- Email: [email protected]
Study of the CN1 peptide of P2 protein using Fourier transform infra-red spectroscopy
Study of the CN1 peptide of P2 protein using Fourier transform infra-red spectroscopy B. H. S t u a r t * School of Biologica/ and Chemical Sciences, University of Greenwich, L ondon SE18 6PF, UK
and E. F. M c F a r l a n e Department of Physical Chemistry, University of Sydney, 2006 Australia Received 22 December 1993; revised 8 March 1994 The conformation of the CN1 peptide, derived from the nervous system P2 protein, has been studied in deuterium oxide solution using Fourier transform infra-red (FTIR) spectroscopy. The peptide was found to be mainly random, with some 0t-helix and fl-structure present. The open structure of CN1 indicates that the constraints necessary for formation of the largely fl-structure in P2 are removed by cleavage of the protein to form peptides. The study produced different quantitative estimations of secondary structures to those previously reported in circular dichroism studies, but the FTIR results are shown to be more reliable.
Keywords: CN1 peptide; P2 protein; FTIR spectroscopy
P2 protein is one of the major extrinsic proteins of the nervous system and has been identified as a possible neuritogenic agent for the induction of experimental allergic neuritis (EAN) ~, a model for the human demyelinating disease, Guillain-Barr6 syndrome 2. P2 is also believed to have a structural role in the myelin sheath. The amino acid sequences of P2 protein for a variety of peripheral nervous system (PNS) sources have been determined, and that of bovine PNS myelin 3 is shown in Figure 1. Peptides of P2 proten have been isolated by cleaving the protein at the methionine residues with cyanogen bromide*. The largest of these peptides has been termed CN1 and comprises residues 21-113 of the protein (Figure 1). The conformation of P2 in aqueous solution has been studied using circular dichroism (CD), nuclear magnetic resonance (NMR) spectroscopy s and Fourier transform infra-red (FTIR) spectroscopy 6. Each of these studies showed that P2 consists largely of fl-structure with a small amount of ~t-helix. The CN1 peptide has also been studied using CD and NMR, with both techniques showing that the peptide is largely random s. However, there appear to be limitations of these two techniques with regard to quantitative analysis of secondary structures. FTIR spectroscopy has developed in recent years as an effective technique for the examination of protein and peptide conformations 7. The amide I mode of proteins and peptides, due mainly to C - - O stretching, has proved
to be the most useful, since this broad mode consists of overlapping components which can be assigned to specific secondary structures. In addition, improvements in instrumentation and analytical methods have allowed accurate quantitative results to be obtained. This paper reports a study of the conformation of the CN 1 peptide in deuterium oxide (D20) using FTIR spectroscopy.
Experimental Extraction of P2 protein P2 protein was isolated from bovine spinal root myelin s. Myelin was obtained by homogenization in sucrose solution and centrifugation of bovine intradural dorsal root tissue (20000 rev rain-1, 30 min, 4°C). The protein was acid-extracted from the myelin at pH 2 with 50% HCI for 18h. The mixture was centrifuged (20 000 rev min- 1, 18 min, 4°C) and the supernatant containing the protein was dialysed against distilled water. The proteins were then lyophilized, dissolved in 0.1 Mammonium acetate, pH 7, and applied to a Cel!ex P
AcSN K F L G T W K L V S S E N F D E Y M K A L G V G L A T R K I , GNLAKPRVIISKKGDIITIRTESPFKNTEISFKLGO EFEETTADNRKTKSTVTLARGSLNOVOKWNGNE TTIKRKLVDGKMVVECKMKDVVCTRIYEKV
Figure 1 The amino acid sequence of bovine P2 protein; letters * To whom correspondence should be addressed
0141-8130/94/030163-03 © 1994Butterworth-HeinemannLimited
represent the one-letter code for amino acids; the CN 1 peptide sequence is underlined
Int. J. Biol. Macmmol. Volume 16 Number 3 1994 163
CN1 peptide structure studied by FTIR spectroscopy." B. H. Stuart and E. F. McFarlane
ion exchange column. The proteins were eluted by increasing the ammonium acetate concentration from 0.1 M to 0.7M, and absorbance of the eluate was determined at 280nm using a Varian UV spectrophotometer. The fractions containing P2 were combined, dialysed against water and lyophilized.
