Eur. J. Biochem. 194,225-236 (1990) 0 FEBS 1990

Isolation and full structural characterisation of six adrenocorticotropin-like peptides from porcine pituitary gland Identification of three novel fragments of adrenocorticotropin and of two forms of a novel adrenocorticotropin-like peptide

Karlheinz VOIGT I , Wolf STEGMAIER I , Gerard Patrick MCGREGOR I , Hannelore ROSCH and Hartmut SELIGER’ Institute of Physiology, University of Marburg, Federal Republic of Germany Sektion Polymere, University of Ulm, Federal Republic of Germany

(Received May 1 O/July 25, 1990) - EJB 90 0528

A partially purified fraction of extracted porcine pituitary glands which possesses lipolytic and adreno- corticotropic activity has been characterised. It consists of six adrenocorticotropin(ACTH)-like peptides (five of which have not been previously described) which were each purified by sequential reverse-phase (rp) HPLC. Their complete primary structures were determined following amino acid compositional analysis, extensive peptide mapping and partial sequencing. Four of the fragments represent the following ACTH fragments; ACTH(1- 31), ACTH(7 - 34), ACTH(7 - 36) and ACTH(7 - 38). By combined analytical rpHPLC and an ACTH radioimmuno- assay (with an antiserum exhibiting full cross-reaction with all six ACTH variants isolated here), evidence was obtained from analysis of extracts of whole pituitary that these fragments of ACTH exist in significant amounts relative to intact ACTH(1- 39). This suggests that ACTH can undergo more extensive differential proteolytic processing than previously thought. These peptides were found to possess reduced or a complete absence of ACTH-like biological activity. Therefore the biological significance of this processing needs to be resolved. The other two fragments also resembled fragments of ACTH but each possessed the same, single amino acid substi- tution: a threonine replacing the arginine at the position corresponding to position 8 in the ACTH sequence and had the structures [Thr8]ACTH(1- 31) and [Thr8]ACTH(7 - 31). They possess little ACTH-like biological activity. If these variants are derived from a variant ACTH, this would be a significant finding in view of the site of the amino acid substitution and the highly conserved nature of the ACTH primary structure. The possible physiological and genetic implications are briefly discussed. In this study attempts were also made to identify the DNA coding for the mutant ACTH sequence.

ACTH is a 39-amino-acid peptide derived from the precur- sor polypeptide, proopiomelanocortin by endoproteolytic cleavage at double basic residues which flank the peptide sequence within the precursor [l-31. Much of the current understanding of the principles of hormonal peptide and neuropeptide biosynthesis have been derived from the studies of the biosynthetic processing of proopiomelanocortin. For example, it is clear that the double basic residues in precursor proteins are not always the targets of endoproteolysis such as the four consecutive basic residues (14- 17) which occur within ACTH. Cell-specific factors appear to influence biosyn-

Correspondence to K. Voigt, Institute of Physiology, University of Marburg, Deutschhausstrasse 2, W-3550 Marburg, Federal Repub- lic of Germany

Note. A preliminary report has been presented at International Narcotic Research Confcrencc, 1986 [Nut. Inst. Drug Abuse Res. Monogr. Ser. 75,279 (1 986)]

Note. The novel amino acid sequence data published here has been deposited with the EMBL sequence data bank.

Abbreviutions. rpHPLC, reverse-phase high-pressure liquid chromatography; DABITC, 4,4’-dimethylaminoazobenzene-4’-iso- thiocyanate; PLF-IID, porcine lypolytic factor subfraction IID; ACTH, adrenocorticotropin; ECSo, effective concentration eliciting 50% maximum response.

Enzymes. Trypsin (EC 3.4.21.43); Staphylococcus uureus protease (EC 3.4.21.19); mouse submaxillary protease (EC 3.4.2.-); proline- specific endopeptidase (EC 3.4.21.26); carboxypeptidase Y (EC 3.4.16.1); T4-polynucleotide kinase (EC 2.7.1.78).

thetic endoproteolysis, since in the intermediate lobe of the pituitary gland these basic residues in the ACTH sequence are the sites of cleavage. Thus, in contrast to the anterior pituitary which produces intact 39-amino-acid ACTH, the intermediate pituitary generates a-melanocyte-stimulating hormone and corticotropin-like intermediate lobe peptides which corre- spond to N-a-acetyl-ACTH( 1 - 13) and ACTH( 18 - 39) re- spectively [4 - 61. In addition, there is evidence that ACTH may undergo further differential processing. ACTH(1- 38) has been identified in bovine pituitary extracts [7], and from the pituitary gland of the porcine ACTH(7 - 39), ACTH- (7 - 38), ACTH(1- 37) and ACTH(1- 38) have been isolated [8]. ACTH(7 - 39) has been isolated from the pituitary gland of man [9] and elephant [lo].

In the studies described here, which were aimed at charac- terising the chemical nature of a previously described lipolytic factor (PLF-IID) [ l l ] derived from porcine pituitary glands, we found evidence for extensive and specific proteolytic pro- cessing of ACTH. We identified that the peptide heterogeneity of PLF-IID was mostly due to the presence of six separate peptides, which may be derived from ACTH. Four of the six peptides represented partial sequences of ACTH. Both of the other two peptide fractions also resembled partial sequences of ACTH with the single difference of a threonine residue in a position corresponding to Arg8 of ACTH. If those two peptides are derived from a variant ACTH, this finding may have important physiological and genetic implications. A

226

100 A -

0 10 20 30 40 Time (min)

B I

10 20 30 4

Fig. 1. Prepurutive rpHPLC of the partially purifiedporcine pituitary gland extract, PLPI ID. See text for chromatographic details. (A) Absorbance (280 nm) profile of the whole fraction following elution with a gradient of acetonitrile (----). The shaded peaks I + 11, I11 and IV were taken for further chromatographic fractionation. I + 11 was further fractionated under similar conditions, and fractions 1 and I1 were separated as shown in (B)

change in the primary sequence at position 8 of ACTH is within the biologically active domain of the hormone and is within a primary structure which appears to have been highly conserved throughout evolution. We undertook some analysis of porcine genomic DNA in order to determine whether these variants have arisen from a point mutation of the pro- opiomelanocortin gene occurring in a subgroup of the pig population or from a second proopiomelanocortin-like gene.

MATERIALS ENZYMES

Trypsin (treated with ~-l-tosyl-2-phenylalanylchloro- methane), mouse submaxilliary protease and carboxy- peptidase Y were obtained from Sigma (Munich, FRG). Staphylococcus aurms protease V8 and proline-specific endo- peptidase were purchased from Miles Laboratories (Slough, England).

CHEMICALS

o-Phthalaldehyde reagent was obtained from Pierce (Rockfort, USA) ready for use as Fluoraldehyde solution. 4,4’-Dimethylaminoazobenzene-4’-isothiocyanate (DABITC) for peptide sequencing was from Fluka (Neu-Ulm, FRG).

rpHPLC FRACTIONATION O F PLF-IID

3 mg partially purified fraction of PLF-IID [l 13, derived from porcine pituitaries, was subjected to a series of rpHPLC

steps. Changing either the type of reverse-phase matrix or the solvent conditions employed provided a change in the selectivity at each chromatographic step. Peptide fractions were identified by ultraviolet absorbance.

The first semipreparative rpHPLC step was performed with a 250 mm x 10 mm column of Ultrasphere octadecyl- silanate, 10 pm (Beckman, USA). Two Kontron 410 pumps and a Kontron gradient controller 200 were used to generate a gradient (see Results) from the two-component solvent system (A, 0.2% trifluoroacetic acid/5% acetonitrile; B, acetonitrile). The flow rate was 2 ml/min and the eluate was monitored with a Uvikon 720 LC variable-wavelength spectrophotometer, set at 280 nm.

