. To one litre of pig blood (containing anticoa- gulating substances) 1 g of sodium thiosulfate and 1 mg of phenylmethylsulfonylfluoride were added. The blood cells were separated by centrifugation and the clear

plasma was eluted through a column with Sepharose 6B (Pharmacia LKB Biotechnology Stockholm), the gel volume corresponding to about 10% of the volume of the solution. After washing with three bed volumes of phosphate buffer- ed saline (PBS = 0.15 M NaCI, 0.05 M sodium phosphate, pH 7.2), the column was eluted with two bed volumes of 1 M α-methyl-D-glucoside dissolved in PBS. The eluate was concentrated and dialysed against water on an "Omega 10k flow through" ultrafilter (Filtron Technology Corp.). The fraction was subsequently fractionated by isoelectric focusing in an ampholine (Pharmacia) gradient pH 4-6 on a 400 ml isoelectrofocusing column (LKB, Sweden). A frac¬ tion having an isoelectric point between 4.7 and 4.9 was collected and dialysed against PBS. Thus, partially puri- fied AF was divided into small aliquotes and used for production of antiserum in rabbits according to a pre¬ viously described method.

The rabbits were immunised and the sera tested for their capacity to stain intracelluiar material in sec- tions of human pituitary gland (method described in Example 6) . Only one of the sera showed specific and distinct intracelluiar staining without staining extra¬ cellular matrix proteins. This antiserum was selected for screening of a cDNA/lambda phage GT11 library from human pituitary gland expressing proteins in E. coll .

Example 2. Screening cDNA libraries from human pituitary gland and brain.

A 5'-stretch cDNA library from normal human pituitary gland, derived from tissues obtained from a pool of nine Caucasians, was purchased from Clontech Laboratories. For screening of the library, phages were plated at 3 x IO-* plaque forming units per 150 mm dish on E. coli Y1090. The previously described rabbit antiserum against porcine AF was absorbed with 0.5 volumes of E. coli Y1090-lysate for 4 hours at 23°C and diluted to a ratio of 1:400 and screening performed according to Young and Davis ( 1 ) .

Alkaline-phosphatase-conjugated goat anti-rabbit anti¬ bodies were used as second antibodies (Jackson). Positive plaques were picked, eluted into phage suspension medium [20 mM Tris-HCl (pH 7.5), 100 mM NaCI, 10 mM MgS0 ₄ , 2% gelatin] , replated, and screened until all plaques tested were positive. cDNA-recloning - Phage DNA from AF recombinants was isolated with Wizard Lambda Preps (Promega) and digested with EcoRl. The inserts were purified with Sephaglas BandPrep Kits (Pharmacia), recloned into pGex-lλT vector (Pharmacia) as described by the manufacturer and trans¬ fected into Epicurian Coli XLl-Blue, Top 1 cells or BL21 cells (all three from Stratagen). rAF or rpeptides were prepared in BL21 cells when not stated otherwise (2) . Amplification of cDNA by PCR - To obtain the missing 5'-end of the cDNA a PCR-based method called RACE (rapid amplification of cDNA ends) was performed. A modified RACE-method that generates 5'-RACE-Ready cDNA with an anchor oligonucleotide ligated to the 3' -ends of the human brain cDNA molecules was purchased from Clontech

Laboratories. The 5 '-end was amplified from a portion of the 5 '-RACE-Ready cDNA in two PCR amplification steps using a 5' primer complementary to the anchor and two nested gene-specific 3' PCR primers A and B (A=base 429-411 and B=base 376-359; Fig. la). Various smaller portions of the RACE fragment was further amplified in order to express the corresponding peptides and test for their biological properties. The position of the base and amino acid at the start and end of these oligonucleotide fragments and their corresponding peptides are shown in Table 1. Porcine and bovine cDNA (Clontech Laboratories) was used as templet for amplifying fragments correspond¬ ing to N3 in Table 1. Variation of the sequence was also inserted artificially by site directed mutagenesis in which method various oligonucleotides corresponding to position 168-193 was synthesised in order to replace one by one of amino acid 35-42 (positions as shown in SEQ ID

No. 1). The amplified DNA fragment was cloned into pGex-lλT vector by using the EcoRl site built into the anchor and the gene-specific primer. To verify the sequence obtained by the RACE method, double stranded cDNA from human pituitary gland and brain (Clontech) were amplified with primer pair C/D containing an extra EcoRl-cleavage site (Fig. lb). The primers were designed to allow the entire open reading frame (ORF) to be ampli¬ fied. The pituitary and brain PCR-products of expected size were digested with EcoRl, isolated and cloned into the plasmid pGex-lλT vector.

DNA sequencing and oligonucleotides - DNA from plas¬ mid pGex-lλT was used as a template for sequencing of the inserts by dideoxy-chain-termination method (15) using the Sequenase version 2.0 kit (U.S. Biochemical Corp.). Initial forward and reverse primers copying regions of pGex-lλT immediately upstream and downstream of inserted DNA were obtained from Pharmacia. Subsequent primers were synthesised (Scandinavian Gene Synthesis AB) on the basis of sequence information obtained. Three different PCR clones were sequenced in order to avoid base-exchange by Tag polymerase in the 5'-RACE method.

