3972
COMMUNICATIONS TO THE EDITOR 70C
60C
50C
400
-bo 30C
20c
100
VOl. a4
tion than L-alanine residues. Also, it is clear that polypeptide helix formation in water solution depends not only on the formation of intramolecular peptide hydrogen bonds but also is aided by “hydrophobic bond” and dispersion force stabilizations’l provided by residues such as methionine and alanine. Further, these new water-soluble polypeptides having high helix contents, can be transformed by reagents known to denature proteins into completely random conformations. (11) For a discussion see W. Kauzmann in “Advances in Protein Chemistry,” Vol. 14, C. B. Anfinsen, Jr., M. L. Anson, K. Bailey and J. T. Edaall, Editors, Academic Press, Inc., New York, N. Y..1959.
CHILDREN’S CANCER RESEARCH FOUNDATION AND DEPARTMENT OF BIOLOGICAL CHEMISTRY R. K. KULKARNI HARVARD MEDICAL SCHOOL E. R. BLOUT 15, MASSACHUSETTS BOSTON RECEIVED AUGUST17, 1962 FLUORINE N.M.R. SPECTROSCOPY. XII. PROOF OF OPPOSITE SIGNS FOR THE “DIRECT” CARBON13 COUPLING CONSTANTS TO HYDROGEN AND T O FLUORINE
Sir : The use of “spin decoupling” (double resonance) 20 40 60 for the determination of relative signs of coupling MOLE % constants in fluorocarbon derivative^^-^ and in C O - A M I N O ACID RESIDUE proton corn pound^^^^ has become important as Fig. 2 -The dependence of bo (calculated from the Moffitt an experimental check on theoretical predictions. equation6 over the wave length range 313 to 578 mp using The work of Lauterbur and Kurland6 was particXO = 212 mp) upon amino acid composition for a series of ularly significant, since i t was a test of the prohigh molecular weight copolypeptides derived from ?-IT- posa17 that “direct” couplings of C13 to protons 12 - morpholinylethyl] - LY,L- glutamamide and L - alanine, have the same sign as vicinal proton-proton 0-0-0: poly - - N - [2-morpholinylethyl] -cu,Lglutamamide couplings. and L-methionine, A-A-A. The solvent in all cases w a s water The predictions of signs and magnitudes of containing 0.2 M NaCl. All measurements were performed couplings involving fluorine, though less certain at 25 f 1’ except those on the copolypeptides containing than those for protons, nevertheless represent a 30, 35 and 40q/,L-methionine with which compounds the further development of the theory. I t is of prime measurements were made at 5 =IC1”. iniportance to test the suppositioii that the “dicent. alanine or methionine were water-soluble. rect” couplings, J(C13F) and J(Cl3H), have the Some rotatory dispersion data of the water-soluble same sign. This has now been done for the particcopolypeptides are shown in Fig. 2, where bo is ularly simple case of CHC12F, dichlorofluoroplotted as a function of mole per cent. coamino acid methane. Experimental.-The basic n.m.r. spectrometer residue content. The solubility of the copolypeptides containing 30 or more per cent. L-methionine and techniques were as previously described.R,9 is increased a t lower temperatures so the dispersion For the double resonance work a Model SD-60 measurements on these compounds were made a t Spin Decoupler’O was employed. It was equipped 5’. The data reveal that, by incorporation of with modules designed for strong irradiation of -40% L-methionine in I, it is possible to obtain an fluorine nuclei a t 37.65 mc./sec., while observing essentially completely helical water-soluble syn- protons a t 40.000 mc./sec. (symbolized as H { F ] decoupling’) and for F{H}decoupling with irradiathetic polypeptide, Other experiment were performed with the tion a t 42.50 mc./sec. With this equipment some copolypeptide containing 30 mole percent L-alanine ten watts of radiofrequency power is available t o which, in 0.2 M NaC1, had a bo value of -331 in- the probe; however, far less than this is required dicating slightly more than 50’% helix content. since the F-H coupling constant in CHCllF is (1) J. D. Baldeschwieler, C k e m . Reus., 63, i n press (1963). When this copolypeptide’s rotatory dispersion was (2) D. P. Evans, Mol. P h y s . , 5, 183 (1962). measured in 8 M urea or in 8.6 M lithium bromide (3) S. L. M a n a t t and D. D. Elleman, J . A m . Chefn. Soc., 84, 1305 solution, the bo value was found to be zero. (1962). From the experiments and the data reported (4) D. D. Ellernan and S. I,. Manatt, J . C h e m . P h r s . , 36, 1945 here we can conclude the following. The helical (1962). (5) S. I,, ivlnnrtt and D. D. Elleman, J . A m . Chent. SOL.,84, 1579 form of poly-7-N- [2-morpholinylethyl]-cr,~-glu(1962). tamamide is not stable in water solution but this (6) P. C. Lauterbur and R. 1. Kurland, ibid., 84,3405 (1962). polypeptide may be converted to a helical conforma(7) M.Karplus, ibid.,84, 2458 (1962). (8) G. Filipovich a n d G. V. D. Tiers, J . Pkrs. Chem., 8S, 761 tion by changing to less polar solvents such as methanol or dioxane. In certain water soluble (1959). (9) G. V. D. Tiers, J. Chem. Phys., 56, 2263 (1962). polypeptides L-methionine residues are more effec( 101 Manufactured by t h e Nuclear Magnetic Resonance Specialties tive in rxomoting helix formation in aqueous solu- Co., Inc., Box 145, Grcensburg Road, New Kensington, Pa.