Four major peaks (I + 11, I11 and IV) of absorbance were obtained (Fig. 1). The fractions 1 and I1 were separated follow- ing a second similar rpHPLC step, this time using a 250 mm x 4.6 mm column of Ultrasphere octadecylsilanate (10 pm), a flow rate of 1 ml/min and different gradient con- ditions (see Results).

Each of these four fractions was then subjected to another rpHPLC fractionation step on a 250 mm x 4 mm Hypersil WP 300 butyl ( 5 pm) with two Gynkotek (Munich, FRG) model 300 C pumps and a Gynkotek gradient controller 250B to generate the selected gradients (see Results) from the following two-component solvent system: (A) 50 mM potassium phos- phate buffer, pH 2.5, with 5% acetonitrile; (B) acetonitrile. The eluate was monitored for absorbance at 210 nm using a Gynkotek SP4 spectrophotometer in order to identify the peptide-containing peak fractions. Six, apparently pure, peptide fractions were obtained and subjected to amino acid analysis, extensive peptide mapping and partial sequencing.

AMINO ACID ANALYSIS

Peptide hydrolysis was performed in tempered-glass tubes of 500 pl volume. Peptide samples (50 - 1250 pmol) were hy- drolyzed in 200 p16 M HCl with 4% 2-mercaptoacetic acid at 105°C for 20 h. Prior to hydrolysis, the tubes were evacuated and sealed. The dry hydrolyzates were reacted with 100 pl Fluoraldehyde solution (Pierce, Rockford, USA) which con- tains o-phthalaldehyde in 1 M potassium borate buffer, pH 10.4, for 90 s, then an aliquot was subjected to HPLC analysis. The rpHPLC separation of the o-phthalaldehyde - amino-acid derivatives was performed on a 125 mm x 4 mm column packed with Shandon octadecylsilanate, 5 pm (Shandon, UK). Eluent A was 12.5 mM sodium phosphate buffer, pH 7.2 and eluent B consisted of methanol/tetra- hydrofuran (97:3, by vol.). The gradient profile was as fol- lows: 18% B for 2min; 18-24% B for 1 min; 24% B for 5min; 24-37% B for 9min; 37-38% B for 8min; 38- 60% B for 6 min. The eluent flow rate was 1.3 ml/min. The separation was performed at 30 “C and the o-phthalaldehyde derivatives were detected by their fluorescence (330 nm exci- tation, 455 nm emission).

PEPTIDE MAPPING

Enzymatic cleavage

Trypsin. 2 - 20 nmol peptide was dissolved in 100 p10.1 M sodium phosphate buffer, pH 8.65, and incubated with trypsin at 37°C for 16 h. The enzyme/substrate ratio varied in the range 1 : 50 - 1 : 30. The resulting solution was freeze-dried.

Staphylococcus aureus protease. 7 - 40 nmol peptide was dissolved in 100 pl 0.1 M ammonium acetate, pH 4.0 and

227

amplirner A1 ampliiner A2

A POMC - CDNA --. ~ G A GAT GGC GTC GCG CCG GGC: CCG C'GC CAG GAC AAG CGC TCC TAC probes P1-PS (X=A,C,G,T) ampiirner B2 I

TCC ATG GAG CAC 1TC ACX TGG GGC AAG CCC GTG GGC AAG AAG CGG CGC CCG CTG CGC

amphmer B1

AAG GTG TAT ccc AAC GGC GCC GLG --- wild-type specific probe GAG CAC TTC CGC TGG GGC AA

mutant specific probes P2: GAG CGC 7TC ACATGG GGC AA P3: GAG CAC TTC ACC TGG GGC AA P4: GAG CAC l T C ACG TGG GGC AA P5: GACCACTTCACTTGGGGCAA

PI:

B ACTH (4-10): ... Met-Glu-His-Phe-Arg-Trp-Gly ... P6: wild-type specific probe ATG GAG CAC TTC CGC TGG GG

[ThrS] ACTH (4-10): ... Met-Glu-€Iis-Phe-Thr-Trp-Gly ...

mutant specific probes P7: ATG GA: CACTTTFACA TGG GG

P8: ATG GAGA CA$*T ACC TGG GG

~ 9 : ATG GAGA CA$TT$ ACG TGG GG

PIO: ATGGA~CA$TT$ACTTGGGG

Fig. 2. Sequences of oligonucleotides used in the DNA analyses. (A) Sequences of the four amplimers A l , A2, B1 and B2 according to the proopiomelanocortin (POMC) cDNA described by Gossard et al. [21] and of the probes P1 -P5 used for the hybridisation against the amplified DNA segment. (B) Sequences of the oligonucleotide probes used for hybridisation against genomic DNA

Staphylococcus aureus protease was added (enzymelsubstrate ratio 1 : 50 - 1 : 25). The solution was incubated at 37 "C for 16 h then frozen at -32°C.

Mouse submaxilliary protease. 6 nmol peptide was dis- solved in 25 p1 water and 100 yl 0.1 M N-methylmorpho- linium acetate, pH 8.0. An aliquot of a solution of mouse submaxilliary protease in the same buffer was added resulting at an enzyme/substrate ratio of 1 : 50. This solution was incu- bated at 37°C for 3.5 h and then freeze-dried.

Proline-specific endopeptidase. 1 - 17 nmol peptide was dissolved in 200 $0.1 M potassium phosphate buffer, pH 7.0. An aliquot of a solution of proline-specific endopeptidase in 0.05 M potassium phosphate, pH 7.0 was added (enzyme/ substrate ratio, 1 : 40 - 1 : 20). The resulting solution was incu- bated at 40°C for 1 h then frozen at -80°C.

Carboxypeptidase Y . Carboxypeptidase Y was dissolved in 0.1 M pyridinium acetate buffer, pH 5.5, at a concentration of 0.25 pg/kl. 2-4 nmol peptide was dissolved in the same buffer and an aliquot of the enzyme solution was added to give an enzyme/substrate ratio of 1 : 20. After brief mixing, a 10-y1 aliquot was taken ( t = 0). During the following incu- bation at 25 "C more aliquots were withdrawn after 5, 10, 15, 20, 30, 40, 60 min, etc. All aliquots were pipetted into 20 pi trifluoroacetic acid in order to stop the reaction. The trifluoroacetic acid was removed in vacuo and amino acid analysis was performed as described without hydrolysis.

Manual sequence determination

Total peptides and enzymatic cleavage products were se- quenced in amounts of I - 10 nmol using the DABITC method [12].

rpHPLC of PLF-IID andporcine whole pituitary gland extracts for radioimmunological analysis

In order to characterise the ACTH-like immunoreactive content of PLF-IID, 100 pg lipophylized preparation (dis-

solved in 50 mM potassium phosphate, pH 2.5) was fraction- ated on a 125 mm x 4 mm Lichrospher-100 RP-18 ( 5 pm) rpHPLC column (Li Chro CART 125-4, Merck). The column was eluted with a binary solvent gradient (see Results) gener- ated from 50 mM potassium phosphate, pH 2.5 (A), and 70: 27 : 3, acetonitrile/water/50 mM potassium phosphate, pH 2.3 (B), using the Gynkotek system described above. The eluate flow rate was 1 ml/min and 1-min fractions were col- lected. An aliquot of each fraction, following a 1 : 100 dilution in the ACTH radioimmunoassay buffer, was assayed for anti- ACTH immunoreactivity as described. A dilution of 1 : 100 was found to be sufficient to dilute out any interference with the assay from the rpHPLC solvent mixture. This obviated the need to dry fractions prior to assay, which can lead to significant peptide losses.

