Information contained in a primary sequence concerned with the helical configuration of the folded protein has previously been shown to be associated with single residues and with pairwise residue combinations. An attempt is made here to obtain a quantitative estimate of these two types of information and to evaluate their separate roles. An information theory approach has been used to obtain helix-forming information from 11 proteins of known sequence and conformation. The validity of these information measures is indicated by the precision of predictions made from them. Using single residue information alone, the observed helices are again predicted but, in addition, some non-helical regions are also predicted as helical. The main effect of adding in the residue pair information is to delete these false helices. It is concluded that residues of positive helix-forming information initiate the formation of helices and, in doing so, call down information in a limited number of pairwise interactions to delete certain of these helices. Such a code would be economical of information, leaving many pairwise (or higher) residue combinations to dictate tertiary folding. These information findings may be explained by a mechanism in which single residue, helix-forming information is expressed in the energy differences between helical and non-helical structure for those residues, dependent presumably on side chain-backbone interactions. From consideration of energy diagrams it is suggested that residue pair information is expressed by loss of a hydrogen bond on the COOH-terminal side of an induced α II conformation, forcing succeeding residues into a lower energy, non-helical conformation. Occurrence of residue pair information at the end of a helical region would act as a sharply defined helix-terminating device. Occurrence at suitable points in a run of potentially helix-forming residues would prevent helix formation from taking place. This mechanism is in keeping with an earlier observation of α II conformation residues at the COOH-termini of helical regions in proteins. The relationship of these findings to the folding of polypeptides is discussed in terms of a micelle hypothesis.
Read full abstract