- Email: [email protected]
Isolation and characterization of corticostatic peptides from guinea pig bone marrow
Vol. 180, No. 2, 1991 October 31, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCHCOMMUNICATIONS Pages 558-565
ISOLATION AND CHARACTERIZATION OF CORTICOSTATIC PEPTIDES FROM GUINEA PIG BONE MARROW Jing Hu*, Hugh P.J. Bennett*, Claude Lazure¶ and Samuel Solomon*# * Endocrine Laboratory, Royal Victoria Hospital, Departments of Medicine, Obstetrics and Gynecology, McGill University, Montreal, Quebec, Canada H3A IA1 ¶ Institut de recherches cliniques de Montr6al, Montreal, Quebec, Canada H2W 1R7 Received August 21, 1991
Summary Three corticostatic peptides (GP-CS1, GP-CS2 and GP-CS3) were purified from extracts of guinea pig bone marrow. Each was identified on the basis of their ability to inhibit the secretion of corticosterone by isolated rat adrenal cells stimulated by ACTH. GP-CS1 and GP-CS2 were found to be 31 residues in length, rich in arginine and to have six cysteines typical of the corticostatin/defensin family of peptides previously purified from phagocytic cells of the immune system. GP-CS1 was found to be identical to GP-CS2 except for having a leucine at position 21 instead of isoleucine. GP-CS3 was also found to be rich in arginine and cysteine but structurally distinct from the other peptides. A combination of endoprotease mapping, ion-spray mass spectrometry and gas-phase sequencing revealed that GP-CS3 was a novel homo-dimer consisting of two 13 amino acid residue subunits cross-linked through eight cysteines in an anti-parellel configuration, o 1991 A c a d e m i c Press, I n c .
The corticostatic peptides and defensins belong to a family of cysteine-rich, cationic peptides of low molecular weight which have been recently purified from rabbit lung (1) and from cells of the immune system of the rabbit, human (2,3) and rat (4). The basic amino acids and the cysteines in these peptides are highly conserved. Most members of this family have been found to have antimicrobial activity by a non-oxygen dependant mechanism (5). Other members of the family have been found to be corticostatic (anti-AC~H) and act by competing for the binding of ACTH to its receptor (6). Rabbit corticostatin 1 (CS1) receptor binding studies indicate that there is a single binding site with a Kd of 2X10-gM and that there are 9000 binding sites/cell (Q.Zhu and S. Solomon, unpublished data). Furthermore ACTHI.z4 could not compete on the rat adrenal cell for the specific binding of CS1. Recently, these corticostatic peptides but not the defensins have been found to induce L-type Ca 2+ channels in jejunal villus enterocytes (7). The corticostatic peptides are highly specific in their action as they do not inhibit angiotensin II stimulated aldosterone production by rat zona glomerulosa cells (1). The only exception is our recent observation that CS1 inhibits a-MSH stimulated aldosterone production in these same cells (J. Hu and S. Solomon, unpublished data). #To whom correspondence should be addressed.
0006-291X/91 $1.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
558
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
It has also been found that an abundant mRNA from mouse jejunum and ileum encodes a 6000
molecularweight peptide named eryptydinwhich has the consensus sequence of six cysteine residues and basic amino acids (8). In preliminary studies with guineapig lungwe noted that there were peptides eluted from a HPLC column that were corticostatic. We then turned to bone marrow to obtain sufficient amounts of the peptides for isolation and sequence analysis. We report here the isolation of GP-CS1, GP-CS2 and GP-CS3 from guinea pig bone marrow. GP-CS1 and GP-CS2 are part of the corticostatin family but GP-CS3 has a unique structure and still retains corticostatic activity.
Materials and Methods
Fifty adult guinea pigs weighing 500 to 750 g were decapitated under ether anesthesia and the femors were removed. Bone marrow from the femors was washed with 0.9% saline and centrifuged at 2000 x g and the cell pellet was resuspended in red blood cell lysis buffer containing 0.15 M ammonium chloride, 0.01M Tris and incubated at 37°C for 40 min. After centrifugation at 2000 x g for 10 minutes, the cell pellet was suspended in the same lysis buffer one more time and extracted by sonication in an acidic medium consisting of 1M HCI/5% formic acid/1% NaCI (wt/vol) 1% trifluoroacetic acid (9). After centrifugation (3300 x g for 15 min), the pellet was reextracted with the same extraction medium and the supernatants rich in peptide were combined for the ODS silica cartridge extraction as previously described (10). Reversed-phase HPLC purification Reversed-phase HPLC purification were carried out as previously described (9). Briefly, the C18 SepPak eluate was applied onto a Waters C1~ laBondapak column and eluted with a linear gradient of 0 to 48% acetonitrile containing 0.1% TFA as counter-ion. The fractions containing corticostatic activity were repurified in the same system using 0.13% HFBA as a counter-ion. The last two HPLC runs were carried out on a Vydac Cls column (Cole Palmer, Chicago, IL) using 0.1% TFA as the counter-ion and a linear gradient as will be described in the legend to the figures. All reversed-phase HPLC was done using a flow rate of 1.5 ml/min. Molecular weight estimation of GP-CS3 was carried out using two Waters 1-125 gel permeation HPLC columns connected in series and eluted isocratically with 40% acetonitrile containing 0.1% TFA at a flow rate of 1 ml/min as described previously (11). Amino Acid Analysis and Microsequencing Aliquots of peptides were lyophilized in borosilicate tubes and hydrolysed in an evacuated reacti-vial for 16 hrs at 105°C with 6N HC1. Amino acid analysis