The ACTH-immunoreactive content of PLF-IID was com- pared with that of a crude extract of porcine pituitary gland. For this, five pituitary glands were obtained fresh from the slaughterhouse and extracted in boiling 0.5 M acetic acid (10 vol./g tissue). The extract was then concentrated and par- tially purified by adsorption onto a Sep-Pak CI8 cartridge (Waters Associates, Milford, MA, USA). The Sep-Pak car- tridge was activated with solvent B of the analytic rpHPLC then washed with solvent A. The adsorbed peptide-containing fraction was then eluted with solvent B, dried under vacuum on a Speed-Vac centrifuge, reconstituted in solvent A, then analysed by rpHPLC and radioimmunoassay as described for the PLF-IID fraction. The purified fractions F1 - F6 were separately run under identical conditions as those described above and their elution times determined by ultraviolet ab- sorbance.

ACTH radioimmunoassay

The radioimmunoassay employed the rabbit antiserum, H8, raised to natural porcine ACTH. It was used at a final dilution of 1 : 150 000 with 2500 cpm of the radiolabelled tracer 8-(3-['251]iodotyrosyl)-ACTH(1 - 39) (IM 183 Amersham,

228 100 z : 9

m

z

ap - 50 .-

C L

m

l.3 0

0 10 20 30 40 50 60 0 10 20 30 Time (min) Time (min)

- 100 E

[ C F5 9

- .... .- .... 50 &

0

.... .... ...... ...... ..... 0.40-

...... u

...... . ...... . ...... . . ..... .. ...... .. ...... .. . ........ . ...... ... ...... .... .. ... .... ........... ..... . . . . ,.. ............ ....... ..... ............ ............ . ...........

0 4) (v

L3 _ _ _ - - --.UIw: - a -_-- - 0 0 10 20 30 40

.... .... .... ...... ...... .....

.... 50 &

0

0.40- ......

u ...... . ...... . ...... . . ..... .. ...... .. ...... .. . ........ . ...... ... ...... .... .. ... .... ........... ..... . . . . ,.. ............ ....... ..... ............ ............ . ...........

0 4) (v _ _ _ - - --.UIw: - a -_-- - 0 0 10 20 30 40

0

D 6E -

.32-

0 4) (v

Q:

0-

z 9

50 & .32

0

i 0 20 30 40 Time (min) Time (min)

Fig. 3. Purifi'catioti o j f izc t ions F I -F6 (shaded peaks) ,following rpHPLC using gradimt elution with acetonitvile (- - - -1

FRG), and the standard was synthetic human ACTH(1- 39) (8740, Peninsula Labs, St. Helens, UK). For the assay, all the reagents were prepared in 50 mM potassium phosphate buffer, pH 7.4, containing 0.2% bovine serum albumin. The assay was incubated overnight at 8°C in a final volume of 0.5 ml. The separation of the bound and free fractions was achieved using a second antibody [goat anti-(rabbit IgG) antibody; 14- GX-25A, Paesel Rr Lorei GmbH, FRG] at a dilution of 1 :20. Incubation with the second antibody was for 3 h at 4°C in the presence of 0.4% rabbit IgG as carrier. The bound complex was then precipitated with 10% poly(ethy1ene glycol). Follow- ing centrifugation at 4000 rpm, the supernatant was aspirated and the precipitates were taken for measuring radioactivity in a LKB Gainmamaster 1277. The assay exhibits full cross- reaction with K T H ( I - 24). I t exhibits no significant crossreaction with natural N-a-acetyl-ACTH(1- 13) (a-mel- anocyte-stimulating hormone) and only 2% cross-reaction with corticotropin-like intermediate lobe peptide.

Testing of tlic hiologicnl cictivity of purifiedpeptides FI - F4

The yield of peptides F1- F4 provided sufficient material for testing their steroidogenic and lipolytic activities. Ster- oidogenic activity was assessed by measuring corticosterone release from isolated rat adrenocortical cells according to the method of Fehm et al. [13]. Lipolytic activity was assayed by measuring glycerol release from isolated fat cells according to the method described by Schleyer et al. [Ill.

DNA analysis: invcwigution o f the possible existence qf'u vuviunt ACTH ,gene

Two possibilities for the genetic origin of the [Thr8]ACTH fragments were considered. First, they may arise from a point mutation in the proopiornelanocortin gene which occurs only in some individuds of the pig population. Alternatively they may arise from a second proopiomelanocortin-like gene.

In order to test the first possibility, the DNA from 100 individual pituitary glands were isolated. Using the primers Al, A2, B1 and I32 (Fig. 2A), which were tailored according to the porcine proopiomelanocortin DNA described by

Gossard et al. [14], the gene segment which codes for pro- opiomelanocortin(l31- 135) was amplified (see [l S] for de- tails). The amplified DNA was incubated under hybridisation conditions with each of the 32P-labelled oligonucleotides (P1 - P5, Fig. 2A; see below for details).

Probing for the existence of a hypothetical second pro- opiomelanocortin gene requires the synthesis of hybridisation probes which include not only the four codons for threonine. but all possible variants of the codons for the neighbouring amino acids. By designing for hybridisation a number of se- quences specific for [ThrS]ACTH(4- lo), it was possible to apply a minimum of four mixed probes P7-P10 as well as the wild-type sequence P6 (Fig. 2B). According to the litera- ture [16], this degeneracy should not significantly influence the hybridisation efficiency.

Oligonucleotide synthesis. The amplimers (primers used for the chain elongation) for polymerase chain reaction (Al, A2, B1, B2) and the probes for dot-blot hybridisation (Pl-P5; Fig. 2 A) were synthesized via an asynchronous simultaneous synthesis strategy [ 171 using hydrogen phosphonate chemistry. The oligonucleotides used as mixed probes (P6 - P10, Fig. 2 B) were prepared in the automatic synthesizer 381A (Applied Biosystems, USA) with protected 2-cyanoethyl phos- phoramidites as monomers. All oligomers were purified by polyacrylamide gel electrophoresis. For hybridisation, the oligonucleotide probes were end-labelled using [;'-"P]ATP (Amersham, UK) and T4 polynucleotide kinase (Boehringer Mannheim).

Polymerase chain reaction. DNA amplification was performed using the method described by Mullis and Faloona [15]. The reaction mixture contained 150 ng DNA, 400 mM of each of the deoxynucleotide triphosphates, 60 pM of each of the amplimers, SO mM KCl, 10 mM TrislC1, pH 8.3, 1.5 mM MgC12, 0.01% (massivol.) gelatine and 2.5 units Tuq polymerase in a total volume of 100 pl. Routinely 10-20 cycles were performed.

