ARCHIVES
OF BIOCHEMISTRY
Vol. 284, No. 1, January,
AND
BIOPHYSICS
pp. 53-57,
1991
Molecular
Mechanism
Zhengyu
Randolph
Dong,
of Molecular
Department
Received
May
of Spider Silk Elasticity
V. Lewis,l
Biology,
30, 1990, and in revised
University
form
August
and C. Russell of Wyoming,
Box
University
Station,
Laramie,
Wyoming
82071
27, 1990
Spider major ampullate (drag-line) silk is an extracellular fibrous protein which has impressive characteristics of strength and elasticity. This silk has been hypothesized to predominantly consist of a single protein, containing regions of antiparallel &sheets which are interspersed with amorphous segments responsible for its elastic properties. A rubber-like mechanism has been suggested to account for this elasticity, but the specific molecular mechanism is unknown. Using Fourier transform infrared spectroscopy (FTIR) we found evidence of either helix formation or reorientation of preexisting helices when axial tension is applied to the spider silk fiber. CD studies of a peptide derived from the silk gene repeat sequence show that it can form &sheets at high temperatures while a-helices are induced in 2,2,2-trifluoroethanol. These results suggest a possible molecular mechanism for the elasticity of spider silk fibers. It is proposed that the elastic process involves the formation and disruption of a-helical Ala-rich regions which are interspersed among stable &sheet domains. 0 1991 Academic Press,
Middaugh2 3944,
Inc.
INTRODUCTION
phous regions contribute substantially to the elastic properties of the fiber (5). In this work we employed Fourier transform infrared spectroscopy (FTIR)3 to probe the molecular mechanism of spider silk elasticity. We detected either the formation of a oc-helicesor a reorientation of helices when the silk fiber is exposed to axial tension. Further studies on a synthetic peptide based on the silk gene repetitive sequence provided supportive evidence for this. MATERIALS
AND METHODS
Materials. The major ampullate silks from the spiders Nephila clawipes or Araneusgemmoides were harvested as described (7) and allowed to dry completely at room temperature. A 3- to 5-cm single fiber of silk was used for each FTIR experiment. Peptides were synthesized on an Applied Biosystem peptide synthesizer, Model 430A, using the Fmoc program, then purified by gel filtration and reversed phase HPLC with a C-18 column. Their identities were verified by amino acid composition analysis. TFE(2,2,2-trifluoroethanol) was purchased from Eastman Kodak Co. and was of spectral grade (99%+). All other reagents were of analytical grade. FTZZ?. Spectra were obtained with an Analect FX 6260 FTIR spectrometer through an Analect Micro-XA FTIR microscope at 2 cm-’ resolution. We compared the effects of different interferograms of 128 scans, 256 scans, and 512 scans and the resulting sample spectral differences were negligible. So interferograms of 128 scans were routinely obtained and spectra were deconvolved by the method of Kauppinen (8) employing Gaussian components with a half-width at half-height (a) of 12 cm-’ and a resolution enhancement factor (K) of 2. Axial extension experiments were carried out by fixing one end of the silk and attaching the other end to small suspended weights of the indicated size. Circular dichroism measurements. CD measurements were carried out with a Jasco spectropolarimeter Model J-500A calibrated with dlo-camphorsulfonic acid. Pathlength cells of 0.5 and 0.2 mm were used for CD measurements from 250 to 195 nm. Temperature was controlled with a NESLAB Endocal circulating water bath.
Major ampullate (drag-line) silk of orb web spiders possessesunique physical properties, combining high tensile strength and substantial elasticity (1,2). As a consequence, there is significant interest in the use of spider silk as a novel biomaterial. Previous investigations suggest that spider silk is composed of a single large protein, consisting primarily of pseudocrystalline regions of stacked P-pleated sheets interspersed with amorphous domains (3-5). The molecular basis for spider silk elasticity is presently unknown, although it has been suggested that RESULTS an entropy driven process like that found in rubber is FTIR analysis. Initially, the uniformity of individual involved (6). It has also been speculated that the amorsilk fibers was examined by obtaining FTIR spectra at 1 To whom correspondence should he addressed. ‘Current address: Merck Sharp & Dohme Research WP 26-331, West Point, PA 19486.