Nucleotide sequence and the deduced protein sequence data were compiled and analysed by using MacVector 4.1 (Eastman Chemical Co.). To predict the corresponding amino acid sequence of the cDNA inserts, codon usage of different reading frames was compared and gave one large open reading frame. Interrogation of DNA and protein sequence data was carried out by use of an Entrez CD-ROM disc (National Center for Biotechnology Information, Bethesda, USA).

Molecular cloning and sequence analysis of cDNA - Polyvalent antisera against AF protein from pig were used for screening cDNA from human pituitary glands. Two clones expressing immunoreactive AF were isolated, rescued from phage lambda and recloned into the EcoRl site of vector pGex-lλT as described in the kit provided

from Pharmacia. Restriction analysis gave insert sizes of 1100 and 900 bp, respectively. DNA-sequencing of the two clones revealed homology to be complete except for one substitution (Fig. 1, C replacing T at position 1011). A sequence upstream of the 5'-end of clone 2 was obtained by means of the RACE method. The fragment had a total length of 376 bp (not including the synthetic nucleotide arm at the 5' -end) . The total reconstructed cDNA contain¬ ed 1309 basepairs followed a poly-A tail, which was pre- ceded by a poly-A signal (Fig. 1, positions 1289-1295). An open reading frame (ORF) of 1146 bp (positions 63-1208) was identified.

Example 3. Expression of mammalian AF protein from recombinant plasmids.

Construction and purification of fusion proteins - The cDNA-clones obtained by immunological screening and by PCR amplification of the entire cDNA were ligated to pGex-lλT. This vector allows expression of foreign pro- teins in E. coli as fusions to the C terminus of the

Schistosoma japonicum 26 kDa glutathione S-transferase (GST), which can be affinity purified under nondenatur- ing conditions with help of the kit provided from Pharmacia. Briefly, overnight cultures of E. coli trans- formed with recombinant pGex-lλT plasmids were diluted in fresh medium and grown for a further 3 h at 37°C. Protein expression was induced by 0.1 mM IPTG (isopropyl-beta-D- thiogalactopyranoside), and after a further 4 h of growth at 30 ^C C, the cells were pelleted and resuspended in PBS. Cells were lyzed by sonication, treated with 1% Triton X-100 and centrifuged at 12000X g for 10 min; the super¬ natant containing the expressed fusion proteins was puri¬ fied by passing the lysates through glutathione agarose (Pharmacia). The fusion proteins were either eluted by competition with free glutathione or were cleaved over¬ night with 10 U bovine thrombin to remove the AF-protein from the GST affinity tail. The entire method of using

the pGex plasmid and purifying the recombinant proteins or peptides was performed by means of the kits provided from Pharmacia.

Sequence and size of recombinant AF- proteins - To confirm the coding sequence, the full-length transcript was isolated by using PCR-amplification of pituitary and brain cDNA. Using the primer pair C/D, 1215 bp identical to the sequence of clone-4 (Fig. 1) was isolated. The open-reading frame encoded 382 amino acids with a calcu- lated molecular mass of 41.14 kDa and a calculated pi of 4.9.

The AF clones-1, 2 and 3 as well as the oligonucleo- tides N1-N5 (Fig. 1 and Table 1) were ligated into the pGEX-lλT plasmid vector so that the ORF was in frame with the glutathione S-transferase (GST) protein. The con¬ structs were transformed into E. coli , and expression of fusion proteins was induced with IPTG. The purified fusion proteins and the thrombin-cleaved AF protein or peptide were subjected to SDS-PAGE and Western blotting using antiserum against porcine antisecretory factor

(Fig. 2). Coomassie brilliant blue staining of the pro¬ teins revealed discrete bands for each protein except for the GST-AF-1 protein which manifested degradation into smaller components. Solid phase peptide synthesis - Smaller peptides

(P7 to Pi8 ⁱⁿ Table 1) was produced (K.J. Ross-Petersen AS) on solid phase in an Applied Biosystems peptide synthesiser. The purity of each peptide was 93-100% as evaluated on reversed phase HPLC on Deltapak C18, 300 A using a linear gradient of 0.1% trifluoro acetic acid in water/acetonitril.

Amino acid sequencing - Protein sequence analysis was performed to further validate the identified ORF. The pure AF proteins were run in 10% macro-slab gel SDS-PAGE (14) and the proteins transferred to a Problot membrane

(Applied Biosystems) by electroblotting (Bio-Rad). Spots, visualised by Ponceau S staining, were excised from the

blot and the first 20 amino acids of the proteins were sequenced by automated Edman degradation on an automatic sequencer (Applied Biosystems).

The N-terminal sequences of clone-2 and clone-3 were determined, and shown to perfectly match amino acids 63-75 and 130-140, respectively, of the predicted se¬ quence (Fig. 1) .

Comparison with other protein sequences available from GenBank revealed that the sequence of rAF (Fig. 1) is unique in all its parts and no similar sequence has been reported.

The first ten residues of the protein appear to be relatively hydrophobic when analysed according to Kyte- Doolittle (22) and might constitute a signal peptide, which is cleaved out prior to exocytosis of the protein. This interpretation is supported by the Western blot ana¬ lyses (Fig. 3) in which the recombinant protein appeared to have a slightly higher molecular mass than the protein extract from pituitary gland. Some of this difference, however, might also be due to the additional five amino acids in the recombinant protein constituting the trombin cleavage site of the fusion protein.

