Effect of Urea on the Circular Dichroism of Lysozyme by luckboy


More Info

THE Joumu, OF BIOLOGICAL CNE~STRY 24i, So 6, Issue of March 25, pp. 1708-1712,



in U.S.A.


of Urea

on the



of Lysozyme*
/Received for publication, October 18, 1971)

KIMBALL P. BARNES,$JOHN R. WARREN,~ ANDJULIUS A. GORDOX From the Department of Pathology, University of Colorado School of Medkirie, Denver, Colorado 802i20

SUMMARY The effect of the protein denaturant urea upon the circular ,dichroism spectrum of lysozyme in aqueous solution has been determined. The near-ultraviolet spectrum of native lysozyme in water contained a prominent negative plateau centered at 258 nm. In the spectrum of lysozyme in 8 M urea this plateau was replaced by a broad shoulder with a value for the molar ellipticity at 258 nm 50% less negative. The shape and intensity of the negative plateau were, however, nearly identical for lysozyme in water or 8 M acetamide, an amide of poor denaturing ability. Thus the extent of change in this plateau correlates directly with change in the stability of lysozyme’s disulfides toward reduction in aqueous urea or acetamide (Warren, J. R., and Gordon, J. A. (1970) J. Biol. Chem. 245, 4097). Alteration in the near-ultraviolet spectrum of lysozyme in urea solutions was unaccompanied by equivalent change in the far-ultraviolet spectrum of the protein. In contrast, both the near- and far-ultraviolet circular dichroism spectra of bovine serum albumin were altered upon exposure of the protein to increasingly concentrated urea solutions. These circular dichroic data are consistent with the following conclusions: (a) the negative plateau in the near-ultraviolet spectrum of native lysozyme arises primarily from the intramolecular disulfide bonds of the protein; (b) upon interaction of urea with lysozyme a conformational rearrangement occurs around the disulfide bond or bonds with, unlike serum albumin, little or no disruption of a! helical structure; (c) lysozyme is not resistant to the denaturing action of urea, although unfolding appears to be restricted to nonhelical portions of the lysozyme molecule.

of native conformation as seen with several other proteins (2, 3). Nevertheless, the interaction of urea with lysozyme was accompanied by a modest but significant increase in intrinsic viscosity and a loss in the native stability of the protein’s intramolecular disulfide bonds toward chemical reduction or oxidation (2). TO clarify further the effect of urea upon this protein, the circular dichroism spectra of lysozyme in both the near- and farultraviolet were compared in water and solutions of urea or acetamide. The data reported in this paper support our earlier contention that a significant rearrangement of conformation occurs upon urea-lysozyme interaction in the region or regions of the intramolecular disulfide bond or bonds without disturbing the minor helical component (2).

Downloaded from www.jbc.org by guest, on January 1, 2010

The behavior
unusual in that

of lysozyme the interaction


concentrated urea solutions is of a large number of urea mole-

cules with the protein, measured in this laboratory by an ultrafilt’ration technique (l), failed to induce the extensive disruption
* This study was supported by Grant GM-11345 from the National Institutes of Health. $ Work performed while a predoctoral student in the graduate program of experimental pathology at the University of Colorado School of Medicine. 0 Recipient of Postdoctoral Fellowship GM-977 from the United States Public Health Service. Present address, United States Army Medical Research Institute of Infectious Diseases, Frederick, Maryland, 21701.

Recrystallized salt-free hen egg white lysozyme (Worthington, LYSF OCQ) and bovine serum albumin (Armour, F71601) preparations were used without, further purification. All organic reagents utilized were the best available commercial products. Solutions for CD1 measurement were obtained by volumetric dilution of concentrated stock solutions of protein with precision microburettes. The solutions were unbuffered and salt-free. Values for the apparent pH ranged between 5.6 in water to 7.6 in aqueous 8 BZ urea for each protein. Protein concentrations were checked on a Cary model 15 spectrophotometer with extinction coefficient values (4%) of 26.4 at 282 nm for lysozyme (4) and 6.7 at 280 nm for bovine albumin (5). Circular dichroism spectra were recorded on a Cary model 60 spectropolarimeter with a model 6001 circular dichroism attachment thermostatted at 27”. A lysozyme concentration of 0.7 g per liter (path length of 1.0 cm) was used in the near-ultraviolet (250 to 310 nm). For the far-ultraviolet (210 to 250 nm) a lysozyme concentration of 0.4 g per liter with a O&m path length was used. BSA was 0.6 g per liter over the entire spectrum with a path length of 2.0 cm in the near-ultraviolet and 0.01 cm in Dhe far-ultraviolet. Base-lines were determined for each solvent, mixture. CD data are presented as reduced mean residue ellipticity, [01x in degrees cm2 decimole-l, as calculated from the expression