0 of
3 Abbreviations used: FTIR, TFE, 2,2,2-trifluoroethanol.
Fourier
transform
infrared
spectroscopy;
53
0003-9861/91$3.00
Copyright All rights
Laboratories,
by Academic Press, Inc. reproduction in any form reserved. 1991
54
DONG,
LEWIS,
several randomly selected sites along the fiber, both parallel and perpendicular to incident polarized radiation. The spectra differed from each other by less than 2 cm-l in peak positions (results not shown). Thus, the silk fiber appears to be spectrally homogeneous. The absorption above 1750 cm-’ in the spectra is associated with the silk fiber and extends over a broad range. Its origin is unknown at present, but is likely to be due to the pseudocrystalline structure of the fiber. To examine the possibility of structural changes in the major ampullate silk of the spider Nephila cluvipes, when the protein responds to axial strain, polarized FTIR spectra of individual silk fibers were obtained through an infrared microscope. Deconvolved infrared spectra of silk fibers, with the ir radiation polarized perpendicular to the fiber axis, are shown in Fig. lA, as a function of axial tension on the fiber. Strong amide I peaks are found at 1694,1630, and 1612 cm-‘, consistent with the high content of P-sheet structure previously detected in other forms of silk by IR spectroscopy (9). Weak bands at 1657 and 1675 cm-’ are also evident and can be assigned to disordered regions and antiparallel P-sheets, respectively (10, 11). Amide II bands are seen at 1520 and 1550 cm-‘, indicating P-sheets and possibly a very small amount of helical regions, respectively (10, 12). When tension is applied to the silk fiber only minor changes in the amide I region are seen with perpendicular incident radiation. In contrast, the peak near 1550 cm-’ in the amide II region decreases in intensity upon applied tension and shifts to 1557 cm-‘. Strikingly different results are observed when the radiation is oriented parallel to the fiber axis (Fig. 1B). While all of the fl-structure peaks are still clearly evident and their positions unchanged, application of tension produces a dramatic increase in the absorbance at 1651 cm-‘. A peak in this region which displays parallel dichroism is strongly indicative of a-helical structure (10, 11). A similar increase is seen in the amide II band at 1559 cm-‘, which also indicates the formation of a helix. A band at 1512 cm-’ is formed as well and splits into two components at 1508 and 1517 cm-’ under tension. An absorbance in this region in proteins is usually assigned to either non-hydrogen-bonded peptide groups (12) or tyrosine residues (10). The tyrosine content of this fiber is only 3-4% (results not shown), so the latter assignment seems less likely. Evidence that the observed changes may be involved in the elastic behavior of drag-line silk comes from the complete return of the original spectra when tension is released (results not shown). Individual fibers display similar but not identical spectra, thus rigorous consistency among these spectra was hard to achieve. However, the spectra show consistent features described above, when compared prior to tension, during the tension and after the release of tension.
AND
MIDDAUGH
1700
1600
1500
I.59 i.og
1
L 1700
1600
1500
WAVENUMBER(cm-‘)
FIG. 1. Deconvolved FTIR spectra of drag-line silk fiber from major ampullate gland of Nephila clauipes as a function of applied force with fiber axis perpendicular (A) and parallel (B) to the polarized radiation. For deconvolution parameters, see Materials and Methods section.
A similar set of tension FTIR experiments was performed with silk from the major ampullate gland of another spider species, Araneus gemmoides. The spectra of a silk fiber with its axis perpendicular and parallel to the incident polarized light are shown in Fig. 2. As seen previously a peak appears near 1650 cm-l with parallel dichroism as tension is applied. Furthermore, in the parallel spectra of both major ampullate silks the peak around 1645 cm-’ assigned to unordered regions dramatically declines as axial tension is applied. To investigate the possibility that the a-helical structure detected during stretching originates from randomly oriented a-helices preexisting prior to applied tension and that the applied force merely reorients the helices into a parallel array, FTIR spectra of silk fiber were measured with unpolarized light. Several strands of silk were randomly oriented on the sample plate without applied tension and spectra were obtained. The result is shown in Fig. 3. No evidence for a a-helices is found in these spectra as would be expected if preexisting a-helices are present in the relaxed state. Repeating sequence of the spider silk gene. The sequence of a substantial portion of the silk protein from
MOLECULAR
MECHANISM
OF
SPIDER
SILK
0.1550~
55
ELASTICITY
1700
1600 WAVENUMBER
in a
1500
(cm-‘)
FIG. 3. FTIR spectra of three strands of drag-line silk from major ampullate gland of Nephilu &wipes being oriented randomly on the sample plate and detected without polarizer. Solid line: original spectrum; Dashed line: deconvolved spectrum.