Results and discussion The amide I mode of the CN1 peptide in the original and resolution-enhanced modes are shown in Figure 2. The results of curve-fitting of the deconvolved amide I band of CN1 in D 2 0 are shown in Figure 3. This shows only very minor differences between the observed and
Preparation of CN1 peptide The CN1 peptide was prepared from P2 protein by cleavage of the methionine residues of the protein with cyanogen bromide 9. A mixture of P2 and cyanogen bromide in 70% formic acid was stirred continuously at room temperature for 24 h in a nitrogen atmosphere in the dark. The reaction was then diluted, lyophilized and applied to a Bio-Gel P-6 column and eluted with 0.1 M acetic acid. The eluate was observed at 280 nm using a Varian UV spectrophotometer, and the first distinctive absorbance peak determined the fractions containing CN1, which were combined and lyophilized. The peptide was allowed to stand in D 2 0 for 1 h at room temperature and then lyophilized. This procedure was repeated twice to ensure satisfactory deuteration and is used routinely for preparation of N M R samples in our laboratories. Unexchanged amide protons are undetectable by a Bruker 4 0 0 M H z proton N M R spectrometer after application of this procedure. The samples were prepared in buffer at a concentration of 10mgm1-1. The buffer used was 60mM 2-1-4-(2hydroxyethyl)-piperazin-l-yl] ethanesulfonic acid (HEPES) in D 2 0 (Aldrich, 99.96%) at pH 7.4.
A
m o.to- ~ ¢u ~ 0.050.~I
I
I
1700 1680 1660
I
1
1640 1620 .avenumOen/cm "~
1600
O°~O--
B o.ts-
i o.1o-
~ 0'.o5o.oo-
-0.05 t700
I
l
t660
I
I
I
1560 1640 1620 1600 wavenumOer/c,="o
Fourier transform infra-red spectroscopy Spectra were recorded using a Digilab FTS-20/80 spectrometer equipped with a liquid nitrogen-cooled mercury cadmium telluride (MCT) detector. For each spectrum, 4000 interferograms were co-added, apodized with a triangular function, and Fourier-transformed to give a resolution of 2 c m - 1. The instrument was purged continuously with dry nitrogen in order to eliminate spectral contributions from atmospheric water vapour. Samples were contained in a demountable infra-red cell with CaF 2 windows and a path-length of 0.050mm. A spectrum of 0 2 0 buffer was recorded under identical conditions and subtracted from the peptide spectrum. Subtraction was carried out by elimination of the D 2 0 modes. The overlapping bands were resolved by Fourier self-deconvolution 1° and derivative spectroscopy 11 using standard Digilab software supplied with the spectrometer. The spectra were deconvolved using a half-width Lorentzian line of 17 cm-1 and a k value of 2.0. The derivative spectra were calculated using a power of 2 and a passband edge of 0.3. The number of component amide I bands and the approximate wavenumbers of these bands were obtained from the resulting deconvolved spectra and the second derivative of the original spectrum.
Computer analysis To obtain the fractions of the various secondary components, a curve-fitting F O R T R A N program was used, based on a Gauss-Newton least-squares minimization procedure 12. Gaussian band shapes were assumed in the program and the relative areas of the component bands gave the required fractions la. Iteration was continued to convergence at a low root mean square error ( ~ 0.001%).
164
Int. J. Biol. Macromol. Volume 16 Number 3 1994
C O. 002Q ¢J
¢ m
0.000-
¢. O D
D I
-0.002I I I [ 1 1700 1680 1660 t640 t620 1600 wavenuamer/c-"
Figure 2 The amide I region of the infra-red spectrum of CNI in DzO: (A) original spectrum; (B) deconvolvedspectrum; (C) derivative spectrum o0
0
.
- .
,
,
16k0
i
i
i
,
i
.
i
1660 1640 wavenumber / cm'l
.
,
i
.
