BSMS205 · Genetics

Gene Regulation

Chapter 28 · Part V · Functional Genetics

Today's central question

Same DNA. Same library.
Why do different cells
read different chapters?

The central dogma · revisited

DNA → RNA → protein
Multiple control points at every step
Transcriptional regulation is the most fundamental
If a gene isn't transcribed, nothing downstream matters

Roadmap

Promoters · the local on-ramp
Pre-initiation complex and pause-release
Enhancers · long-range control
Combinatorial logic and cell-type specificity
3D chromatin looping · CTCF and cohesin
TADs and enhancer hijacking
The IGF2 / H19 case study

§ 1

Promoters

What a promoter is

DNA region immediately upstream of a gene
Where RNA polymerase II assembles to begin transcription
Contains short binding motifs · TATA box · Inr · CpG island
The on-ramp to the gene

Common promoter elements

Element	Position	Role
TATA box	~30 bp upstream	Recruits TBP · classic but not universal
Initiator	Overlaps TSS	Positions Pol II in TATA-less promoters
CpG island	±200 bp around TSS	Marks housekeeping genes

Building the pre-initiation complex

TFIID (with TBP) binds the TATA box
TFIIB joins · then RNA Pol II + TFIIF
TFIIE and TFIIH arrive · TFIIH unwinds DNA
Pol II begins transcription

Promoter-proximal pausing

After ~20–60 bp, Pol II pauses
Like a runner waiting in the starting blocks
Released by P-TEFb phosphorylation of Pol II CTD
Enables fast activation in response to signals

§ 2

Enhancers ·
Long-Range Control

What an enhancer is

Short DNA element · 50 – 1500 bp
Binding platform for multiple transcription factors
Recruits coactivators (Mediator, p300/CBP)
Loops through 3D space to contact a promoter

Distance independence

Can be 1 megabase from the gene
Upstream · downstream · or in an intron
Works in either orientation
Example: SHH gene · ZRS enhancer >1 Mb away

Combinatorial logic

Multiple TFs must bind together for an enhancer to fire.

Muscle enhancer: MYOD + MEF2 + SRF
Neuron enhancer: NEUROD + REST + SOX
Active only in cells with the right TF combination

Histone marks identify active enhancers

Mark	Meaning
H3K4me1	Marks all enhancers (active and poised)
H3K27ac	Marks active enhancers
eRNA	Short bidirectional transcripts at active enhancers

Rapid evolution · functional conservation

Enhancer sequences evolve faster than coding regions
Human-mouse enhancer similarity: ~50–60% (vs 85–90% for coding)
But essential developmental enhancers are deeply conserved
Species differences (e.g. human brain) emerge from enhancer changes

§ 3

3D Chromatin Looping

The cast of characters

CTCF · DNA-binding protein · loop anchor · ~40 – 60K sites genome-wide
Cohesin · ring-shaped complex · embraces DNA · extrudes loops
Cohesin slides until stopped by CTCF
Convergent CTCF orientation (><) → stable loop

Two architectural protein systems do most of the loop-forming work. C T C F is a D N A-binding protein with about forty to sixty thousand binding sites scattered across the human genome. It acts as a loop anchor and as an insulator. Cohesin is a ring-shaped protein complex that physically embraces D N A and slides along it, extruding loops bidirectionally until it hits something that stops it. The thing that stops cohesin is C T C F — but only when C T C F is in a specific orientation. The C T C F binding site is asymmetric, and the protein blocks cohesin only from one direction. So loops form when two C T C F sites point toward each other — convergent orientation, drawn as right-arrow then left-arrow. Over ninety percent of chromatin loops have convergent C T C F anchors.

The convergent orientation rule

CTCF convergent orientation determines loop formation — Cohesin extrudes until stopped by convergent CTCF anchors → enhancer-promoter loop forms.
Osato 2019, *bioRxiv*. CC BY 4.0.

Topologically Associating Domains (TADs)

Genomic neighbourhoods · 100 kb to a few Mb
Enhancer–promoter contacts within the same TAD
Boundaries: convergent CTCF sites
Crossing TAD boundaries is rare

Enhancer hijacking · when boundaries break

Chromosomal rearrangement disrupts a TAD boundary
Strong enhancer ends up next to the wrong gene
Classic example: T-cell leukaemia · TCR enhancer near MYC
MYC overexpression drives the cancer

What happens when T A D boundaries break? Enhancer hijacking. A chromosomal rearrangement — translocation, deletion, inversion — can disrupt the convergent C T C F sites that define a boundary. Suddenly an enhancer that was confined to its native T A D can reach across and activate a gene in a previously separate domain. The classic example is in T-cell acute lymphoblastic leukaemia. A translocation between chromosomes eight and fourteen brings a T-cell receptor super-enhancer into proximity with the M Y C oncogene. M Y C, a master regulator of cell proliferation, becomes massively overexpressed. The result is leukaemia. Enhancer hijacking is now recognised as a common mechanism in many cancers, especially where the protein-coding sequence of an oncogene is unchanged but its regulation is rewired.

§ 4

Case Study ·
IGF2 / H19

The IGF2 / H19 setup

Two genes near each other · share a distal enhancer
Between them: a differentially methylated region (DMR)
DMR can recruit CTCF — but only when unmethylated
Genomic imprinting · which gene is on depends on parent of origin

Paternal allele · IGF2 ON

DMR is methylated
CTCF cannot bind · no insulator
Enhancer loops to the IGF2 promoter
IGF2 expressed · H19 silent

Maternal allele · H19 ON

DMR is unmethylated
CTCF binds · creates an insulator
Insulator blocks the enhancer from reaching IGF2
Enhancer loops to H19 instead · H19 expressed · IGF2 silent

The architecture in one figure

IGF2/H19 chromatin loops — Paternal: methylated DMR · enhancer reaches IGF2.
Maternal: unmethylated DMR + CTCF · enhancer redirected to H19.
Merkenschlager & Odom 2013, *Cell*.

§ 5

Summary

What to take away

Cell identity = gene expression, not DNA sequence
Promoters: local on-ramp · pre-initiation complex · pause-release
Enhancers: long-range, combinatorial, cell-type specific
3D looping (CTCF + cohesin) brings them together
TADs insulate · breaking them causes enhancer hijacking
IGF2 / H19 — same DNA, opposite outcomes by 3D geometry

Next lecture

How do we measure
gene regulation in actual cells?

Chapter 29 · Gene Regulation — Methods and Applications