1700
1600
1500
WAVENUMBER(cm-‘) FIG. 2. Deconvolved FTIR spectra of drag-line silk from major ampullate gland of Aruneus gemmoides as a function of applied force with fiber axis perpendicular (A) and parallel (B) to the polarized radiation. For deconvolution parameters, see Materials and Methods section.
helix formation of 30P. The results are shown in Fig. 5b. A dramatic increase in the value of [0]220nm is proportional to the concentration of TFE% (v/v) in the 20 to 40% range, suggesting nearly complete induction of a a-helix as indicated by the characteristic double minima at 220 and 208 nm. Previous studies of a a-helical poly(Ala) peptides show the negative minima at 207-208 nm to be stronger than that those at 220-221 nm (17), as is observed here. The results imply a significant helix forming capability for 30P. DISCUSSION Antiparallel P-Sheet Is Silk’s Major Component
Nephila clavipes has been determined from a silk gland The unpolarized FTIR spectra of silk fiber (Fig. 3) cDNA clone (25). A consensusrepetitive motif of 30 amino demonstrate that this drag-line silk is predominantly acids (Fig. 4) is found. It contains a poly(Ala) like segment composed of antiparallel ,&sheets. The bands at 1637 and flanked by two Gly-rich regions. On the basis of an analogy 1691 cm-’ represent two vibrational modes of the carbonyl with Bombyx mori fibroin, the Gly-rich region presumably group in an antiparallel P-sheet (18). The shoulder at 1612 is found in a P-sheet conformation (13). In contrast, the cm-’ originates from the same vibrational modes as the poly(Ala) like region is probably present in the amorphous 1637 cm-’ amide I absorbance in antiparallel P-pleated regions of silk fiber. sheets (19). Also present is a shoulder at 1666 cm-‘, inCD measurements. Apeptide (M, = 2331), designated dicating the presence of the expected turns in P-sheets. 30P with the silk repeat sequence shown in Fig. 4, was During stretching the P-sheet regions undergo only minor, synthesized and subjected to CD analysis. When the con- if any, changes. This suggests a rigid nature for the /3centration of 30P is at or above 2.0 mg/ml, it displays a sheet regions of the fiber. The repetitive sequence of the P-sheet like spectrum at high temperature (Fig. 5a) with drag-line silk gene demonstrates a similarity to other a single broad negative peak near 215 nm. At lower con- forms of silk with a high content of glycine residues precentrations, however, the peptide exists in a random coil sumably constituting P-sheet regions. The peptide 30P conformation (200 nm minimum) throughout this tem- based upon the silk gene repeat sequence exhibits a pperature range (results not shown). Another example of a P-sheet forming peptide is poly(Lys), which adapts the AGRGGQGAGAAAAAAGGAGQGGYGGLGGQG P-form after being heated at 50°C for 10 min (14,15). We used TFE, an organic solvent known to promote helix FIG. 4. Consensus sequence from spider Nephela &wipes drag-line formation in peptides (16), to test the potential for a (Y- silk gene derived from cDNA.
56
DONG,
LEWIS,
FIG. 5.
(a) CD spectra of 30P in 20 mM phosphate buffer, pH 7.0, at different temperatures; (b) CD spectra of 30P at 0°C in 20 mM phosphate buffer, pH 7.0, with different TFE percentages (v/v).
sheet CD spectrum at 80°C demonstrating the anticipated P-sheet forming potential. The concentration dependence of P-sheet formation indicates that this structure is primarily intermolecular, as expected for short peptides. Aligned a-Helices Are Induced by Axial Tension The FTIR results suggest that a-helix formation may be involved in the elastic process of spider silk fiber. The presence of a peak near 1650 cm-’ with parallel dichroism when axial tension is applied to a fiber indicates hydrogen bond formation along the axis of a helical structure. Release of stress appears to return the a-helix content to an undetectable level, probably replaced by an unordered structure (1645 cm-‘). Repeated experiments on silk fibers showed a consistent pattern of results, indicating a-helix formation in the silk protein when tension is applied. Combining early X-ray crystallographic results with our gene sequence suggests that the amorphous regions are very likely to be composed of poly(Ala) like segments with the remainder of the repeating motif containing the crystalline P-sheet regions of the silk. Previous studies (17, 20, 21) have shown that poly(Ala)-containing peptides can form stable a-helices in aqueous solution. Most strikingly, a regular a-helical structure has been obtained by stretching a poly(Ala) fiber (22). Recently short alaninebased peptides have been found to form unusually stable a-helical structures (23). Alanine residues are thought to possesshigh helical potential (23). Thus, the alanine rich segments in the amorphous regions of major ampullate drag-line silk are the most likely candidates for the observed tension-induced formation of u-helix. Some confirmation of this hypothesis comes from our CD results, where it is obvious that the 30P peptide can
AND
MIDDAUGH
form a-helices in the presence of an organic solvent. Gly is known to be a helix destabilizer and the proportion of Gly residues in the proposed helical domains is significantly below average (24). Thus, the Gly-rich regions are unlikely to form a-helices. In addition, studies with the peptide 15P, the C-terminal 15 residues of 3OP, shows no helix formation in TFE (unpublished results). Thus, the alanine-rich regions appear to be the best candidates for helix formation. We can’t yet rigorously exclude the possibility that preexisting c+helices, whose spectral features are hidden in the broad amide I peak, become reoriented upon stretching to give rise to the dichroic helical signals. However, we have been unable to find any evidence for their presence. On the basis of the previous work of others and these data, major ampullate spider silk can be most simply pictured as semicrystalline regions of interlocking P-sheets, which give the fiber its remarkable strength. These regions appear to change little when force is applied along the fiber axis. The P-sheet regions of individual polypeptide chains are interspersed with short, alanine-rich domains which are disordered when the fiber is in a relaxed state. Application of tension induces these regions to become helical with the energy for helix formation arising at least partially from the applied mechanical forces. When tension is relaxed the ordered regions are then entropically driven back to a more disordered state, producing the observed elasticity. This clearly contrasts with the cy to p transformation seen upon stretching in some other silks (13). This force-induced formation of ordered structure is unusual but may be of general relevance to other biochemical systems. ACKNOWLEDGMENT We thank Purnima Ray and Mark Colgin for silk collection. This work was supported by grants from ONR and NIH to R.V.L. and C.R.M.
