3699 trogen furthest removed from the COO- group, generating 6a rather than 6b.26,28
7032I N
[RNH, I
CH,--CO,6a
I
CH,--CO,6b
Acknowledgement. W e wish to thank the United States Public Health Service, Grants A M 17323 (G.L.K.) and G M 20564 (P.A.K.), for financial support. References and Notes (1) Recipient of a Career Development Award, AM-00014, from the National Institute of Arthritis, Metabolism and Digestive Diseases. (2) Department of Pharmaceutical Chemistry. (3) United States Public Health Service Predoctorai Fellow, 1970-1974, Feilowship No. GM 41089. (4) Recipient of a Career Development Award, GM-70718, from the National Institute of General Medical Sciences. (5) A. L. Lehninger, “Bioenergetics”, 2nd ed, W. A. Benjamin, Menlo Park, Calif., 1973, pp 213-215. (6) G. L. Rowley, A. L. Greenleaf. and G. L. Kenyon. J. Am. Chem. Soc., 93, 5542 (1971).
(7) A. C. McLaughlin, M. Cohn. and G. L. Kenyon, J. Biol. Chem., 247, 4382 (1972). (8) H. Kessier and D. Liebfritz. Tetrahedron Left.. 427 (1969). (9) H. Kessler and D. Liebfritz, Tetrahedron, 25, 5127 (1969). (10) H. Kessler and D. Liebfritz, Tetrahedron, 26, 1805 (1970). (11) V. J. Bauer, W. Fulmor, G. 0. Morton, and S. R. Safir, J. Am. Chem. Soc., 90, 6846 (1968). (12) I. 0. Sutheriand, “Annual Reports on NMR Spectroscopy”, Vol. 4, Academic Press, New York. N.Y.. 1971, p 83 ff and 106 ff. (13) F. A. Bovey, “Nuclear Magnetic Resonance Spectroscopy”, Academic Press, New York, N.Y., 1969, pp 183-188. (14) D. A. Klein and G. Binsch, J. Am. Chem. SOC.,92, 3787 (1970). (15) J. A. Pople and G. A. Segal, J. Chem. Phys., 44, 3289 (1966). (16) D. P. SantryandG. A. Segal, J. Chem.Phys., 47, 158(1967). (17) J. R. Herriott and W. E. Love, Acta Crystallogr., 24, 1014 (1968). (18) 0. Meyerhof and K. Lohmann, Biochem. Z., 196, 49 (1928). (19) W. D. Kumler and J. J. Eiler, J. Am. Chem. SOC.,65, 2355 (1943). (20) G. W. Allen and P. Haake, J. Am. Chem. SOC., 95, 8080 (1973). (21) P. A. Kollman, J. McKelvey, and P. Gund, J. Am. Chem. SOC.,97, 1640 (1975). (22) H. Mendel and D. C. Hodgkin, Acta Crystallogr., 7, 443 (1954). (23) L. H. Jensen, Acta Crystallogr., 8, 237 (1955). (24) P. A. Kollman, W. F. Trager, S. Rothenberg, and J. E. Williams, J. Am. Chem. Soc., 95, 458 (1973). (25) For example, in the intramolecularly H-bonded conformation the H--O distance is 2.17 A: assigning partial changes to the 0 (-0.45) and guanidinium H (+0.3) from the Mulliken populations, one calculates an electrostatic attraction of 21 kcal/mol. (26) G. E. Struve and G. L. Kenyon, submitted for publication. (27) T. L. James and M. Cohn, J. Biol. Chem., 249, 2599 (1974). (28) We are currently examining, by both theoretical and experimental methods, the relative stabilities of derivatives of 2-iminoimidazolidine (without the CHzCOO- group) phosphorylated on the primary and secondary nitrogen.
Phosphorus-3 1 Nuclear Magnetic Resonance Studies on Nucleoside Phosphates in Nonaqueous Media Richard J. Labotka,* Thomas Glonek, and Terrell C. Myers Contributionfrom the Research Resources Center and the Department of Biological Chemistry, Unitiersity of Illinois at the Medical Center, Chicago, Illinois 6061 2. Receitied May 27, 1975
Abstract: 31Pnuclear magnetic resonance chemical shift and coupling constant data obtained from tetra-n-butylammonium nucleoside phosphates in water and anhydrous tetramethylurea support an early proposal by Albert Szent-Gyorgi, where it was suggested that the phosphate side chain of the adenosine triphosphate molecule was folded so as to lie upon the adenine ring, the conformation being stabilized by the formation of a hydrogen bond between the /3 phosphorus and the 6-amino nitrogen of the ring. This conformation, which predominates in anhydrous media, is observed for all of the common nucleoside diand triphosphates.