was performed using a model 6300 Analyser (Beckman Instruments, Palo Alto, CA). For microsequencing the derivatized peptides were submitted for analysis with an Applied Biosystem gas-phase sequenator (Model 470A) as previously described (12) but using a sequence program adapted from Speicher (13). The resulting phenylthiohydantoin (PTH) amino acids were analysed by reverse-phase HPLC on the on-line PTHanalyser (Applied Biosystem Model 120A) and/or a stand alone Vydac HPLC unit as described previously (12). The PTH yields for each standard was normalized according to a PTH-Nor Leucine internal standard while the initial and repetitive yields were obtained by linear regression from the yields of selected stable PTH derivatives. To determine the cysteine bridge orientation of GP-CS3, 20 nmoles of the purified peptide was digested using a mixture of 2 I~g chymotrypsin and 2~tg trypsin in 50 nM Tris HC1 buffer (pH 7.5) at 37 ° for 18 h. The fragments obtained were separated by HPLC on a C18 I~ Bondapak column eluted for 1 h using a linear gradient of 0 to 40% aeetonitrile containing 0.1% TFA as the counter-ion. Each fragment was subjected to amino acid analysis. Rat Adrenal Cell Bioassay The dispersed rat adrenal cell bioassay was a modification of the method of Sayers (14) as previously described (1). Prior to use, all media were prewarmed to 37°C in a shaking water bath (Dubnoff incubator, Precision Scientific Co., Chicago, USA) in an atmosphere of 95% 02/5% CO z. A total of 10 to 26 Sprague-Dawley male rats (Charles River Breeding Laboratories) weighing 150 to 250 grams were sacrificed by decapitation, and adrenals were immediately decapsulated
Vol. 180, N o. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
and quartered in Ham's F-12 medium containing 0.5% BSA. The tissue was then incubated for 1 hr in 10 ml of the medium with 2 mg/ml collagenase and 250 Ixg/ml DNase. At the end of the incubation, the dispersal of the cells was completed by pipetting the medium up and down 50 times using a Pasteur pipette. The cell suspension was centrifuged at 150 x g for 7 min, and the cell pellet was resuspended and washed twice with 12 ml of the same medium. Finally, the cell pellet was suspended in 2 ml of the medium and filtered through prewetted nylon gauze (100 I~m) and layered on top of 8 ml of Ham's F-12 medium containing 2.5% BSA prior to gradient centrifugation. After centrifugation at 170 x g for 10 rain, 6 ml of the upper layer was aspirated and the cells in the lower 4 ml were diluted with 10 ml of incubation medium (Ham's F-12 medium containing 0.5% BSA and 7 mM Ca 2+) before determining the cell number. The viability of the cells was determined by the trypan blue exclusion method. The cell suspension was adjusted to a final concentration of 400,000 cells per ml and 0.5 ml was added to the incubation tubes. The cells were preincubated for 60 rain at 37°C on a shaking water bath under an atmosphere of 95% Oz/5% CO 2. Then, a 0.5 ml aliquot of incubation medium or incubation medium containing either synthetic human or synthetic porcine ACTH (ACqlq1.39) alone or ACTH plus test material was added. The mixture was incubated for an additional 2 hr. After incubation the tubes were centrifuged at 120 x g for 10 min, and the supernatant was decanted into borosilicate culture tubes. Steroids were extracted with methylene chloride and corticosterone in the extract were determined by radioimmunoassay as previously described. The radioimmunoassay for corticosterone was carried out according to the instructions supplied by the manufacturer (BioMega, Montreal).
Results
The first HPLC purification step used in the isolation of guinea pig corticostatin is shown in Fig. 1A. There were three or four peaks eluted with corticostatic activity. The first peak (fractions 48-55, Fig. 1A) which was also the most abundant, was further purified on a second HPLC column using 0.13% H F B A as the counter-ion (Fig. IB) and the rat adrenal cell bioassay also indicated one main bioactive peak. This corticostatic material was further purified on a C18 Vydac using 0.1% T F A as the counter-ion (Fig. 1C). The fractions constituting the front of the peak (Fig. 1C) were collected separately from those at the tail of the peak and both were rechromatographed twice more until homogeneous materials were obtained (data not shown). Amino acid analysis (Table 1) of the material from both peaks indicated that two distinct peptides (GP-CS1 and GP-CS2) were present that differed only by two amino acids, leucine and isoleucine. A total of 0.5 nM of the pyridylethylated GP-CS1 and GP-CS2 were submitted for sequence analysis and the sequences obtained are shown in Table 2. The second corticostatic peak of maternal eluted on Fig. 1A (fractions 56-63) was chromatographed on HPLC as shown in Fig. 2A and the bioactive material eluted (fractions 75-76) was rechromatographed on HPLC to give a homogeneous material (Fig. 2B). The amino acid composition of this material revealed a lot of cysteine and arginine as shown in Table 1 and it was named GP-CS3. When 0.5 nM of the pyridylethylated GP-CS3 was submitted for sequence analysis the following sequence was obtained: RRPRCFCRLHCRC Gel permeation HPLC analysis indicated that the mass of GP-CS3 was too high for this peptide to be composed of a monomer of 13 residues (data not shown). From chromatographic evidence it seemed that this peptide could exist as a dimer. Ionspray Mass Speetrocopy performed by Dr. Kcmi~hi of the Bioteehnolom/Research Institute, Montreal indicated a mass of 3403.74 (charge of
Vol. 180, No. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
/
0.3
-80
6O
E "40
'~
-
o.1-
<
-20 0
g
~
25 ° °o
,.I 0
20
.
40
60
Ill
80
100
120
80
0.0 0.2
E
~
40 i
o4
° !1i1
.
0
.
.
20
.
.
J!lJ
J, L_ ,
.