REFERENCES 1. Denny, M. W. (1976) J. Exp. Biol. 65, 483-506. 2. Lucas, F. (1964) Disco~lery 25, 20-26. 3. Warwicker, J. 0. (1960) J. Mol. Biol. 2, 350-362. 4. Lucas, F., Shaw, J. T. B., and Smith, S. G. (1955) J. Text. Inst. 46, T440-T452. 5. Hepburn, H. R., Chandler, A. D., and Davidoff, M. R. (1979) Insect Biochem.
9,69-77.
6. Gosline, J. M., Denny, M. (London) 309,551-552. 7. Work, R. W., and Emerson,
W., and Demont, P. D. (1982)
M. E. (1984)
J. Arachnol.
Nature
10, l-10.
8. Kauppinen, J. K., Moffatt, D. J., Mantsch, H. H., and Cameron, D. G. (1986) Appl. Spectrosc. 35, 271-276. 9. Suzuki, E. (1967) Specfrochim. Acta 23A, 2303-2308. 10. Fraser, R. D., and MacRae, T. P. (1973) Conformation in Fibrous Proteins, pp. 95-123. Academic Press, New York/London. 11. Byler, D. M., and Susi, H. (1986) Biopolymers 26, 469-487.
MOLECULAR 12. Cantor, C. R., and Schimmel, Part II, pp. 466-472, Freeman, 13. Lucas, F., and Randall, (Florkin, M., and Stotz, Amsterdam/London/New 14. Sakar, 987.
P., and Doty,
15. Greenfield, 4115. 16. Nelson, &net.
MECHANISM
P. R. (1980) Biophysical San Francisco.
Chemistry,
Proc. N&l.
Acad.
N., and Fasman,
G. D. (1969)
J. W., and Kallenbach,
N. R. (1986)
Biochem Elsevier,
Sci. USA 55,981-
Bioc/zemistry Proteins
Struct.
8, 4108Futzct.
1,211-217.
17. Quadrifalio,
2755-2760.
F., and Urry,
D. W. (1968)
J. Amer.
SPIDER 18. Susi,
SILK
57
ELASTICITY
H., Timasheff,
S. N., and Stevens,
L. (1967)
J. Biol.
Chen.
242,5460-5467.
K. M. (1968) in Comprehensive E. H., Eds.), Vol. 26B, pp. 475-558, York.
P. (1966)
OF
Chem.
Sot. 90,
19. Wantyghem, J. et al. (1990) Biochemistry 29, 6600-6609. 20. Gratzer, W. B., and Doty, P. (1963) J. Amer. Chem. Sot. 85,11931197. 21. Ingwell, R. T. et al. (1968) Biopolymers 6, 331-368. 22. Elliott, A. (1967) in Poly-a-Amino Acids (Fassman, G. D., Ed.), pp. l-67, Dekker, New York. 23. Marqusee, S., Robbins, V. H., and Baldwin, R. L. (1989) Proc. N&l. Acad. Sci. USA 86, 5286-5290. 24. Fraser, R. D., and MacRae, T. P. (1973) Conformation in Fibrous Proteins, pp. 285-286, Academic Press, New York/London. 25. Xu, M., and Lewis, 7120-7124.
R. V. (1990)
Proc.
N&l.
Acad.
Sci.
