The lac operon is an example of an inducible system in the sense that the synthesis of an enzyme is induced by the presence of its substrate. Repressible systems also exist, in which an excess of product leads to a shutdown of the production of the enzymes that synthesize that product. Such a control system has been identified for a cluster of genes controlling enzymes in the pathway for tryptophan production. Synthesis of tryptophan is shut off when there is an excess of tryptophan in the medium. Jacob and Monod suggested that the cluster of five trp genes in E. coli forms another operon, differing from the lac operon in that the tryptophan repressor will bind to the trp operator only when it is bound to tryptophan (see Figure 11-17). (Recall that the lac repressor binds to the operator except when it is bound to lactose.) A second control pathway also modulates tryptophan biosynthesis at the level of enzyme activity. This is termed feedback inhibition. Here, the first enzyme in the pathway, encoded by the trpE and trpD genes, is inhibited by tryptophan itself.

As with the lac operon, further analysis of the trp operon revealed yet another level of control superimposed on the basic repressor–operator mechanism. In studying mutant strains (carrying a mutation in trpR, the repressor locus) that continue to produce trp mRNA in the presence of tryptophan, Yanofsky found that removal of tryptophan from the medium leads to almost a tenfold increase in trp mRNA production in these strains, even though the Trp repressor was inactive, and thus could not account for the increase through normal derepression of the operator because of low tryptophan levels. Furthermore, Yanofsky identified the region responsible for this increase by isolating a totally constitutive mutant strain that produces trp mRNA at this tenfold maximal level, even in the presence of tryptophan, and he showed that this mutation has a deletion located between the operator and the trpE gene (see the map in Figure 11-17).

Yanofsky was able to isolate the multigene trp operon mRNA. On sequencing it, he found a long sequence, termed the leader sequence, of approximately 160 bases at the 5′ end before the first triplet in the trpE gene. The deletion mutant that always produces trp mRNA at maximal levels has a deletion extending from base 130 to base 160 (Figure 11-18). Yanofsky called the element inactivated by the deletion the attenuator, because its presence apparently leads to a reduction in the rate of mRNA transcription when tryptophan is present. Figure 11-19 shows the position of these elements in the trp operon. But what is the role of the leader sequence in bases 1 to 130? A surprising observation provides the key to solving this problem.

While studying mRNAs transcribed by the trp operon (using trpR⁻ mutants), Yanofsky discovered that, even in the presence of high levels of tryptophan (which should cause the attenuator region to reduce the rate of transcription tenfold), the first 141 bases of the leader sequence were always transcribed at the maximal rate, though the full-length mRNA was found, as expected, at levels only one-tenth as great. Another way of stating this is that, even in the presence of high concentrations of tryptophan, the first 141 bases are transcribed in maximal numbers but, because of some attenuating mechanism in this region, only 1 in 10 of the mRNAs can be transcribed farther (to completion). This suggests that the end of the attenuator acts as an mRNA chain terminator that, in the presence of tryptophan, halts transcription of 9 of 10 mRNAs. In the absence of tryptophan, this attenuator is somehow deactivated, and every mRNA goes to completion—hence, the tenfold increase. In those trpR⁻ mutants in which the attenuator also is deleted, there is no block to extension of the mRNA, so transcription is carried through in every case (that is, production is at its maximal rate) regardless of whether tryptophan is present.

What causes the interference with termination at the attenuator in the absence of tryptophan? Figure 11-20 presents a model based on alternative secondary structures formed by the mRNA in the leader region. The model proposes that one of the two conformations favors transcription termination and that the other favors elongation. Translation of part of the leader sequence would promote the conformation that favors termination.

It is known that a part of the leader sequence near the beginning is in fact translated and yields a short peptide of 14 amino acids. There are two tryptophan codons in the translated stretch of the leader mRNA (Figure 11-21). When excess tryptophan is present, there is a sufficient supply of Trp-tRNA to allow efficient translation through the relevant part of the leader mRNA. The mRNA passes through the ribosome at a sufficiently fast rate that segment 2 of the leader section is drawn into the ribosome before it can form a stem-loop structure with segment 3, as shown in Figure 11-20b. Segment 3 is thus able to form a transcriptiontermination stem-loop structure with segment 4. In conditions of low tryptophan levels, however, translation is slowed in segment 1 at the Trp codons by the relative unavailability of Trp-tRNA. As Figure 11-20c shows, this condition allows the stem-loop structure between segments 2 and 3 to form, preventing the formation of the transcription-terminating loop between segments 3 and 4. Consequently, under conditions of low tryptophan, transcription is not stopped by the attenuator. In this manner, an additional tenfold range of tryptophan biosynthetic enzymes is superimposed on the normal range that is achieved by repressor–operator interaction. The analysis of numerous point mutations in the trp leader sequence that favor or disfavor the respective secondary structures lends strong support to Yanofsky’s attenuation model.

Several operons for enzymes in biosynthetic pathways have attenuation controls similar to the one described for tryptophan. For instance, the leader region of the his operon, which encodes the enzymes of the histidine biosynthetic pathway, contains a translated region with seven consecutive histidine codons. Mutations at outside loci that result in lowering levels of normal charged His-tRNA produce partly constitutive levels of the enzymes encoded by the his operon.