Example 4. Production and testing antisera against rAF Antisera against recombinant GST-AF fusion protein -

Antibodies against the purified fusion proteins GST-AF-1, GST-AF-2 and thrombin-cleaved pure AF-1 protein (=rAF) for use in ELISA, Western blot and immunohistochemical studies were produced in rabbits. Each rabbit was given 100 μg of antigen in 1 ml PBS mixed with an equal volume of Freund's complete adjuvant; each immunisation was distributed in 8-10 portions injected in the back intra- cutaneously. Two booster doses with 50 μg antigen were injected at 3 and 5 weeks, the last one without Freund's complete adjuvant. The rabbits were bled 6 days after last booster and sera were prepared and stored at -20°C. The sensitivity of the antiserum was tested with a dot

blot assay. GST-AF-2 was applied on an ECL nitrocellulose membrane in 1/5 dilutions, and the antiserum diluted 1:1000. The membrane was blocked with 1% bovine serum albumin (BSA) in PBS at 4°C for 16 h, and then incubated for 1 1/2 h with a 1:800 dilution of rabbit anti-GST-AF or porcine AF antiserum. The blot was developed with alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin followed by 5-bromo-4-chloro-3-indolyl phosphate and p-nitro blue tetrazolium (Boehringer Mannheim). The estimated limit for antigen detection was about 1 ng in this test.

SDS- poly acrylamide gel electrophoresis and immuno- blotting - SDS-polyacrylamide gel electrophoresis (SDS- PAGE) of human and porcine pituitary gland extracts and pure AF-proteins was performed in 10% acrylamide minislab gels, essentially as described by Laemmli (4) with the modification that bis-acrylamide as a cross-linker was replaced by N _v"-diallyltartardiamide with the corre¬ sponding molarity. Pyronin Y (Sigma) was used as a marker of the electrophoretic front. Prestained molecular weight reference were purchased from BDH. Proteins were then either stained with Coomassie brilliant blue or transfer¬ red electrophoretically to 0.45 mm pore-size ECL nitro¬ cellulose (Amersham) for immunoblotting. The subsequent incubations with BSA, conjugated anti-IgG and alkaline phosphatase substrate were the same as for the dot blot assay described above.

As stated above Coomassie Brilliant Blue staining revealed no discrete band for the GST-AF-1 protein, which was probably due to proteolytic degradation into smaller components. However, in the Western blot analyses the full length protein gave a much stronger signal than the degradated products (Fig. 2b) . The strong reaction with the antiserum against porcine AF indicated that the recombinant proteins indeed have the same immunoreacti- vity as AF. The molecular weight of the full length pro¬ tein appeared to be about 60 kDa which is higher than the

true mol. wt of 41139 Da estimated from the amino acid composition. Furthermore, the proteins were also immuno- blotted and probed with antiserum raised against GST-AF- 2, which bound to the thrombin-cleaved proteins (Fig. 3). Antiserum against recombinant GST-AF-2 reacted with the naturally occurring AF protein of an apparent mol mass of 60 kDa, and with some smaller components, proba¬ bly enzymatic degradation products (Fig. 3 a).

ELISA for determination of AF- concentrations - ELISA assays were performed using anti-AF-1 and anti-AF-2 according to a previously described method (5). As shown in Fig. 3b the sensitivity of the test with the crude antiserum was between 1-10 μg protein whereas the test with the affinity purified antibody had a sensitivity between 5 and 50 ng protein.

Example 5. Northern blot analysis of RNA from pituitary gland

Northern Jlot analysis - Human pituitary glands were obtained postmortem from Sahlgrenska Hospital (permission given by Swedish Health and Welfare Board; 2§ transplan- tationslagen, 1975:190). To obtain RNA, pituitary glands were extracted with guanidinium thiocyanate RNA according to Chomczynski and Sacchi (6). Polyadenylated RNA was selected by means of a commercial kit (Pharmacia) using columns with oligodT-cellulose. In addition, a pool of human pituitary mRNA from 107 individuals purchased from Clontech was used. Five μg of each sample of poly(A+)RNA was glyoxal-treated and electrophoresed in a 1.2% agarose gel (7). After capillary alkaline transfer for 3 h in

0.05 M NaOH to Hybond N+ nylon membranes (Amersham), pre- hybridisation and hybridisation were carried out for 24 h each at 42°C. The hybridisation solution contained 50% formamide, 5xSSPE, lOXDenhard's solution with 250 μg/ml denaturated low-MW DNA and 50 μg/ml polyadenylic acid.

The blots were probed with four different antisense 28 bp oligonucleotides comprising the positions 132-105

(primer E), 297-270 (primer F), 748-721 (primer G) and 833-806 (primer H) of the sequence (Fig. 1); the probes were 3'-end labelled with terminal transferase (Boehringer Mannheim) plus [α ³² P]ddATP (Amersham) and purified on Nick columns (Pharmacia). Five postwashes in 5XSSPE/0.1% SDS - 0.5XSSPE/0.1% SDS were made at 42°C for 30 min each time, with a repeat of the last wash. Filters were exposed to Hyperfilm MP (Amersham) for 7 days.

Expression in pituitary gland - Northern blot ana- lyses were performed with a mixture of four oligonucleo¬ tide probes hybridising with different sequences along the cloned cDNA (Fig. 4). The probes hybridised with a single band of about 1400 bp in the separated mRNA from pituitary gland. The strongest signals were obtained with the human material, but the porcine material also cross- reacted.