40
00
80
100
C 0.08
40
~o.~
E
~° ~ ~ 20
o 004 0 o
lO I 0
,
, ,~ , 20 40 FRAGTION NUMBER
,Lo eO
Figure IA. Isolation of GP-CSI and GP-CS2. HPLC purification of a guinea pig bone marrow extract. Bone marrow from fifty guinea pigs was extracted as described in the methods. The extract was loaded onto a Waters C18 pBondapak reversed-phase column which was eluted using a linear gradient of 0 to 48% acetonitrile in 0.1% TFA for the first 90 rain and 48 to 80% acetonitrile in the last 30 min. One minute fractions were collected and submitted for bioassay. Figure lB. Fractions 48-55 from Fig. 1A were combined and applied onto the same column as above and eluted with a linear gradient of 0 to 48% acetonitrile in 0.13% HFBA over the first 90 min and 48-65% acetonitrile in the last 10 min. One minute fractions were collected and submitted for bioassay. Figure 1C. Fractions 67-74 from Fig. 1B were loaded onto a Vydac reversed-phase HPLC column and the column was eluted using a linear gradient of 15-30% acetonitrile in 0.1% TFA in water, throughout. Fractions were collected by hand. As can be seen from Fig. 1C the front of the peak (fraction 40) was collected separately and the tail of the peak (fraction 43) was kept separate. Fraction 40 and fraction 43 were chromatographed separately using the same Vydac column and a gradient of 20-27% acetonitrile in 0.1% TFA for 5 to 40 min. Each peak of material was rechromatographed twice in the same system until both were homogeneous. The material in fraction 40 turned out to be GP-CS1 and in fraction 43, GP-CS2.
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Table 1. Amino acid composition of GP Corticostatins after hydrolysis
CORTICOSTATIC
PEPTIDES
AMINO ACID
GP-CSI
GP-CS2
Aspartic acid Asparagine Threonine Glutamic acid Glutamine Proline Glycine Cysteine Valine Isoleucine Leucine Tyrosine Phenylalanine Histidine Arginine
1.2
1.2
4.7 i.i 1.0 1.2 5.4 1.2 2.0 1.2 2.0 2.7 7.4
(0) (i) (5) (0) (i) (i) (i) (6) (i) (2) (i) (2) (3) (7)
4.7 i.i 1.0 1.2 4.7 i.i 2.0 2.1 2.0 2.9 7.2
(0) (i) (5) (0) (i) (i) (i) (6) (i) (i) (2) (2) (3) (7)
GP-CS3
i.i
(i)
3.5
(4)
i.i
(i)
1.1 o.9
(i) (i)
4.8
(5)
* Numbers in parentheses are calculated from sequence analysis data.
6) and 3403.46 (charge of 5) and the molecular weight of the monomer was calculated to be 1702.0986. The calculated mass of the dimer was 3404.1972. The slight discrepancy in mass is due to small alignment problem with the instrument when the sample was run. In addition GP-CS3 was subjected to endoprotease of digestion using a mixture of chymotrypsin and trypsin. The fragments obtained were subjected to reveresed-phase HPLC (data not shown). Three major fragments were obtained which had the following amino acid compositions (molar ratios shown in parenthesis). Fragment 1, Pro (1), Arg (3); Fragment 2, Cys (2), Phe (1) and Fragment 3, Cys (2), Leu (1), His (1), Arg (2). All of this data suggests that GP-CS3 is a dimer having an anti-parallel configuration (Table 2). Table 2. G u i n e a pig c o r t i c o s t a t i n s and their I.D. 50 for the i n h i b i t i o n of A C T H induced c o r t i o o s t e r o n e p r o d u c t i o n in rat adrenal cell s u s p e n s i o n s
Peptide
Sequence
I.D.
50
(nM) CSI
G-I-C-A-C-R-R-R-F-C-P-N-S-E-R-F-S-GY-C-R-V-N-G-A-R-Y-V-R-C-C-S-R-R
27
GP-CSI
R-R-C-I-C-T-T-R-T-C-R-F-P-Y-R-R-L-GT-C-I-F-Q-N-R-V-Y-T-F-C-C
250
GP-CS2
R-R-C-I-C-T-T-R-T-C-R-F-P-Y-R-R-L-GT-C-L-F-Q-N-R-V-Y-T-F-C-C
250
GP-CS3
R-R-P-R-C-F-C-R-L-H-C-R-C C-R-C-H-L-R-C-F-C-R-P-R-R
2000
Vol. 180,
No. 2, 1991
BIOCHEMICAL ANDBIOPHYSICAL RESEARCHCOMMUNICATIONS
A
E
0.2 1 ~ 0.1
0'81
~ 60
40
r~ I.--
~ 0.4
I
0 0
~,f,D ,5 ~ m u.
-
0
0
2,0
4'0
6'0
, 80
0.4 ~ 40
0"081
,a 0.2
o
004
I 20
0 0
2'0
,'o
FRACTION
6'0 NUMBER
o uu
80
Figure 2A. Isolation of GP-CS3. Fractions 56-63 from Fig. 1A were combined and the material was applied onto a Waters reversed-phase pBondapak column which was eluted using a linear gradient of 0-56% acetonitrile in 0.13% HFBA in water over a 100 min period. One minute fractions were collected and submitted to a bioassay. Figure 2B. Corticostatic fractions 75-76 of Fig. 2A were subjected to another reversed-phase HPLC step using a Vydac column eluted using a gradient of 15-30% aceonitrile in 0,1%.TFA. 1.5 ml fractions were taken over an 80 rain period. In Fig. 3 is shown the comparison of the biologic activity of the three guinea pig corticostatins with the most active of the rabbit corticostatins, CS1. It can be seen from this data GPC1 and GPC2 have approximately 1/10 of the activity and GPC3 approximately 1/80 of the activity of CS1. Discussion
We have recently isolated corticostatic peptides from tissues and i m m u n e system cells of the rabbit (1), h u m a n (3) and rat (4).