Oct. 20, 1962
COMMUNICATIONS TO THE EDITOR
only 53 c./sec., for the collapse of which HZ= 20 milligauss is entirely adequate. For the exact measurement of changes in irradiation frequency, an EC-60 Counter Adapter, lo which produces an easily-countable beat frequency by means of a standard crystal, was employed. A BC-221-J frequency meter (Zenith Radio Corp.) was used to establish the direction of increase. Dichlorofluoromethane was obtained from the Matheson Co. ; its doublet fluorine resonance, centered at +80.881 c$*, f 0.003* (5 vol. % in CClaF, 0.5y0 Mersi, 24.5') had J ( H F ) = 53.65 f 0.16 c./sec. I n the proton spectrum of this solution the doublet center is a t 2.587 r , f 0.002. The neat liquid, in a 4 mm. i d . thin-walled tube, was required for the study of the C13 compound a t natural abundance. The fluorine isotope shiftg due t o C13, A+(C13F)-(C12F), was found to be +0.156 f 0.003 p.p.m., and t h a t for H, 4-0.004 f 0.001s p.p.m.; while J(C13F) = 293.8 f 0.2 c./sec., and J(C13H) = 220.00 f 0.l3c./sec. Though three dissimilar nuclei are employed here, the spin decoupling procedure and the interpretation as to relative signs are essentially the same as in the case previously describede6 Upon irradiation of the low-field (high frequency) C13 satellite doublet in the proton spectrum, the high-field C13 doublet in the fluorine spectrum drew together. When the irradiation frequency was lowered by about 512 c./sec., the low-field C13-F doublet narrowed instead. Had the coupling constants been of like sign, i t would have been necessary to raise the irradiation frequency by about 75 c./sec. to achieve this result. Analogous observations were made for H (F j decoupling, further confirming the assignment of unlike signs for J(C13F) and J(C13H). I thank Emmett B. Aus for the careful n.m.r. spectral work reported here.
3973
X=F
XzOCH3
n=0
I
2
3
Fig. l.-JCH as a function of n f o r two series CH(r - %,X,,. If the substituent effects were additive, the points for n = 2 and ?t = 3 should fall on the extrapolated lines through those for n = 0 and n = 1, shown as solid lines in the figure.
deviations from colinearity, though smaller. also is usually deexceed the experimental error. JCH termined with an uncertainty not exceeding fl C.P.S. from proton spectra, although values derived from C13 ~ p e c t r amay ~ , ~ be subject to larger errors. TABLE I COMPARISON OF MEASURED COUPLING CONSTANTS AND PREDICTIONS BASEDON ZETAVALUES JCH (C.P.9.)
Compound
Observed
Predicted
Difference
CsHsCHtF 151" 150 + I CHzFa 18d 173 12 CHFI 238 197 $41 CHFClz 220 203 17 CHFzCl 23 1 200 +31 CHz( COOH)z 132* 136 - 4 CONTRIBUTION KO.242 FROM THE CH2ClCOOH 152 157 - 5 CENTRAL RESEARCH DEPARTMENT OF THE CHC12COOH 181 184 - 3 MINNESOTA MININGAND MFG.COMPANY, CHaOCHs 140 [140]" [OI ST. PAUL19, MINNESOTA GEORGE V. D. TIERS CHz(OCH,h 162 155 $ 7 RECEIVED SEPTEMBER 17, 1962 CH(OCH3)a 186 170 16 D. T. Carr, Thesis, Purdue University, 1962. * Reference 8. e Used to evaluate zeta for the methoxy group as NON-ADDITIVE SUBSTITUENT EFFECTS ON C"56.6 C.P.S. PROTON SPIN-SPIN COUPLINGS'
+ +
+
Sir: It has been reported2 that CIa-proton couplings in the nuclear magnetic resonance spectra of substituted methanes conform to "a simple, but precise, additivity relation." In the belief that a further experimental test of this generalization was desirable, we have measured coupling constants for a number of compounds not previously studied. The results, presented in Table I together with coupling constants calculated using zeta-values, show that the additivity relation is not generally valid. The accompanying figure is a plot of JCH against n for the two series CH(4-,,X, in which X is either F or OCH3. Clearly, the points are not colinear as would be required by the additivity relation. If the data3for CH