The Nature of Mutations II

J Pediatr Endocrinol Metab. 2002 Dec;15 Suppl 5:1279-88. Related Articles, LinksImprinting disorders: non-Mendelian mechanisms affecting growth.Butler MG.Section of Medical Genetics and Molecular Medicine, Children 's Mercy Hospitals and Clinics and The University of Missouri-Kansas City School of Medicine, Kansas City, MO 64108, USA. mgbutler@cmh.eduMost autosomal genes are expressed from both maternal and paternal alleles. However, imprinted genes are an example of non-Mendelian genetics, in which only one member of the gene pair is expressed and expression is determined by the parent of origin. Imprinted genes may account for 0.1-1% of all mammalian genes. At least 50 imprinted genes have been identified in humans, and imprinted genes frequently cluster under the control of an imprinting center. Many imprinted genes contribute to growth, either as growth factors, such as insulin-like growth factors (IGF2 in Beckwith-Wiedemann syndrome), or as growth inhibitors, such as the GRB10 gene in Russell-Silver syndrome. Imprinted genes have evolved over time in mammals to fine-tune the growth of the fetus. Paternally expressed genes generally enhance growth, whereas maternally expressed genes appear to suppress growth. In addition, normal and abnormal genomic imprinting and loss of heterozygosity contribute to a wide range of malignancies. A common process for controlling gene activity is methylation, which can be changed during male or female gametogenesis. Examples of classic human disorders related to genomic imprinting are Beckwith-Wiedemann syndrome (chromosome 11), Prader-Willi/Angelman syndromes (chromosome 15), Russell-Silver syndrome (chromosome 7), and Albright hereditary osteodystrophy (chromosome 20). Several of these disorders are discussed and illustrated.If you look in the abstract, it says that methylation can be changed during gametogenesis...so like a DNA mutation, methylation mutations can occur during gametogenesis and cause phenotypic change.

Page not found - The University of North Carolina at Chapel HillThis link (unless I am mis-understanding -- which is highly
likely) suggests to me that the 'switches' are mapped off of
a different area of the genome of the cell ... that is they
are cell products -- so they must get 'built' from DNA ...????This abstract also made me think this way ...Disruption of an imprinted gene cluster by a targeted
chromosomal translocation in miceMichele A. Cleary1, Catherine D. van Raamsdonk1,
2, John Levorse1, Binhai Zheng3, Allan Bradley4 &
Shirley M. Tilghman1Genomic imprinting is an epigenetic process in which the
activity of a gene is determined by its parent of origin.
Mechanisms governing genomic imprinting are just beginning
to be understood. However, the tendency of imprinted genes
to exist in chromosomal clusters suggests a sharing of
regulatory elements. To better understand imprinted gene
clustering, we disrupted a cluster of imprinted genes on
mouse distal chromosome 7 using the Cre/loxP recombination
system. In mice carrying a site-specific translocation
separating Cdkn1c and Kcnq1, imprinting of the genes
retained on chromosome 7, including Kcnq1, Kcnq1ot1, Ascl2,
H19 and Igf2, is unaffected, demonstrating that these genes
are not regulated by elements near or telomeric to Cdkn1c. In
contrast, expression and imprinting of the translocated
Cdkn1c, Slc22a1l and Tssc3 on chromosome 11 are affected,
consistent with the hypothesis that elements regulating both
expression and imprinting of these genes lie within or
proximal to Kcnq1. These data support the proposal that
chromosomal abnormalities, including translocations, within
KCNQ1 that are associated with the human disease
Beckwith-Wiedemann syndrome (BWS) may disrupt CDKN1C
expression. These results underscore the importance of gene
clustering for the proper regulation of imprinted genes.

Certain sequences are more likely to be imprinted than others is all. so that is the association with a "switch" or a DNA sequence. However, an imprinting mutation would not be observable as a change in DNA sequence.

