Biology Dept
Kenyon College
Chapter 10:
Eukaryotic Gene Control
Fall Section Spring Section 1 Spring Section 2
Eukaryotic RNA polymerases; in vitro transcription
Enhancer Elements on DNA
Transcription Factors and Transcriptional Activation
Post-transcriptional Control
Advanced Topic: Research on the Gridlock Transcription Factor

Eukaryotic Gene Control
Eukaryotic control sites include promoter consensus sequences similar to those in bacteria.

However, there can be many control sequences, called enhancers and silencers, responsive to many different signals.  Enhancers were defined by cis/trans complementation experiments, in which their activation only occurred when they were present on the same DNA helix with the gene under their control.  Thus they were originally called cis-acting elements; this terminology is still used in experiments defining new regulatory sites.

Three RNA Polymerases in Eukaryotes
Review from before break: Eukaryotes have three different RNA polymerases, which transcribe three different classes of genes.  RNA pol II transcribes hnRNA (precursor to mRNA).  RNA pol I and III transcribe functional RNAs such as rRNAs and tRNAs.

Initiation of RNA pol II transcription requires multiple basal transcription factors.  Most of these were identified initially through biochemical approaches, i.e. fractionation of nuclear extracts (by chromatography or density gradient centrifugation) and reconsitution of transcription in vitro.

For example, in this experiment, different purified basal transcription factors (TBP, TFIIB, IIF, IIE, IIH) and RNA polymerase II were mixed and matched to see which would support transcription from the adenovirus major late promoter.

(Shilatifard A, Haque D, Conaway RC, Conaway JW.  J Biol Chem 1997 Aug 29;272(35):22355-63)

Discovery of Enhancers:  Using recombinant DNA transfected into cultured cells.

Susumu Tonegawa:  Transcription of the human antibody heavy chain gene is under control of enhancer elements in Intron 1.

[Figure from Freeman, S. (2002)  Biological Science]

Assessment of enhancer elements using recombinant reporter genes in transgenic mice: 
Testis-specific Lactate Dehydrogenase C promoter.


Three different constructs contained different portions of the LDHC promoter coupled with beta galactosidase reporter gene.


X-gal staining indicates beta-galactosidase activity driven by DNA regulatory elements in the indicated construct. Which portion of the promoter supported the greatest testis-specific expression?

Tissue-specific function of regulatory elements in the LDHC promoter:
Panel A:  Compares beta-galactosidase activity in construct A (black bars) and construct B (white bars).

Panel B:  Compares beta-galactosidase activity in construct A (black bars) and construct C (white bars)

Note the difference in the scales on Y-axes in the two graphs.

From:  Li, S, W. Zhou, L. Doglio, and E. Goldberg (1998)  Transgenic Mice Demonstrate a Testis-specific Promoter for Lactate Dehydrogenase, LDHC  J. Biol. Chem. 273:31191-31194.

How do enhancer elements work to regulate transcription of specific genes in specific times and places?  By serving as binding sites for transcription factors--proteins that regulate transcription.

[Figure from Freeman, S. (2002)  Biological Science]

DNA:Protein and Protein:Protein interactions are important for transcription factor function.  Note modular structure of transcription factors:  one part of the protein is responsible for DNA binding, another for dimer formation, another for transcriptional activation (i.e. interaction with basal transcription machinery).

Dimer formation adds an extra element of complexity and versatility.  Mixing and matching of proteins into different heterodimers and homodimers means that three distinct complexes can be formed from two proteins.

[Figure from Freeman, S. (2002)  Biological Science]


Although they are cis-acting, the enhancers and silencers can be strung out across 10-20 kilobases (thousands of base pairs) of DNA upstream.  Some signals can even be downstream of the coding gene, or even found within introns (!)   How can this be possible?  Long regions of the DNA can loop over to enable the regulatory connections.

Based on Robert Tjian, "Molecular Machines that Control Genes," Scientific American.

  • Activators bind to enhancer sites, controlled by hormones or other signals.  They increase transcription of the regulated gene.
  • Repressors bind to silencer sites, controlled by hormones or other signals.  They decrease transcription of the regulated gene, possibly by interfering with activators.
  • Coactivators bind to activators and/or repressors  (at one end) and to basal factors (at the other end).  The coactivators somehow communicate the signal from activators and/or repressors to the RNA polymerase.
  • Basal factors act similarly to bacterial sigma factors.  They enable RNA polymerase to initiate transcription.  However, they require interaction with coactivators.