Example 6. Distribution of AF in sections of pituitary gland Species and tissues - Human pituitary glands were obtain- ed postmortem from Sahlgrenska Hospital (permission given by the Swedish Health and Welfare Board; §2 transplanta- tionslagen, 1975:190). Glands were kept frozen at -70°C, except those used for histological examination which were fixed for 24 h in 4% paraformaldehyde dissolved in phos- phate-buffered saline (PBS=0.15 M NaCI, 0.05 M sodium phosphate, pH 7.2) and thereafter transferred to 7.5% sucrose in PBS. Pituitary glands from pigs, 5-7 months old, obtained from a slaughter house, were placed on dry ice during transport and kept frozen at -70°C until used. Sprague-Dawley rats, 2-3 months old, were obtained for bioassay from B ~ K Universal AB, Sollentuna, Sweden. Rabbits (New Zealand White) for immunisations were obtained from Lidkδping Kaninfarm, Sweden.

Immunohiεtochemistry - The fixed pituitary glands were frozen in liquid nitrogen, and cryo sections, 7 μm thick, were prepared. From each sample 5-10 sections comprising different parts of the gland were fastened

to microscope slides. The sections were blocked in 5% fat-free dried milk and incubated with primary rabbit antiserum (anti-GST-AF-2 fusion protein) diluted 1:4000- 1:8000 in a humid chamber overnight at 4°C. After rinsing in buffer, the specimens were incubated for 1 h at 23°C with alkaline phosphatase-conjugated swine anti-rabbit immunoglobulins diluted 1:50 (Dako A/S). The immunoreac- tion was visualised with phosphatase substrates as described elsewhere (8). Control sections were incubated with immune serum absorbed with an excess of GST-AF-2 protein or with all incubation steps except the primary antibody.

Distribution of AF in sections of pituitary gland. The distribution of AF in sections of human pituitary glands was studied with immunohistochemical techniques (Fig. 5). In all specimens investigated, a moderate number of cells in the adenohypophysis were stained; the immunostained material appeared to be located in granules in the cytoplasm; preabsorption of the immune serum with an excess of GST-AF-2 protein abolished the signal. No staining was observed in the posterior part (neurohypo- physis).

The distribution of immunoreactive material in the pituitary gland demonstrated solely intracelluiar distri- bution of AF in secreting cells of the anterior lobe

(adenophypophysis). The proteins emanating from this lobe include growth hormone, thyrotropin, corticotropin, pro- lactin and luteinising hormone. The passage of these hor¬ mones from intracelluiar localisation to the vascular system is triggered by releasing factors produced by neuroendocrinic cells in the hypothalamus.

Example 7. Biological activity of rAF Antisecretory activity - The antisecretory activity was measured in a rat intestinal loop model previously described (9). A jejunal loop was challenged with 3 μg of cholera toxin. Either different doses of purified

AF-1-proteins or PBS (control) was injected before or after the challenge with cholera toxin. The weight of the accumulated fluid in the intestinal loop (mg/cm) was recorded after five hours. Each AF preparation was tested in at least six rats. Fisher's PLSD was used for statis¬ tical analysis of the data.

Biological activity of rAF protein - The biological activity of the pure rAF protein of clone-1 produced in E. coli was tested in a rat model. The capacity of the rAF to inhibit intestinal fluid secretion when injected intravenously 20-30 sec before intestinal challenge with cholera toxin is shown in Fig. 6. In control animals injected with buffer only, the cholera toxin caused a pronounced secretion, 412f9 mg fluid per cm intestine. The pure rAF caused dose-dependent inhibition of the cholera secretion which was significantly different from the response to the buffer (p<0.01, n=6). Nine ng of clone-1 protein is sufficient to reduce the response by 34%, whereas 44 ng ( 10~ ¹² mol) and 220 ng reduced it by 46% and 78%, respectively. The biological activity of recombinant AF is greater than that of any enterotoxin known to us and greater than that of any intestinal hor¬ mone or neuropeptide modifying water and electrolyte transport. Moreover, the level of activity of human rAF in rat is surprisingly high which probably reflects a ubiquitous structure conserved in rAF molecules from different species. This hypothesis is supported by the cross-reactivity between human and porcine material obtained in the Western blot and Northern blot analyses. The capacity of 0.5 μg of rAF to inhibit intestinal secretion when injected intravenously 20-30 sec before and 90 min after cholera toxin challenge was compared (Fig. 7). Both administrations gave significant inhibition compared to control animals (p<0.01, n=6). Thus, in contrast to natural AF, the recombinant protein was also efficient when given after toxin challenge which make rAF useful for therapeutic treatment of diarrhea.

3 μg rAF was injected in a 8-10 cm long loop placed immediately proximal to the loop which was challenged with cholera toxin. The rAF was either induced 20-30 sec before or 90 min after the toxin-challenge. In Fig. 8 it is shown that both test groups obtained a significant reduction of the fluid secretion compared to controls (p<0.01, n=6); no difference was observed between the two test groups. This experiment suggests that rAF is active after oral administration and might be used as an addi- tive in animal feed provided that no serious side effect is obtained.

