Regulation of Protein Synthesis: An Alternative to the Repressor-Operator Hypothesis

REGULATION OF PROTEIN SYNTHESIS: AN ALTERNATIVE TO THE REPRESSOR-OPERATOR HYPOTHESIS* M. GRUBER and R. N. CAMPAGNE\ The synthesis ofproteins is often regulated by specific "effectors," i.e., inducers and "co-repressors." These substances—usually compounds of low molecular weight—can elicit or inhibit the co-ordinated synthesis of a group of metabolically related proteins, e.g., the enzymes of a metabolic pathway. This, and genetic evidence, ledJacob and Monod [i, 2] to propose the Operon, a stretch of DNA comprising several structural genes, thus coding for a number of polypeptide chains. At one end of this Operon, eitherjust outside or just within the first structural gene, an operator was assumed switching the operon on and off. The effectors of protein synthesis, or possibly their immediate derivatives, were assumed to interact with a "repressor," preventing the transcription ofthe operon into RNA. The stereospecific interaction with the effector would change the repressor by an allosteric transition [3], thus inducing or destroying its affinity for the operator and thereby preventing or allowing the transcription ofthe operon. The repressor was assumed to be the product of a regulator gene which is not necessarily spatially or structurally related to the structural genes ofthe operon. This concept and its elaboration had a very stimulating effect on research in the field of regulated protein synthesis and created order in a mass of seemingly unrelated phenomena. One of the predictions of the analysis byJacob and Monod was the existence ofmessenger RNA, now a concept well established in molecular biology. Subsequent studies by many authors have shown that the operon is a reality. In a number ofcases, the clustering on the chromosome and the * An earlier version was published in Koninkl. Ned. Akad. Wetenschap., Proc, Series C, 68: i, 1965. t Professor and Associate Professor, respectively, Biochemistry Department of the University, Groningen, The Netherlands. 125 common regulation of genes for enzymes acting in the same metabolic pathway have been demonstrated [4]. The operon is probably transcribed into a single polycistronic messenger RNA which has only one starting point for the ribosome (polarity effects). It has been shown that the order offormation ofthe different enzymes ofan operon upon induction occurs sequentially beginning at the operator end [5]. Further, evidence has been obtained that regulation by effectors occurs at the level of the transcription of DNA into RNA, which does not mean that additional regulation at the level ofpolypeptide formation is excluded or that there cannot exist a coupled regulation at both levels [6]. Nature ofthe Repressor The nature [7] ofthe repressor, however, or indeed its existence in the sense ofJacob and Monod's hypothesis, has remained uncertain. Whereas genetic analysis has unequivocally demonstrated the existence of regulator genes, i.e., genes which through a cytoplasmic product control the induction or repression process, the identification of this product as a repressor directly interacting with an operator region has not been possible experimentally. Such a repressor should have two properties which, in our opinion, cannot be ascribed to any single class of substances: (1) The repressor should interact stowspecifically with the effector, and by this interaction should undergo conformational changes. (2) The repressor should interact se^HeMce-specifically with a rather long polynucleotide sequence. The first property requires that the repressor is a protein, since only proteins are known to possess such spatial specificity; it excludes DNA and RNA. The second property requires that the repressor is a polynucleotide, since proteins of a reasonable molecular weight cannot interact specifically with more than a few bases (or base pairs). In E. coli each operator should consist ofat least ten base pairs in order to specify each individual Operon;1 in higher organisms the number should be even larger. Such numbers of 1 The number for E. coli can be derived from either of the following calculations: (i) n base pairs are needed to discriminate between each individual operator and any other base sequence in the chromosome: 4 X io6 (total number ofbase pairs in the E. coli chromosome) ^ 411 (number of possible combinations with four bases at 11 sites); (2) five base pairs for an unequivocal signal indicating "starting point ofan operon" (any sequence offour base pairs will occasionally occur among neighboring...