Dot-blot hyhridisution. 2 ml amplified DNA solution was spotted on to nylon filters (Schleicher & Schull, Dassel, FRG). The filters were prehybridized overnight ai SO "C in S x buffer A (buffer A: 10 mM sodium phosphate, pH 7.0, 1x0 mM NaCI, 1 mM EDTA), 7% SDS, 100 mg/ml denatured salmon

229

Table 1. Amino mid compositions offiactions F1- F6 The values indicate the relative amounts of the listed amino acids with, in parenthesis, the nearest integral value. For each fraction the sum of the integral values is also given. Also indicated for each fraction is the assigned ACTH variant. The amino acid composition of ACTH is also listed. n.d. not determined

Amino acid -

Amino acid content of fraction

F1 F2 F3 F4 F5 F6 Porcine ACTH(7-31) [Thr8]ACTH(7-31) ACTH(1-31) [Thr8]ACTH(1-31) ACTH(7-36) ACTH(7-38) ACTH

Ala

Asx Glx

His Ile Leu LYS Met Phe Pro Ser Thr

Arg

GlY

TrP TYr Val

Sum

1.0 (1)

2.0 (2) 2.0 (2)

0.2 (0) 0.1 (0) 1.1 ( 1 ) 3.8 (4) 0.1 (0) 1.3 (1) n.d. (3) 0.3 (0) - (0) 1.3 (1) 1.0 (1) 2.8 (3)

3.0 (3)

3.2 (3)

25

1.0 (1) 1.9 (2) 1.9 (2) 1.9 (2) 3.3 (3) 0.2 (2) 0.1 (0) 1.0 ( 1 )

0.1 (0) 1.0 (1) n.d. (3) 0.2 (0) 0.8 ( 1 ) 1.2 (1) 0.9 (1) 2.7 (3)

25

3.8 (4)

1.1 (1)

2.0 (2) 2.9 (3)

3.0 (3) 2.9 (3) 0.9 (1) 0.1 (0) 1.1 (1)

1.0 (1) 1.1 (1) n.d. (3) 2.0 (2) 0.1 (0) 1.0 (1) 1.9 (2)

3.8 (4)

2.8 (3)

31

m o 1 / m o 1

1.1 (1) 2.0 (2) 2.0 (2) 3.0 (3) 3.3 (3) 1.0 (1) - (0)

1.0 (1) 1.1 (1) n.d. (3) 2.1 (2) 0.9 (1) 1.1 (1) 2.0 (2)

1.4 (1) 3.1 (4)

2.8 (3)

31

3.0 (3) 3.1 (3) 2.0 (2) 3.0 (3) 3.0 (3) 0.2 (0) 0.1 (0) 1.2 (1) 4.2 (4) 0.1 (0) 2.3 (2) n. d. (4) 0.2 (0) 0.1 (0)

1.0 (1) 0.9 (1)

2.9 (3)

30

3.0 (3) 3.0 (3) 2.2 (2) 4.3 (4)

0.1 (0) - (0) 2.2 (2)

0.1 (0) 2.2 (2)

3.0 (3)

4.0 (4)

n.d. (4) 0.4 (0) 0.3 (0) 1.0 (1) 1.1 (1) 3.1 (3)

32

3 3 2 5 3 1 0 2 4 1 3 4 2 0 1 2 3

39

sperm DNA and 5 x Denhardt's solution, and subsequently hybridized for 4 h at 50°C in the same mixture containing 2 pmol 32P-labelled oligonucleotide probe (P1 -P5). The fil- ters were washed with 2 x buffer A containing 0.1 % SDS for 5 min at room temperature, then with 5 x buffer A containing 0.1% SDS for 20 min at 50°C. Finally, stringent washing was carried out with 5 x buffer A containing 0.1% SDS for 10 min at 64°C.

Hybridisation against genomic DNA. 20 mg isolated DNA (from a mixture of about 30 pituitary glands) was digested with Hind111 or EcoRI (Boehringer Mannheim) and electrophoresed on a 0.6% agarose gel. The gel was dried in vacuo and used for hybridisation with the mixed oligo- nucleotide probes P6 - P10. Prehybridisation was done in 3 M (CH3)4NC1, 50 mM Tris/Cl, pH 8.0, 2 mM EDTA, 100 mg/ ml denatured salmon sperm DNA, 0.1 % SDS, 1 x Denhardt's solution for 4 h at 50"C, followed by incubation with the 32P-labelled probes in the same mixture at 50°C overnight. The washing was performed as described above for the nylon filters.

RESULTS

Purification of peptide,fractions from PLF-IID

Fig. 1 a shows the ultraviolet absorbance profile of 3 mg of the partially purified fraction, PLF-IID from porcine pituitary glands, following semi-preparative rpHPLC. The separately shaded peak fractions were taken for further purification. The complex peak labelled I + I1 was further resolved under similar rpHPLC conditions, but with a modified gradient. The absorbance profile obtained is shown in Fig. 1 b. The fractions from peak I (shaded) and those from peak I1 (shaded) were pooled separately.

The fractions I - IV were each subjected to one further rpHPLC purification step and the apparently pure fractions F1 -F6 were obtained (Fig. 3). These rpHPLC separations were all achieved using a C4 column with phosphate as the counter ion and seems to provide sufficiently different selec- tivity from the initial semipreparative separation using a CI8 column and trifluoroacetate as the counter ion. The homogen- eity of fractions F1 -F6 were tested by rechromatography under identical conditions.

Amino acid analysis ofpeptide pactions F l - F6

The results of the amino acid analyses of F1 -F6 are summarized in Table 1 . The molar ratios are given, and for all individual amino acids the values are close to integral. This is further demonstration of the purity of the six fractions. The amino acid compositions are compared in the table to the known composition of ACTH(1- 39). Proline and cysteine are not detected by the o-phthalaldehyde derivatisation method used here. Preliminary and previously reported [I 11 amino acid analysis of PLF-IID using ninhydrin did not reveal the presence of cysteine. The amino acid compositions of F1 -F6 are strikingly similar to each other and to that of ACTH(1- 39), but note that the analysis of F2 and F4 con- tains a threonine which is not present in the other fractions nor in ACTH(1- 39).

Peptide mapping of fractions F l - F6

The results of the peptide mapping of fractions F1 -F6 are shown in the miniprint supplement. The fragments from each digest were separated on rpHPLC and their amino acid compositions were determined. These data are not included. A partial N-terminal sequence was determined for each of the

230

I 10 20 30 39 p~~~~ ~ S Y S M E ~ F R W G K P V G K K R R P V K V Y P N G A E D E S A E A F P L E F ~

I I I I I I I I

F3 ACTH (1-31) I I I I I I 1 I I I I

I I I I I I I I

I I I I I I

I

F4[Thrq A C T H ( l - 3 1 ) F T I I

F1 ACTH (7-31) I I I /

F2phr8] ACTH (7-31) I=.- I I

F5 ACTH (7-36) I I / / I I I I I

I I I I

I I I I I I

I I F6 ACTH ( 7 38)

I

Fig. 4. Primary structures of ,fractions Fl - F6 as elucidated f rom the amino acidanalysi.c, peptide mapping andpartialsequencing. The amino acid sequence of porcine ACTH is given at the top in single-letter code. The solid bars are aligned to the porcine ACTH sequence to indicate the exact extent of the shared truncated primary sequence of each fraction. For fractions F2 and F4, T indicates the threonine residue which substitutes for Arg8 (R) in the porcine ACTH sequence. Each structure is assigned ACTH with numbers indicating the corre- sponding positions of the N- and C-terminal residues within the ACTH sequence

fractions, F1 -F6, and for some of the fragments generated by enzymic cleavage. A partial C-terminal sequence was deter- mined from the kinetic analysis of the carboxy-peptidase Y digestion of whole fraction or of a C-terminal fragment gener- ated from one of the enzymic cleavage reactions. The overlap- ping fragments align with the sequence of porcine ACTH. The amino acid composition data were consistent with this alignment, and this was confirmed by the sequence analysis of the intact peptide fractions and of their fragments. In sum- mary this data reveals that fractions F1, F3, F5 and F6 corre- spond to fragments of porcine ACTH. F2 and F4 represent fragments of porcine ACTH, but each has the single substi- tution at position 8, of a threonine for an arginine residue. Furthermore. synthetic porcine ACTH was also subjected to the same series of enzymic digestions (see miniprint sup- plement) and the pattern of peptide mapping obtained is con- sistent with the assigned primary structures of F1 -F6 shown in Fig. 4.