In the Examples described above, the rAF was produc¬ ed in Epicurian Coli XL-1 cells. In these cells much of the produced rAF was degradated into smaller peptides. When rAF was produced in BL21 cells only a small portion of the rAF was degradated while in Top 1 cells no degra¬ dation was observed. Surprisingly the biological activity was proportional to the extent of degradation, i.e. more degradation resulted in higher activity. Therefore various shorter fragments were produced in order to test for their possible biological activity.

As shown in Table 1, these fragments were tested intravenously prior to cholera toxin challenge in the same way as described above for the intact rAF. The pep- tides expressed by clone 2 and 3 tested in amounts of

0.1, 1 and 10 μg had no effect on the toxin response. In contrast one microgram of the peptide expressed by the RACE fragment (clone 4) had a pronounced effect. A lot of shorter constructs were made from the RACE fragment and expressed in pGex-1-lambda. As shown in Table 1, the active site was found to be situated between amino acid residue 35 to 51. In order to determine more exactly the active site three small peptides were made by solid phase peptide synthesis. Two of them were active, peptide 35-46 (P3) and peptide 35-42 (PI); the latter octapeptide

IVCHSKTR (PI) was active in a dose less than 1 ng being almost as active in a molar basis as the intact rAF. In

contrast a shorter hexapeptide VCHSKT (P2) exerted no effect when tested in doses between 1 ng and 10 μg.

A peptide X1VCX2 3KX4R corresponding to the human fragment PI but with certain changes and/or deletions, have also been produced by site directed mutagenesis and tested for biological activity. Comparison was also made of sequences from bovine and porcine cDNA. These studies suggested the following changes and/or deletions: Xl is I or none X2 is H, R or K

X3 is S, L or another neutral amino acid X4 is T or A.

Table 1

* Position of the basepair in SEQ ID No. 1 which was expressed in the construct made from the AF cDNA and pGex-1-lambda ** Positions of the amino acid-residues (SEQ ID No. 1) in AF at the start and end of the synthesised pep¬ tide. *** Inhibition of cholera toxin induced fluid secretion in a ligated rat intestinal loop; the amount (pmol) causing halfmaximal inhibition (ED50) is noted for active peptides.

**** These peptides were produced by solid phase syn¬ thesis.

The effect of rAF on inflammation in the intestinal mucosa was also tested in the rat intestinal loop model. Thus, 20 rats were challenged with 0.5 μg of toxin A from Clostridium difficile (10) and the inflammatory and fluid secretion measured after 2.5 and 5 hours, respectively (10 + 10 rats). Half of the rats in each group received 100 ng of rAF intravenously 30 sec prior to the chal¬ lenge; the other half received PBS buffer as control. After killing the rat, the loops was dissected out, and the middle 2-3 cm part of the loops were frozen on dry ice. The frozen specimens were then sectioned in 8 μm thick sections by use of a Leica cryostat. The sections were stained to demonstrate alkaline phosphatases by enzyme histochemistry. Alkaline phosphatases are express- ed by the intestinal epithelial cells and the staining allows an assessment and of the integrity of the intesti¬ nal epithelium.

The results revealed (Fig. 9) that the control rats developed extensive damage of the intestinal mucosa: after 2.5 h shedding of epithelial cells from the basal membrane was observed together with necrotic tissue, whereas extensive bleeding was observed after 5 h. In contrast, animal treated with rAF developed no shedding, necrosis or bleeding. The toxin A-induced fluid secretion was also inhibited from 199^+4 to 137+5 mg/cm after 2.5 h (p<0.01) and from 421+3 to 203+ _^ 6 mg/cm after 6 h (5 rats/ group, p<0.01).

A similar experiment was performed with 0.5 μg of the peptide IVCHSKTR (=P1) replacing the rAF protein. The octapeptide achieved the same effect on toxin A-induced intestinal inflammation and fluid secretion as shown in Fig. 9.

Toxicity - In order to test the toxicity of rAF it was injected in a high dose, 50 μg per rat. No obvious toxic reaction was registered during an observation period of one week.

Example 8. Biological activity of rAF on intestinal per¬ meability

In order to evaluate the effect of rAF on the permeabili¬ ty of an organic substance dissolved in the blood a test with Evans blue dye was performed according to a pre¬ viously described method (11). The experiment was ini¬ tially performed as described above in Example 7 and Fig. 5 with intravenous injection of rAF prior to cholera toxin challenge. However, no fluid secretion was measured but 90 min after toxin challenge Evans blue dye, 1 ml of a 1.5% solution in PBS, was injected intravenously. The dye was allowed to circulate for a 5 min long period. Thereafter the rat was subjected to transcardial perfu¬ sion via the left ventricle - right atrium (using a peristaltic pump [Cole Parmer Instruments, Chicago, 111., USA]) with 200 ml of 4°C PBS/Alsevier's (1/1 ratio) solu¬ tion during a period of some 150 sec, performed under ether anaesthesia. This procedure was undertaken in order to remove all of the EB present in the vascular system, leaving only the EB in the interstitial tissue to be detected by the formamide extraction of the dye.

The results in Table 2 demonstrate that CT-challenge significantly (p<0.001) increases the amount of EB that can be extracted from the intestinal tissue with some 43%, while an intravenous injection of 1 BrT prior to cholera toxin challenge prevent this increase, i.e. the amount of EB extracted from the tissue in group 1 (con¬ trol) did not differ from that in group 3 (1 rAF + CT)

The results shown in Figs 10 and 11 demonstrate the extravasation of the azo dye Evans blue in the small intestine and in the corresponding plexus choroideus from the lateral ventricles of the brain after intestinal challenge with cholera toxin, with and without previous treatment of the rats with PI (IVCHSKTR).

