Differential gene expression

Concept

Differential gene expression

T-061Home BU-207Threads information · regulation

Statement

One genome gives rise to many cell types.

Why it matters

This result is the central organising principle of developmental biology: it resolves the apparent paradox that a neuron and a liver cell, despite carrying essentially identical DNA, are entirely different in structure and function. embryonic-induction and morphogen-gradients describe the signals that trigger this differential expression in specific cells at specific times, and hox-genes-body-plan describes one particularly well-studied category of the genes being differentially expressed; all three depend on the more basic claim established here, that a single genome is capable of producing many distinct expression outcomes depending on context.

Hypotheses

Essentially all somatic cells of an organism retain a complete, unaltered copy of the genome (genomic equivalence).Without this assumption, cell-type differences could simply be explained by different cells carrying different genes, which would make "differential expression" unnecessary as an explanation; genomic equivalence is what forces the explanation to lie in gene regulation rather than gene content. Which genes are transcribed, and how much, can be controlled independently in different cells, via regulatory DNA sequences and the transcription factors that bind them.This is the regulatory machinery differential expression actually runs on: promoters, enhancers, and the transcription factors (themselves gene products) that bind them, allowing an identical genome to be read out differently depending on which transcription factors happen to be present and active in a given cell. Once established, a cell's particular pattern of gene expression can be stably maintained and passed on to its daughter cells through further divisions, even after the original inducing signal is gone — a form of cellular memory distinct from, but related to, the heritable chromatin modifications treated separately under epigenetic-inheritance.

Proof

\text{Nuclear transfer from a fully differentiated somatic cell into an enucleated egg can produce a complete, normal organism.}

This classic experimental result (most famously in frogs, and later extended to mammals) directly demonstrates genomic equivalence (Hypotheses): if the somatic donor nucleus had lost or permanently altered any genes required for development, the resulting organism could not develop normally across all cell types, yet it does. A

\text{Different cell types express markedly different sets of genes, despite sharing this identical genome (Step 1).}

Direct measurement of which genes are transcribed in different tissues (historically by hybridisation-based methods, now routinely by RNA sequencing) confirms that only a fraction of the genome is actively transcribed in any given cell type, and that this active fraction differs substantially between cell types — the observation the entire result exists to explain. A

\text{Regulatory DNA sequences (promoters, enhancers) and the transcription factors that bind them determine which genes are transcribed in a given cell.}

A gene's regulatory sequences act as a docking site for a specific combination of transcription factors; only in a cell where the right combination of factors is present and active does RNA polymerase get efficiently recruited to that gene's promoter, explaining how expression can differ between cells sharing identical DNA sequence at every gene, regulatory sequences included. A

\text{Some of the genes turned on in a given cell are themselves transcription factors, which go on to activate or repress a further set of genes.}

This creates cascading regulatory networks, in which an initial, relatively small difference in signal or transcription-factor activity between two cells (established, for example, by an inductive signal or a cytoplasmic determinant) is amplified through successive rounds of transcription-factor activity into a large, stable, and self-reinforcing difference in overall gene-expression profile. B

Result

\text{Identical genome} + \text{differing regulatory-factor activity} \to \text{differing gene-expression profile} \to \text{differing cell type}

Reading. Cell-type diversity is not explained by cells carrying different genetic information; it is explained by an identical genetic information set being read differently, cell by cell, according to which transcription factors happen to be active where.

Scope. Applies to essentially all normal somatic differentiation in multicellular organisms; the small number of documented exceptions (programmed genomic rearrangement in some immune and a few other specialised cell lineages) are notable precisely because they represent a genuine departure from genomic equivalence, rather than the general rule.

Corollaries & converses

embryonic-induction and morphogen-gradients both describe upstream mechanisms that generate the initial, cell-to-cell differences in transcription-factor activity that Step 4's cascades then amplify into full, stable differentiation.
hox-genes-body-plan concerns a specific, evolutionarily conserved category of transcription factors regulating other genes exactly as described in Step 4, applied specifically to specifying regional identity along an embryo's main body axis.
Converse: two cells found to differ in phenotype but shown to have an identical, unaltered genome must differ in gene regulation (Steps 3–4), not gene content — the logical basis on which nuclear-transfer and cloning experiments are interpreted as evidence for this result.

Fails without

Drop genomic equivalence (Hypotheses): if different somatic cell types genuinely carried different genetic content, nuclear transfer from a fully differentiated cell into an enucleated egg would be expected to fail, or to produce a defective organism missing whatever genes that particular donor cell type had lost; the actual success of such experiments (Step 1) is direct evidence against this alternative.
Drop independent, sequence-specific regulatory control (Hypotheses): without regulatory sequences capable of responding differently to different transcription-factor combinations, there would be no mechanism by which an identical genome could be read out differently in different cells, and every cell with the same DNA would necessarily express the same genes — directly contradicted by observed cell-type diversity (Step 2).