USA
87,
OF BIOCHEMISTRY
Vol. 284, No. 1, January,
AND
BIOPHYSICS
pp. 53-57,
1991
Molecular
Mechanism
Zhengyu
Randolph
Dong,
of Molecular
Department
Received
May
of Spider Silk Elasticity
V. Lewis,l
Biology,
30, 1990, and in revised
University
form
August
and C. Russell of Wyoming,
Box
University
Station,
Laramie,
Wyoming
82071
27, 1990
Spider major ampullate (drag-line) silk is an extracellular fibrous protein which has impressive characteristics of strength and elasticity. This silk has been hypothesized to predominantly consist of a single protein, containing regions of antiparallel &sheets which are interspersed with amorphous segments responsible for its elastic properties. A rubber-like mechanism has been suggested to account for this elasticity, but the specific molecular mechanism is unknown. Using Fourier transform infrared spectroscopy (FTIR) we found evidence of either helix formation or reorientation of preexisting helices when axial tension is applied to the spider silk fiber. CD studies of a peptide derived from the silk gene repeat sequence show that it can form &sheets at high temperatures while a-helices are induced in 2,2,2-trifluoroethanol. These results suggest a possible molecular mechanism for the elasticity of spider silk fibers. It is proposed that the elastic process involves the formation and disruption of a-helical Ala-rich regions which are interspersed among stable &sheet domains. 0 1991 Academic Press,
Middaugh2 3944,
Inc.
INTRODUCTION
phous regions contribute substantially to the elastic properties of the fiber (5). In this work we employed Fourier transform infrared spectroscopy (FTIR)3 to probe the molecular mechanism of spider silk elasticity. We detected either the formation of a oc-helicesor a reorientation of helices when the silk fiber is exposed to axial tension. Further studies on a synthetic peptide based on the silk gene repetitive sequence provided supportive evidence for this. MATERIALS
AND METHODS
Materials. The major ampullate silks from the spiders Nephila clawipes or Araneusgemmoides were harvested as described (7) and allowed to dry completely at room temperature. A 3- to 5-cm single fiber of silk was used for each FTIR experiment. Peptides were synthesized on an Applied Biosystem peptide synthesizer, Model 430A, using the Fmoc program, then purified by gel filtration and reversed phase HPLC with a C-18 column. Their identities were verified by amino acid composition analysis. TFE(2,2,2-trifluoroethanol) was purchased from Eastman Kodak Co. and was of spectral grade (99%+). All other reagents were of analytical grade. FTZZ?. Spectra were obtained with an Analect FX 6260 FTIR spectrometer through an Analect Micro-XA FTIR microscope at 2 cm-’ resolution. We compared the effects of different interferograms of 128 scans, 256 scans, and 512 scans and the resulting sample spectral differences were negligible. So interferograms of 128 scans were routinely obtained and spectra were deconvolved by the method of Kauppinen (8) employing Gaussian components with a half-width at half-height (a) of 12 cm-’ and a resolution enhancement factor (K) of 2. Axial extension experiments were carried out by fixing one end of the silk and attaching the other end to small suspended weights of the indicated size. Circular dichroism measurements. CD measurements were carried out with a Jasco spectropolarimeter Model J-500A calibrated with dlo-camphorsulfonic acid. Pathlength cells of 0.5 and 0.2 mm were used for CD measurements from 250 to 195 nm. Temperature was controlled with a NESLAB Endocal circulating water bath.
Major ampullate (drag-line) silk of orb web spiders possessesunique physical properties, combining high tensile strength and substantial elasticity (1,2). As a consequence, there is significant interest in the use of spider silk as a novel biomaterial. Previous investigations suggest that spider silk is composed of a single large protein, consisting primarily of pseudocrystalline regions of stacked P-pleated sheets interspersed with amorphous domains (3-5). The molecular basis for spider silk elasticity is presently unknown, although it has been suggested that RESULTS an entropy driven process like that found in rubber is FTIR analysis. Initially, the uniformity of individual involved (6). It has also been speculated that the amorsilk fibers was examined by obtaining FTIR spectra at 1 To whom correspondence should he addressed. ‘Current address: Merck Sharp & Dohme Research WP 26-331, West Point, PA 19486.