In this paper we report the isolation of three corticostatic
peptides GP-CS1, GP-CS2 and GP-CS3 from the bone marrow of the guinea pig. GP-CS1 and GPCS2 belong to the family of peptides known as defensins (5), corticostatins (1) and cryptydins (8) but GP-CS3 is a unique 13 amino acid peptide which occurs as dimer as shown in Table 2. The 563
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
g -o o
m
100-
•
-S~-%LL
75" 50
g .:,-E 25"
/
/
I ..4 ,/ f
,f".,~
[
/ .... .....
+*"
+Jy
c
~ - - - . - . 7. .a"
O-
101
102
103
104
CORTICOSTATINS(nM)
Figure 3. A comparison of the corticostatic activity of GP-CS1, GP-CS2, GP-CS3 and rabbit CS1. A total of 33 mM of hACTH was used to stimulate 200,000 cells/ml in the presence of increasing concentrations of rabbit CS1 (solid circles) GP-CS1 (open squares) and GP-CS2 (open circles) and GP-CS3 (solid squares). Corticosterone was measured by RIA. Values are the mean - standard deviation of four separate experiments.
evidence that it is a dimer came from the ion spray mass spectrum. The structure shown in Table 2 indicates that the cysteines in the two chains are coupled to one another.
This is only one
possibility, the other would have four cysteines coupled to form the dimer while the remaining four cysteines formed two internal rings. The anti-ACTH activity of the three guinea pig peptides are considerably lower than the most active rabbit corticostatin, CS1.
The activity of GP-CS3 is
comparable to the human corticostatic peptide, HP4 (3). It is surprising to us that GP-CS3 has corticostatic activity because it lacks the backbone of cysteines and arginines which characterize the corticostatins. GP-CS1 was isolated from guinea pig neutrophils and found to be antimicrobial (16) and GP-CS1 as well as GP-CS2 were isolated from guinea pig neutrophils and found to release histamine from rat mast cells (17). Very recently the genes for GP-CS1 and GP-CS2 have been cloned (18) and found to differ only by two amino acids, one in the pre-pro protein and one in the mature peptide as is shown in Table 2.
It is of interest that GP-CS1 and GP-CS2 are only
biosynthesised in bone marrow cells and not in neutrophils although they are abundent in neutrophils.
GP-CS3 has up until now not been described and its biologic activity in the
antimicrobial or the histamine release assays is not known.
Acknowledgments The authors wish to acknowledge the expert technical assistance of Mariette Houle. This study was supported by a grant (PG2) from the Medical Research Council (MRC) of Canada and operating grants MT-1658 and MT-6733 from the MRC. Jing Hu is supported by a grant from the Fonds pour la Formation de Chercheurs et l'Aide ~ la Recherche, Hugh Bennett and Claude Lazure are supported by a Chercheur-boursier from the Fond de la Recherche en Sant6 du Qudbec. References 1.
Zhu, Q., Hu, J., Mulay, S., Esch, F., Shimasaki, S. and Solomon, S. (1988) Proc. Natl. Acad. Sci. U S A 85, 592-596. 564
Vol. 180, No. 2, 1991
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Zhu, Q., Singh, A., Bateman, A., Esch, F. and Solomon, S. (1987) J. Steroid Biochem. 27, 1017-1022. Singh, A., Bateman, A., Zhu, Q., Shimasaki, S., Esch, F. and Solomon, S. (1988) Biochem. Biophy. Res. Commun. 155, 524-529. Belcourt, D., Bateman, A., Singh, A., Lazure, C., Bennett, H.P.J. and Solomon, S. (1990) The Endocrine Society 72nd Annual Meeting, Atlanta, Georgia, Abs. 1009. Selsted, M.E., Brown, D.M., Delange, R.J. and Lehrer, R.I. (1983) J. Biol. Chem. 258, 14485-14489. Zhu, Q., Bateman, A., Singh, A. and Solomon, S. (1989) Endocr. Res. 15, 129-150. MacLeod, R.J., Hamilton, J.R., Bateman, A., Belcourt, D., Hu, J., Bennett, H.P.J. and Solomon, S. (1991) Proc. Natl. Acad. Sci. USA 88, 552-556. Ouellette, A.J., Greco, R.M., James, M., Frederick, D., Naftilan, J. and Fallon, J.T. (1989) Cell Biol. 108, 1687-1695. Bennett, H.P.J., Browne. C.A. and Solomon. S. (1981) Biochemistry, 20, 4530-4538. Bennett, H.P.J. (1986) J. Chromatography, 359, 383-390. Bennett, H.P.J., Browne, C.A. and Solomon, S. (1983) Anal. Biochem. 128, 121-129. Lazure, C., Saayman, H.S., Naud6, R.J., Oelofoen, W. and Chr6tien, M. (1989) Int. J. Peptide Res., 33, 46-58. Speicher, O.W. (1989) In Techniques in Protein Chemistry (T. Hugh, Ed.), pp. 24-35, Academic Press, San Diego, CA. Sayers, G., Swallow, R.L. and Giordano, N.D. (1971) Endocrinology, 88, 1063-1068. Bennett, H.P.J., Brubaker, P.L., Seger, M.A. and Solomon, S. (1983) J. Biol. Chem. 258, 8108-8112. Selstedt, M.E. and Harwig, S.S.L. (1987) Infect. Immun. 55, 2281-2286. Yamashita, T. and Saito, K. (1989) Infect. Immun. 57, 2405-2409. Magaoka, I., Someya, A., Iwabuehi, K. and Yamashita, T. (1991) FEBS Letters, 280, 287-291.