,
-
,
1620
1600
Figure 3 The deeonvolvedamide I mode with best-fittedcomponent modes for CNI in D20. The symbols ~, ~, T, R and S represent ~-helix, fl-structure, turns, random coil and side-chain contributions, respectively. The dashed line represents the difference between the observed and calculated values
CN1 peptide structure studied by FTIR spectroscopy." B. H. Stuart and E. F. McFarlane
Table 1 Wavenumbers and fractional areas of the components of the amide I band of CN1 in D20 Wavenumber (cm- 1)
Percentage area
'1624 1640 1654 1664 1679
14 49 4 28 5
Standard deviations: wavenumber: ___1.5 cm- 1; area: ___1%,
Table 2 Secondary structure determinations for P2 protein and the CN1 peptide
CN1 (FTIR) CN1 (CD) (Ref 5) P2 (FTIR) (Ref 6)
a-helix ,//-structure
Turns
Randomcoil
4 5 11
28 14 19
49 72 9
19 9 61
calculated spectra. Table 1 lists the wavenumbers and fractional areas of each component band. The secondary structure assignments are based on a study of 21 globular proteins (of known X-ray crystal structure) in D 2 0 using F T I R spectroscopy 13. The major component of the amide I mode of CN1 occurs at 1640 c m - 1 and represents 49% of the area of the band. The band is most likely to represent random coil in the peptide. The band at 1654cm-1 is characteristic of ~-helix and the band at 1664 cm-~ is characteristic of turns. The mode at 1624 c m - t represents low frequency fl-sheet, and there is a corresponding higher frequency component at 1679 c m - 1. There is also a small mode at 1607 c m - 1 which may be assigned to the arginyl side-chain vibrations 14. Thus, analysis of the F T I R spectrum of CN1 in D 2 0 gives 4 % or-helix, 19% fl-structure, 28% turns and 49% random coil (Table 2). The CN1 peptide in aqueous solution has been studied previously using C D and proton N M R 5. Both techniques showed CN1 to contain a large random component, with a marked loss in fl-structure compared to P2 protein. Analysis of the C D spectrum found 5% or-helix, 9% fl-structure, 14% turns and 72% random coil (Table 2). Although F T I R and C D both show CN1 to be mainly random, there are significant differences in the analyses, particularly in the prediction of the amount of random coil in the peptide. Such a discrepancy was also observed for the results of a C D study of P2 protein 6. The proportion of fl-structure in P2 was notably underestimated in the C D studies and the relative amounts of secondary structures determined depend very much on the database chosen for analysis. It is thought that the F T I R values are more reliable because this method does not rely directly on a chosen database 6. In addition, the C D database consists of proteins, and there is some doubt as to whether these are appropriate when the technique is applied to peptides. An X-ray study of P2 protein by Jones et al.15 shows two segments of ~-helix in the P2 sequence at residues 15-24 and 27-35. The CN1 peptide encompasses 13 of these residues, indicating that 14% of the CN1 sequence has at least the propensity to adopt an or-helical structure. However, it is unlikely that the regions of CN1 predicted
to form ~t-helix by the X-ray structure of P2 will actually be able to form the structure. Although the first four residues of the sequence are ~t-helical in the protein, it is doubtful whether four suitable residues are sufficient to form a stable helical segment in the absence of the constraints imposed by the remainder of the protein. Elimination of the first helix leaves nine residues, 9% of the sequence, to form ~-helix. CN1 has one low frequency fl-sheet component at 1624cm -1, a frequency associated with water-bridged fl-structure 6, which is highly accessible to hydrogenbonding .water molecules. This is to be expected since any fl-sheet in the peptide would have more access to the aqueous environment than in the protein. The majority of the remaining structure is made up of turns and random coil. The mainly open structure of CN1 shows that the constraints necessary for the formation of a fl-structure reside in the regions which are cleaved off during cyanogen bromide treatment.
Conclusions The CN1 peptide was studied in D 2 0 and was found to be mainly random (49%), but with some a-helix (4%) and fl-structure (19%). A C D study of C N I found considerably more random coil, but the F T I R results are considered to be more acceptable as there is some doubt as to whether the C D databases, made up of globular proteins, can be applied to peptides. The results indicate a much more open structure for CN1 than for the protein and demonstrate that the constraints necessary for formation of the largely fl-structure in P2 are removed by cleavage of the protein to form peptides.