0 of
3 Abbreviations used: FTIR, TFE, 2,2,2-trifluoroethanol.
Fourier
transform
infrared
spectroscopy;
53
0003-9861/91$3.00
Copyright All rights
Laboratories,
by Academic Press, Inc. reproduction in any form reserved. 1991
54
DONG,
LEWIS,
several randomly selected sites along the fiber, both parallel and perpendicular to incident polarized radiation. The spectra differed from each other by less than 2 cm-l in peak positions (results not shown). Thus, the silk fiber appears to be spectrally homogeneous. The absorption above 1750 cm-’ in the spectra is associated with the silk fiber and extends over a broad range. Its origin is unknown at present, but is likely to be due to the pseudocrystalline structure of the fiber. To examine the possibility of structural changes in the major ampullate silk of the spider Nephila cluvipes, when the protein responds to axial strain, polarized FTIR spectra of individual silk fibers were obtained through an infrared microscope. Deconvolved infrared spectra of silk fibers, with the ir radiation polarized perpendicular to the fiber axis, are shown in Fig. lA, as a function of axial tension on the fiber. Strong amide I peaks are found at 1694,1630, and 1612 cm-‘, consistent with the high content of P-sheet structure previously detected in other forms of silk by IR spectroscopy (9). Weak bands at 1657 and 1675 cm-’ are also evident and can be assigned to disordered regions and antiparallel P-sheets, respectively (10, 11). Amide II bands are seen at 1520 and 1550 cm-‘, indicating P-sheets and possibly a very small amount of helical regions, respectively (10, 12). When tension is applied to the silk fiber only minor changes in the amide I region are seen with perpendicular incident radiation. In contrast, the peak near 1550 cm-’ in the amide II region decreases in intensity upon applied tension and shifts to 1557 cm-‘. Strikingly different results are observed when the radiation is oriented parallel to the fiber axis (Fig. 1B). While all of the fl-structure peaks are still clearly evident and their positions unchanged, application of tension produces a dramatic increase in the absorbance at 1651 cm-‘. A peak in this region which displays parallel dichroism is strongly indicative of a-helical structure (10, 11). A similar increase is seen in the amide II band at 1559 cm-‘, which also indicates the formation of a helix. A band at 1512 cm-’ is formed as well and splits into two components at 1508 and 1517 cm-’ under tension. An absorbance in this region in proteins is usually assigned to either non-hydrogen-bonded peptide groups (12) or tyrosine residues (10). The tyrosine content of this fiber is only 3-4% (results not shown), so the latter assignment seems less likely. Evidence that the observed changes may be involved in the elastic behavior of drag-line silk comes from the complete return of the original spectra when tension is released (results not shown). Individual fibers display similar but not identical spectra, thus rigorous consistency among these spectra was hard to achieve. However, the spectra show consistent features described above, when compared prior to tension, during the tension and after the release of tension.
AND
MIDDAUGH
1700
1600
1500
I.59 i.og
1
L 1700
1600
1500
WAVENUMBER(cm-‘)
FIG. 1. Deconvolved FTIR spectra of drag-line silk fiber from major ampullate gland of Nephila clauipes as a function of applied force with fiber axis perpendicular (A) and parallel (B) to the polarized radiation. For deconvolution parameters, see Materials and Methods section.
A similar set of tension FTIR experiments was performed with silk from the major ampullate gland of another spider species, Araneus gemmoides. The spectra of a silk fiber with its axis perpendicular and parallel to the incident polarized light are shown in Fig. 2. As seen previously a peak appears near 1650 cm-l with parallel dichroism as tension is applied. Furthermore, in the parallel spectra of both major ampullate silks the peak around 1645 cm-’ assigned to unordered regions dramatically declines as axial tension is applied. To investigate the possibility that the a-helical structure detected during stretching originates from randomly oriented a-helices preexisting prior to applied tension and that the applied force merely reorients the helices into a parallel array, FTIR spectra of silk fiber were measured with unpolarized light. Several strands of silk were randomly oriented on the sample plate without applied tension and spectra were obtained. The result is shown in Fig. 3. No evidence for a a-helices is found in these spectra as would be expected if preexisting a-helices are present in the relaxed state. Repeating sequence of the spider silk gene. The sequence of a substantial portion of the silk protein from
MOLECULAR
MECHANISM
OF
SPIDER
SILK
0.1550~
55
ELASTICITY
1700
1600 WAVENUMBER
in a
1500
(cm-‘)
FIG. 3. FTIR spectra of three strands of drag-line silk from major ampullate gland of Nephilu &wipes being oriented randomly on the sample plate and detected without polarizer. Solid line: original spectrum; Dashed line: deconvolved spectrum.