The role of adenosine triphosphate (ATP) as the energy mediator in a large number of biological reactions has been known since its discovery in 1929.’ However, the conformation of the molecule in the context of these reactions, i.e., the nature of the transition state, has not been established with certainty. In 1957, Albert Szent-Gyorgi2 speculated that the ability of the aromatic ring of the adenine base of the nucleotide to “quench” long-lived excited states of organic systems by energy absorption through K-K* transitions allows the adenine ring to participate in the activated complex in some biological reactions, e.g., oxidative phosphorylation. The capture or release of energy in reactions of this type is associated with the formation and cleavage of P - 0 - P bonds, usually the terminal P - 0 - P linkage of the tripolyphosphate side chain. However, these two functional groups, the adenine ring and the polyphosphate chain, are separated by the sugar ribose “which has no conjugated double bonds and no T electrons”* through which energy interactions could be effected. Szent-Gyorgi postulated that the ribose might function as a “hinge” to allow the polyphosphate chain to double back over the ring, thus creating a close physical and chemical rela-
tionship between these two components. The structure would be stabilized by the formation of a hydrogen bond between the 6-amino nitrogen of the ring and the p phosphate of the chain, leaving the terminal phosphate group free for cleavage.2 Alternate structures proposed1-I0 included mono- or polyvalent cations, particularly Mg2+ and Ca2+ with these ions forming coordination complexes incorporating the y phosphate of the chain and the 7 nitrogen of the adenine base. Several investigations of A T P in aqueous solution^'^^-^ have produced varying results, with some observers concluding that little interaction exists between the ring and the chain in the absence of metal ions,8 while other^',^,^,' have found such associations when monovalent metal ions such as N a + and K+ were used to titrate solutions. Still others have reported the formation of numerous complexes between nucleotides and divalent cations. 1 ~ 5 , 6 , 9 Szent-Gyorgi’s original and bold proposal has also inspired a number of mathematical investigations of this nucleotide, and these have lent credence to the concept of “folded” ~ ~ p . 8 ~ 1 0
Intracellular, and certainly intramitochrondrial environLabotka et al.
/ 3 1 PN M R on Nucleoside Phosphates
3700
Table 1. Nucleoside Triphosphate Chemical Shifts and Coupling Constants in Water and in Anhydrous Tetramethylurea at H+ Ion Concentrations Corresponding to pH 7.40 Chemical shiftso Compd Solvent ATP dATP ITP GTP CTP UTP
H20 TMU H2O TMU H20 TMU H2O TMU H2O TMU H2O TMU
a
p
Y
11.45 12.25 11.43 13.47 11.49 12.34 11.49 12.04 11.51 12.28 11.50 12.36
22.66 23.17 22.83 23.46 22.84 23.03 22.85 23.12 22.77 22.74 22.92 22.70
7.33 10.64 7.71 10.46 7.63 10.87 7.70 10.97 7.25 10.45 7.52 10.45
Coupling constants, Hz Jm-o
19.75 25.6 20.50 28.3 20.75 25.1 19.25 25.6 20.25 25.1 20.53 25.1
JE-v
19.75 23.6 20.50 24.6 20.75 23.6 19.25 25.0 20.25 23.1 20.53 23.8
Chemical shifts in parts per million relative to 85% orthophosphoric acid. ments, where most biochemical reactions occur, are not simple aqueous solutions, however. Indeed, the surfaces of enzyme proteins or biomembranes are best considered as nonaqueous, and in certain instances aproteic as well. In such environments even water molecules, when present, participate in ordered structures.*,’ Data on such nonaqueous systems are scant and, thus, the present 31Pnuclear magnetic resonance ( N M R ) investigation into the properties of A T P in an anhydrous, aproteic medium was undertaken, and the data were interpreted in terms of the conformation of A T P and of other nucleotides. In vivo 3 1 P spectra of A T P in intact frog gastrocnemius muscle were also obtained and the data parallel to a considerable degree those observed in the anhydrous system.