565
BIOCHEMICAL AND BIOPHYSICAL RESEARCHCOMMUNICATIONS Pages 558-565
ISOLATION AND CHARACTERIZATION OF CORTICOSTATIC PEPTIDES FROM GUINEA PIG BONE MARROW Jing Hu*, Hugh P.J. Bennett*, Claude Lazure¶ and Samuel Solomon*# * Endocrine Laboratory, Royal Victoria Hospital, Departments of Medicine, Obstetrics and Gynecology, McGill University, Montreal, Quebec, Canada H3A IA1 ¶ Institut de recherches cliniques de Montr6al, Montreal, Quebec, Canada H2W 1R7 Received August 21, 1991
Summary Three corticostatic peptides (GP-CS1, GP-CS2 and GP-CS3) were purified from extracts of guinea pig bone marrow. Each was identified on the basis of their ability to inhibit the secretion of corticosterone by isolated rat adrenal cells stimulated by ACTH. GP-CS1 and GP-CS2 were found to be 31 residues in length, rich in arginine and to have six cysteines typical of the corticostatin/defensin family of peptides previously purified from phagocytic cells of the immune system. GP-CS1 was found to be identical to GP-CS2 except for having a leucine at position 21 instead of isoleucine. GP-CS3 was also found to be rich in arginine and cysteine but structurally distinct from the other peptides. A combination of endoprotease mapping, ion-spray mass spectrometry and gas-phase sequencing revealed that GP-CS3 was a novel homo-dimer consisting of two 13 amino acid residue subunits cross-linked through eight cysteines in an anti-parellel configuration, o 1991 A c a d e m i c Press, I n c .
The corticostatic peptides and defensins belong to a family of cysteine-rich, cationic peptides of low molecular weight which have been recently purified from rabbit lung (1) and from cells of the immune system of the rabbit, human (2,3) and rat (4). The basic amino acids and the cysteines in these peptides are highly conserved. Most members of this family have been found to have antimicrobial activity by a non-oxygen dependant mechanism (5). Other members of the family have been found to be corticostatic (anti-AC~H) and act by competing for the binding of ACTH to its receptor (6). Rabbit corticostatin 1 (CS1) receptor binding studies indicate that there is a single binding site with a Kd of 2X10-gM and that there are 9000 binding sites/cell (Q.Zhu and S. Solomon, unpublished data). Furthermore ACTHI.z4 could not compete on the rat adrenal cell for the specific binding of CS1. Recently, these corticostatic peptides but not the defensins have been found to induce L-type Ca 2+ channels in jejunal villus enterocytes (7). The corticostatic peptides are highly specific in their action as they do not inhibit angiotensin II stimulated aldosterone production by rat zona glomerulosa cells (1). The only exception is our recent observation that CS1 inhibits a-MSH stimulated aldosterone production in these same cells (J. Hu and S. Solomon, unpublished data). #To whom correspondence should be addressed.
0006-291X/91 $1.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
558
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
It has also been found that an abundant mRNA from mouse jejunum and ileum encodes a 6000
molecularweight peptide named eryptydinwhich has the consensus sequence of six cysteine residues and basic amino acids (8). In preliminary studies with guineapig lungwe noted that there were peptides eluted from a HPLC column that were corticostatic. We then turned to bone marrow to obtain sufficient amounts of the peptides for isolation and sequence analysis. We report here the isolation of GP-CS1, GP-CS2 and GP-CS3 from guinea pig bone marrow. GP-CS1 and GP-CS2 are part of the corticostatin family but GP-CS3 has a unique structure and still retains corticostatic activity.
Materials and Methods
Fifty adult guinea pigs weighing 500 to 750 g were decapitated under ether anesthesia and the femors were removed. Bone marrow from the femors was washed with 0.9% saline and centrifuged at 2000 x g and the cell pellet was resuspended in red blood cell lysis buffer containing 0.15 M ammonium chloride, 0.01M Tris and incubated at 37°C for 40 min. After centrifugation at 2000 x g for 10 minutes, the cell pellet was suspended in the same lysis buffer one more time and extracted by sonication in an acidic medium consisting of 1M HCI/5% formic acid/1% NaCI (wt/vol) 1% trifluoroacetic acid (9). After centrifugation (3300 x g for 15 min), the pellet was reextracted with the same extraction medium and the supernatants rich in peptide were combined for the ODS silica cartridge extraction as previously described (10). Reversed-phase HPLC purification Reversed-phase HPLC purification were carried out as previously described (9). Briefly, the C18 SepPak eluate was applied onto a Waters C1~ laBondapak column and eluted with a linear gradient of 0 to 48% acetonitrile containing 0.1% TFA as counter-ion. The fractions containing corticostatic activity were repurified in the same system using 0.13% HFBA as a counter-ion. The last two HPLC runs were carried out on a Vydac Cls column (Cole Palmer, Chicago, IL) using 0.1% TFA as the counter-ion and a linear gradient as will be described in the legend to the figures. All reversed-phase HPLC was done using a flow rate of 1.5 ml/min. Molecular weight estimation of GP-CS3 was carried out using two Waters 1-125 gel permeation HPLC columns connected in series and eluted isocratically with 40% acetonitrile containing 0.1% TFA at a flow rate of 1 ml/min as described previously (11). Amino Acid Analysis and Microsequencing Aliquots of peptides were lyophilized in borosilicate tubes and hydrolysed in an evacuated reacti-vial for 16 hrs at 105°C with 6N HC1. Amino acid analysis was performed using a model 6300 Analyser (Beckman Instruments, Palo Alto, CA). For