484
COMMUNICATIONS TO THE EDITOR
but at p H values below 7 its change is too small to be reported without full statistical analysis. Optical rotation differences between C T and modified forms are greatest at low and at high p H values. Since bo differences become large at p H values above 8, i t is probable that two different domains of the protein dominate our observations--one a t low pH, the other at high. Above p H 8 reliable data for H C binding have not been obtained since H C greatly increases the autolysis rate of CT. In the pH region below 7.5 a. differences appear to consist of a large pH-independent change supplemented by a smaller pH-dependent variation which reflects the p H dependence of a. in unmodified CT. .Is measured by a. the normal conformation for catalysis appears to be that at pH 7.5 and the larger change mentioned above is either a consequence of acylation and substrate binding or a necessary preliminary process. Acetyl-D-tyrosine ethyl ester (9-D-TEE) produces no change in rotation a t p H 2.4. IVith A T E E rotation change shows a first-order dependence on substrate concentration with an equilibrium constant at p H 2.5 of approximately 2 X 10-3 M. Dialysis-equilibrium sludies? show only one binding site. Hence rotation changes are to be attributed to the binding of one molecule of XTEE and not to any indirect effect of B T E E on solvent properties. Rotation changes due t o H C a t pH 4 are also first-order in H C concentration with an M in good agreement equilibrium constant of (3 x with constants from fluorescence s t u d i e ~ , ~from ~'~ virtual substrate binding studiesZand from some studies of H C as a competitive inhibitor of CT Hence, again we can attribute the effect to binding of substrate. In current interpretations, changes in a0 without corresponding bo changes are due to changes in freedom of vicinal groups or environment composition at asymmetric carbons. Hence a. variations are to be attributed to conformation changes, though these may involve only alterations in structural rigidity and conformational relaxation times. There is no proof in this work that substrates influence conformational properties through binding to groups of the catalytic site. However, though a second binding site may be involved, there is thus far no evidence which makes this a probable situation. Substrate binding at low p H can be divided into two parts: (1) a change in conformational properties and ( 2 ) direct interaction of substrate and protein a t the binding site. A t p H 7.5 the change in conformation indicated by uo is a minimum and the direct interaction appears to dominate the thermodynamics. However, fluorescence studies to be reported suggest that this analysis is too simple. The conformational changes rnay prove to be trivial or too sluggish to be related to fast catalytic processes.1° On the other hand, the relationship between A H or A S for virtual substrate binding' and the pH dependence of a. for C T indicates an important relationship between the pH-dependent conformational processes associated with the forination of the specific nucleophilic site for catalysis and virtu11 substrate binding. There are relatively small but essential readjustments in the protein fabric which convert C T from a passive binding agent to a catalyst and these are directly or indirectly coupled to substrate binding. The effects of substrate binding are similar to and, where comparisons are possible, as large as those produced by specific acylation. The data suggest t h a t if specific acylation occurs a t the catalytic site, so does the binding of more ( 9 ) S Yanari, R . 1,umry and F. Uovey, t o be published. ( I O ) J. 1 1 Sturtevant, B i d t e i n . Hiophvs. R e s . Cominz~~z,, 8 , :1?1 f l c j f i ? ) . (1 1) 1%.S e i , r a t h , J . A Gladnrr and G . Ilehlaria, J . Bioi C h r n , 188, -107 (1951). (19) R . J . ICoster a n d C. Nirman, .I. A m . Chem. . S o . , 7 1 , X170 (IC1,>.7).