in which t&, is t,he observed ellipticity at wave length X, X0 the mean residue weight of the protein, 1 the path length in centimeters, and c, protein concentration in grams per ml. The
1 The abbreviations bovine serum albumin. used are: CD, circular dichroism; BSA,


Issue of March 25, 1972

K. P. Barnes, J. R. Warren, and J. A. Gordo~~






i I I

-80 -1



-80 t -160 -120 t -160

Downloaded from www.jbc.org by guest, on January 1, 2010


270 x(w)




310 I 250 270 A(w) / 290 I I 310 I

Near-ultraviolet circular dichroism spectra of lysozyme in water and aqueous urea: water (-) ; 2 M (-a----), 4 M (--), 6 M (-..----j, and 8 M (---) urea. Variation in values of [0] did not exceed A2.6 degrees cm2 decimole-1 for three separate determinations of each spectrum, as indicated by vertical bars.
FIG. 1.

3. Near-ultraviolet circular dichroism spectra of lysozyme in water (-) , 8 M acetamide (- - -), and 8 M urea (-). Reproducibility of [0] within f5 for each spectrum

(see vertical


- 2000

mean residue weight employed for lysozyme was 111, for BSA 114. Each CD spectrum was repeated a minimum of three times with separate solutions. Unfavorable signal to noise ratios prevented CD spectra determinations below 210 nm for lysozyme in water or albumin in aqueous urea, below 220 nm for lysozyme in 8 M urea, and below 250 nm in 8 M acetamide due to the high absorbance of this solvent.

- 6000


I 210

1 230 X(wl

I 250


I 270



FIG. 2. Far-ultraviolet circular dichroism spectra of lysozyme in water (-) and 8 M urea (-). Maximal variation in values of [e] less than ~150 (indicated by vertical bars).

The near-ultraviolet CD spectrum of lysozyme in water (Fig. 1) is characterized by three positive peaks at 294, 288, and 282 nm and a negative plateau between 262 to 255 nm. This circular dichroism pattern closely resembles that previously reported for native lysozyme by other investigators (6, 7). The addition of urea to aqueous solutions of lysozyme resulted in significant changes with increasing urea concentration as follows: (a) the three positive bands show a progressive enhancement unaccompanied by change in the position of their maxima (Fig. 1) ; (b) the cross-over point occurs at progressively shorter wave lengths; and (c) the negative plateau centered near 258 nm in the spectrum of the native protein decreases. A well defined plateau is detectable in 2 M urea, but the decrease in the value of the molar ellipticity at 258 nm is already one-half of that ultimately seen in 8 M urea. At urea concentrations above 2 M this plateau becomes a broad sholder. In contrast to the near-ult,raviolet spectrum, the ultraviolet




of Lysoxyme

Vol. ‘47,

No. G

CD spectra of lysozyme below 250 nm in water and 8 M urea are nearly identical within experimental certainty (Fig. 2). The effect of acetamide, a weak protein denaturant. (8), upon the near-ultraviolet CD spectrum of lysozyme was compared to 8 M urea (Fig. 3). In 8 M acetamide a very large enhancement of the three positive bands is present, about twice that occurring in 8 M urea, with cross-over points into the negative spectrum




0.: -2 YE? A I -160

Downloaded from www.jbc.org by guest, on January 1, 2010


250 0.d

270 X(w)



0 360



-80 1


‘0 r X z U











4M 6M I 0.7 I 0.8 8M I 0.9








I 0.6




FIG. 4. The reduction of lysozyme by 0.1 M 2-mercaptoethanol in 4 M (O), 6 M (Cl), and 8 M (A) urea: a, first order kinetic plotsof change in molar ellipticity at 260 urn; b, effect of urea concentration on the forward rate constant of reduction.

