Folate And Methylation - (Mar/31/2005 )

Dear Nick,
I would like to predict and discuss (in the most general way) the possible results, when someone investigating methylation , because I see some possible dificulties in the interpretation of results in the lots of such type experiments. I've met lots of results like this - 40% cancer cell samples had mutated and inactivated p53 or 70% promoters were hypermethylated and so on. I was thinking about the meaning of such results the years and would like to try to present the general model - how such nondeterminism appears in reality using the known facts (simply more order).
Well, we know now - thanks for the all Experimentators - DNA methylation pattern is something like cell memory state and on this depends lots of events in development and normal/abnormal cel functioning, but not the all.
Disregulation of methylation mechanism could occure when any chain of One carbon unit metabolism is affected by some reasons, even by special nutrition - it was obtained using experimental rats. So, some enzyme important for one carbon units providing carbon for methyl group, repression, carier folate, B12 insufficiency, providers Choline, Methionine insuficience, Serine Glycine metabolism changes could disregulate normal cell DNA, histones... methylation system. During the last years the role of different DNA methylases was evaluated too. I've studied, mainly, the problem - how S-adenosylmethionine insufficiency could appear, starting from some nutrition models, when Methionine...was eliminated, and continuing on facts about specific carcinogenes, that affect DNA methylation system (great puzzle and passians of facts). When significant disregulation of cell DNA methylation system happens, first the global DNA CpG hypomethylation appears, DNA changes it's configuration and some hypermethylated regions appears.
I've made prediction that such changes could depend on insufficience of substrate s-adenosylmethionine too, and that they are random - stochastic - probabilstic (facts proven by SUSAN CLARK - yeh)...and NOW the main point one more time: this depends on the chance which areas of DNA were affected by methylation changes:
if some oncogenes triggered on by hypomethylation or/and tummour supressor genes switched off by hypermethylation by chance, then malignant process starts...
but if some repression mutation deletion of tummour supressors appears by chance, for example, after carcinogenes action or radiation or...., then malignant process starts again...
The probability of those both situations occuarence is aproximately 5/1000 - cancer rate.
And this is the reason why the samples never gives unfortunately exact 100% answer about abnormal methylation or mutations. Some different mechanisms of switching (on or/and out) exists and experimentators usually measure only one of them sometimes obtaining confusing results. The process of malignisation seems to be not linear sequence of events too- this is one more great problem in the malignant growth process identification....(this is very simplified -sorry).
If methylation abnormality appears on 6q24 some gene, then neonatal diabetes mellitus starts - not cancer...but mutations of this gene are possible too... - great results.
if some abnormal methylation appears in embryo cell development regulating genes, then NTD could appear (hypothesis).
When studying NTD samples, it is interesting, how often abnormal histone methylation appear. If 75%, this could mean only that 25% were caused by some other reasons - not methylation problems.
The systematic point of view, using known facts, I would like to present in

The most interesting question for me is about the genes which can cause both developmental anomalies and cancer - PAX3(2q35), KIT (4q12), I've mentioned above - PTC(Patched) (9q22), RET (10q11), WT1 (11p13) - THE TIME dimension (moment of epigenetical/genetical change - before or after neurulation) problem we discussed above.
The malignant process starting cell in human, acording to my opinion, suffers some partial changes simmilar to embryo cell developmental program and starts to behave like being in pre/embryo, but cell's surrounding is absurd - no ectoderm mesoderm endoderm neighbours and no positive meaning. This opinion is supported by the facts about methylation role in embryo development and malignisation process, even by results about B12 forms...
Prof. Susan Clark group considering both developmental and cancer problems, succesfully working on methylation processes too, have choosed the great way, so You Nick, sharing Your job and contacting with those skillfull researchers are going to solve great problems, I hope. Good luck.