The experiments were performed in the following way: Male Sprague-Dawley rats, weighing 350 g, were starved for 18 h prior to the experimental procedure, but had free excess to water. The rats were used in groups of six. The peptide PI, cholera toxin (CT), and PBS were administrated according to Table 3.

* Pl(iv. ) injection 1 were given in a volume of 2 ml PBS, the peroral (po. ) injection were given in a volume of 5 ml, the intravenous injection 2 consisted of 1.5 ml of 3% Evans blue dissolved in PBS. Ether was used for anaesthesia during the performance of all injections. The i.v. injection of PI (0.5 μg) or of PBS were performed 10-15 sec before the peroral challenge with 100 μg CT or with PBS; 60 min after the peroral chal¬ lenge, the rats were anaesthetised with ether and inject- ed iv. with Evans blue. The dye was allowed to equili¬ brate for another 30 min, whereafter the rats were again anaesthetised with ether and perfused intracardially via the left ventricle with 250 ml of Alsevers solution/PBS = 50/50, in order to remove all dye present in the vascular system. After this perfusing treatment, performed during some 2-3 min, the fluroescence registrated should repre¬ sent dye present only outside the vascular system.

The brain and a part of the small intestine were sampled and frozen on dry ice and cryostat sections, 8 μm thick, were prepared. The sections were air-dried and mounted in a xylene-containing mounting media. The sections were viewed in a Zeiss fluorescence microscope, using a filter combination identical to that used for rhodamin-emitted fluorescence.

The results in Figs 10 and 11 demonstrate that the fluorescent intensity (white colour) is of a similar mag- nitude in both the small intestine (Fig. 10) and in the plexus choroideus (Fig. 11) in group A (PI iv + CT po) and C(PBS iv + PBS po). Compared to the high fluorescent intensity in the small intestine as well as in the plexus choroideus in group B (PBS iv + CT po), the results clearly demonstrate that injection of the octapeptide prior to toxin challenge inhibits the CT-induced extra- vascular penetration of Evans blue. The results suggest that this holds true not only in the vascular system of the small intestine, but also in the plexus choroideus of the lateral ventricles of the brain.

In conclusion: the effect of intravenous octapeptide IVCHSKTR administration inhibits cholera toxin-induced extravascular penetration of Evans blue in the small intestine as well as in the plexus choroideus in the central nervous system. Thus, the action of rAF and its peptide derivatives is not confined to the small intes¬ tine only, but influences also the permeability of blood vessels in the central nervous system. These findings indicate that rAF and its peptide derivatives can be used to reverse pathological intracranial pressure, pressure alteration in the middle ear and various forms of permeability changes in blood vessels.

REFERENCES [1] Young, R.A. and Davis, R.W. (1983) Proc. Natl.

Acad. Sci. USA. 80, 1194-1198 [2] Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular cloning: a laboratory manual, pp 1.74- 1.84, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. [3] Frohman, M.A. , Dush, M.K., and Martin, G.R. (1988) Proc. Natl. Acad. Sci,. USA 86, 8998-9002. [4] Laemmli, U.K. (1970) Nature 227,680-685.

[5] Zachrisson, G. , Lagergard, T. and Lonnroth, I. (1986) Acta path, microbiol. immunol. scand. C, 94, 227-231. [6] Chomczynski, P., Sacchi, N. (1987) Analyt. Biochem. 162, 156-159.

[7] Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular cloning: a laboratory manual, pp 7.40- 7.42, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. [8] Jennische, E., Matejka, G.L. (1992) Acta Physiol. Scand. 146,79-86. [9] Lange, S. (1982) FEMS Microbiol. Lett. 15, 239-242. [10] Torres, J.F., Jennische, E., Lange, S. and Lonnroth, I. (1990) Gut 781-785 [11] Lange, S., Delbro DS, Jannische E. Evans Blue per¬ meation of intestinal mucosa in the rat. Scand J Gastroenterol 1994, 29:38 - 46.

FIGURE LEGENDS

Fig. la and continued on Fig. lb. Nuclear acid sequence and deduced amino acid sequence of the new human protein. The confirmed amino acid sequence is underlined. Fig. lc. Horizontal map showing cloned cDNA and oligonu¬ cleotide primers.

SUBSTITUTESHEET

Fig. 2. Coomassie brilliant blue-stained SDS-polyacryl- amide minigel (A) and immunoblot probed with antisera against porcine AF (B) . Lanes with unprimed numbers con¬ tain glutathione-agarose-purified GST-AF fusion proteins AF-1, AF-2 and AF-3, whereas lanes with primed numbers contain the fusion proteins cleaved with thrombin. Molecular weight references (R), (BDH), are indicated on the left. The GST-AF-1 fusion protein is highly degraded but the immunoblot analysis shows only the detection of a full-length protein and spontaneous thrombin cleavage product. There is a 26 kDa product in the GST-AF-3 pro¬ tein, probably the glutathione S-transferase-tail that has been independently expressed.

Fig. 3a. Western blot using antiserum against recombi¬ nant protein AF-2. To the left, porcine (P) and three human (HI, H2, H3) pituitary glands; and to the righ , the three recombinant proteins AF-1, AF-2 and AF-3 (see Fig. 2) were applied; in the centre the molecular weight standard (R).