1700
1600
1500
WAVENUMBER(cm-‘) FIG. 2. Deconvolved FTIR spectra of drag-line silk from major ampullate gland of Aruneus gemmoides as a function of applied force with fiber axis perpendicular (A) and parallel (B) to the polarized radiation. For deconvolution parameters, see Materials and Methods section.
helix formation of 30P. The results are shown in Fig. 5b. A dramatic increase in the value of [0]220nm is proportional to the concentration of TFE% (v/v) in the 20 to 40% range, suggesting nearly complete induction of a a-helix as indicated by the characteristic double minima at 220 and 208 nm. Previous studies of a a-helical poly(Ala) peptides show the negative minima at 207-208 nm to be stronger than that those at 220-221 nm (17), as is observed here. The results imply a significant helix forming capability for 30P. DISCUSSION Antiparallel P-Sheet Is Silk’s Major Component
Nephila clavipes has been determined from a silk gland The unpolarized FTIR spectra of silk fiber (Fig. 3) cDNA clone (25). A consensusrepetitive motif of 30 amino demonstrate that this drag-line silk is predominantly acids (Fig. 4) is found. It contains a poly(Ala) like segment composed of antiparallel ,&sheets. The bands at 1637 and flanked by two Gly-rich regions. On the basis of an analogy 1691 cm-’ represent two vibrational modes of the carbonyl with Bombyx mori fibroin, the Gly-rich region presumably group in an antiparallel P-sheet (18). The shoulder at 1612 is found in a P-sheet conformation (13). In contrast, the cm-’ originates from the same vibrational modes as the poly(Ala) like region is probably present in the amorphous 1637 cm-’ amide I absorbance in antiparallel P-pleated regions of silk fiber. sheets (19). Also present is a shoulder at 1666 cm-‘, inCD measurements. Apeptide (M, = 2331), designated dicating the presence of the expected turns in P-sheets. 30P with the silk repeat sequence shown in Fig. 4, was During stretching the P-sheet regions undergo only minor, synthesized and subjected to CD analysis. When the con- if any, changes. This suggests a rigid nature for the /3centration of 30P is at or above 2.0 mg/ml, it displays a sheet regions of the fiber. The repetitive sequence of the P-sheet like spectrum at high temperature (Fig. 5a) with drag-line silk gene demonstrates a similarity to other a single broad negative peak near 215 nm. At lower con- forms of silk with a high content of glycine residues precentrations, however, the peptide exists in a random coil sumably constituting P-sheet regions. The peptide 30P conformation (200 nm minimum) throughout this tem- based upon the silk gene repeat sequence exhibits a pperature range (results not shown). Another example of a P-sheet forming peptide is poly(Lys), which adapts the AGRGGQGAGAAAAAAGGAGQGGYGGLGGQG P-form after being heated at 50°C for 10 min (14,15). We used TFE, an organic solvent known to promote helix FIG. 4. Consensus sequence from spider Nephela &wipes drag-line formation in peptides (16), to test the potential for a (Y- silk gene derived from cDNA.
56
DONG,
LEWIS,
FIG. 5.
(a) CD spectra of 30P in 20 mM phosphate buffer, pH 7.0, at different temperatures; (b) CD spectra of 30P at 0°C in 20 mM phosphate buffer, pH 7.0, with different TFE percentages (v/v).
sheet CD spectrum at 80°C demonstrating the anticipated P-sheet forming potential. The concentration dependence of P-sheet formation indicates that this structure is primarily intermolecular, as expected for short peptides. Aligned a-Helices Are Induced by Axial Tension The FTIR results suggest that a-helix formation may be involved in the elastic process of spider silk fiber. The presence of a peak near 1650 cm-’ with parallel dichroism when axial tension is applied to a fiber indicates hydrogen bond formation along the axis of a helical structure. Release of stress appears to return the a-helix content to an undetectable level, probably replaced by an unordered structure (1645 cm-‘). Repeated experiments on silk fibers showed a consistent pattern of results, indicating a-helix formation in the silk protein when tension is applied. Combining early X-ray crystallographic results with our gene sequence suggests that the amorphous regions are very likely to be composed of poly(Ala) like segments with the remainder of the repeating motif containing the crystalline P-sheet regions of the silk. Previous studies (17, 20, 21) have shown that poly(Ala)-containing peptides can form stable a-helices in aqueous solution. Most strikingly, a regular a-helical structure has been obtained by stretching a poly(Ala) fiber (22). Recently short alaninebased peptides have been found to form unusually stable a-helical structures (23). Alanine residues are thought to possesshigh helical potential (23). Thus, the alanine rich segments in the amorphous regions of major ampullate drag-line silk are the most likely candidates for the observed tension-induced formation of u-helix. Some confirmation of this hypothesis comes from our CD results, where it is obvious that the 30P peptide can
AND
MIDDAUGH
form a-helices in the presence of an organic solvent. Gly is known to be a helix destabilizer and the proportion of Gly residues in the proposed helical domains is significantly below average (24). Thus, the Gly-rich regions are unlikely to form a-helices. In addition, studies with the peptide 15P, the C-terminal 15 residues of 3OP, shows no helix formation in TFE (unpublished results). Thus, the alanine-rich regions appear to be the best candidates for helix formation. We can’t yet rigorously exclude the possibility that preexisting c+helices, whose spectral features are hidden in the broad amide I peak, become reoriented upon stretching to give rise to the dichroic helical signals. However, we have been unable to find any evidence for their presence. On the basis of the previous work of others and these data, major ampullate spider silk can be most simply pictured as semicrystalline regions of interlocking P-sheets, which give the fiber its remarkable strength. These regions appear to change little when force is applied along the fiber axis. The P-sheet regions of individual polypeptide chains are interspersed with short, alanine-rich domains which are disordered when the fiber is in a relaxed state. Application of tension induces these regions to become helical with the energy for helix formation arising at least partially from the applied mechanical forces. When tension is relaxed the ordered regions are then entropically driven back to a more disordered state, producing the observed elasticity. This clearly contrasts with the cy to p transformation seen upon stretching in some other silks (13). This force-induced formation of ordered structure is unusual but may be of general relevance to other biochemical systems. ACKNOWLEDGMENT We thank Purnima Ray and Mark Colgin for silk collection. This work was supported by grants from ONR and NIH to R.V.L. and C.R.M.