microsequencing the derivatized peptides were submitted for analysis with an Applied Biosystem gas-phase sequenator (Model 470A) as previously described (12) but using a sequence program adapted from Speicher (13). The resulting phenylthiohydantoin (PTH) amino acids were analysed by reverse-phase HPLC on the on-line PTHanalyser (Applied Biosystem Model 120A) and/or a stand alone Vydac HPLC unit as described previously (12). The PTH yields for each standard was normalized according to a PTH-Nor Leucine internal standard while the initial and repetitive yields were obtained by linear regression from the yields of selected stable PTH derivatives. To determine the cysteine bridge orientation of GP-CS3, 20 nmoles of the purified peptide was digested using a mixture of 2 I~g chymotrypsin and 2~tg trypsin in 50 nM Tris HC1 buffer (pH 7.5) at 37 ° for 18 h. The fragments obtained were separated by HPLC on a C18 I~ Bondapak column eluted for 1 h using a linear gradient of 0 to 40% aeetonitrile containing 0.1% TFA as the counter-ion. Each fragment was subjected to amino acid analysis. Rat Adrenal Cell Bioassay The dispersed rat adrenal cell bioassay was a modification of the method of Sayers (14) as previously described (1). Prior to use, all media were prewarmed to 37°C in a shaking water bath (Dubnoff incubator, Precision Scientific Co., Chicago, USA) in an atmosphere of 95% 02/5% CO z. A total of 10 to 26 Sprague-Dawley male rats (Charles River Breeding Laboratories) weighing 150 to 250 grams were sacrificed by decapitation, and adrenals were immediately decapsulated
Vol. 180, N o. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
and quartered in Ham's F-12 medium containing 0.5% BSA. The tissue was then incubated for 1 hr in 10 ml of the medium with 2 mg/ml collagenase and 250 Ixg/ml DNase. At the end of the incubation, the dispersal of the cells was completed by pipetting the medium up and down 50 times using a Pasteur pipette. The cell suspension was centrifuged at 150 x g for 7 min, and the cell pellet was resuspended and washed twice with 12 ml of the same medium. Finally, the cell pellet was suspended in 2 ml of the medium and filtered through prewetted nylon gauze (100 I~m) and layered on top of 8 ml of Ham's F-12 medium containing 2.5% BSA prior to gradient centrifugation. After centrifugation at 170 x g for 10 rain, 6 ml of the upper layer was aspirated and the cells in the lower 4 ml were diluted with 10 ml of incubation medium (Ham's F-12 medium containing 0.5% BSA and 7 mM Ca 2+) before determining the cell number. The viability of the cells was determined by the trypan blue exclusion method. The cell suspension was adjusted to a final concentration of 400,000 cells per ml and 0.5 ml was added to the incubation tubes. The cells were preincubated for 60 rain at 37°C on a shaking water bath under an atmosphere of 95% Oz/5% CO 2. Then, a 0.5 ml aliquot of incubation medium or incubation medium containing either synthetic human or synthetic porcine ACTH (ACqlq1.39) alone or ACTH plus test material was added. The mixture was incubated for an additional 2 hr. After incubation the tubes were centrifuged at 120 x g for 10 min, and the supernatant was decanted into borosilicate culture tubes. Steroids were extracted with methylene chloride and corticosterone in the extract were determined by radioimmunoassay as previously described. The radioimmunoassay for corticosterone was carried out according to the instructions supplied by the manufacturer (BioMega, Montreal).
Results
The first HPLC purification step used in the isolation of guinea pig corticostatin is shown in Fig. 1A. There were three or four peaks eluted with corticostatic activity. The first peak (fractions 48-55, Fig. 1A) which was also the most abundant, was further purified on a second HPLC column using 0.13% H F B A as the counter-ion (Fig. IB) and the rat adrenal cell bioassay also indicated one main bioactive peak. This corticostatic material was further purified on a C18 Vydac using 0.1% T F A as the counter-ion (Fig. 1C). The fractions constituting the front of the peak (Fig. 1C) were collected separately from those at the tail of the peak and both were rechromatographed twice more until homogeneous materials were obtained (data not shown). Amino acid analysis (Table 1) of the material from both peaks indicated that two distinct peptides (GP-CS1 and GP-CS2) were present that differed only by two amino acids, leucine and isoleucine. A total of 0.5 nM of the pyridylethylated GP-CS1 and GP-CS2 were submitted for sequence analysis and the sequences obtained are shown in Table 2. The second corticostatic peak of maternal eluted on Fig. 1A (fractions 56-63) was chromatographed on HPLC as shown in Fig. 2A and the bioactive material eluted (fractions 75-76) was rechromatographed on HPLC to give a homogeneous material (Fig. 2B). The amino acid composition of this material revealed a lot of cysteine and arginine as shown in Table 1 and it was named GP-CS3. When 0.5 nM of the pyridylethylated GP-CS3 was submitted for sequence analysis the following sequence was obtained: RRPRCFCRLHCRC Gel permeation HPLC analysis indicated that the mass of GP-CS3 was too high for this peptide to be composed of a monomer of 13 residues (data not shown). From chromatographic evidence it seemed that this peptide could exist as a dimer. Ionspray Mass Speetrocopy performed by Dr. Kcmi~hi of the Bioteehnolom/Research Institute, Montreal indicated a mass of 3403.74 (charge of
Vol. 180, No. 2, 1 9 9 1
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
/
0.3
-80
6O
E "40
'~
-
o.1-
<
-20 0
g
~
25 ° °o
,.I 0
20
.
40
60
Ill
80
100
120
80
0.0 0.2
E
~
40 i
o4
° !1i1
.
0
.
.
20
.
.
J!lJ
J, L_ ,
.
