Cells with similar DNA can make different products because regulatory conditions differ.
By the end: Separate a regulator acting through molecules from a sequence element that must be on the affected DNA.
A promoter is DNA involved in recruiting transcription machinery. A regulatory protein can bind DNA and change transcription. In eukaryotes, DNA wraps around histones and forms chromatin. Accessibility, transcription-factor binding, distant regulatory elements, and molecular contacts affect which genes are expressed.
A cis-regulatory change acts through the DNA region physically linked to the affected gene copy. A trans-acting regulator is a diffusible product, such as a transcription factor, that can affect multiple suitable targets. These terms describe how an effect acts, not whether it is strong or beneficial.
A regulatory protein encounters DNA through molecular motion. It does not read a chromosome as a person reads a sentence. Its surface can form favorable contacts with particular DNA shapes and chemical groups. Some sequences therefore retain that protein more often than others. Occupancy means how often, or in what fraction of a collection, a site has a protein bound. Changing the protein concentration or its binding properties can change occupancy without changing the DNA sequence. Transcription then depends on what that bound protein does to the machinery at the site.
A repressor can obstruct a step needed for transcription, while an activator can help recruit or stabilize transcription machinery. Neither description requires permanent attachment. Proteins bind and leave, and transcription initiation can occur repeatedly. A drawing of a bound repressor represents a condition that favors repression, not a photograph showing that every gene copy is blocked at every instant. This explains why a strongly repressed gene can still have basal expression, meaning a small background level. Stronger repression and complete absence of transcription are different claims that would need different evidence.
Consider two copies of an invented gene in one cell, called copy A and copy B. Both have suitable binding sites for the same activator. Now suppose one base changes in the promoter of A, weakening recruitment there. The altered DNA remains physically attached to A. It does not move to B and replace B's promoter. If the activator and all other conditions stay the same, the direct effect belongs to A. This is a cis effect. The comparison concerns the two linked DNA regions, not two different versions of the activator protein.
Now keep both promoter sequences unchanged, but reduce the amount of functional activator in the shared cellular space. Both copies can receive fewer productive encounters. The effect acts in trans because a diffusible molecule mediates it. Diffusible means able to move away from where it was made and reach other molecules. The activator's own gene occupies a particular place on a chromosome, but that location does not make the protein's action cis. The classification follows how the regulatory effect reaches a target, rather than where the regulator was originally encoded.
A third target, copy C, helps separate trans action from indiscriminate action. Suppose C lacks a suitable binding site for this activator. Changing the activator can directly affect A and B while leaving C largely unaffected through that particular mechanism. A diffusible regulator does not have to affect every gene. Conversely, a cis element can influence more than one nearby transcription unit through local contacts. Counting affected genes is therefore a poor definition of cis or trans. Physical linkage and the molecule that carries the effect give the more useful distinction.
These examples predict directions under stated assumptions, not numerical expression levels. A weaker binding site might have little visible effect when the regulator is abundant, because sufficient binding still occurs. It might matter more when the regulator becomes scarce. Another factor could also compensate at the promoter. To argue for a binding mechanism, evidence must connect the changed sequence or protein to occupancy and then to transcription. A final RNA measurement alone leaves RNA stability as an alternative explanation. Even a correct cis or trans label does not supply those missing steps.
In the bacterial lac example, a repressor can bind an operator and reduce transcription. An inducer derived from lactose, allolactose, reduces this binding. When glucose is scarce, cAMP-associated CAP activation can favor transcription. Relief from repression and positive activation are separate requirements for strong output in the simplified model.
The controls specify effective inducer presence and glucose availability. With no inducer, output is repressed rather than mathematically zero. With inducer but abundant glucose, output is low. With inducer and scarce glucose, output is high. This is a qualitative state model, not measured rates. It omits transport, inducer exclusion, metabolism, and feedback.
In the lac example, the operator is a DNA region recognized by the repressor. The inducer and the operator are different kinds of objects. Allolactose is a small molecule that binds the repressor protein, rather than a replacement stretch of DNA. Binding at one part of a protein can change its behavior at another part. This is allosteric regulation. Here the change reduces the repressor's ability to occupy the operator. The DNA sequence can remain unchanged while the chemical conditions change which protein state is favored and how strongly repression acts.
This mechanism differs from destroying the repressor. A cell can contain repressor protein while much less of it binds the operator. An assay that measures total repressor amount would not necessarily detect the regulatory change. The relevant question concerns the protein's binding state under those conditions. Likewise, calling allolactose an inducer describes its effect in this system, not a rule that every small sugar activates every gene. Its molecular interaction with this particular repressor supplies the explanation. Different proteins can respond to different ligands, meaning molecules that bind them.