Acknowledgements The authors are grateful to Dallas McFarlane assistance with preparation of the CN1 peptide, to Dr Jeff Reimers for use of his curve-fitting plotting programs. The research was supported Sydney University Research Grant Funds.
for and and by
References 1 Brostoff,S.W., Levitt,S. and Powers,J.M. Nature 1977,268, 752 2 Ballin, R.H.M. and Thomas, P.K.J. Neurol. Sci. 1969, 8, 1 3 Kitamura, K., Suzuki, M., Suzuki, A. and Uyemura, K. FEBS Lett. 1980, 115, 27 4 Isobe,T. and Okuyama, T. Eur. J. Biochem. 1978, 89, 379 5 Shin, H.-C. and McFarlane, E.F. Biochem. Biophys. Acta 1987, 913, 155 6 Stuart, B.H. and McFarlane, E.F. Aust. J. Chem. 1991, 44, 1523 7 Jackson,M. and Mantsch, H.H. Spectrochim. Acta 1993, 15, 53 8 James,G.E. and Moore, W.J.J. Neurochem. 1980, 34, 1334 9 Cahnmann,H.J., Arnon, R. and Seta, M. J. Mol. Biol. 1966, 241, 3247 10 Kauppinen,J.K., Moffat, D.J., Mantsch, H.H. and Cameron, D.G. Appl. Spectrosc. 1981, 35, 271 11 Susi, H. and Byler, D.M. Biochem. Biophys. Res. Commun. 1983, 115, 391 12 Fraser,R.D.B. and Suzuki, E. Anal. Chem. 1969, 41, 37 13 Byler,D.M. and Susi, H. Biopolymers 1986, 25, 269 14 Chirgadze, Y.N., Fedorov, O.V. and Trushina, N.P. Biopolymers 1975, 14, 679 15 Jones,T.A., Bergfors,T., Sedzik, J. and Unge, T. EMBO J. 1988, 7, 1597
Int. J. Biol. Macromol. Volume 16 Number 3 1994
165
and E. F. M c F a r l a n e Department of Physical Chemistry, University of Sydney, 2006 Australia Received 22 December 1993; revised 8 March 1994 The conformation of the CN1 peptide, derived from the nervous system P2 protein, has been studied in deuterium oxide solution using Fourier transform infra-red (FTIR) spectroscopy. The peptide was found to be mainly random, with some 0t-helix and fl-structure present. The open structure of CN1 indicates that the constraints necessary for formation of the largely fl-structure in P2 are removed by cleavage of the protein to form peptides. The study produced different quantitative estimations of secondary structures to those previously reported in circular dichroism studies, but the FTIR results are shown to be more reliable.
Keywords: CN1 peptide; P2 protein; FTIR spectroscopy
P2 protein is one of the major extrinsic proteins of the nervous system and has been identified as a possible neuritogenic agent for the induction of experimental allergic neuritis (EAN) ~, a model for the human demyelinating disease, Guillain-Barr6 syndrome 2. P2 is also believed to have a structural role in the myelin sheath. The amino acid sequences of P2 protein for a variety of peripheral nervous system (PNS) sources have been determined, and that of bovine PNS myelin 3 is shown in Figure 1. Peptides of P2 proten have been isolated by cleaving the protein at the methionine residues with cyanogen bromide*. The largest of these peptides has been termed CN1 and comprises residues 21-113 of the protein (Figure 1). The conformation of P2 in aqueous solution has been studied using circular dichroism (CD), nuclear magnetic resonance (NMR) spectroscopy s and Fourier transform infra-red (FTIR) spectroscopy 6. Each of these studies showed that P2 consists largely of fl-structure with a small amount of ~t-helix. The CN1 peptide has also been studied using CD and NMR, with both techniques showing that the peptide is largely random s. However, there appear to be limitations of these two techniques with regard to quantitative analysis of secondary structures. FTIR spectroscopy has developed in recent years as an effective technique for the examination of protein and peptide conformations 7. The amide I mode of proteins and peptides, due mainly to C - - O stretching, has proved
to be the most useful, since this broad mode consists of overlapping components which can be assigned to specific secondary structures. In addition, improvements in instrumentation and analytical methods have allowed accurate quantitative results to be obtained. This paper reports a study of the conformation of the CN 1 peptide in deuterium oxide (D20) using FTIR spectroscopy.
