Operons: the lac operon

Why a bacterium bothers to switch genes off

Making proteins costs energy and raw materials. An E. coli cell that made lactose-digesting enzymes all the time would waste resources whenever there was no lactose around. So bacteria switch many genes on only when they're useful, and they mostly do it at the first possible step: starting transcription.

Bacteria often group genes that work together into an operon: a cluster of genes that share a single promoter and are transcribed together into one polycistronic mRNA. The operon also includes a control switch called the operator, a short DNA sequence overlapping the start of transcription. Whatever sits on the operator decides whether RNA polymerase can get through.

The parts of the lac operon

• lacZ codes for β-galactosidase, the enzyme that splits lactose into glucose and galactose.
• lacY codes for lactose permease, a membrane protein that carries lactose into the cell.
• lacA codes for a transacetylase. Its role in lactose use is minor, and the pathway lab leaves it out.
• lacO, the operator, is a DNA site. It doesn't code for anything; it's where the repressor binds.
• The promoter, where RNA polymerase binds, sits next to a binding site for an activator protein called CAP.
• lacI is a separate gene nearby with its own promoter. It codes for the lac repressor (LacI), a protein that's made all the time.

Negative control: the repressor and its inducer

With no lactose around, the lac repressor binds tightly to the operator and blocks RNA polymerase from transcribing lacZYA. This is negative control: a protein that switches genes off by binding DNA.

The repressor has a second binding site, for a small molecule called an inducer. When the inducer binds, the repressor changes shape (an allosteric change) and lets go of the operator. Now RNA polymerase can transcribe the genes. An operon that's normally off and gets switched on by a signal molecule is called inducible.

Here's a surprise: the natural inducer isn't lactose itself. It's allolactose, a close relative of lactose made by a side reaction of β-galactosidase. That creates a chicken-and-egg puzzle. To make allolactose, the cell needs some β-galactosidase, and to get lactose inside it needs some permease, but those are the very proteins the repressor is holding back.

The answer is that repression is never perfect. Every so often the repressor falls off long enough for one transcript to slip through, so an uninduced cell keeps a handful of permease and β-galactosidase molecules. This trickle is called basal expression, and it's what lets lactose get in and be turned into the first allolactose.

Positive control: CAP and cAMP

Removing the repressor isn't enough on its own. The lac promoter is weak: RNA polymerase binds it poorly. To get strong transcription, an activator is needed. This is positive control.

The activator is CAP (catabolite activator protein). CAP binds DNA next to the promoter only when it's holding a small signalling molecule, cyclic AMP (cAMP). Bound CAP–cAMP helps RNA polymerase attach, boosting transcription many times over. And cAMP levels are high when glucose is scarce and low when glucose is plentiful.

Glucose also acts a second way. While glucose is being taken up, one of the proteins of the glucose transport system blocks lactose permease, so less lactose gets in and less allolactose is made. This is called inducer exclusion. Current research suggests inducer exclusion, more than the cAMP effect, is the main reason E. coli uses glucose before lactose.

Putting the two controls together

The two controls answer two different questions. The repressor asks, "Is there lactose?" CAP asks, "Is there glucose?" Strong expression needs the answers "yes" and "no".

• No lactose, no glucose: repressor bound. Only basal expression, even though CAP–cAMP is high.
• No lactose, glucose present: repressor bound and little CAP–cAMP. Basal expression only, the lowest of all.
• Lactose, no glucose: repressor released and CAP–cAMP high. Full expression.
• Lactose and glucose: some repressor released, but little CAP–cAMP and less inducer getting in. Low expression, well below full but above basal.

Now Monod's diauxic growth curve makes sense. With both sugars, the cells use glucose first and the lac operon stays low. When the glucose runs out, cAMP rises and more lactose gets in, but it takes time to make enough permease and β-galactosidase. That pause is the flat step in the curve. Then growth resumes on lactose.

Mutants: reading the switch by breaking it

Jacob and Monod worked out the logic by studying mutants. Geneticists describe a strain's lac phenotype by what happens to β-galactosidase with and without inducer. Inducible means made only with inducer, like the wild type. Constitutive means made all the time. Uninducible means it stays at the basal level even with inducer.

• I⁻ (no working repressor): nothing blocks the operator, so the operon is constitutive.
• Iˢ (super-repressor): the repressor still binds the operator but can't bind inducer, so it never lets go. Uninducible, even with IPTG.
• Oᶜ (constitutive operator): the operator sequence is changed so the repressor can't bind it. Constitutive.
• Z⁻: no working β-galactosidase. Lactose also can't induce the permease, because no allolactose is made. IPTG still induces permease normally.
• Y⁻: no working permease, so lactose can't get in to induce anything. IPTG, which doesn't need permease, still induces β-galactosidase.
• crp⁻ (no CAP): the operon can be induced, but only weakly, whatever the glucose level.

Cis and trans: merodiploids

I⁻ and Oᶜ look the same in a normal cell: both are constitutive. To tell them apart, Jacob and Monod used a trick. They gave cells a second copy of the lac region on a small extra DNA molecule called an F′ plasmid. A cell with two copies of a region like this is a partial diploid, or merodiploid. It's written chromosome / F′, for example I⁻ Z⁺ / F′ I⁺ Z⁺.