Fig. 3b Enzyme linked immuno-assay (ELISA) of rAF using crude antiserum and affinity purified antibodies raised in rabbit.

Fig. 4. Autoradiogram of Northern blots of RNA from a human and porcine pituitary gland (p = pooled and i = individual material) . Five μg of purified mRNA was applied in each basin; 3' -end ³² P-labelled oligonucleo- tide probes were used and the autoradiogram developed after 7 days.

Fig. 5 Cryosections of adenohypophysis stained with antiserum against recombinant protein GST-AF-2. A. Sec- tions incubated with immune serum showing scattered cells with varying degrees of positive immunoreactivity (solid arrows). Many cells completely lack staining (open

arrows) . B. Serial sections to A incubated with immune serum preabsorbed with excess of recombinant protein GST- AF-2, There is no specific staining of the cells. C and D. Larger magnifications of immunopositive cells demon- strating cytoplasmatic staining of the endocrine cells, n = nucleus, c = cytoplasma.

Fig. 6. Biological activity of recombinant protein AF-1 testing inhibition of cholera toxin-induced fluid secre- tion. Graded doses of the protein were injected intrave¬ nously in rat; three μg of cholera toxin was injected into an intestinal loop; after five hours the accumulated fluid (mg/cm intestine) in the loop was measured. Each value represents the mean +_ S.A.E. of a group of six animals.

Fig. 7. Biological activity of intravenously injected rAF-1; 0.5 μg of rAF was administrated 20-30 sec before or 90 min after challenge with 3 μg of cholera toxin in an intestinal loop of rat.

Fig. 8. Biological activity of intraluminarly injected rAF-1; 3 μg of rAF was injected 20-30 sec before or 90 min after challenge with 3 μg of cholera toxin in an intestinal loop of rat; the rAF was injected about 5 cm proximate to the loop in which the toxin was injected.

Fig. 9 A (x 2.5) is control (PBS) loops showing cellu¬ lar debris in the intestinal lumen (L), but no staining of the remaining mucosa, which suggests a total destruc¬ tion of the epithelial lining. B (0.5 μl of PI prior to toxin challenge) shows a clearly delineated epithelial lining forming villi, suggesting a conserved and normal intestinal mucosa. L = intestinal lumen. Bars = 500 μm. C (x 10) shows the destructed mucosa in the PBS-treated control group, and D shows the corresponding mucosa in the experimental (Pl-treated) group. The black arrow

point at the epithelial lining, LP = lamina propria, mm = muscularis mucosa, open arrow point at the crypt cells. Bars = 100 μm. E (x 25) shows the destructed mucosa in the control (PBS-treated) group, and F shows a corresponding magnification from a rat subjected to PI treatment prior to toxin challenge. Bars = 50 μm.

Fig. 10. Evans blue fluorescence in jejunal specimens from three groups of rats treated with cholera toxin (CT) or control buffer (PBS); pretreatment with antisecretory peptide PI or control buffer (PBS). LP = lamina propria. Black arrow indicating epithelial cell lining; open arrow head indicating crypt cells. Bars = 100 μm.

Fig. 11 Evans blue fluorescence in plexus choroideus specimens from the rats shown in Fig. 10. Bars = 50 μm.

Sequence lasting

SEQ ID NO: 1 .

SEQUENCE TYPE: Nucleotide with corresponding protein

SEQUENCE LENGTH: 1309 base pairs plus poly(A) tail

STRANDEDNESS: single

TOPOLOGY: Liniar

MOLECULE TYPE: cDNA

ORIGINAL SOURCE ORGANISM: human

IMMEDIATE EXPERIMENTAL SOURCE: Pituitary gland

FEATURES: from 63 to 1208 bp mature protein from 1289-1295 bp poly(A) signal sequence

AATTGGAGGAGTTGTTGTTAGGCCGTCCCGGΛGΛCCCGGTCGGGAGGGΛG

GAAGGTGGCAAG ATG GTG TTG GAA ΛGC ACT ΛTG GTG TGT GTG GAC AAC AGT 101 Met Val Leu Glu Scr Thr Met Val Cys Val Asp Asn 5er> 5 10

GAG TAT ATG CGG AAT GGA GAC TTC TTA CCC ACC AGG CTG CAG GCC CAG 149 Glu Tyr Met Arg Asn Gly Asp Phc Leu Pro Thr Λrg Leu Gin Ala Glπ> 15 20 25

UG GAT GCT GTC AAC ATA GTT TGT CAT TCA AAG ΛCC CGC AGC AAC CCT 197 Gin Asp Ala Val Asn He Val Cys His Scr Lys Thr Arg Scr Asn Pro 30 35 40 45

GAG AAC AAC GTG GGC CTT ATC ACA CTG GCT AAT GAC TGT GAA GTG CTG 245 Glu Asn Asn Val- Gly Leu He Thr Leu Ala Asn Asp Cys Glu Val Leu> 50 55 60

ACC ACA CTC ACC CCA GAC ACT GGC CGT ΛTC CTG TCC ΛAG CTA CAT ACT 293 Thr Thr Leu Thr Pro Asp Thr Gly Arg He Leu Scr Lys Leu His Thr 65 70 75