REFERENCES 1. Denny, M. W. (1976) J. Exp. Biol. 65, 483-506. 2. Lucas, F. (1964) Disco~lery 25, 20-26. 3. Warwicker, J. 0. (1960) J. Mol. Biol. 2, 350-362. 4. Lucas, F., Shaw, J. T. B., and Smith, S. G. (1955) J. Text. Inst. 46, T440-T452. 5. Hepburn, H. R., Chandler, A. D., and Davidoff, M. R. (1979) Insect Biochem.
9,69-77.
6. Gosline, J. M., Denny, M. (London) 309,551-552. 7. Work, R. W., and Emerson,
W., and Demont, P. D. (1982)
M. E. (1984)
J. Arachnol.
Nature
10, l-10.
8. Kauppinen, J. K., Moffatt, D. J., Mantsch, H. H., and Cameron, D. G. (1986) Appl. Spectrosc. 35, 271-276. 9. Suzuki, E. (1967) Specfrochim. Acta 23A, 2303-2308. 10. Fraser, R. D., and MacRae, T. P. (1973) Conformation in Fibrous Proteins, pp. 95-123. Academic Press, New York/London. 11. Byler, D. M., and Susi, H. (1986) Biopolymers 26, 469-487.
MOLECULAR 12. Cantor, C. R., and Schimmel, Part II, pp. 466-472, Freeman, 13. Lucas, F., and Randall, (Florkin, M., and Stotz, Amsterdam/London/New 14. Sakar, 987.
P., and Doty,
15. Greenfield, 4115. 16. Nelson, &net.
MECHANISM
P. R. (1980) Biophysical San Francisco.
Chemistry,
Proc. N&l.
Acad.
N., and Fasman,
G. D. (1969)
J. W., and Kallenbach,
N. R. (1986)
Biochem Elsevier,
Sci. USA 55,981-
Bioc/zemistry Proteins
Struct.
8, 4108Futzct.
1,211-217.
17. Quadrifalio,
2755-2760.
F., and Urry,
D. W. (1968)
J. Amer.
SPIDER 18. Susi,
SILK
57
ELASTICITY
H., Timasheff,
S. N., and Stevens,
L. (1967)
J. Biol.
Chen.
242,5460-5467.
K. M. (1968) in Comprehensive E. H., Eds.), Vol. 26B, pp. 475-558, York.
P. (1966)
OF
Chem.
Sot. 90,
19. Wantyghem, J. et al. (1990) Biochemistry 29, 6600-6609. 20. Gratzer, W. B., and Doty, P. (1963) J. Amer. Chem. Sot. 85,11931197. 21. Ingwell, R. T. et al. (1968) Biopolymers 6, 331-368. 22. Elliott, A. (1967) in Poly-a-Amino Acids (Fassman, G. D., Ed.), pp. l-67, Dekker, New York. 23. Marqusee, S., Robbins, V. H., and Baldwin, R. L. (1989) Proc. N&l. Acad. Sci. USA 86, 5286-5290. 24. Fraser, R. D., and MacRae, T. P. (1973) Conformation in Fibrous Proteins, pp. 285-286, Academic Press, New York/London. 25. Xu, M., and Lewis, 7120-7124.
R. V. (1990)
Proc.
N&l.
Acad.
Sci.
USA
87,