Kestutis Urba
P.S. I remember the nice time, I've spent. discussing and working together with biochemists - making calculations for cancer diagnosis and then trying to understand the theoretical meaning of obtained results about B12 forms usage. My speciality is applied mathematics, I'm playing not only piano, but chess not looking at them too, so when I say, that the problem is too complex even for me - I'm laughing little at myself - studies of math were extremly difficult, they 've learned to deal with a great amount of facts logically, but reading of biochemistry books and the articles later was impressive for me too. Bill Gates said: if I'm young I study gene engineering... but what to do with Google found 2 570 000 pages of web, mentioning DNA methylation? Real information digging technologies are needed...I've read: human ussually uses 7% of his possibilities, if 9% - famous proffesor, if 11% - genius, genome limit reached. My sister says, that I'm genius, but I can laugh only, because in Greek this means ...having the beard. I hope, I use 8.5% and trying to moove these problems for some mm forward. Molecular oncology is not the only my hobby, I hope the main ideas and facts I've posted are understandable and usefull, but needs the great criticism (where is it). Moderation of this forum will help to reach Your 9% soon, I wish - wery wide field of problems You consider gave feeling of intelect abbilities. Use more, than 9%...
I'm going to work more on these probabilistic-stochastic- random mechanisms evaluation, that control DNA methylation changes. The total human cell's Genome length is distance to the moon x 6000. What is the probability of fatal DNA methylation changes - great enough - the price for pollutions, ozon damage, but not greater than 5/1000.
It's not the big problem to find computer time four some hours in Vilnius, but some people from our organisation (like free, more art, than science institution) don't want to understand, that my book CARCINOGENES, DNA METHYLATION DISORDERS AND A MALIGNANT GROWTH (carcinogenesis theory)I'm rewritting again manuscript, having the CONTENT:1. DNA methylation disorders 2. DNA methylation system 3. SAM fluxes redistribution mechanisms 4. Carcinogenes and Met cycles disturbances 5. Disregulation of Met cycle and subsystems related to it in amalignant cell 6. Hemoblastoses, pool of folates and expression of genome 7. Malignant grotw as a self stimulating process 8. Markers of a malignant growth related to a DNA methylation changes 9. Regulation of a DNA methylation system by drugs and diet 10. General model of Cancerogenesis initiation Discussion has some chance to move those problems, especially 3-5, 7, 9-10 I hope. Few days ago I've listened the lecture about Biotechnologies at The Academy of Science, Lithuania and was very glad - the half of listeners were schoolchildren! In the scientific orbit Lithuania is famous by laser production providing them for 174 countries and situation is going to change in the other fields!

---

Kio ora, Hi,
I would like to discuss important methodological question -
CANCER INITIATION IDENTIFICATION PROBLEM related to DNA methylation disorders

It is well known, that DNA methylation pattern depends on One carbon units metabolism and s-adenosylmethionine (SAM) is the final and direct DNA methyl (CH3-) donor.


DNA methylation process is stochastic (experimentaly prooven by Susan Clark), so it is really random, depends on chance too and cancer cells differs when comparing methylation patterns. The general fact:: the global DNA hypomethylation and hypermethylation of some regulatory regions - it is usuall characteristic of cancer cell. In the cancer initiation model by methyl deficient diet - SAM level decreases, the global hypomethylation appears (this changes expression of some oncogenes, if affected...) and later - hypermethylation (this triggers off some tummour supressor genes, if affected by chance). The huge amount of experiments was made trying to reveal the reason of malignancies initiation: folate system investigated, Met cycle, methyl defficient diets(rat hepatocarcinogenesis), some carcinogenes disturbing methylation- arsen, TPA, DMBA, and finally - DNA methylases, ....but...still no clear answer, no possibility to control malignant process.
Theoretically - the damage of any chain of the DNA methylation system (together with One carbon units metabolism), could change the character of DNA methylation and initiate the malignant growth. How many chains there are (I am mathematician solving cancer puzzle)? Let the specialists say, but I've calculated more than enough of them to make identification of cancer initiation task like unsolvable, if not stressing this circumstance:
DNA methylation closely depends aproximately on 170 substrates, metabolites, enzymes, the other SAM acceptors... (see bellow The descript of OCU metabolism)
Possibly, I hope, there exist some main reasons of OCU metabolism disorders, which change DNA methylation character, but if they all distributed aproximately unique - the chance of appearance in single experiment is less than 1%!!! So, to conclude: Cancer initiation identification problem, if investigate only DNA methylation system, could be of hardest degree, and results about different phenomenas (possible reasons) describng phenotype of malignances with a rate 1%,...3%,... 5%,...7%,.........20%,.........30%...............could be very valuable!

Good luck!
Kestutis Urba
Lithuania

P.S. When investigating DNA methylation system and discussing these questions we must choose something like this to be basic (sorry- my speciality is applied mathematics....)

THE DESCRIPT OF THE ONE CARBON UNITS METABOLISM
Input:
Ser - serine, Gly - glycine, Met - methionine, Chol - choline, Bet - betaine, His - histydine, Trp -tryptophane
(input means – dietary and/or degradating protein..., synthesys by metabolic path)
OCU Carier Folate system
Folate, 7,8-DHF, 5,6,7,8-THF, 5,10 -MethylenTHF, 5-Methyl-THF, 5,10 -Methenyl THF, 10-Formyl-THF, 5-Formyl-THF, 5-Formimino-THF
OCU Cariers
Betaine, Sarcosine, Histydine, Tryptophane, sHmc, Hmc, SAM – DNA methyl donor
The other important SAM methyl acceptors
Histone, polyamine...
....
Total: 80 SAM acceptors!!
Events
DNA hypomethylation
DNA hypermethylation
Histone hypomethylation
Histone methylation...