Vol. 85
normal substrates responsible for the changes in ao. It is important to note, however, that there is no basis in this work upon which to postulate the occurrence of important conformation changes during normal catalysis. Of the substances tested only diisopropylfluorophosphate produces significant uo changes a t pH 7.5 and there is no reason t o suppose that the interaction of this substance with C T parallels the processes of normal catalysis. Current high-speed kinetic studies and static fluorescence studies rnay provide soiiie basis for postulating conformational changes during normal cataipis and these will be reported. Several authors hav: proposed that primary bond rearrangement within the Irotein is essential to the formation of the specific nuc1r:ophilic site.I3 I t is our thesis that an equally essc:r!tial aspect of the development of this site is the disi.ortion of primary bonds which provides the necessar local electronic properties for catalysis.I4 \\'e shz 11 defend this thesis and present details of our results in forthcoming papers. 'Experimental.-Rotatory dispersion measurements rw re made with a Rudolph Model 200 spectropolarirnet :r modified to increase ease of reading and precision. P..otein concentrations were usually 0.3"i.;. The molecu'.ar weight of the protein was taken as 24,500. The rmtine standard deviation in a. is 3' but variations in jrotein concentration often inqease this error to 5 " . 'The Xo parameter was 2390 A , , the best-fit value found by B r a n d t ~ . Worthington ~ C T and chymotrypsinogen were used throughout without further purification. TMACT, acetylCT and D I P C T were made by well-known procedures. l5 This work was supported by the National Institutes of Health Grant A-35853-32. (13) S. Bernhard, 142nd National Meeting, American Chemical Society, Atlantic City, S. J., September, 1962, Abstract 2 6 - 0 ; H. N. Rydon, Naluve, 182, 928 (1958); J. A. Cohen, R. A. Oosterbaan, H. S. Jansz and F. Rerends, J . Cell Lonip. Physiol., 64, Suppl. 1, 231 (1959). (13) R . Lumry and H. Eyring, J . Phys. Chem., 68, 110 (1954). (15) A. K , Balls and F. L. Aldrich, Proc. N u l l . A c a d . S c k . U . S., 41, 100 (1965); A. K. Balls a n d H. N . Wood, J . B i d . C h e w , 219,245 (1956).
DEPARTMENT O F CHEMISTRY HELEI\'PARKER RUFUSLCIIRY UNIVERSITYOF MINNESOTA MINSEAPOIJS 14, MINNESOTA RECEIVED NOVEMBER 24, 1962
H ATOM ADDUCTS-NEW
FREE RADICALS?
Sir : The question as to whether molecules such as water or ammonia in the gas phase have an affinity for a hydrogen atom is of considerable interest since it implies the stability of free radicals of the type H A where A is a saturated proton acceptor molecule. An answer to this question may be obtained by considering the bond dissociation energies involved in a conventional hydrogen bonded system such as X-H. . .A The bond dissociation energy in the isolated hydrogen donor molecule XH in the gas phase is given by D(X
-
+ A H j ( H ) - AH,(SH)
H ) = AH,r(X)
(1)
where AH, is the heat of formation in the gas phase I n the complex D(X-H-%)
=
AH,(S)
+ ~ l f I ~ ( H . 4-)
AH,(XI-I-\)
(2)
the enthalpy of hydrogen bond formation may be written b s -AH
=
D(SH-A)
AHj(X)
+ AH,(SH) - AH,(SI-SA)
( t i )
To construct a cycle we may write the bond dissociation energy of the HA radical as D(H-A)
=
AH,(H)
+ A H j ( . A ) - AHr(H.1)
(4)
so that from these equations one readily obtains the
COMMUNICATIONS TO THE EDITOR
Feb. 20, 1963
485
The infrared spectrum in the CsBr region of a benzene solution containing a 1: 1 mixture of ethyllithium-7 and t-butyllithiuni-7 exhibits a strong broad band Since - A H is positive, D(H-A) is greater than zero if centered around 500 c m - l , not found in the spectra D(X-H) - D(X-HA) is either positive or negative of either of the components. Ethyllithium and tbut less than -AH. An approximation of D(X-H) butyllithiuni have strong absorption bands a t 538 and D(X-HA) may be obtained from the Morse poand 480 an.-',respectively, attributed to vibrations tential for X H and X-HA treated as diatoms and the of the carbon-lithium framework. Summation of vibrational frequencies v(X-H) and v(X-HA). For optical densities of spectra of ethyl- and t-butyllithiumOH bonds for example i t turns out that D(X-H) 7 show that the 500 cm.