4198
pprn which is considered t o be characteristic of the acylated H 2 C 0 group of glycerol. l 4 Additional signals are: a triplet at 0.89 ppni, accounting for the terminal methyl groups; a single peak at 1.28 ppm. associated with the internal methylene groups of the aliphatic chains; a triplet at 5.37 ppni. representing the isolated olefinic groups; and a n apparent doublet corresponding to the CH, groups in the position cy to the carbon double bonds centered at 2.03 ppm. The doublet which appears at 2.27 ppni can be correlated with an cy-CH? group of the acyl functions. The apparent doublet occurring at 3.83 ppm can be assigned to the CH?O group of glycerol connected to the ether linkage. The signal of the acylated H C O group of glycerol near 5.17 ppm is partially obscured by the triplet representing the isolated carbon double bonds. A signal nenr 2.9 ppm, characteristic of a methylene group between two double bonds,'.' is absent. Approximately two out of three aliphatic chains per molecule are monounsaturated.
+
+
II
0- C -R"
+
+
Table 1. Melting Transitions for O l i g o d e o ~ y t h ~ i n i d y l a t e Pol! deoxyadenylate Complcxesrz Ohgo-P CoI-ICIl.
21
Sol\ent
x
A
8
14.0
2.88
4.57
55.0 13.5 14.5
2.88 2.97 2.88
3.13
10,
~~
NaCI, 0.15 .M; sodium citrate, 0.015 ,\I (SSC) Potassium pliosphate, 40 m.W, pH 7 Potassium phosphate; 30 mM. pH 7; with 8 m:24\.lgCI?
0 H-C-
raphy on diethylaminoethylcellulose. The melting behakior of the complexes d(pA),,, d(pT),,, d(pA),i d(pT),,,, and d(pA),, d(pT),, formed by mixing the constituents in a suitable solvent, was then observed by ultraviolet absorbance changes. The stoichiometry of the interactions was determined from mixing curves or (more directly) by gel filtration on Sephadex G-R~OO.~ The interaction of d(pA),, with d(pTjrfLfollows the B / m and the stoichionirelationship 1000/Tl = A etry is 1 : 1 in equimolar mixtures in all solvents used (Table I). This interaction is a multimolecular process
~
4.78
1.26
+
(1 T h e relationship. 1000 T I = A B / m is observed. T h e values in the table are A a n d B for this equation, in different solvents a n d at different oligodeoxynticleotide coiiceiitratioiis (expressed in nucleotide residues). Polymer is present in an equivalent a m o u n t .
H-C*-O-C-R'
I
Chemical reactions, specific optical rotations, and spectroscopic data prove that the fraction (1) isolated from the liver of Hjdrolugrrs colliei consisted of D ( + ) 1-0-cis-alk- 1 '-enyldiglycerides. (14) C. Y . Hopkiiis in "Progress i n the Chemistry of Fats a n d Other Lipids," Vol. 8, R . T. Holman, Ed., Pergamon Press, Ne\v York, N. Y . , p 213.
H. H. 0. Schmid, \V. J. Baumann, H. K . Xlangold L>iicersi/.v of hfiiiiiemto, Thr Hoi~inel/ii.\iiriii A l l S f in, .Mime.t I I ~ O 554 I 2 Rccrii.ec/ Jiinr 30. 1967
Oligodeoxynucleotide-Pol ydeoxgnueleotide Interactions. Adenine-Thymine Base Pairs'
Sir: Quantitative data on the interaction of structurally defined oligonucleotides with complementary polynucleotides are of value in practical and theoretical studies on nucleic acid structure. A n investigation of oligodeoxynucleotide-polydeoxynucleotideinteractions, covering a broad range of chain lengths in both members of the interacting pairs, displays partial confirmation of theory a n d some anomalous behavior. A series of oligodeoxynucleotides, d(pA),,l and d(pT),,' with m = 2-25, was prepared by D N a s e I degradation of d(pA),i and d(pT),i followed by chromatog(I)Supported by Grant CA-08487 from the U. S . Public Health
Service.
( 2 ) Abbreviations uscd: m , oligonucleotide chain length: 17, ;i pol\mer chain length greater than 300 iiucleotides; Ti !, temperature a t the midpoint of the ultraciolct thermal transition i n degrees ldroxyl group; d(pA),,. p o l y d e o x ~ a d c n y l a t s ; ~ ( P T ) polydeoxythymidylate: ~~, DNase I, pnncrcatic deox)rihonuclcasc: ribonucleotide polymers are prefixcd by r.