Radioimmunoassuy o f rpHPLC analysis of PLF-IID and whole pituitary cwract

The antiserum employed in the radioimmunoassay ex- hibits full cross-reaction with all six of the variants (Fig. 5 ) . The radioimmunoassay was able to identify six major peaks of ACTH immunoreactivity following separation of PLF-IID on an analytical CI8 rpHPLC column (Fig. 6A). These six peaks appear to represent the six fractions purified from PLF- IID, based on comparison of elution times of the individual fractions analysed under identical conditions. In addition, it was possible to find, following identical analysis of a crude extract of pituitary gland, peaks of ACTH immunoreactivity with elution times corresponding to those of the six purified ATCH variants (Fig. 6B).

-1 JACTH (1-131, aMSH 1.0-

0.8-

0.6-

0.2-

0-

+ ACTH (1-31) A[Thr 81 ACTH (1-31) 0 ACTH ( 7 - 3 6 )

0.001 0.01 0:1 l:o i o loo iooo compound (pmolltube)

Fig. 5. Displacement curves generated under the rudioimmunoassay conditions (see text) by porcine ACTH (pACTH) and the various A C T H fragments shown each separately incubated at serial dilution with the A C T H antiserum. H8, and radioiodinated human ACTH. Hu- man and porcine ACTH differ at only one residue: human ACTH, Ser31 ; porcine ACTH, Leu31. aMSH, a-melanocyte-stimulating hor- mone; CLIP, corticotropin-like intermediate lobe peptide

A F1 F2 F 3 F4 F5 F6pACTH

200 T T T T T T T 100 n

Time (min)

Fig. 6. The profile of ACTH-like immunoreactivity qf a sample of PLF- IID ( A ) and a sample of porcine pituitary gland extract ( B ) resolved under identical conditions, on rpHPLC (see text for conditions). The arrows indicate the elution times of the ACTH-like fractions F1 - F6 and of porcine ACTH chromatographed under identical conditions. (-- --) Gradient of acetonitrile employed

ACTH, whereas ACTH(1- 37) had between 30% and 40% of ACTH activity. The substitution by a threonine for an arginine at position 8 has a profound effect on biological activity. This is shown by [Thr8]ACTH(1-31) having an ac- tivity one order of magnitude lower than ACTH(1- 31).

Biological activity of purif icd,fractions FI - F4

Fig. 7 shows that all four variants tested were less active than ACTH(1- 39) in both bioassays. Their relative potencies were similar in both systems. The peptides lacking the N- terminal sequence of ACTH exhibited 2% of the activity of

23 1

sl€floIDooEIssIs

A

LPOLYSIS

1 10 100 1000 peptide concentration [nM ]

Fig. 7. Steroidogenic andlipolytic activity of ACTH variants. Compari- son of the mean ECS0 values (n = 6) for corticosterone-releasing ac- tivity of four of the ACTH variants compared to porcine ACTH(1- 39) (A). The relative values are expressed as a percentage of the steroidogenic activity of porcine ACTH(1- 39). Similar relative po- tencies were found for lipolytic activity and the results obtained for three of the peptides (B)

DNA analysis

Positive hybridisation was always observed with the wild- type probe, P1, which codes for ACTH(5 - 10). However, in none of the 100 DNA samples tested was a positive hybridis- ation signal obtained with any of the mutant probes, P2 - P5. Each of these contain a codon for threonine at the position corresponding to 8 in ACTH(5 - 10). A sample of these results is shown in Fig. 8. This result practically eliminates the possi- bility of a point mutation in the proopiomelanocortin gene of individual porcine pituitary glands.

In experiments with the mixed probes designed to detect a hypothetical second proopiomelanocortin gene, no positive hybridisation was obtained, but, surprisingly, the wild-type specific sequence also failed to give a positive signal in these experiments; so clearly the signal detection level was too low in these experiments. This would be overcome by amplifi- cation of a pertinent segment of the hypothetical second pro- opiomelanocortin gene. However such extensive studies were beyond the scope of these current investigations.

DISCUSSION

The primary structure of the six peptide fractions (Fig. 4) isolated from the porcine pituitary fraction PLF-IID suggests

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2

P1

P2

P3

w

P5

Fig. 8. Hybridisation of wild-type-specific (PI) and mutant-specific (P2- P5) oligonucleotide probes with the amplified DNA obtained from 12 individual porcine pituitary glands

that they may arise during proteolytic posttranslational pro- cessing of ACTH. However, the amino acid sequences of two of these each contain a single amino acid substitution, and represent novel ACTH-like structures. It is well established that in the pituitary intermediate-lobe ACTH undergoes pro- teolytic cleavage to generate a-melanocyte-stimulating hor- mone and corticotropin-like intermediate lobe peptide. The results of this study indicate that ACTH may undergo more extensive and specific, differential proteolytic processing. The finding of ACTH(1- 38) confirms the finding of Ekman et al. [18] who, in addition, isolated ACTH(1- 37) and ACTH(7 - 39). We did not identify ACTH(1- 37) and ACTH(7 - 39), but found five novel ACTH fragments. Those differences may be due to Ekman’s group and ourselves having started with different partially purified subfractions of porcine pituitary extracts. Alternatively, different strains of pig may exhibit differences in the pattern of processing of ACTH.

Evidence was obtained for the presence of each of the six variants in significant amounts in a crude extract of whole pituitary glands. This was achieved by combining analytical rpHPLC with the ACTH radioimmunoassay which exhibited full cross-reaction with each of the variants. Peaks of ACTH immunoreactivity were found which had elution times corre- sponding to those of the purified variants run separately under identical conditions. Therefore the proportion of total ACTH immunoreactivity represented by the variants is significant and suggests that the degree of processing of ACTH into these variant forms is similarly significant. However the precise origin of the variants is unclear at this stage. If fractions F1, F3, F5 and F6 are derived from ACTH(1- 39) during the posttranslational processing of proopiomelanocortin, then novel biosynthetic proteolysis must be involved.

The isolation of the two novel threonine-containing ACTH analogues, [Thr8]ACTH(7 - 38) and [Thr8]ACTH(1- 31) is especially interesting. They may both arise from the same precursor, which is quite distinct from proopiomelanocortin, but which contains the ACTH-like sequence. Gene duplication may have given rise to such homology, but there is no direct evidence for a second porcine proopiomelanocortin gene. However, the duplicated gene may have undergone sufficient divergence to have so far escaped detection. To further eluci- date this, investigations outside the scope of these studies are required. Conversely these variants are derived from a mutated form of the proopiomelanocortin gene or from a mutant allele present in a subpopulation of the pigs used here.

232

Boileau and coworkers [19] also considered this possibility having found that in vitro translation of mRNA from porcine pituitary glands generates two forms of proopiomelanocortin. The results of our DNA analyses suggest, however that the threonine-containing fragments have not arisen from a mutant proopiomelanocortin gene.

The finding of a variant ACTH primary structure must be considered unexpected given the remarkable conservation of the ACTH sequence through evolution. The sequences of pro- opiomelanocortin from six mammalian species, ox [20], pig [14,21], rat [22], mouse [23], monkey [24] and human [25], and two non-mammalian species, Xenopus [26] and salmon [27], have now been determined by sequencing cloned cDNA de- rived from mRNA extracted from the pituitary glands of these species. The sequence ACTH(2- 12) is identical across all these species including the two non-mammalian species and ACTH(1- 24) is conserved in all six mammalian species with relatively minor amino acid substitutions in the Xenopus and salmon sequences. The putative substitution of Arg8 by thre- onine, as represented by the peptides F2 and F4, lies within the sequence Glu-His-Phe-Arg-Trp [ACTH(5 - 9)] which oc- curs at two further sites within proopiomelanocortin as well as within the ACTH sequence. Such duplication together with its evolutionary conservation can be considered evidence that this motif possesses a particular biological relevance.