Common errors

Assuming differentiated cells have permanently lost the genes they no longer express; in the overwhelming majority of cases the genes are retained but transcriptionally silent, not deleted (Step 1, Hypotheses).
Treating "gene regulation" as a vague, unspecified process rather than the concrete, sequence-specific mechanism of Step 3: a defined regulatory DNA sequence, bound by a defined combination of transcription factors.
Assuming a single transcription factor alone typically determines a full cell fate decision; most differentiation outcomes depend on combinations of multiple factors acting together, and on the cascading amplification described in Step 4.
Confusing differential gene expression (which genes are transcribed and how much, at any given moment) with epigenetic-inheritance's specific concern, the heritability of chromatin-based regulatory states across cell divisions or generations — related but conceptually distinct.

Discussion

John Gurdon's nuclear-transfer experiments in the 1950s and 1960s, transplanting nuclei from differentiated frog intestinal cells into enucleated frog eggs and obtaining viable, normal tadpoles, provided the decisive early experimental evidence for genomic equivalence (Step 1); this work, together with Shinya Yamanaka's later demonstration that mature cells could be reprogrammed back to a pluripotent state using a defined set of transcription factors, was jointly recognised with the Nobel Prize in Physiology or Medicine in 2012.

A cell's expression state, once established, is often reinforced and stabilised by chromatin-level modifications (DNA methylation, histone modification) that make the relevant regulatory sequences more or less accessible to transcription factors; this additional layer of stability, and its capacity to be inherited through cell division independently of the original inducing signal, is the specific subject of epigenetic-inheritance.

Common misconception: that cloning an organism from a differentiated somatic cell nucleus (Step 1) implies the donor cell was somehow "reset" to an embryonic genetic state before development could proceed. The genome itself requires no genetic alteration at all; what actually happens is a resetting of the regulatory (expression) state, achieved by the egg cytoplasm's own reprogramming factors acting on an already genetically complete and unaltered donor genome.

Worked examples

\text{A liver cell and a neuron, from the same individual, are compared by RNA sequencing.}

Genes encoding liver-specific metabolic enzymes are found highly transcribed in the liver cell and essentially silent in the neuron, while genes encoding neuron-specific ion channels and synaptic proteins show the reverse pattern, despite both cells carrying an identical genomic DNA sequence, including at these very genes' own regulatory regions. A

\text{same genome, largely non-overlapping active transcriptomes}

Reading. The measured expression difference between the two cell types is precisely the observable signature the Result predicts: not different genetic information, but different regulatory readout of identical genetic information.

Scope. The same comparison, repeated for any two differentiated cell types from one individual, is expected to show the same qualitative pattern — substantial, largely non-overlapping active gene sets from an identical shared genome.

Problems

A researcher claims that muscle cells must have lost the genes for haemoglobin production, since muscle cells never make haemoglobin. Using Step 1, evaluate this claim and propose a more accurate explanation.
Solution
The claim is inconsistent with genomic equivalence (Step 1): nuclear-transfer experiments show that differentiated somatic nuclei, including from cell types that never make haemoglobin, retain a complete genome capable of directing development of an entire organism, haemoglobin-producing cells included. The more accurate explanation is that muscle cells retain the haemoglobin gene but keep it transcriptionally silent (Step 3), because the specific combination of transcription factors required to activate it is never present in muscle cells, not because the gene itself is absent.
Two cell types are found to differ in the expression of 500 genes. Investigation shows that only 3 "master" transcription-factor genes differ substantially in activity between them, with the remaining 497 genes downstream of those 3. Explain this pattern using Step 4.
Solution
This is the expected signature of a cascading regulatory network (Step 4): a small initial difference in the activity of a few master transcription factors is amplified through their downstream regulatory targets (which may themselves be further transcription factors) into a much larger overall difference in gene-expression profile, without requiring an equally large number of independent, primary regulatory differences.
Explain why reprogramming a differentiated adult cell back into a pluripotent state (as in induced pluripotent stem cell technology) does not require altering the cell's DNA sequence, using the Hypotheses.
Solution
Because the differentiated cell already retains a complete, unaltered genome (genomic equivalence, Hypotheses), the pluripotent state it is being returned to is already fully genetically encoded within it; what differs between the pluripotent and differentiated states is only which genes are actively transcribed (Step 3). Reprogramming therefore only needs to reset the pattern of transcription-factor activity to that characteristic of pluripotent cells, not to introduce or restore any missing genetic sequence.