The glucose input uses a separate route. Cyclic AMP, abbreviated cAMP, is a small signaling molecule associated here with glucose availability. It binds catabolite activator protein, or CAP. Under glucose scarcity, the cAMP-associated CAP state can bind regulatory DNA and favor transcription machinery recruitment. CAP does not remove the operator or convert the repressor into RNA polymerase. It supplies positive regulation at a distinct site. This separation lets a cell combine information about an alternative sugar with information about glucose, rather than treating either signal as the whole explanation for strong expression.
Work through the abundant-glucose comparison first. With effective inducer absent, operator occupancy by the repressor is favored and output is repressed. Add effective inducer while keeping glucose abundant. The repressor-binding obstacle is reduced, but strong CAP-associated activation is still missing. The simplified prediction is low output. The difference between these cases shows why releasing repression can increase expression without producing the high-output state. Nothing in the qualitative labels tells us how many times larger low output is than repressed output, and both can be greater than zero.
Next compare the scarce-glucose cases. Without effective inducer, the display allows CAP-associated activation but retains repression. Its output remains repressed. With effective inducer present, repression is relieved and activation is available, so output is high. The first case is a counterexample to the claim that glucose scarcity alone guarantees high lac expression. The second satisfies both conditions in the model. Treat these as statements about a specified bacterial regulatory arrangement. Human genes use many other combinations of factors and cannot be assigned this truth table merely because they also have promoters.
| Effective inducer | Glucose | DNA-binding state | Output |
|---|---|---|---|
| Absent | Abundant | Repressor favored; CAP absent | Repressed, not zero |
| Absent | Scarce | Repressor favored; CAP bound | Repressed, not zero |
| Present | Abundant | Repressor reduced; CAP absent | Low |
| Present | Scarce | Repressor reduced; CAP bound | High |
The term effective inducer matters because the controls start after several biological steps have been compressed. Lactose outside a cell is not identical to allolactose acting on a repressor inside it. Entry, metabolism, and the state of transport proteins can affect that connection. The widget deliberately supplies the effective condition rather than calculating it. Basal expression also makes the real system different from a perfect digital switch with an impossible zero-to-on transition. Keeping these distinctions prevents the simple four-case comparison from becoming an unsupported claim about the full response to nutrients.
Site positions are schematic. Binding conditions, not molecule sizes, carry the comparison.
Repressed, not absolute zero
Repressor binding suppresses output even when an activator is available.
| Effective inducer | Glucose | Qualitative transcription |
|---|---|---|
| Absent | Either state | Repressed, with possible basal expression |
| Present | Abundant | Low |
| Present | Scarce | High |
Use the regulation model to make two paired comparisons, rather than selecting a high-output state alone. Begin with inducer absent and glucose abundant. Then select glucose scarce without changing inducer. CAP changes from absent to bound, while the repressor stays bound and the output stays repressed. The unchanged output does not mean that the glucose control failed. The molecular readout changed, but repression still constrains transcription in this simplified model. This distinction between an intermediate state and the final output is why both panels need to remain visible during the comparison.
Return to abundant glucose and select inducer present. Now the repressor becomes absent from its DNA site and output becomes low. Keep inducer present and select scarce glucose. CAP binds and output becomes high. Comparing these last two states isolates the activation input when repression is relieved. Comparing the earlier pair isolates that same input while repression remains. The effect of an input depends on the other condition. This is a conditional relationship, not a contradiction and not evidence that CAP changes its identity between button presses.
The visible bound and absent labels are simplified state assignments. Real binding is reversible, and a collection of molecules can have mixed occupancy. The model does not calculate a binding probability or a distribution of RNA counts across individual cells. Its high, low, and repressed categories are ordered descriptions with no supplied numerical distance between them. Drawing a bar twice as tall for high output would invent a fold change. A suitable visual instead shows the two binding sites and marks the qualitative outcome in words, keeping the missing quantitative information visible.
Consider a paper counterfactual that the controls do not implement. Suppose the operator on copy A can no longer bind the repressor, while the operator on copy B remains intact in the same cell. Removing effective inducer would no longer restore repression through A's damaged site, although it could restore repression at B. This is a cis change in the response to a shared trans-acting protein. Under abundant glucose, A still lacks strong CAP-associated activation. Loss of operator repression therefore need not produce the original model's high-output condition by itself.
A different counterfactual removes functional repressor protein while leaving both operators intact. Both suitable copies now lack repression through that protein. This is a trans change, but it still does not guarantee high output without the activation condition. Comparing these two counterfactuals helps locate the mechanism. One changes a DNA target; the other changes a molecule shared by targets. Neither case requires a changed coding sequence in the regulated genes. These are reasoning extensions, not hidden widget modes, and their predictions assume that no other regulator compensates for the stated change.
Finally, a button press is not a biological time course. The display replaces conditions immediately to show the selected regulatory state. Existing RNA and protein would not disappear merely because transcription became repressed. Their amounts would depend on continued synthesis and removal over time. A delayed protein response would therefore be compatible with an early change in DNA binding. The next lesson develops that distinction quantitatively. Here the defensible conclusion concerns which regulatory conditions favor transcription, while the timing and magnitude of accumulated products remain outside this model's predictions.