GTC CAA CCC AAG GGC AAG ATC ACC TTC TGC ACG GGC ATC CGC GTG GCC 341 Vol Gin Pro Lys Gly Lys lie Thr Phc Cys Thr Gly He Arg Val Ala> 80 85 90

CAT CTG GCT CTG AAG CAC CGA CAA GGC AAG AAT CAC AAG ATG CGC ATC 389 His Leu Ala Leu Lys His Arg Gin Gly Lys -Asn His Lys Met Arg Ilo 95 100 105

ATT GCC TTT GTG GGA AGC CCA GTG GAG GAC AAT GAG AAG GAT CTG GTG 437 He Ala Phc Val Gly Scr Pro Val Clu Asp Asn Glu Lys Asp Leu Val> 110 115 120 125

AAA CTG GCT AAA CGC CTC AAG AAX GAG AAA GTA AAT GTT GAC ATT ATC 485 Lys Leu Ala Lys Arg Leu Lys Lys Glu Lys Val Asn Val Asp He Ile>

130 135 140

AAT TTT GGG GAA GAG GAG GTG AΛC ACA GAA AAG CTG ACA GCC TTT GTΛ 533 Asn Phc Gly Glu Glu Clu Val Asn Thr Glu Lys Leu Thr Λla Phc Val> 145 150 155

AAC ACG TTG AAT GGC AAA GAT GGA ACC GGT TCT CAT CTG GTG ΛCΛ GTG 581 Asn Thr Leu Asn Gly Lys Asp Gly Thr Gly Ser His Leu Vol Thr Val> 160 165 170

CCT CCT GGG CCC ΛGT TTG GCT GΛT GCT CTC ΛTC ΛGT TCT CCG ΛTT TTG 629 Pro Pro Gly Pro Scr Leu Ala Asp Λla Leu He Scr Scr Pro He Lcu> 175 180 185

GCT GGT GAA GGT GGT GCC ATG CTG GGT CTT GGT GCC AGT GAC TTT GΛΛ 677 Ala Gly Glu Gly Gly Ala Met Leu Gly Leu Gly Λla Scr Asp Phc Glu> 190 195 200 205

TTT GGA GTA GAT CCC ΛGT GCT GAT CCT GAG CTG GCC TTG GCC CTT CGT 725 Phe Gly Val Asp Pro Scr Λla Λsp Pro Glu Leu Alo Leu ^* Λla Leu Arg> 210 215 220

GTA TCT ATG GAA GΛG CAG CGG CΛC GCΛ GGA GGA GGA GCG CGG CGG GCΛ 773 Val Scr Met Glu Glu Gin Arg His Λlo Gly Gly Gly Ala Arg Arg Alo 225 230 235

GCT CGA GCT TCT GCT GCT GAG GCC GGG ATT GCT ACG ACT GGG ACT GAA 821 Alo Arg Ala Scr Ala Ala Glu Λla Gly lie Ala Thr Thr Gly Thr Glu> 240 245 250

GAC TCA GAC GAT GCC CTG CTG AAG ΛTG ACC ATC AGC CAG CAA GAG TTT 869 Asp Scr Asp Asp Ala Leu Leu Lys Met Thr He Scr Gin Gin Glu Pho 255 260 265

GGC CGC ACT GGG CH CCT GAC CTA AGC AGT ATG ACT GAG GAA GAG CAG 917 Gly Arg Thr Gly Leu Pro Asp Leu Scr Scr Met Thr Glu Glu Glu Glπ 270 275 280 285

ATT GCT TAT GCC ATG CAG ATG TCC CTG CΛG GGA GCA GAG TTT GGC CAG 965 He Ala Tyr Ala Met Gin Met Scr Leu Gin Gly Λla Glu Phe Gly Glπ 290 295 300

GCG GAA TCA GCA GAC ATT GAT GCC AGC TCA GCT ATG GAC ACA TCT CAG 1013 Ala Glu Scr Ala Asp He Asp Ala Scr Scr Ala Met Asp Thr Ser Glu> 305 310 315

CCA GCC AAG GAG GAG GAT GAT TAC GAC GTG ΛTG CAG GΛC CCC GAG TTC 1061 Pro Ala Lys Glu Glu Asp Asp Tyr Asp Val Met Gin Asp Pro Glu Phe> 320 325 330

CH CAG AGT GTC CTA GAG AAC CTC CCA GGT GTG GΛT CCC ΛAC AAT GAA 1109 Leu Gin Scr Val Leu Glu Asn Leu Pro Gly Val Asp Pro Asn Asn Glu> 335 340 345

GCC ATT CGA AAT GCT ATG GGC TCC CTG CCT CCC AGG CU CCA AGG ACG 1157 Ala He Arg Asn Ala Met Gly Scr Leu Pro Pro Arg Pro Pro Arg Thr> 350 355 360 365

GCA AGA AGG ACA AGA AGG AGG AAG ACA ACA AGT GAG ACT GGA GGG ΛAA 1205 Ala Arg Arg Thr Arg Arg Arg Lys Thr Arg Scr Glu Thr Gly Gly Lys>

370 375 380

GGG TAGCTGAGTCTGCTTAGGGGACTGCATGGGAAGCACGGAATΛTAGGGTTAGATGTGTGT Gly

TATCTGTAACCATTACAGCCTAAATAAAGCπGGCAACTTTTAAAAΛ/vAAΛAAA AAAAAAΛ