An important step in proving the principal hypothesis of the operon model was the isolation of the Lac repressor and the demonstration of its expected properties. Walter Gilbert and Benno Mullier-Hill succeeded in isolating the Lac repressor from E. coli extracts.They fractionated proteins by standard techniques and then assayed individual fractions for their ability to bind [¹⁴C]IPTG (one of the gratuitous inducers). Binding was detected by equilibrium dialysis.
    The Lac repressor is a homotetramer with a molecular mass of 154 kDa. Each subunit, which is made of 360 amino acids, can bind one molecule of IPTG. Crude cell extracts bind about 20 to 40 molecules of IPTG per cell, so there are roughly 5 to 10 repressor molecules per cell. Support for the idea that the IPTG-binding protein is the lac repressor comes from the observation IPTG-binding protein is absent in extracts of ldcl- mutants. Still stronger support comes from the observation that lacI mustants, which introduce amino acid substitutions in lac repressor, alter the repressor's affinity for IPTG.
    Because the number of repressor molecules is extremely small, these molecules must be translated from no more than one or two repressor mRNA molecules transcribed per generation time. The number of mRNA molecules is so small that either repressor synthesis itself is regulated or the mRNA is transcribed from a weak promoter. Both mechanisms have been observed for regulation of repressor synthesis in other operons, but for the Lac repressor the second explanation is correct, that is, repressor mRNA is transcribed constitutively from a weak promoter. The reason for the small number of repressor molecules is made clear from the properties of several mutants in which the weak lacI promoter is converted to a strong promoter. These mutants are noninducible because it is not possible to fill a cell with enough inducer to overcome repression.
    Repressor-overproducers have been extremely valuable experimentally because high concentrations of repressor (about 1% of the cellular protein) have in turn meant that very large amounts of repressor could be purified, providing sufficient amounts for physical study and characterization. With purified repressor the specific binding of repressor to the operator sequence and the inhibition of this binding by an inducer have been demonstrated.
    An important procedure for studying repressor-operator binding is the nitrocellulose filter assay. Proteins stick to these filters but DNA does not. If a mixture of repressor and radioactive lac DNA is passed through such a filter radioactivity will be retained on the filter if the protein and the lac DNA form a complex. The data were obtained by means of this test. The results indicate that Lac repressor binds to DNA with a normal lac operator but fails to bind to DNA with a lacO^c mutant operator. Furthermore, IPTG prevents the Lac repressor from binding to DNA with a normal lac operator. These studies confirm the major predictions of the operon model.