40
00
80
100
C 0.08
40
~o.~
E
~° ~ ~ 20
o 004 0 o
lO I 0
,
, ,~ , 20 40 FRAGTION NUMBER
,Lo eO
Figure IA. Isolation of GP-CSI and GP-CS2. HPLC purification of a guinea pig bone marrow extract. Bone marrow from fifty guinea pigs was extracted as described in the methods. The extract was loaded onto a Waters C18 pBondapak reversed-phase column which was eluted using a linear gradient of 0 to 48% acetonitrile in 0.1% TFA for the first 90 rain and 48 to 80% acetonitrile in the last 30 min. One minute fractions were collected and submitted for bioassay. Figure lB. Fractions 48-55 from Fig. 1A were combined and applied onto the same column as above and eluted with a linear gradient of 0 to 48% acetonitrile in 0.13% HFBA over the first 90 min and 48-65% acetonitrile in the last 10 min. One minute fractions were collected and submitted for bioassay. Figure 1C. Fractions 67-74 from Fig. 1B were loaded onto a Vydac reversed-phase HPLC column and the column was eluted using a linear gradient of 15-30% acetonitrile in 0.1% TFA in water, throughout. Fractions were collected by hand. As can be seen from Fig. 1C the front of the peak (fraction 40) was collected separately and the tail of the peak (fraction 43) was kept separate. Fraction 40 and fraction 43 were chromatographed separately using the same Vydac column and a gradient of 20-27% acetonitrile in 0.1% TFA for 5 to 40 min. Each peak of material was rechromatographed twice in the same system until both were homogeneous. The material in fraction 40 turned out to be GP-CS1 and in fraction 43, GP-CS2.
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Table 1. Amino acid composition of GP Corticostatins after hydrolysis
CORTICOSTATIC
PEPTIDES
AMINO ACID
GP-CSI
GP-CS2
Aspartic acid Asparagine Threonine Glutamic acid Glutamine Proline Glycine Cysteine Valine Isoleucine Leucine Tyrosine Phenylalanine Histidine Arginine
1.2
1.2
4.7 i.i 1.0 1.2 5.4 1.2 2.0 1.2 2.0 2.7 7.4
(0) (i) (5) (0) (i) (i) (i) (6) (i) (2) (i) (2) (3) (7)
4.7 i.i 1.0 1.2 4.7 i.i 2.0 2.1 2.0 2.9 7.2
(0) (i) (5) (0) (i) (i) (i) (6) (i) (i) (2) (2) (3) (7)
GP-CS3
i.i
(i)
3.5
(4)
i.i
(i)
1.1 o.9
(i) (i)
4.8
(5)
* Numbers in parentheses are calculated from sequence analysis data.
6) and 3403.46 (charge of 5) and the molecular weight of the monomer was calculated to be 1702.0986. The calculated mass of the dimer was 3404.1972. The slight discrepancy in mass is due to small alignment problem with the instrument when the sample was run. In addition GP-CS3 was subjected to endoprotease of digestion using a mixture of chymotrypsin and trypsin. The fragments obtained were subjected to reveresed-phase HPLC (data not shown). Three major fragments were obtained which had the following amino acid compositions (molar ratios shown in parenthesis). Fragment 1, Pro (1), Arg (3); Fragment 2, Cys (2), Phe (1) and Fragment 3, Cys (2), Leu (1), His (1), Arg (2). All of this data suggests that GP-CS3 is a dimer having an anti-parallel configuration (Table 2). Table 2. G u i n e a pig c o r t i c o s t a t i n s and their I.D. 50 for the i n h i b i t i o n of A C T H induced c o r t i o o s t e r o n e p r o d u c t i o n in rat adrenal cell s u s p e n s i o n s
Peptide
Sequence
I.D.
50
(nM) CSI
G-I-C-A-C-R-R-R-F-C-P-N-S-E-R-F-S-GY-C-R-V-N-G-A-R-Y-V-R-C-C-S-R-R
27
GP-CSI
R-R-C-I-C-T-T-R-T-C-R-F-P-Y-R-R-L-GT-C-I-F-Q-N-R-V-Y-T-F-C-C
250
GP-CS2
R-R-C-I-C-T-T-R-T-C-R-F-P-Y-R-R-L-GT-C-L-F-Q-N-R-V-Y-T-F-C-C
250
GP-CS3
R-R-P-R-C-F-C-R-L-H-C-R-C C-R-C-H-L-R-C-F-C-R-P-R-R
2000
Vol. 180,
No. 2, 1991
BIOCHEMICAL ANDBIOPHYSICAL RESEARCHCOMMUNICATIONS
A
E
0.2 1 ~ 0.1
0'81
~ 60
40
r~ I.--
~ 0.4
I
0 0
~,f,D ,5 ~ m u.
-
0
0
2,0
4'0
6'0
, 80
0.4 ~ 40
0"081
,a 0.2
o
004
I 20
0 0
2'0
,'o
FRACTION
6'0 NUMBER
o uu
80
Figure 2A. Isolation of GP-CS3. Fractions 56-63 from Fig. 1A were combined and the material was applied onto a Waters reversed-phase pBondapak column which was eluted using a linear gradient of 0-56% acetonitrile in 0.13% HFBA in water over a 100 min period. One minute fractions were collected and submitted to a bioassay. Figure 2B. Corticostatic fractions 75-76 of Fig. 2A were subjected to another reversed-phase HPLC step using a Vydac column eluted using a gradient of 15-30% aceonitrile in 0,1%.TFA. 1.5 ml fractions were taken over an 80 rain period. In Fig. 3 is shown the comparison of the biologic activity of the three guinea pig corticostatins with the most active of the rabbit corticostatins, CS1. It can be seen from this data GPC1 and GPC2 have approximately 1/10 of the activity and GPC3 approximately 1/80 of the activity of CS1. Discussion
We have recently isolated corticostatic peptides from tissues and i m m u n e system cells of the rabbit (1), h u m a n (3) and rat (4).