Schwyzer (281 attempted to delineate the functional do- mains of the ACTH primary structure and assigned this pentapeptide sequence as the site of the hormone’s intrinsic activity. Thus, i heoretically, this structure interacts directly with the ACTH receptor on the steroid-producing cells of the adrenal cortex. ‘The substitution of Arg by Thr at position 8 should have a profound effect on the biological activity of ACTH. This must be highly likely given the fidelity of the primary Structure of the amino terminal half of ACTH through evolution and the importance of ACTH(5 -9) for intrinsic activity. The results of our biological testing of frac- tions confirms this.

Measurement of the steroidogenic and lipolytic activities of four of the isolated fractions revealed that they were all less potent in these two systems than ACTH(1-39). The importance of the N-terminal sequence of ACTH for its ster- oidogenic and lipolytic activities is apparent from the relative potencies of those fractions with an intact N-terminus and those fractions with an N-terminal sequence commencing with the residue 7 of ACTH. The substitution of Arg by Thr has an even more profound effect on biological activity.

Further pharmacological testing is necessary in order to determine if any of the ACTH fragments possess separate non-adrenocortrcotropic activities. The biological relevance of the differential processing of ACTH described here may be analogous to what occurs in the pituitary intermediate lobe where proteolytic processing of ACTH generates two peptides which have biological properties quite separate from those of ACTH.

If the threonine-containing analogues represent forms of a mutant ACTH sequence, this may be of pathophysiological significance. This would be analogous to examples in the human population where single-amino-acid substitutions have been found to be responsible for certain disease states. A well-documented example in endocrinology is the occur- rence in some individuals of single-amino-acid substitutions in insulin giving rise to diabetes [29].

The authors ;ire grateful to Dr Schleyer of the University of Ulm, Germany for providing the starting material and also to Dr

Schrezenmayer of the University of Ulm for performing the bioassays. The work was supported by Deutsche Forscliungsgemeinschaft (grant numbers Vo 253/5-I, Se 168/19-1).

REFERENCES 1. Mains, R. E., Eipper, B. A. & Ling, N. (1977) Proc. Yatl Acad.

2. Roberts, J . L. & Herbert, E. (1977) Proc. Nut1 Acad. Sci. USA

3. Chretien, M., Benjannet, S., Gossard, F., Gianoulakis, C., Crine, P., Lis, M. & Seidah, N. G. (2979) Can. J . Biochem. 57, 11 I1 - 1121.

4. Roberts, J. L.. Phillips, M., Rosa, P. A. & Herbert. E. (1978) Biochemistry 17, 3609 - 361 8.

5. Mains, R. E. & Eipper, B. A. (1981) J . B id . Cliern. 256, 5683- 5688.

6. Scott, A. P.. Ratcliffe, J. G., Rees, L. H., Landon, J.. Bennett, H. P. J.. Lowry, P. J. & MeMartin. C. (1973) A’ature New Biol.

7. Brubaker, P. L., Bennett, H. P. I., Baird, A. C. & Solomon, S. (1980) Biochem. Biophys. Res. C‘omrnun. 96, 2441 - 1448.

8. Ekman, R., Noren, H., Hakanson, R . & Tornvall, H. (1984) Regul. Pept. 8, 305-314.

9. Li, C. H., Chung, D., Yamashiro, D. & Lee, C. Y. (1978) Proc. Nut1 Acad. Sci. USA 75,4306 - 4309.

10. Li, C. H., Oosthuizen, M. M. J. & Chung, D. (1988) lnt. J . Pepf. Protein Rex 32, 573 - 578.

1 1 . Schleyer, M., Straub, K., Faulhaber, J . D. & Pfeiffer, E. F. (1969) Horm. Metuh. Res. I , 286-2289,

12. Wittmann-Liebold, B., Kimura, M. & Walker, J. M. (eds) (1984) Methods in molecular hiolog-v, Humana Press, Clifton.

13. Fehm, H. L., Voigt, K. H., Lang, R., Ozyol. B. & Pfeiffer, E. F. (2973) FEBS Lett. 16, 364-371.

14. Gossard, F. J., Chang, A. C. Y. & Cohen. S. N. (1986) Biochim. Biophys. Acta 866, 68 - 74.

15. Mullis, K. B. & Faloona, F. A. (1987) Methods Enzyrnol. 155,

16. Petrukhin, K. E., Grishin, A. V. , Arsnyan, S. G., Brade, N. E., Grinkevich, V. A,, Fillippova. L. Y., Severtsova, J . V. & Modyanov, N. N. (1985) Bioorg. Khin7. 11, 1636-1641.

17. Seliger, H., Ballas, K., Herold. A., Kotschi, U., Lyons, J., Eisenbeiss, F., Sinha, N . D. & Talwar. G . P. (1986) Chern. Sci.

18. Ekman, R., Noren, H., Hakanson. R. & Tornvall. H. (1984)

19. Boileau, G., Gossard, F., Seidah, N. G. & Chretien, M. (1983)

20. Nakanishi, S., Inoue, A., Kita, T., Nakamura, M., Chang, A. C.

21. Boileau, G., Barbeau, C., Jeanotte, L., Chriticn, M. & Drouin,

22. Drouin, J., Chamberland, M., Charron. J., Jeanotte, L. & Nemer,

23. Notake, M., Tobimatsu, I., Watanabe, I., Takanashi, H.,

24. Patel, P. D., Sherman, T. G. &Watson, S. J. (1988) DNA 7,627-

25. Takahashi, H., Teranishi, Y., Nakanishi, S. & Numa, S. (2981)

26. Martens, G. J. M. (1986) Nucleic Acid. Res. 14, 3791 -3798. 27. Kawauchi, M. (1983) Arch. Biochem. Biophys. 227, 343-350. 28. Schwyzer, R. (1980) Trends Pharmacol. Sci. 1, 327-331. 29. Shoeison, S., Haneda, M., Blix, P., Nanjo, A., Sanke, T., Inouye,

K., Steiner, D., Rubenstein. A. & Tager, H. (1983) Nature 302,

Sci. USA 74, 3014-3018.

74,4826-4830.

244,65 - 67.

335 - 350.

6, 569 - 577.

Regul. Pept. 8, 305-314.

Can. J . Biochem. Cell Biol. 61, 333 - 339.

Y., Cohen, S. N. & Numa, S. (1979) Nature 278, 423-427.

J. (1983) Nucleic Acids Rex 11, 8063-8071.

M. (1985) FEBS Lett. 193, 54-58.

Mischina, M. & Numa, S. (1983) FEBS Lett. 156, 67-71.

635.

FEBS Lett. 135, 97- 102.

540 - 543. 30. Geiger, T. & Clarke, S. (1987) .1. Bid . (‘heni. 262, 785-794. 31. Aswad, D. W. &Johnson, B. A. (1987) Trends Biochem. Sci. 12,

32. Schenkein, I., Levy, M., Franklin, E. C. & Frangione, B. (1977) 155 - 158.

Arch. Biochem. Biophju. 182, 64 -70.