OK..... so there are enzymes (?) in the germ cells, that
persist in the zygote, but are not manufactured within the
zygote.These are then replicated for each cell during
division, but some cells get different 'regulators' (somehow).Man that's complicated!!! So complicated I cannot image that
it was designed at all [This message has been edited by Peter, 07-09-2003]

These enzymes are not 'replicated'. In some organisms it is the case that maternal mRNAs may code for proteins expressed in the embryo up until a certain point, such as the mid blastula transition in frogs. In mammals however zygotic transcription begins as soon as first cleavage has ocurred at the 2 cell stage.The cells do not get different patterns of methylation, 'regulators', until they are much more advanced as part of the normal process of development.

So how are these 'enzymes' (or whatever is the approriate
term) copied into the 'new' cell after a division?

It is the same as with any other proteins, some will be inherited directly from the parent cell and some will be produced de novo in the new cell, provided they are receiving suitable regulatory cues for expression of the appropriate gene. A cell does not lose its entire protein complement directly before cell division, any number of cytoplasmic and membrane bound proteins may persist.This varies from protein to protein, and off the top of my head I don't know the exact behaviour of all the methyltransferases, but I'm sure the information is out there.

So the only way that these enzymes can be 'different'
is if they are damaged during the division.They may cause different patterns of expression, but they
can only be considered to have 'mutated' if they were
damaged during the cell's 'creation' ... ???

No one is saying that the enzymes are different. What is different is the inherited pattern of methylation on the CpG residues or alternatively patterns of methylation/acetylation on the nucleosomal proteins.

Oh, now I understand ... sort of. Thanks.If the methylation state is inherited from the parent,
unchanged then it cannot be considered a mutation in any
sense.It's just like inheriting any other feature from a parent.If an unchanged set of whatever-ases can cause phenotypic differences
it leads me back to having a super-class of heritable variation.
[But ultimately it's no different to having two brown-eyed parents
having a blue eyed child -- phenotype is different but it's not
becuase of a 'mutation']If there has been no chemical change then there had been no
mutation (in the sense of 'unexpected change' rather than my
preference).[This message has been edited by Peter, 07-11-2003]

Having a superclass of 'heritable variation' is fine, the entire point we have been discussing is using 'mutation' as a term for this superclass with specific subclasses such as genetic and epigenetic. There are some really freaky types of heritable variation, such as those caused by an endocytic parasite which clearly couldn't be considered a mutation but genetic and epigenetic variation I would say could both arguably be classified as mutation and both of which are adressed as such in the scientific literature. You say perhaps I shouldn't have used the word inherited, what I mean is that the methylation status of the zygote is dependent on factors that have affected the germ cells methylation status. It is non-sensical to talk about an embryo inheriting its parents entire pattern of gene methylation as methylation patterns, unlike genes, vary hugely throughout differing tissues.Once again the methyltransferases are not causing the phenotypic changes, the changes are due to the effect of methylation on the regulation of the genes. Obviously the methylation of cytosine is a chemical change associated with the DNA, but it is not a genetic base change.[This message has been edited by Wounded King, 07-11-2003]

Here is quite an interesting paper, they agree with you Peter that an epigenetic change is not a mutation, instead they call it an epimutation, and I have to say, I quite like 'epimutation' and 'epimutant'. It also might help you understand epigenetics better, something I seem to be singularly failing to do.

Thanks for that ... it has made things a little
less murky for me It leads me to think in terms of using 'mutation' as
a qualifying term that means something like 'unpredicted change'.Heritable variation then has sub-classes of genetic mutation
and epigentic mutation ... and possibly others as yet
undiscovered.For me, then, the ultimate source of heritable variation is
the somewhat vague notion of a 'cell copy-error'. Either a segment
of DNA gets reproduced differently to the parent cell (genetic
mutation) or the epigenetic markers get 'copied' differently (epigentic mutation).... which is probably no more helpful than
a sack full of possums doing the tango.....hmmm maybe Peter Borger should get let back in I'd be interested
to see what he does with this information

Why 'unpredicted change'? Why not just 'change'?

Just to cover the possibility that there may be discovered
at some time in the future mechanisms within cells that
cause changes in a predictable fashion.I don't think there will be, but you never know.