In this paper we report the isolation of three corticostatic
peptides GP-CS1, GP-CS2 and GP-CS3 from the bone marrow of the guinea pig. GP-CS1 and GPCS2 belong to the family of peptides known as defensins (5), corticostatins (1) and cryptydins (8) but GP-CS3 is a unique 13 amino acid peptide which occurs as dimer as shown in Table 2. The 563
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
g -o o
m
100-
•
-S~-%LL
75" 50
g .:,-E 25"
/
/
I ..4 ,/ f
,f".,~
[
/ .... .....
+*"
+Jy
c
~ - - - . - . 7. .a"
O-
101
102
103
104
CORTICOSTATINS(nM)
Figure 3. A comparison of the corticostatic activity of GP-CS1, GP-CS2, GP-CS3 and rabbit CS1. A total of 33 mM of hACTH was used to stimulate 200,000 cells/ml in the presence of increasing concentrations of rabbit CS1 (solid circles) GP-CS1 (open squares) and GP-CS2 (open circles) and GP-CS3 (solid squares). Corticosterone was measured by RIA. Values are the mean - standard deviation of four separate experiments.
evidence that it is a dimer came from the ion spray mass spectrum. The structure shown in Table 2 indicates that the cysteines in the two chains are coupled to one another.
This is only one
possibility, the other would have four cysteines coupled to form the dimer while the remaining four cysteines formed two internal rings. The anti-ACTH activity of the three guinea pig peptides are considerably lower than the most active rabbit corticostatin, CS1.
The activity of GP-CS3 is
comparable to the human corticostatic peptide, HP4 (3). It is surprising to us that GP-CS3 has corticostatic activity because it lacks the backbone of cysteines and arginines which characterize the corticostatins. GP-CS1 was isolated from guinea pig neutrophils and found to be antimicrobial (16) and GP-CS1 as well as GP-CS2 were isolated from guinea pig neutrophils and found to release histamine from rat mast cells (17). Very recently the genes for GP-CS1 and GP-CS2 have been cloned (18) and found to differ only by two amino acids, one in the pre-pro protein and one in the mature peptide as is shown in Table 2.
It is of interest that GP-CS1 and GP-CS2 are only
biosynthesised in bone marrow cells and not in neutrophils although they are abundent in neutrophils.
GP-CS3 has up until now not been described and its biologic activity in the
antimicrobial or the histamine release assays is not known.
Acknowledgments The authors wish to acknowledge the expert technical assistance of Mariette Houle. This study was supported by a grant (PG2) from the Medical Research Council (MRC) of Canada and operating grants MT-1658 and MT-6733 from the MRC. Jing Hu is supported by a grant from the Fonds pour la Formation de Chercheurs et l'Aide ~ la Recherche, Hugh Bennett and Claude Lazure are supported by a Chercheur-boursier from the Fond de la Recherche en Sant6 du Qudbec. References 1.
Zhu, Q., Hu, J., Mulay, S., Esch, F., Shimasaki, S. and Solomon, S. (1988) Proc. Natl. Acad. Sci. U S A 85, 592-596. 564
Vol. 180, No. 2, 1991
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Zhu, Q., Singh, A., Bateman, A., Esch, F. and Solomon, S. (1987) J. Steroid Biochem. 27, 1017-1022. Singh, A., Bateman, A., Zhu, Q., Shimasaki, S., Esch, F. and Solomon, S. (1988) Biochem. Biophy. Res. Commun. 155, 524-529. Belcourt, D., Bateman, A., Singh, A., Lazure, C., Bennett, H.P.J. and Solomon, S. (1990) The Endocrine Society 72nd Annual Meeting, Atlanta, Georgia, Abs. 1009. Selsted, M.E., Brown, D.M., Delange, R.J. and Lehrer, R.I. (1983) J. Biol. Chem. 258, 14485-14489. Zhu, Q., Bateman, A., Singh, A. and Solomon, S. (1989) Endocr. Res. 15, 129-150. MacLeod, R.J., Hamilton, J.R., Bateman, A., Belcourt, D., Hu, J., Bennett, H.P.J. and Solomon, S. (1991) Proc. Natl. Acad. Sci. USA 88, 552-556. Ouellette, A.J., Greco, R.M., James, M., Frederick, D., Naftilan, J. and Fallon, J.T. (1989) Cell Biol. 108, 1687-1695. Bennett, H.P.J., Browne. C.A. and Solomon. S. (1981) Biochemistry, 20, 4530-4538. Bennett, H.P.J. (1986) J. Chromatography, 359, 383-390. Bennett, H.P.J., Browne, C.A. and Solomon, S. (1983) Anal. Biochem. 128, 121-129. Lazure, C., Saayman, H.S., Naud6, R.J., Oelofoen, W. and Chr6tien, M. (1989) Int. J. Peptide Res., 33, 46-58. Speicher, O.W. (1989) In Techniques in Protein Chemistry (T. Hugh, Ed.), pp. 24-35, Academic Press, San Diego, CA. Sayers, G., Swallow, R.L. and Giordano, N.D. (1971) Endocrinology, 88, 1063-1068. Bennett, H.P.J., Brubaker, P.L., Seger, M.A. and Solomon, S. (1983) J. Biol. Chem. 258, 8108-8112. Selstedt, M.E. and Harwig, S.S.L. (1987) Infect. Immun. 55, 2281-2286. Yamashita, T. and Saito, K. (1989) Infect. Immun. 57, 2405-2409. Magaoka, I., Someya, A., Iwabuehi, K. and Yamashita, T. (1991) FEBS Letters, 280, 287-291.
565