233

Supplementary material to

Isolation and full structural characterisation of six adrenocorticotropin-like peptides from porcine pituitary gland Identification of three novel fragments of adrenocorticotropin and of two forms of a novel adrenoeorticotropin-like peptide

Karlheinz VOlGT ', Wolf STEGMAIER', Gerard Patrick MCGREGOR', Hannelore ROSCH' and Hartmut SELIGER'

' Institute of Physiological, University of Marburg, Federal Republic of Germany * Sektion Polymere, University of Ulm, Federal Republic of Germany

Figs 9- 15 show in summary the results of the peptide mapping of each of the fractions FI-F6 and of porcine ACTH. For each fraction, the corresponding truncated sequence of ACTH is given. The results of the partial N-terminal sequencing and for F5 and F6 of the partial C-terminal sequencing are given immediately below the assigned primary sequences.

N-terminal sequence analysis of F l - F6

Table 2 shows the N-terminal sequences of F1 - F6 as determined using the manual sequencing method. The sequences are shown aligned to the N-terminal sequcnce of porcine ACTH. There are two ambiguities in the analyses of F5 which are clearly due to cross contamination between successive steps. Failure to identify the amino acid residue occurred twice. Comparison with the N-terminal se- quence of porcine ACTH reveals that F3 and F4 share the identical N-terminus with porcine ACTH but with the single doubt about the identity of the first residue in F4. The N-terminal sequences of F1, F2, F5 and F6 can be aligned to the porcine ACTH sequence beginning at residue 7. The single anomaly is the Thr as residue 2 in the F2 sequence. All these analyses are consistent with the results of the amino acid composition determinations

Tryptic digestion and partial sequencing

Partial N-terminal sequence analysis of the tryptic fragments of peptide fractions F1- F6, provided further extensive sequence information. The analysis of fragment T6 from the digest of fraction 2 revealed a threonine as the second residue, corresponding to position 8 (Arg) in porcine ACTH. Similarly, a threonine was identified in the corresponding position in fraction 4 following N-terminal analysis of tryptic fragment, T5. The tryptic digestion of all six peptide fractions (FI - F6) and of porcine ACTH generated the sequence correspond- ing to ACTH(17-21) and a peptide fragment, differing in length between the different fractions, but with an identical N-terminal sequencing commencing with the valine at position 22 in porcine ACTH.

Digestion with Staphylococcus protease V8

This digestion was carried out at pH 4.0. This directed the enzy- matic cleavage to the carboxy side of glutamic acid residues and no cleavage occurred after aspartic acid. However the digestions were not always complete, such as at Glu28 and Glu38 in porcine ACTH. The single residue, leucine, was generated from F1- F4 indicating this as the C-terminal residue of these fractions following the glutamic acid corresponding to position 30 in porcine ACTH. From each of the digests of porcine ACTH, F3-F6 up to five fragments were isolated, differing markedly in retention time but possessing identical amino acid compositions. All these fragments include the sequence ACTH(22 -27). The probable reason for the heterogeneity is that the sequence contains the dipeptide Asn-Gly. It has been well documented

that such sequences readily undergo deamidation then isomerisation and racemization due to an G( to p rearrangement at the Asn residue [30, 311. These alterations would change rpHPLC retention times without effecting the results of amino acid analysis [30].

Mouse submaxillary pro tease digestion

The intact fraction of F1 and some of the fragments generated by the Staphylococcus V8 digestion, the SP2 fragments from F2 - F4, SP4 fragment of F5 and SP3 fragment of F6, werc treated with the mouse submaxillary protease. The fragments were separated on rpHPLC and their amino acid compositions were determined. This revealed that the fragments overlapped with the fragments generated by trypsin and the V8 protease. Cleavage occurred at the C-terminal side of arginines, but it is clear that cleavage also occurred at lysine residucs [32].

Proline-specific endopeptidase digestion

The preceding enzymic digestions allowed the elucidation of a long stretch of the N-terminal sequences of Fl-F6. To begin to elucidate the C-terminal sequences, porcine ACTH, F5 and the follow- ing tryptic fragments, the respective T5 fragments of F1 -F4 and F6, were treated with proline-specific endopeptidase. The fragments so generated were separated by rpHPLC, and their amino acid compo- sition determined and partially sequenced. In each case the sequence corresponding to ACTH(22 - 24) was generated and sequenced. A second fragment was also generated in each case which shared an identical N-terminus commencing with an Asn corresponding to resi- due 25 of porcine ACTH. The complete sequence of the second frag- ment was elucidated for F1- F4 thus completing the sequence deter- minations of these fractions.

Carboxypeptidase Y digestion

In order to sequence the very C-terminal portion of F5 and F6 both peptides were treated with carboxypeptidase Y. The kinetics of these reactions is shown in Fig. 16. The curves show the change in the concentration of amino acids generated during enzymic digestions. The closer the residue is to the terminal position, the more rapidly it is generated, and so this kinetic data indicates the amino acid sequence. According to these kinetic data, F5 terminates with thc sequence Glu- Ala-Phe (Fig. 16A). However, proline was proven to be the terminal residue by derivdtising the hydrolysates of SPZ and SP3 with phenyl isothiocyanate. The resulting phenyl isothiocyanate derivatives were separated by means of HPLC and proline was detected qualitatively. Additional proof for C-terminal proline comes from the coelution of PSE6 of F5 and PSE2 of F6 (data not shown). The position of proline in PSE2 of F6 was determined from the sequence analysis. The kinetics of the carboxypeptidase Y digestion of F6 indicates the C-terminus of F6 to be Leu-Glu (Fig. 16B).

234

12 1 4 T3 T5 P h e - 9 i ig-GlJ-~-Pro- '%GI Lys ~ - $ - P r o - V a l - L y s V a l - I r Pro-Asn-GI

77 7 1-, -4-e 311 - Pro 4-

..

Glu SP!

MSL 31

MS2 171110 MS3 31 i9 I, MSI 15

Fig. S 9. Summary ofresults ofpeptide mapping ofpeptide FI [ACTH( 7- 31)] . The half-arrows indicate the sequential digestion of the terminal sequence analysis. TP, total peptide. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the letters indicate the enzyme responsible as follows: T, trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline- specific endopeptidase. The numbers indicate the order of elution of the fragments on rpHPLC (data not shown)

7 10 20 30 Phe-lhr- lrp-Gly-Lys-Pro- i V d Gly i ~ L y s ~ L y s ~ Arg- Arg-Pro-Val -Lys-Val - lyr-Pro-Asn-Gly - Alo- Glu- As-Glu-Leu

16 I3 15 Phe-Thi-'~-Gly-Lys-Pro (Val. Gly I-Lys Arg-Aig-Pro-Vol-LyS Val-lyr-Pio-Asn-Gly-Alo 31 I 7 7 T h r 7 7 7 - - 7 7 - 7 7 7 7 7

T L 15 10 T I 211

tI5 301 MS3

MS 2 30

M51 30 )I8

Fig. S 10. Summary o f resulth ofpeptide mapping ofpeptide F2 {[ThrS]ACTH(7 -31)}. The half-arrows indicate the sequential digestion of the terminal sequence analysis. TP, total peptide. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the letters indicate the enzyme responsible as follows: T, trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline-specific endopeptidase. The numbers indicate the order of elution of the fragments on rpHPLC (data not shown)

20 Se! - l y r - Ser - Met-Glu - HIS- Phe- Arg- Trp -GI;- Lys-Pro - Val - Gly -Lys-Lys- Arg- AT-Pro - Val -Lys-Val- lyr -Pro -L\sn-Gly - Ala - Glu- Asp-G?%Leu

Ser- Iyi-Ser-Met-Glu-His TP 311 7 7 7 7 7 7

t ! 13 12 1 5 1 T5* 15 Arg I rp-Gly-Ly~-Pro-V~I-Gly-Ly~ Ly5-Arg-Arg-Pro-Vol-Ly. Vol-Tyr-Pro 31