-l band does not result from D(X-H9) is about 5 to 25 kcal./mole in agreement with overlapping of the bands of these components. The the expectation that the X H bond in the complex is 500 cm-l band in the EtLi7-t-BuLi7 complex shifts to longer and weaker than X H in the free molecule. about 514 ern.-' in a 1 : l solution of EtLP and tExperimental values for -AH probably are in the range B u L ~ . This ~ band may be tentatively assigned to of 2 t o 8 kcal.,'niole (if the solution values can be taken carbon-lithium framework vibrations in one or more as a guide) so that the bond dissociation energy of the new organolithium species containing both ethyl and free radical HA is about 7 to 33 kcal./mole in the gas t-butyl groups. phase. The characteristic triplet and quartet pattern in the Perhaps low temperature hydrogen atom reactions hydrogen n.m.r. spectrum of ethyllithiurn? in benzene is in the gas phase or e.s.r. experiments on H atoms not changed by addition of t-butyllithium, and the peak trapped in an inert matrix containing a diluted proton positions are only slightly shifted. This finding is not acceptor could reveal the presence of such radicals. inconsistent with complex formation, but does suggest If the Morse function is a valid description for that, if ethyl groups exist in different environments in D(X-H) and D(X-.HA), equation 5 gives a relationthe solution, exchange must be rapid compared to the ship between AH and the difference in stretching frerelaxation time. On the other hand, the singlet a t quencies v(X-H) - v(X-HA). This is being investi7 = 9.01 in the proton n.m.r. spectrum of pure t gated in more detail both experimentally and theobutyllithium in benzene is split into a doublet in soluretically. tion containing ethyllithium. A 1: 1 mixture of the Helpful discussions with Drs. S. Bauer, K . Kutschke two compounds gives t-butyl peaks a t r = 9.08 and and F. P. Lossing are gratefully acknowledged. r = 9.31, with relative intensity about 7 : 2 . Thus DIVISION OF PURECHEMISTRY t-butyl groups appear to be present in at lest two NATIONAL RESEARCH COUNCIL H. J. BERNSTEIN environments. OTTAWA, C A N A D A relation
D(H-A)
D(X-H)
RECEIVED
-
D(X-HA)
-
AH
(5)
DECEMBER 1, 1962
COMPLEX FORMATION BETWEEN ETHYLLITHIUM AND l-BUTYLLITHIUM Sir :
Cosolution of ethyllithium and t-butyllithium in benzene leads to the formation of complex organolithium compositions, differing from either pure component, containing both types of alkyl groups bonded to lithium. Pure ethyllithium1~2 and t - b ~ t y l l i t h i u mexist ~ as six-fold and four-fold polymers, respectively, in benzene solution. Apparently carbon-lithium bond breaking takes place in this solvent, leading to exchange of alkyl groups between polymeric organolithium molecules when both compounds are present. The new compounds which are produced are believed to be electron-deficient polymers of the type (EtLi),(t-BuLi),-,, where rn is a small number such as 4 or 6. Ethyllithium alone is sparingly soluble in cold benzene, but its solubility is greatly enhanced in benzene solutions of t-butyllithium. Evaporation of the benzene from a solution of both compounds leaves a lowmelting white solid residue. The white solid is highly soluble in pentane, unlike ethyllithium which is virtually insoluble in this solvent. The complex is also much more volatile than ethyllithium. I t distils readily a t 70' and 0.1 mm., forming colorless crystals. Gas chromatographic analysis of the hydrocarbons formed by hydrolysis of several samples of the distilled material show that its composition can vary. The ratio of ethyl to t-butyl groups in the resublimed composition is nearly the same as that in the original solution. Benzene solution with ethyllithium : t-butyllithium ratios of 1: 1 and 1.8:1 gave distilled products with ethyl: tbutyl ratios of 1.1:1 and 1.7:1, respectively: ni.p. 68.-72' and 56-59'. (1) T. L. Brown a n d M. T. Rogers, J . A n i . Chem. .9oc., 79, 1859 (1957). (2) R . West a n d W. Glaze, ibid., 83, 3580 (1981). (3) h l . Weiner, G. Vogel nnd I