FIG. 5. Circular dichroism spectra of bovine serum albumin in water and aqueous urea: water (---) ; 2 M (----), 4 M (-----), 6 M (-.. -), and 8 M (----) urea. a, near-ultraviolet spectra ([O] reproducible within f3,
indicated error less by vertical . than f480 bars); b, far-ultraviolet ^ _ ., for values of LSj). spectra (maximal

Issue of March

25, 1972

K. P. Barnes, JO R. Warren,

and J. A. Gordon


about the same for the two solvents. Most importantly, the spectra from 265 to 255 nm in 8 M acetamide resembles water alone and differs from urea. The change in molar ellipticity at 260 nm following the addito lysozyme in 4, 6, or 8 M urea tion of 0.1 M mercaptoethanol was followed with time until a constant value was obtained. Assuming a two-state transition between unreduced and fully reduced lysozyme, the fraction of unreduced lysozyme, f.+~, was calculated at any given time in each urea solution with the expression

fN, = b]R - bit bin - [elf+
where [0], is the molar ellipticity of reduced lysozyme, [0],~ the molar ellipticity of unreduced lysozyme, and [0] t the ellipticity the disapat time t. At each of the three urea concentrations pearance of unreduced lysozyme showed first order kinetics (Fig. 4~) and was proportional to the 3.6th power of urea (Fig. 4b). BSA behaved significantly different in aqueous urea when compared to lysozyme. As shown in Fig. 5, the spectrum of albumin in water is negative throughout with two smaller extrema at 268 and 262 nm and two distinct extrema at 222 and 208 nm. The circular dichroism spectrum of BSA in pH 7.5 borate buffer recently reported by Gorbunoff shows similar detail between 255 to 270 nm and two intense negative maxima at 208 and 220 nm (9). With increasing urea concentra6ion the near-ultraviolet spectrum becomes progressively less negative; the fine structure in the near-ultraviolet of the native spectrum is maintained, however, in the most concentrated urea solution studied (Fig. 5~). The far-ultraviolet spectrum also becomes much less negative at higher concentrations of urea with, in addition, virtual elimination of the 222-nm extremum in 8 M urea (Fig. 5b). The broad negative shoulder from 215 to 230 nm in the spectra of BSA in 6 or 8 M urea is characteristic of proteins in a denatured but constrained state lacking long range order (10). The fine structure observed in the near-ultraviolet CD spectrum of albumin in 8 M urea (Fig. 5a) could thus be accounted for by focal areas of ordered conformation enveloping amino acid side chain chromophores contributing to the near-ultraviolet.

To interpret the ultraviolet circular dichroism spectra of a protein, it is necessary to know the molecular origin of the various ellipticity bands. Asymmetrical peptide bond conformations with repeatable order make major contributions to CD spectra at wave lengths shorter than 240 nm (11). Ellipticity bands from 250 to 310 nm are generated by asymmetries of the local environment imposed upon various side chain chromophores by the native conformation of the protein (11). Variation in both the near- and far-ultraviolet spectrum during the thermal or chemical denaturation of a protein implies an unfolding of the ordered peptide chain which affects local interactions of side chain chromophores. The urea denaturation of BSA is an example of change induced in the near-ultraviolet by extensive protein unfolding (Fig. 5). In contrast, although the near-ultraviolet spectrum of lysozyme is markedly altered by urea, the far-ultraviolet spectrum remains essentially unchanged even in 8 M urea (Figs. 1, 2). Thus the change in the near-ultraviolet cannot,