The key difference is what each part is made of. The repressor is a protein that diffuses through the cytoplasm, so it can reach operators on any DNA molecule in the cell. It acts in trans. The operator is a stretch of DNA that can only control the genes physically attached to it on the same molecule. It acts in cis.

• I⁻ / F′ I⁺: inducible. The repressor made from the plasmid's I⁺ gene swims over and binds both operators. I⁺ is dominant over I⁻.
• Iˢ / F′ I⁺: uninducible, even with IPTG. The Iˢ repressor sits on every normal operator in the cell and ignores inducer, even while the normal repressor is being released. Iˢ is dominant over I⁺.
• Oᶜ Z⁺ / F′ O⁺ Z⁻: β-galactosidase is constitutive. The working lacZ is next to the broken operator.
• Oᶜ Z⁻ / F′ O⁺ Z⁺: β-galactosidase is inducible. Oᶜ only frees the genes on its own molecule, and here that's a broken lacZ. The working lacZ is next to a normal operator. Oᶜ is cis-dominant: it's dominant only for genes on the same DNA.

1. 1Is there an Iˢ anywhere in the cell? If so, every O⁺ operator stays blocked, with or without inducer.
2. 2If not, is there at least one I⁺? If so, every O⁺ operator is repressed without inducer and released with it. If there's no I⁺ (all I⁻), every operator is always free.
3. 3Now look at each DNA molecule on its own. An Oᶜ operator is always free, whatever the repressors are doing.
4. 4Finally, a gene's protein is made only if its operator is free and the gene itself works. Check lacZ and lacY on each molecule separately, then add up.
5. 5Remember lactose's extra requirement: it can only induce if the cell has at least some working permease and some working β-galactosidase. IPTG has no such requirement.

A repressible contrast: the trp operon

The lac operon is off until a food molecule turns it on. The trp operon, which codes for the enzymes that make the amino acid tryptophan, works the other way round. It's normally on, and it's switched off when tryptophan is plentiful. That makes it repressible.

Its repressor is made in an inactive form that can't bind the operator. When tryptophan builds up, tryptophan binds the repressor and activates it, and the complex shuts the operon down. A small molecule that switches a repressor on like this is a corepressor. The logic fits the job: a cell doesn't need to make tryptophan when it already has plenty.

The trp operon also has a second, finer control called attenuation. It depends on bacteria translating an mRNA while it's still being made. If tryptophan is plentiful, a ribosome moves quickly through a short leader sequence at the start of the mRNA, and the RNA behind it folds into a hairpin that stops transcription early. If tryptophan is scarce, the ribosome stalls at tryptophan codons in the leader, the RNA folds a different way, and transcription carries on.

Key terms

Operon

A group of bacterial genes transcribed together from one promoter and controlled by one operator.

Operator

DNA site where the repressor binds; it controls only the genes on the same molecule (cis).

Lac repressor (LacI)

Diffusible protein that binds the operator and blocks transcription until inducer binds it.

Allolactose / IPTG

The natural inducer, made from lactose by β-galactosidase / a synthetic inducer that needs no permease and isn't broken down.

CAP–cAMP

Activator complex, abundant when glucose is low, that helps RNA polymerase bind the lac promoter.

Inducer exclusion

Glucose uptake inhibits lactose permease, so less inducer enters the cell.

Cis-acting / trans-acting

Affects only the DNA molecule it's on (e.g. operator) / acts on any copy in the cell (e.g. a repressor protein).

Merodiploid

A bacterium with a second copy of part of its genome, e.g. on an F′ plasmid.

Inducible / repressible operon

Normally off, switched on by an inducer (lac) / normally on, switched off by a corepressor (trp).

Check yourself

Try answering in your head before you open each answer.

• 1.A haploid strain is lacY⁻ but otherwise wild type. Predict β-galactosidase levels with lactose and with IPTG (no glucose in either case).Show answerHide

  With lactose, it stays at basal level: without permease, lactose can't get in, so no allolactose forms. With IPTG, which enters without permease, the repressor is released and β-galactosidase is fully induced.

• 2.Predict β-galactosidase for Iˢ Z⁺ / F′ I⁺ Oᶜ Z⁺ with no inducer and with IPTG.Show answerHide

  Constitutive in both. The Iˢ repressor keeps the chromosome's O⁺ operator shut even with IPTG, so its lacZ stays basal. But the plasmid's Oᶜ can't bind any repressor, so the plasmid's Z⁺ is always transcribed. Oᶜ is the one thing Iˢ can't overcome.

• 3.In a diauxic growth experiment with glucose and lactose, a researcher adds cAMP to the medium at the start. Would the lag between the two growth phases vanish completely? Explain.Show answerHide

  Probably not. Extra cAMP would restore CAP activation, but glucose would still inhibit lactose permease through inducer exclusion, so less allolactose would form while glucose lasts. Expression during the glucose phase would rise somewhat but stay below full, so the preference for glucose would weaken rather than disappear.