Experimental Extraction of P2 protein P2 protein was isolated from bovine spinal root myelin s. Myelin was obtained by homogenization in sucrose solution and centrifugation of bovine intradural dorsal root tissue (20000 rev rain-1, 30 min, 4°C). The protein was acid-extracted from the myelin at pH 2 with 50% HCI for 18h. The mixture was centrifuged (20 000 rev min- 1, 18 min, 4°C) and the supernatant containing the protein was dialysed against distilled water. The proteins were then lyophilized, dissolved in 0.1 Mammonium acetate, pH 7, and applied to a Cel!ex P
AcSN K F L G T W K L V S S E N F D E Y M K A L G V G L A T R K I , GNLAKPRVIISKKGDIITIRTESPFKNTEISFKLGO EFEETTADNRKTKSTVTLARGSLNOVOKWNGNE TTIKRKLVDGKMVVECKMKDVVCTRIYEKV
Figure 1 The amino acid sequence of bovine P2 protein; letters * To whom correspondence should be addressed
0141-8130/94/030163-03 © 1994Butterworth-HeinemannLimited
represent the one-letter code for amino acids; the CN 1 peptide sequence is underlined
Int. J. Biol. Macmmol. Volume 16 Number 3 1994 163
CN1 peptide structure studied by FTIR spectroscopy." B. H. Stuart and E. F. McFarlane
ion exchange column. The proteins were eluted by increasing the ammonium acetate concentration from 0.1 M to 0.7M, and absorbance of the eluate was determined at 280nm using a Varian UV spectrophotometer. The fractions containing P2 were combined, dialysed against water and lyophilized.
Results and discussion The amide I mode of the CN1 peptide in the original and resolution-enhanced modes are shown in Figure 2. The results of curve-fitting of the deconvolved amide I band of CN1 in D 2 0 are shown in Figure 3. This shows only very minor differences between the observed and
Preparation of CN1 peptide The CN1 peptide was prepared from P2 protein by cleavage of the methionine residues of the protein with cyanogen bromide 9. A mixture of P2 and cyanogen bromide in 70% formic acid was stirred continuously at room temperature for 24 h in a nitrogen atmosphere in the dark. The reaction was then diluted, lyophilized and applied to a Bio-Gel P-6 column and eluted with 0.1 M acetic acid. The eluate was observed at 280 nm using a Varian UV spectrophotometer, and the first distinctive absorbance peak determined the fractions containing CN1, which were combined and lyophilized. The peptide was allowed to stand in D 2 0 for 1 h at room temperature and then lyophilized. This procedure was repeated twice to ensure satisfactory deuteration and is used routinely for preparation of N M R samples in our laboratories. Unexchanged amide protons are undetectable by a Bruker 4 0 0 M H z proton N M R spectrometer after application of this procedure. The samples were prepared in buffer at a concentration of 10mgm1-1. The buffer used was 60mM 2-1-4-(2hydroxyethyl)-piperazin-l-yl] ethanesulfonic acid (HEPES) in D 2 0 (Aldrich, 99.96%) at pH 7.4.