How were all these control proteins figured out?  Robert Tjian explains some experiments:

  • The researchers tested human cell extracts for a sigma-like protein: one that (1) bound to DNA and (2) stimulated RNA transcription in the test tube.  They tested many, many proteins, and found one: SP1.
  • SP1 only increased transcription when the DNA contained "GC box" sequence (an enhancer).  Without GC box, only basal (low-level) transcription occurred.
  • The "zinc finger" domain of SP1 was essential for binding to GC box.  The "glutamine-rich domain" was not needed for DNA binding, but was needed to increase transcription.  The researchers guessed that the glutamine-rich domain bound to basal factors needed for low-level transcription, and converted it to high-level transcription.
  • Basal "Factor D" (known to bind TATA box) was suspected to be the target of SP1.  To Tjian's surprise, however, when Factor D was better purified, SP1 failed to increase transcription.  Therefore he guessed that Factor D included the TATA Binding Protein plus some other factor.  The other factor(s) turned out to be eight coactivators.


Splicing of hnRNA to make mRNA
The first transcript of RNA from a eukaryotic gene is not yet ready for transcription.  It is called hnRNA,  for high-molecular-weight nuclear RNA.  In order for the RNA to exit the nucleus, and for  proteins to be translated by ribosomes in the cytoplasm, the following processing steps must first occur:

  • Capping of the 5' sequence with 5' methyl-7-guanidine (the "m-7-G cap")
  • Addition of a run of adenine nucleotides to the 3' OH end (the "poly-A tail")
  • Splicing out of the intron sequences
Interestingly, retroviruses such as HIV which use an RNA genome have a "cap" and "tail," enabling them to mimic harmless messenger RNA.

Post-transcriptional control

Degradation of mRNA.  Certain hormones can stimulate (or retard) the rate of degradation of mRNA, thereby decreasing (or increasing) its availability fortranslation to protein.
Translational repression.Translation of mRNA can be repressed.  For example, when iron is low, in human blood, a translational repressor protein binds to the mRNA encoding the iron carrier protein ferritin, and prevents translation of the iron carrier.

Post-translational control

Protein cleavage and/or splicing.  The initial polypeptide can be cut into different functional pieces, with different patterns of cleavage occurring in different tissues.  In some cases, different pieces may be spliced together.
Chemical modification.  Protein function can be modified by addition of methyl, phosphoryl, or glycosyl groups.
Signal sequences direct packaging and secretion.  Some proteins have "signal sequences" which direct their packaging in the Golgi and movement through the endoplasmic reticulum (ER) to be secreted.  The signal sequences usually end up cleaved off.

Problem.  Gene expression is subject to many levels of control.  Outline the mechanisms at all the levels of control, from gene sequence to final product.  Explain why control is useful at different levels.

Research Advanced topic:
"Gridlock," a Developmental Regulator in Zebrafish

Zebrafish is a major model system for vertebrate development.
The "gridlock" gene grl was discovered as a major developmental signal distinguishing between arteries and veins in the early vertebrate embryo.  (Zhong et al, 2000, Science 287:1820).

Genotype and phenotype of grl.

  • Chemical mutagenesis of a large population of zebrafish yielded some deformed embryos.
  • One deformed embryo lacked circulation to the back and tail, due to a blocked arterial junction--"gridlock".
  • By positional cloning (see below) the mutation was mapped and sequenced to a gene named grl.
  • A point substitution resulted in partial loss of function of the gene product.

  • When grl mRNA was injected, the mutant embryo developed normal arterial circulation!

                                                                    Zhong et al, 2000, Science 287:1820

How was grl mapped, located, and sequenced?
It is no small task to find one gene in a vertebrate genome of perhaps 50,000 genes, buried within 20X as much non-coding DNA.  Friday afternoon's Advanced Topic session will present the details.

  • Classical recombinant mapping (meiotic crossover analysis) between hybrid grl carriers and fish with various genetic markers.  These markers are sequence polymorphisms detectable by SSLP PCR.
  • The position of the mutation was narrowed down to ever smaller chunks of DNA by radiation hybrid cloning, yeast artificial chromosomes (YACs), bacterial artificial chromosomes (BACs), and P1 phage clones (PACs).
  • Bioinformatic analysis revealed exons and introns.  BAC and PAC clones were used to screen embryonic cDNA libraries for genes expressed during early development.
  • One gene was found which:
    • Was expressed ONLY in the embryonic aorta
    • Contains a point substitution of lysine instead of a stop codon--thus the protein extends 44 extra amino acids

                                                                    Zhong et al, 2000, Science 287:1820
What is the function of grl?
Bioinformatic analysis of the grl sequence identifies it as a transcription factor of the Helix-Loop-Helix family.  This protein motif (short conserved protein sequence) is found in many Drosophila homeotic developmental genes (more later.)   To find out about protein families and domains, see Procite. Click for Chimeview--Helix-Loop-Helix Model

The terminal ends of the protein form a clamp that binds DNA.  Major or minor groove?  Check the Chime model!

The protein encoded by grl appears to be a transcriptional repressor that distinguishes certain populations of aortic angioblasts (precursor cells of arterial structures).