by rotary evaporation at 35 O C , redissolved in anhydrous acetonebenzene ( 1 : l ) and evaporated at least three times to render the preparation anhydrous. After dissolution of this preparation in a volume of anhydrous tetramethylurea equivalent to the volume of the original aqueous solution, the sample was again analyzed by 31P spectroscopy. Evaporations from anhydrous acetone-benzene were repeated until the 31Pchemical shifts no longer changed upon further treatment. Such final solutions showed no H2O signal in the ‘H NMR spectrum, and also proved to be nonreactive to CaH2. In these anhydrous preparations, the concentrations of nucleotide and protons were the same as in the original aqueous solutions. To check for changes in the proton concentration (pH) resulting from the evaporative procedure, the tetramethylurea solutions were lyophilized and redissolved in water, and the 31Pshifts and pH values were remeasured. It was found that, using reasonable care, the 31P shifts and pH values of the solutions were reproducible within experimental precision. The above outlined procedure ensures that the ratio of acid protons to nucleotide was identical in both the aqueous and nonaqueous solutions; however, the nature of the solvation state of the protons in the anhydrous system is not known, and it is not meant to be implied by the use of the term “pH” that this state is analogous to that in water.
Results and Discussion 31PChemical Shifts and POP Coupling Constants. Triphosphates. The shifts and coupling constants ( J ) of several
common nucleoside triphosphates in water and tetramethylurea ( p H 7.40) are presented in Table I. In all cases the y phosphate experiences a pronounced upfield shift of ca. 3 ppm in tetramethylurea relative to that in water. Similarly, the a phosphate is shifted upfield, but to a lesser degree. For the ribose nucleotides, this shift is about 0.8 ppm; for deoxy-ATP this shift is more pronounced (2.04 ppm). The 0 groups undergo the smallest shift changes between aqueous and nonaqueous media (ca. 0.3 ppm). This behavior is not due solely to a solvent effect on the N M R parameters, for when different anhydrous solvents such as dimethylformamide or pyridine were employed, the A T P spectra were virtually identical with Materials and Methods those in tetramethylurea. Furthermore, this effect is not due Instrumentation. Samples were examined by the method of 31P to intrinsic behavior of the tripolyphosphate chain, because nuclear magnetic r e s ~ n a n c e . ’The ~ . ~spectrometer ~ employed was a chemical shift changes observed for inorganic p o l y p h ~ s p h a t e s ~ ~ Bruker HFX-5 operating at 36.43 MHz for 3 1 P and 90 MHz for IH, do not parallel those found for nucleoside di- and triphosphates. and containing Fourier transform and broad-band proton decoupling It should be noted in Table I that, while there are essentially capabilities. I n this work, protons were routinely decoupled from no differences in /3 group chemical shifts between the purine phosphorus by broad-band irradiation techniques, to permit precise and pyrimidine phosphates in water, a shift difference for this measurement of the a-phosphate resonances. The proton-coupled grouping is observed in the anhydrous solvent. On the average, spectra were not sufficiently resolved to permit precise measurements the pyrimidine /3 groups lie 0.37 ppm to lower field than those of the POCH couplings, although it was clear that these parameters of the corresponding purines. also reflected the same differences observed with greater precision The coupling constants between the two pairs of phosphates, through the POP couplings. Chemical shift (6) data are reported in parts per million (ppm) a-0 and 0-7,are always equal in aqueous solution, and smaller relative to the usual standard of 85% H3P0412with chemical shifts than their corresponding values in tetramethylurea. Furtherincreasing with increasing field intensity, as is customary in 31PNMR. more, in tetramethylurea the a-0 coupling is larger than the The primary standard, however, was a capillary (1-mm) containing 0-r coupling by about 2.0 Hz. either acetone-ds, or 1 .O M methylenediphosphonic acid Diphosphates. The chemical shifts and coupling constants [(H0)20PCH2PO(OH)2] in DlO (pD 9.5, Na+ counter-cation in of muscle studies) coaxially mounted in the sample t ~ b e . ’ ~ . ~ ~ ~ ’ ~ some nucleoside diphosphates in water and in tetramethylurea (pH 7.40) aregiven in Table 11. Again, the chain terminal Preparation of Solutions. The solvent chosen was N,N,N‘.N’-te(now the p) group experiences an upfield shift in the anhydrous tramethylurea (TMU),16which contributes no free protons, and whose solvent relative to that in water (ca. 2 ppm). In contrast to the monoprotonated species exhibits a pK value of
