Premium
Thermodynamics of the various forms of the dodecamer d(ATTACCGGTAAT) and of its constituent hexamers from proton nmr chemical shifts and UV melting curves: Three‐state and four‐state thermodynamic models
Author(s) -
Pieters Jane M. L.,
Mellema JanRemt,
Van Den Elst Hans,
Van Der Marel Gijs A.,
Van Boom Jacques H.,
Altona Cornelis
Publication year - 1989
Publication title -
biopolymers
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.556
H-Index - 125
eISSN - 1097-0282
pISSN - 0006-3525
DOI - 10.1002/bip.360280304
Subject(s) - dodecameric protein , chemistry , duplex (building) , thermodynamics , ionic strength , ionic bonding , enthalpy , stacking , aqueous solution , molecule , dna , ion , organic chemistry , biochemistry , physics
Chemical shifts of base and H1′ protons of the single‐stranded hexamers d(ATTACC) and d(GGTAAT), of the 1 : 1 mixtures of these complementary hexamers, and of the self‐complementary dodecamer d(ATTACCGGTAAT) were measured at various temperatures in aqueous solution. Four different sample concentrations were used in the case of the dodecamer and of the mixture of the complementary hexamers; the individual hexamers were measured at two different DNA concentrations. Absorbance temperature profiles at five different NaCl concentrations were measured for the dodecamer in order to quantify the effect of the ionic strength on the duplex formation. Under suitable conditions of nucleotide concentration, temperature, and ionic strength, the dodecamer adopts either a B‐DNA duplex or a hairpin‐loop structure. Chemical shift vs temperature profiles, constructed for all samples, were used to obtain thermodynamic parameters either for the various stacking interactions in the single strands or for the duplex or the hairpin‐loop formation. In the analysis of the duplex formation of the hexamers, a two‐state approach appeared too simple, because systematic deviations were revealed. Therefore, a new three‐state model (DUPSTAK) was developed. In order to investigate the magnitude of error arising from the use of the two‐state approach in cases where the DUPSTAK model appears more appropriate, a series of test calculations was made. The magnitude of error in the enthalpy and in the entropy of duplex melting is found to depend linearly upon the actual melting temperature and not upon the individual Δ H O dand Δ S O dvalues. Thermodynamic analysis of the chemical shift vs temperature profiles in D 2 O solution (no added salt) yields an average T md value of 341 K (1 M DNA) and Δ H O dof −121 kJ · mol −1 for the dimer/random‐coil transition of the hexamer duplex d(ATTACC) · d(GGTAAT). For the duplex ⇄ random‐coil transition of the 12‐mer d(ATTACCGGTAAT) an average T md value of 336 K (1 M DNA) and Δ H O dof −372 kJ · mol −1 are found. The hairpin/random‐coil transition of d(ATTACCGGTAAT) is characterized by a rather large Δ H O hvalue, −130 kJ · mol −1 , and an average T mh value of 304 K.