like serum albumin, follow from large alterations in ordered peptide bond conformations. Tryptophan, tyrosine, and cystine residues contribute to the near-ultraviolet CD spectrum of lysozyme (11). It appears that tryptophan and tyrosine in the native protein are responsible for the three positive bands observed between 275 and 300 nm (6, 7). The enhancement of these bands seen upon the interaction of urea with lysozyme (Fig. 1) could originate from some disruption of protein conformation in the vicinity of the tyrosine and tryptophan residues or, as acetamide induced an even greater increase in the ellipticity of these bands than urea (Fig. 3), a nonspecific, solvent perturbation effect (12), or both. Since Ikeda and Hamaguchi have shown that the CD spectrum of lysozyme in the 250- to 270-nm region is relatively unaffected by oxidation of indole or phenol groups, it is unlikely that tryptophan or tyrosine participate to any extent in the observed changes in the negative plateau below 270 nm (13). Beychok has demonstrated that disulfide-containing compounds, such as cystine or oxidized glutathione, show a large, negative band in their CD spectra at about 260 nm (14). Also, CD spectra of proteins or small peptides with a high disulfide content but few or no aromatic residues possess bands near 250 nm (15, 16). Thus the disulfide bond has significant optical activity in the 250- to 270-nm portion of the near-ultraviolet. We recently confirmed that upon the interaction of urea with lysozyme the intramolecular disulfide bonds of the protein become accessible to reduction by thiols (2). Since this increased disulfide bond reactivity was demonstrated to be independent of nonspecific solvent, pH, or oxidation-reduction potential effects, we suggested a conformational rearrangement in proximity to the disulfides (2). The occurrence of such a conformational change is further supported by the urea dependence of the CD spectrum observed between 250 to 270 nm (Fig. 1). The CD spectrum of lysozyme in 8 M acetamide or water (Fig. 3) is consistent with the previously observed native stability toward reduction in 8 M acetamide (2). The forward rate of unfolding of lysozyme in aqueous ureamercaptoethanol (unfolding detected by change in the mean residue rotation at 578 nm) was reported by us as directly proportional to the 3.9th power of urea concentration (a), which compares to the 3.6th power shown by the rate of disulfide bond reduction (Fig. 4b). This close correspondence suggests strongly that the rate-limiting step lies in the reduction of disulfide or disulfides and not unfolding of the reduced protein. Therefore it appears that accessibility of the intramolecular disulfide bonds of lysozyme is hindered by the (steric) arrangement of the protein molecule around the disulfides in the absence of urea. Early reports of the “urea-resistance” of lysozyme (17, 18) were based on the failure of urea to disrupt the a-helix of this protein. However, the results obtained from measurement of intrinsic viscosity (2)) disulfide bond accessibility to mercaptans (2), and the near-ultraviolet circular dichroism spectrum clearly indicate that nonhelical regions of the lysozyme molecule unfold in aqueous urea. Investigation of the effect of urea on the conformation of lysozyme has thus demonstrated that any molecular model proposed for a “urea-denatured” protein should describe the state of nonhelical as well as helical portions of the protein. Acknowledgment-We are indebted to Dr. John R. Cann for the use of the Cary model 60 spectropolarimeter.

Downloaded from www.jbc.org by guest, on January 1, 2010

REFERENCES J. A., -END W~RRES, 1. GORDON, 5663-5669 J. R., AND GORDON: 2. WARREN, 4097-4104 J. R., AND GORDON, 3. WARREN, Acta 229, 216 K. C., AND TANFOED, 4. AUNE, 4585 M. D., AND FOSTER, 5. STERMAN, Sot. 78,3652 K., AND HAMAGUCHI, 6. IKEDA, 66,513 V. I., KAY, C. M., 7. TEICHBERG, Biochem. 16, 55 J. R. (1968)



of Lysozyme

Vol. 247, So. 6

J. Biol. J. Biol. Biochim.

Chem. Chem.

243. 246,

J. A. (1970) J. A. C. (1971)

Biophys. 8, 4579-

(1969) J. F.

Biochemistry (1956) J. J. Amer.

Chem. (Tokyo) Eur. J.






8. GORDON, J. A., AND JENCKS, W. P. (1963) Biochemistry 2, 47 M. J. (1970) Arch. Biochew,. Biophys. 138, 684 9. GORBUNOFP, G. D., HOVING, H., .~ND TIMA~EEFF’, S. N. (1970) 10. FASMAN, Biochemistry 9, 33163324 11. BEYCHOK, S. (1966) Science 164, 1288-1299 12. WOODY, R. W. (1968) J. Chem. Phys. 49, 4797 K., AND HAMAGUCHI, K. (1970) J. Bzochem. (Tokyo) 13. IKEDA, 68, 785 14. BEYCHOK, S. (1965) Proc. A-d. Acad. Sci. CT. S. A. 63, 999 S., AND BRESLOV-, E. (1968) J. Biol. Chem. 243, 15. BEYCHOK, 151-154 16. BRESLOW, E. (1970) Proc. ,$-at. Acad. Sci. U., S. A _ 67, 493 17. HAMAGUCHI, K., AND KUROXO, A. (1963) J. Biochem. (Tokyo) 64, 111 18. HASHIZUME, H., SHIRAICI, M., AICD IMAHOHI, K. (1967) J. Biothem. (Tokyo) 62,543

Downloaded from www.jbc.org by guest, on January 1, 2010

To top