A
m o.to- ~ ¢u ~ 0.050.~I
I
I
1700 1680 1660
I
1
1640 1620 .avenumOen/cm "~
1600
O°~O--
B o.ts-
i o.1o-
~ 0'.o5o.oo-
-0.05 t700
I
l
t660
I
I
I
1560 1640 1620 1600 wavenumOer/c,="o
Fourier transform infra-red spectroscopy Spectra were recorded using a Digilab FTS-20/80 spectrometer equipped with a liquid nitrogen-cooled mercury cadmium telluride (MCT) detector. For each spectrum, 4000 interferograms were co-added, apodized with a triangular function, and Fourier-transformed to give a resolution of 2 c m - 1. The instrument was purged continuously with dry nitrogen in order to eliminate spectral contributions from atmospheric water vapour. Samples were contained in a demountable infra-red cell with CaF 2 windows and a path-length of 0.050mm. A spectrum of 0 2 0 buffer was recorded under identical conditions and subtracted from the peptide spectrum. Subtraction was carried out by elimination of the D 2 0 modes. The overlapping bands were resolved by Fourier self-deconvolution 1° and derivative spectroscopy 11 using standard Digilab software supplied with the spectrometer. The spectra were deconvolved using a half-width Lorentzian line of 17 cm-1 and a k value of 2.0. The derivative spectra were calculated using a power of 2 and a passband edge of 0.3. The number of component amide I bands and the approximate wavenumbers of these bands were obtained from the resulting deconvolved spectra and the second derivative of the original spectrum.
Computer analysis To obtain the fractions of the various secondary components, a curve-fitting F O R T R A N program was used, based on a Gauss-Newton least-squares minimization procedure 12. Gaussian band shapes were assumed in the program and the relative areas of the component bands gave the required fractions la. Iteration was continued to convergence at a low root mean square error ( ~ 0.001%).
164
Int. J. Biol. Macromol. Volume 16 Number 3 1994
C O. 002Q ¢J
¢ m
0.000-
¢. O D
D I
-0.002I I I [ 1 1700 1680 1660 t640 t620 1600 wavenuamer/c-"
Figure 2 The amide I region of the infra-red spectrum of CNI in DzO: (A) original spectrum; (B) deconvolvedspectrum; (C) derivative spectrum o0
0
.
- .
,
,
16k0
i
i
i
,
i
.
i
1660 1640 wavenumber / cm'l
.
,
i
.
,
-
,
1620
1600
Figure 3 The deeonvolvedamide I mode with best-fittedcomponent modes for CNI in D20. The symbols ~, ~, T, R and S represent ~-helix, fl-structure, turns, random coil and side-chain contributions, respectively. The dashed line represents the difference between the observed and calculated values
CN1 peptide structure studied by FTIR spectroscopy." B. H. Stuart and E. F. McFarlane
Table 1 Wavenumbers and fractional areas of the components of the amide I band of CN1 in D20 Wavenumber (cm- 1)
Percentage area
'1624 1640 1654 1664 1679
14 49 4 28 5
Standard deviations: wavenumber: ___1.5 cm- 1; area: ___1%,
Table 2 Secondary structure determinations for P2 protein and the CN1 peptide
CN1 (FTIR) CN1 (CD) (Ref 5) P2 (FTIR) (Ref 6)
a-helix ,//-structure
Turns
Randomcoil
4 5 11
28 14 19
49 72 9
19 9 61
calculated spectra. Table 1 lists the wavenumbers and fractional areas of each component band. The secondary structure assignments are based on a study of 21 globular proteins (of known X-ray crystal structure) in D 2 0 using F T I R spectroscopy 13. The major component of the amide I mode of CN1 occurs at 1640 c m - 1 and represents 49% of the area of the band. The band is most likely to represent random coil in the peptide. The band at 1654cm-1 is characteristic of ~-helix and the band at 1664 cm-~ is characteristic of turns. The mode at 1624 c m - t represents low frequency fl-sheet, and there is a corresponding higher frequency component at 1679 c m - 1. There is also a small mode at 1607 c m - 1 which may be assigned to the arginyl side-chain vibrations 14. Thus, analysis of the F T I R spectrum of CN1 in D 2 0 gives 4 % or-helix, 19% fl-structure, 28% turns and 49% random coil (Table 2). The CN1 peptide in aqueous solution has been studied previously using C D and proton N M R 5. Both techniques showed CN1 to contain a large random component, with a marked loss in fl-structure compared