Textbook:    Strachan and Read, Chapters 9, 10; JJ Pasternak Chapter 11                               

SLIDES A

SLIDES B

I. INTRODUCTION

1.a The discovery of the phenomenon:

1b. relationship, if any, to X-inactivation??

1c. so far observed only in eutherian (placental) mammals: "a phenomenon in search for a reason";

.. IMPRINTING ..

3. the big questions:

4. Does evolution teach us about a rationale for imprinting in eutherians?

5. What is the role of methylation in imprinting? What signals initiate de novo methylation?

Some specific examples of phenotypes in mice and humans

a. mouse   see Jaehnisch et al. ( TIG 13:323 (1997))

The phenotype is that described for the knockout mice, i.e. the absence of function; the same phenotype would be seen in heterozygous (+/-) mice if the mutated allele was on the chromosome from which it is expressed.
One could go into the specific function of each of these genes and discuss the developmental biology/physiology and try to explain the phenotype, but this is not relevant for this particular discussion.
One could also ask if the type of genes/functions involved give any clue about the rationale for imprinting: is it a type of dosage compensation? (see later)

b. in humans there are several examples, but the most famous one is the Angelman syndrome and the Prader Willi syndrome ; other examples: the Beckwith-Wiedeman syndrome (involves the IGF2 gene); pediatric tumors adult cancers

PWS and AS:

Since the deletion is cytologically visible, it has to include a significant amount of DNA (~ 2 Mb), and probably more than one gene (now established by molecular methods). A simple early idea was that there were two imprinted genes, A and B, one expressed from the paternal chromosome, the other expressed from the maternal chromosome; if the deletion was in the paternal chromosome, B was still expressed, hence PWS; if the deletion was in the maternal chromosome, A was still expressed, hence AS.

-it is a little more complicated.

II. METHYLATION OF MAMMALIAN DNA

1. the phenomenon

2. methyltransferases

III. ESTABLISHMENT OF METHYLATION PATTERNS DURING DEVELOPMENT

Consider: (Fig. 1 from Jaehnisch)

methylation level:

-- high in gametes and zygote,
-- low in blastocyst and PGC (primordial germ cells)
-- intermediate in placenta and yolk sac (YS),
-- very high in somatic lineages and adult cells,

-- in other words, during the early cleavage divisions methylation is erased, but when the blastocyst starts to differentiate and implanting occurs, one sees a dramatic rise in the methylation of DNA (many genes) in the somatic cell lineages, but not in the lineage leading to primordial germ cells and gonads

methylation plays a role in control of gene expression in somatic cells but not in gametogenesis

1. How is methylation related to imprinting? If methylation was the only story, and if methylation is totally erased during zygotic divisions, the distinction between the parental chromosomes would be lost; -- thus, one has to postulate that a specific region of parent-specific methylation (the imprinting box) is resistant to de-methylation

2. Experimental approach to study methylation during early embryogenesis:

Find an imprinted gene, eg. Igf2 or Xist

use of methylation-sensitive restriction enzymes to detect methylation in DNA examples:

isolate DNA from small number of embryonic cells;
-- restrict samples with either enzyme,
-- then use two oligonucleotides complementary to flanking regions to perform PCR: if the DNA is cut, there will be no product; if it is protected against the methylation-sensitive restriction enzyme, one will see an expected PCR product

Methylation cannot be established from DNA sequencing!! Methods using restriction sites and methylation-sensitive restriction enzymes are very tedious and in fact depend on the existence of such sites, (not all methylated Cs are part of such sites)

Current hypothesis (from the study of a few known genes in the mouse):

- all imprints are erased during primordial germ cell development
- male specific and female-specific de novo methylases are expressed during spermatogenesis or oogenesis, respectively
- alternatively, there must be specific chromatin conformations introduced at specific sites (genes) that constitute the imprint, but    have nothing to do with methylation

3. How does methylation repress transcription? {9548}

Definition of a CpG island:   >200 bp >0.5% G+C
- found by searching data bases
- many CpG islands are found in promoter regions of genes (~ 1kb upstream), but many are also found in coding regions,       introns and far downstream of the promoter
- methylation can occur at all such locations;
-- methylation in the promoter regions is frequently associated with silencing,
-- methylation in down stream segments is often associated with the active gene (eg. on the active X-chromosome);
-- methylation may be correlated with increased rather than decreased level of gene expression

1. reduced binding of transcription factors (eg. CREB) to methylated recognition sites

2. specific factors recognizing and binding to methylated segments of DNA independent of sequence in association with other factors affect on chromatin structure
- two such binding proteins have been isolated: MeCP1 and MeCP2

Recent References

- Ben-Porath and H. Cedar (2000) Imprinting: focusing on the center. Curr. Opin. in Genet. and Developm.10: 550 - 554
- Ohta, T., Gray, T.A., Rogan, P.K., Buiting,K., Gabriel, J.M., Saitoh, S., Muralidhar, B., Bilienska, Krajewska, W.M., Driscoll, D.J., et.al., (1999) Am.J. Hum. Genet. 64: 397-414
- Nicholls, R.D., Saitoh, S., and B. Horsthemke (1998) Imprinting and Prader Willi and Angelman Syndromes. Tends Genet. 14: 194-200