Joitmul of the Americnn Chemicrrl Sociefj, I 89.18
and thus T , . \ d u e s 3 (Table I) and the slope of the transition are quite sensitive to oligonucleotide concentration. These results are in accord with statistical thermodynamic treatment of oligo and polymer interactio n s. The interaction of equimolar amounts of d(pA)fnwith d(pT),, i n SSC is complicated by a change in the stoichiometry of the I ~ C I J ' O P complex present to 1A--2T when ~n < 16. The double-stranded (1 : 1) complex is not detected in the interaction in\,olving d(pA)7, but it is clearly evident as a n early rise in absorbance at 260 nip (or a decrease at 284 mp) when d(pA),,,, rn = 8-12, complexes with d(pT),, are melted. The presence of complexes of different stoichiometry in the mixture could result from a slow approach to equilibrium. Repeated analysis on a n equimolar mixture of d(pA)s and d(pT),i o \ e r a period of 3 months demonstrated that the amounts of double helix and triple helix, detected by melting, remain unchanged. We conclude that a n equilibrium mixture of double and triple helices are present in this case. The triple-stranded complexes d(pA)!n-2d(pT),, (with 112 < 16), obtained by mixing the ( 3 ) F. J. Bollum, "Procidurcs i n Nucleic Acid Rcsedrch," G . Cantoni D. D,i\ics, E d . , H a r p e r l i ~ i c ! R o u , I I ~ c .W , e n York, N.Y . , 1966, 111) 577-592. x 100 cm coIuI11II 011 Scphtiticx ( 4 ) Seg;lrations \ \ e r e mC1deon 1' 1 G-700 equilibrated \\it11 the 'ippropriate solvent and dc\clopetl a t d temperature about 1 5 ' helo\\ the T I of the complex to be tested. Dct,iiletl results \\ 111 be presented i n ii f u t u r e publicdtioii. ( 5 ) M. N.Lippsctt, L. A . Heppel, a n d D. F. Bradley, Biochiu~,Biophi I . 4 c t a 41, 175 il960), h,i\c o b s e r \ c d thc dcpcndcncc upon t h e total iiuclcotidc coiicciitratioii iii ii s t u d y of thc intcractions of r(pA),,zkrith pol! rU. ( 6 1 i d ) \V,S . hl,igi.e, J r . , J . H . Gibbj, ,Ind 13. H . Zimnl, Biopoll.)?lers, 1, I33 (1963). S o t c t h a t t h e iiuthors IiilLe c:illctl rittciitioii to the depetidciicc on absolute c l c t i \ i t y of the oligonuclcotidc i n e q 19, a n d i i l . ~ ~ c o r c l ~ i nith i c c their trc:itment o u r B = B ' 111 C D , \\ here B ' = slope ,it u n i t 'icti\ity, C = ,ibsolute a c t i v i t y of the m-~iicr,m i d D = the ratio of tu L> intcrnal p,irtition functions. We iissume t h a t the absolute acti\itb is cssi.ntinll!. constant for all m-niers 'it the iiuclcotide residue coiiceiitr,itioii uscd ( I S x 10-8 M), (b) W . S . Magee, J r . , J . H . Gibbs, 'itid G. F. Kc\\cll,,/, Chem Ph,is., 43, 2115 (1965). liiiii
Airgirsr 30, 1967
constituents in 1 :2 proportion, undergo a single-step thermal transition observed a t 260 and 284 mp. The latter wavelength is specific for the triple strand melting.7 Values of l/T1,,?,when m = 6-9, for a n oligonucleotide residue concentration of 10 p M d o follow a linear relationship, but A = 3.05 and B = 2.20 (Figure 1). The 1 : I complex for equimolar mixtures of d(pA), and d(pT), is completely formed only when m > 16 in SSC, and these double-stranded complexes have TI,? similar (within the range of our experimental error, =t0.5 ") t o those formed by d(pA),, and d(pT), a t equal values of rn.8 l/Tl,/lvalues now fit on the same line (Figure 1). The anomalous stoichiometry described above for a Naf concentration of 0.19 M (SSC) is also exhibited in 40 m M phosphate buffer, p H 7.0, with or without the presence of 8 m M MgC12(data not presented here).g The over-all reactions that may be postulated t o occur in our studies are given in eq 1-3 (major complex at equilibrium is underlined) when equimolar amounts of d(pA),
+ d(pT), I-d(pA),,-d(pT)n
+ d(pA)n
(1)
d ( ~ A ) n - 2 d ( ~ T ) m d(pA)n
+
(2)
+ d(pA)m -
(3)
d(pA),-2d(pT)n dCPA)n
d(pA),a
+ d(pT)m s d(pA)n-d(& + d(pT)n z1.0 M ) where all the d(pA,,) or d(pA),, form a triple-stranded complex with d(dT),,. The complexities observed in the present case require that the contribution of the chain length to the stability of oligo-polymer interactions can be evaluated only after careful determination of the stoichiometry of the complexes formed. In the "critical region," where stoichiometry values are changing, the exact molecular stoichiometry will be i n de t e r m i n a t e. ( I I ) Department of Genetics, Univcrsity of Pn\ia, I"i\ia,
Itcll),
G . Cassani,ILF. J. Bollurn De/lurr/,ic./l/ of' BiOC/IC~/lli.\/ r y L'/ricersity of Ktritrrcky, Lrsirigroii. Kwrritck): Receicc4 .Lfiiy 4. 1967
Di-t-butyl Trioxide and Di-t-butyl Tetroxide
~
(7) M . Riley, B. Maling, and M. J. Chamberlin, J . Mol. Eiol., 20, 359 (1966); F. J. Bolluni, unpublished data, 1965. (8) A . M. Michelson, Boll. Soc. Chem. Eiol., 47, 1 5 5 3 (1965), has sho\\n that complcxcs of different stability are formed when r(A), or r(U),a of thc. same chain length are mixed hith the compleinentary ribopolymer, but the stoichiometry of the complexes formed was not discussed. (9) The choice of solvents m a y be considered arbitrary, but in fact the Mg?+ phosphate buft'er is uscd for the enzymatic replication of homopolydeoxyiiucleotidcs. This investigation was a result of problems encountered i n the rcplicarion study. (10) R . D. Blake a n d J. R . Fresco, J . Mol. Eiol., 19, 145 (1966).
Sir : Kinetic, tracer, and product studies of autoxidat i ~ n ' - support ~ the sequence of reactions 1-3 as controlling chain termination. Experimental observations
(I) (2) (3) (4)
H. S. Blanchard, J . Ani. Chcnr. SOC.,81, 4548 (1959). P. D. Bnrtlett and T. G . Traylor, ibid., 85, 2407 (1963). J. R. Thomas, ibid., 87, 3935 (1965). G. A. Russell, ibid., 79, 3871 (1957).