RNA stability, translation, and protein degradation also regulate output. Similar total RNA can hide different splice products or translation rates. A chromatin mark associated with low expression does not by itself prove that adding that mark caused the reduction. Marks can be consequences or parts of larger regulatory processes.
Epigenetic regulation concerns states beyond DNA sequence, with definitions differing over whether persistence or inheritance is required. Do not infer transmission across generations of organisms from a short-lived cellular response. A useful claim specifies the measured mark, cell type, duration, and tested causal mechanism.
In a eukaryotic nucleus, a suitable DNA sequence can be present but difficult for a regulatory protein to reach. DNA wraps around histone protein assemblies to form nucleosomes. A binding site facing or contacting the histone surface is not equivalent to the same site exposed on an accessible stretch. Chromatin remodeling means changing this organization, for example by moving a nucleosome relative to the DNA. Such changes can expose a site without rewriting its bases. Accessibility supplies an opportunity for binding, but the needed factor must still be present and functional.
Chemical modifications can also alter chromatin interactions. Acetylation adds an acetyl group to selected protein sites. On certain histone lysines, it reduces positive charge and can affect interactions with negatively charged DNA. Modified histones can also recruit proteins that recognize the modification. This is more specific than saying that every mark opens or closes every gene. The effect depends on the modified residue and the proteins present. DNA methylation adds a methyl group to particular bases without substituting a different sequence letter. Its regulatory interpretation likewise depends on genomic and cellular context.
Suppose two cell populations have the same sampled promoter sequence, but one shows lower RNA and more of a particular chromatin mark. The association permits several explanations. The mark might help reduce transcription, a repressor might recruit the machinery that adds it, or a common change might produce both observations. A sequence match rules out a difference only in the sequence actually examined. It does not establish that the mark is the cause. A stronger causal account needs evidence that distinguishes these alternatives and connects accessibility or recruitment to new RNA production.
Regulation can continue after the RNA leaves its DNA template. A protein can recognize a sequence in an untranslated region of mature mRNA and affect its stability or access to translation machinery. Untranslated means not decoded into the main peptide, not absent from the RNA. A change there can alter protein output without changing the encoded amino-acid sequence. Short regulatory RNAs called microRNAs can help associated protein complexes recognize target messages through base pairing. Their effects can involve translation and RNA removal, so a protein change alone does not identify which occurred.
A persistence claim adds another question. Imagine a low-expression state that remains in daughter cells after the original signal disappears. That observation concerns memory within a cell lineage. It does not by itself show transfer through reproductive cells to an organism's offspring. Continued expression of a regulator, restoration of chromatin organization after DNA copying, and other cellular processes can contribute to persistence. Naming a state epigenetic does not choose among those mechanisms. The useful description specifies what persists, in which cells, over what interval, and whether the initiating signal is still present.
These mechanisms can combine rather than replace one another. Exposing a promoter can permit a transcription factor to act, while an RNA-binding protein later shortens the message's persistence. The final protein amount can conceal these opposing changes. A claim that a gene is on or off therefore needs a named readout. DNA occupancy, new RNA production, accumulated RNA, protein amount, and functional activity are different quantities. Matching the evidence to the proposed regulatory step gives a testable explanation without assuming that a chromatin mark, a sequence difference, or a single endpoint identifies the whole cause.
Effective inducer is present, but glucose remains abundant. Does this simplified model predict maximal transcription?
There is also a positive regulator.
| Condition | State |
|---|---|
| Repressor binding | Reduced |
| Activator binding in this model | Absent |
| Output | Low |
No. Repression is relieved, but strong CAP-associated activation is absent. Output is low rather than high. Removing one obstacle is not equivalent to supplying every activating condition.
A promoter change affects one gene copy. A defective transcription factor changes several genes with compatible binding sites. Which is cis and which is trans?
Locate the altered DNA element or diffusible product.
| Change | Mechanism |
|---|---|
| Promoter on one allele | Cis, linked DNA element |
| Transcription-factor product | Trans, acts at compatible targets |
The promoter change acts in cis on the linked copy. The transcription-factor change acts in trans through its product on multiple targets. The number of affected genes alone is not the definition; physical linkage versus a diffusible regulator is.
Selected textbook passages support the mechanisms below. The prose and diagrams are original, and examples are constructed rather than measured. These references are not a full literature review or independent specialist validation. Optional links are not runtime dependencies.
Repressors, activators and lac regulation. The qualitative model omits inducer transport and feedback.
DNA accessibility and chromatin. Association with a mark does not establish a universal causal rule or inheritance across organisms.
RNA-binding proteins, microRNAs and RNA stability. Half-life calculations here are explicit first-order teaching models.