7 7 7 7 7 7 7 7 7 7 7 7 7

T I ~ - A ~ - P + - V a I - L y s

Glu-Leu SPL

SPI Ser-x-Ser-Met -Glu l 6 - 7

151 F A r g 1' MS2 MS 3 30 P

8 PSE I PSE2 9 - T z - P r o AZ-G~ - % -Gb - 9 - C k - L e u

13 PSE3

Fig. S 11. Summary of results of peptide mapping of peptide F3 [ACTH(I -31) ] . The half-arrows indicate the sequential digestion of the terminal sequence analysis. TP, total peptide. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the lctters indicate the enzyme responsible as follows. T, trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline-specific endopeptidase. The numbers indicate the order of elution of the fragments on rpHPLC (data not shown)

235 I I 20 30

S e r - T y r - S e r - M ~ I - G l u - H l s - P h e - T h r - T r p - G l ~ - L y ~ - P ~ ~ - ~ G l y , V ~ l ~ - L y ~ - ~ ~ - A ~ g - A ~ g - P ~ ~ - V ~ l - L y ~ - ~ ~ l - Tyr-Pio-Aan-Gly- Alo-Glu- Asp-Glu-Leu

X -Y-Ser-Mel-Clu-His TP 31 7 7 7 7

I'

t l

T5 T2 7(Thr-Trp-Gly- X -Pro- X - X -Lys L 5 Ar Ar Pro-Vol-Lys Val-Tyr-Pro

TB/ TI0

T6 T7 311 -7- 7 4-4-4-7 777

T1 II 151 s-s-2-va1-Lys

T9 111112 74 15 I,

Glu SP3 SPI

Ser-Tyr-Ser-Mel-Uu l6 7 7 7 7

16 MSI 919 us4 17,118 MS5 304

MS7 161117 MS6 30 [

us3 MS2 301 I

I ,

t6 II

19 v2 +-+-&!$ I

PSE2 Ilp(Gly. Lysl-Pro

PSEI* PSE2* &-Tyr-Pre 7 As~G~-AI . -GIu-AG(u ' I~u 77 7

m Fig. S 12. Summary of results ofpeptide mapping ofpeptide F4 {[Thr8]ACTH(1- 31)). The half-arrows indicate the sequential digestion of the terminal sequence analysis. TP, total peptide. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the letters indicate the enzyme responsible as follows. T, trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline-specific endopeptidase. The numbers indicate the order of elution of the fragments on rpHPLC (data not shown)

36 P h i - Arg-Trp- GI;- Lys- Pro-Val - Gly- Lys-Lys- Arg- Arg -Pro - V;?. Lys -Val - Tyr -Pro - Am-Gly - Ala-Glu - AS+-G?-L~U-A~~-GIU - Ala- Re- Pro

g!l-glp-F%&-PlO P h - ~ - r l p - ~ - ~ - P r o - V o l - ~ - X -Lys-,s TP 7 7 7

T1 36 I T3 72 T7* %-A2 T ~ - ~ - ~ - ~ - l G l y . V O l i - L y s Ar Ar Pro Val Lys Val-Tyr-Pro-Am-Gly-Alo-Glu 4-4-7- - 7 7 7 7 7 7 7

SP3

17 30{ L+-$-Gl$Ala.Phe. Pro1 SPL/SP5/SP6

SP1 % iPhe.Pral I lo SP2 30

i9 US2 171 118 MS3 30 I1

l9 MSI 15

17 PSE5 121 ~ - ~ - " " - G I ~ - A S ~ - G I ~ - ~ - ~ - ~ ~ ~ PSE7

vo PsE3 2 L 1 127 PSE6 36

Fig. S 13. Summary of results of peptide mapping of peptide FS (ACTH(7-36)]. The half-arrows indicate the sequential digestion of the terminal sequence analysis. TP, total peptide. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the letters indicate the enzyme responsible as follows. T, trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline-specific endopeptidase. The numbers indicate the order of elution of the fragments on rpHPLC (data not shown)

P h i - A r g - i r p - G I ~ - L y s - P r o - V o l - G l y - L y s - L y s - A r g - A r g - P l o - ~ ~ - L y s - V o l - T y ~ - P ~ ~ - A ~ ~ - G l ~ ~ A l ~ - G l ~ - A ~ ~ - ~ ~ - L ~ ~ - A I o - G l u - A l a - P h e - P r a - L ~ ~ - ~ ~

Leu Glu Phe-Arg-Trp-Gly TP 7 - 7 - 7 7

T5* 381 T3 T2

T ~ P - G ~ Y - L Y ~ - P ~ ~ - ~ - G I Y - L Y ~ L Y ~ - % - ~ - P ~ ~ - V O I - L ~ ~ V.,I-T~~-P~~- x -tiy-~io - 7 7 7 P l 0 7 7 -7 7 7 7 7 7

T1 Arg-Arg-Pro-Vol-Lys 7 7 7

i7 SPI SP2

SP3/SP4/SP5/SP6/SP7 3 o I ~ - A l o - G l u A + - ~ - P P l G l u . L e u l

19 MS1 15, 117 MSL 30 I I

I t0 us3 30

120 PSE3 2L1125 PSE7 39 I

122 PSEl 241 125 PSEWPSE2' 361

Vol-Tyr-Pro Am-Gly-Ah-Glu-As Glu 30 PSEI* PSEL*

7 7 7 7 7 4- 127 PSE3' 361

Fig. S 1 4 . Summary of results of peptide mapping of peptide F6 [ACTH(7-38)]. The half-arrows indicate the sequential digestion of the terminal sequence analysis. TP, total peptide. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the letters indicate the enzyme responsible as follows. T, trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline-specific endopeptidase. The numbers indicate the order of elution of the fragments on rpHPLC (data not shown)

236

I' 'bu 2 2 Th2 33,

Fig. S 15. Sunnnurj' of resiilts ofpeptide mapping of standard porcine ACTH. The half-arrows indicatc the sequential digestion of the terminal sequence analysis. The fragments generated by the various peptidase digestions are aligned to the ACTH sequence and the letters indicate the enzyme responsible as follows: T. trypsin; SP, Staphylococcus protease V8; MS, mouse submaxillary protease; PSE, proline-specific endopcptidase. The iitnnbers indicate the order of elution of the fragments on rpHPLC (data not shown)

0 " . . r I , , , 0 5 10 15 20 30 40 50 60

t e

-0 5 10 15 30 40 60 Time (min)

Fig. S 16. Kinetics o f t l w cnrho.Kyyeptidase Y treatment ofpeptide fractions FS ( A ) and F6 ( B )

Table 2 Amino-tr rmrriul sequence determination of fractions F1 -M A postulated ACTH variant is assigned to each fraction and for comparison the N-terminal sequence of ACTH is given

Frdction Proposed peptide N-terminal sequence

F1 ACTH(7- 31) Phe- Arg-Trp-Gly F2 [l'hr8]ACTH(7 - 31) Phe-Thr-Trp F ? ACTH(1 -31) Scr-Tyr-Ser-Met-Glu-His F 4 [l hrg]ACTH(Z -31) Xaa-Tyr-Ser-Met-Glu-His 1- 5 AC1 H(7 - 36) Phe-Arg-Trpi Arg-Gly/Trp-Lys-Pro-Val-Gly-Xaa- Lys-Arg r6 AC1 H(7 - 38) Phe-Arg-Trp-Gly

ACT H Ser-Tyr-Ser-Met-CTlu-His-Phc-Arg-Trp-Gly-Lys-Pro-Val-Gly-Lys-Lys