to P2 protein. Analysis of the C D spectrum found 5% or-helix, 9% fl-structure, 14% turns and 72% random coil (Table 2). Although F T I R and C D both show CN1 to be mainly random, there are significant differences in the analyses, particularly in the prediction of the amount of random coil in the peptide. Such a discrepancy was also observed for the results of a C D study of P2 protein 6. The proportion of fl-structure in P2 was notably underestimated in the C D studies and the relative amounts of secondary structures determined depend very much on the database chosen for analysis. It is thought that the F T I R values are more reliable because this method does not rely directly on a chosen database 6. In addition, the C D database consists of proteins, and there is some doubt as to whether these are appropriate when the technique is applied to peptides. An X-ray study of P2 protein by Jones et al.15 shows two segments of ~-helix in the P2 sequence at residues 15-24 and 27-35. The CN1 peptide encompasses 13 of these residues, indicating that 14% of the CN1 sequence has at least the propensity to adopt an or-helical structure. However, it is unlikely that the regions of CN1 predicted
to form ~t-helix by the X-ray structure of P2 will actually be able to form the structure. Although the first four residues of the sequence are ~t-helical in the protein, it is doubtful whether four suitable residues are sufficient to form a stable helical segment in the absence of the constraints imposed by the remainder of the protein. Elimination of the first helix leaves nine residues, 9% of the sequence, to form ~-helix. CN1 has one low frequency fl-sheet component at 1624cm -1, a frequency associated with water-bridged fl-structure 6, which is highly accessible to hydrogenbonding .water molecules. This is to be expected since any fl-sheet in the peptide would have more access to the aqueous environment than in the protein. The majority of the remaining structure is made up of turns and random coil. The mainly open structure of CN1 shows that the constraints necessary for the formation of a fl-structure reside in the regions which are cleaved off during cyanogen bromide treatment.
Conclusions The CN1 peptide was studied in D 2 0 and was found to be mainly random (49%), but with some a-helix (4%) and fl-structure (19%). A C D study of C N I found considerably more random coil, but the F T I R results are considered to be more acceptable as there is some doubt as to whether the C D databases, made up of globular proteins, can be applied to peptides. The results indicate a much more open structure for CN1 than for the protein and demonstrate that the constraints necessary for formation of the largely fl-structure in P2 are removed by cleavage of the protein to form peptides.
Acknowledgements The authors are grateful to Dallas McFarlane assistance with preparation of the CN1 peptide, to Dr Jeff Reimers for use of his curve-fitting plotting programs. The research was supported Sydney University Research Grant Funds.
for and and by
References 1 Brostoff,S.W., Levitt,S. and Powers,J.M. Nature 1977,268, 752 2 Ballin, R.H.M. and Thomas, P.K.J. Neurol. Sci. 1969, 8, 1 3 Kitamura, K., Suzuki, M., Suzuki, A. and Uyemura, K. FEBS Lett. 1980, 115, 27 4 Isobe,T. and Okuyama, T. Eur. J. Biochem. 1978, 89, 379 5 Shin, H.-C. and McFarlane, E.F. Biochem. Biophys. Acta 1987, 913, 155 6 Stuart, B.H. and McFarlane, E.F. Aust. J. Chem. 1991, 44, 1523 7 Jackson,M. and Mantsch, H.H. Spectrochim. Acta 1993, 15, 53 8 James,G.E. and Moore, W.J.J. Neurochem. 1980, 34, 1334 9 Cahnmann,H.J., Arnon, R. and Seta, M. J. Mol. Biol. 1966, 241, 3247 10 Kauppinen,J.K., Moffat, D.J., Mantsch, H.H. and Cameron, D.G. Appl. Spectrosc. 1981, 35, 271 11 Susi, H. and Byler, D.M. Biochem. Biophys. Res. Commun. 1983, 115, 391 12 Fraser,R.D.B. and Suzuki, E. Anal. Chem. 1969, 41, 37 13 Byler,D.M. and Susi, H. Biopolymers 1986, 25, 269 14 Chirgadze, Y.N., Fedorov, O.V. and Trushina, N.P. Biopolymers 1975, 14, 679 15 Jones,T.A., Bergfors,T., Sedzik, J. and Unge, T. EMBO J. 1988, 7, 1597
Int. J. Biol. Macromol. Volume 16 Number 3 1994
165