Enzymes
DNA methylases:
DNAmt1, DNAmt3a, DNAmt3b
Met cycle
Hmc ( 2.1.1.13 -turn into Met, - Met resynthesys) 5-CH3-THF-homocistein-CH3-transferazė
1.Hmc ( 2.1.1.5 -turn into Met, - Met resinthesys), Betain-homocistein metiltransferazė
2.Met (2.5.1.6 -turn into- SAM-Met), metionin-adenoziltransferazė
3.SAM (2.1.1.37 -turn into - SHmc, -DNR methylation) DNR 5-citozinmetilazė
4.sHmc (3.3.1.1 -turn into Hmc), S-adenosyl-l-homocisteine hidrolazė
5.Hmc, Ser (4.2.1.22 Cistathionin sinthesys, - Met cycle output), Cistationin-beta sintetazė

Folate system
Folate (1.5.1.3 - turn into 7,8-DHF)
7,8-DHF (1.5.1.3 - nonreversable one direction turn into 5,6,7,8-THF)
5,6,7,8-THF (2.1.2.1 and 2.1.2.10 – turn into 5,10 -MethylenTHF)
5,6,7,8-THF (2.1.1.13 - turn into 5-Methyl-THF)
7,8-DHF( 2.1.1.45 - turn into 5,10 -MethylenTHF)
5,10 -Methylen THF (1.1.1.5 and 1.1.1.15 - turn into 5,10 -Methenyl THF)
5,10 -Methylen THF (1.5.1.20 - turn into 5-Methyl-THF)
5,10 -Methenyl THF ( 3.5.4.9 - turn into 10-Formyl-THF)
5,10 -Methenyl THF ( 2.1.2.10 - turn into 5-Formyl-THF)
5,10 -Methenyl THF ( 4.3.1.4 - turn into 5-Formimino-THF)
5,10 -Methenyl THF (2.1.2.5 - turn into 5,6,7,8-THF)
10-Formyl-THF( 2.1.2.1 and 2.1.2.2 and 6.3.4.3 and 1.1.5.6 and 2.1.2.9 - turn into
5,6,7,8-THF )
5-Formyl-THF (2.1.2.5 - turn into 5,6,7,8-THF)
5-Formyl-THF ( 6.3.3.2 - nonreversable one direction turn into 5,10 -Methenyl THF )
5-Formimino-THF( 2.1.2.5 - turn into 5,6,7,8-THF)
Total of humane folate system enzyme -21.
Ser synthesys
phosphoserine phosphatase [EC:3.1.3.3]
Ser metabolism -catabolism
serine dehydratase [EC:4.3.1.17]
serine-pyruvate aminotransferase) [EC:2.6.1.44, 2.6.1.51]
cystathionine-beta-synthase [EC:4.2.1.22]
seryl-tRNA synthetase 2 [EC:6.1.1.11]
serine hydroxymethyltransferase 1 (soluble) [EC:2.1.2.1]
phospholipid synthesis...
Gly synthesis
serine hydroxymethyltransferase 1 (soluble) [EC:2.1.2.1]
glycine amidinotransferase (L-arginine:glycine amidinotransferase [EC:2.1.4.1]
glycyl-tRNA synthetase [EC:6.1.1.14]
pipecolic acid oxidase [EC:1.5.3.1 1.5.3.7]
sarcosine dehydrogenase [EC:1.5.99.1]
glycine N-methyltransferase [EC:2.1.1.20]
aminolevulinate, delta-, synthase 1 [EC:2.3.1.37]
glycine C-acetyltransferase (2-amino-3-ketobutyrate coenzyme A
ligase) [EC:2.3.1.29]
aminolevulinate, delta-, synthase 1 [EC:2.3.1.37]
glycine dehydrogenase (decarboxylating; glycine decarboxylase,
glycine cleavage system protein P) [EC:1.4.4.2]
alanine-glyoxylate aminotransferase (oxalosis I; hyperoxaluria I;
glycolicaciduria; serine-pyruvate aminotransferase) [EC:2.6.1.44
2.6.1.51]
D-amino-acid oxidase [EC:1.4.3.3]

Chol – Gly path
Chol (1.1.99.1 – nonreversable one direction turn into Betaine aldehyde)
Betaine aldehyde (1.1.99.1 nonreversable one direction turn into Betaine)
Betaine ( 2.1.1.5 – nonreversable one direction turn into Dimethylglycine)
Dimethylglycine (1.5.99.2 – nonreversable one direction turn into Sarcosine)
Sarcosine(1.5.3.1 and 1.5.99.1 – nonreversable one direction turn into Glycine)
Glycine (2.1.1.20 – Sarcosine)

His metabolism
histidine decarboxylase [EC:4.1.1.22]
dopa decarboxylase (aromatic L-amino acid decarboxylase)
[EC:4.1.1.28]
histamine N-methyltransferase [EC:2.1.1.8]
Trp metabolism
tryptophan hydroxylase 2 [EC:1.14.16.4]
dopa decarboxylase (aromatic L-amino acid decarboxylase)
[EC:4.1.1.28]
monoamine oxidase A [EC:1.4.3.4]
indolethylamine N-methyltransferase [EC:2.1.1.49]
dopa decarboxylase (aromatic L-amino acid decarboxylase)
[EC:4.1.1.28]
indolethylamine N-methyltransferase [EC:2.1.1.49]
tryptophan 2,3-dioxygenase [EC:1.13.11.11]

Coenzymes
B12, B6, B2
FADP(H), NADP(H)
Microelements
Zn, Mg2+


Diseases marked by DNA methylation abnormalities
Beckwith–Wiedemann, Prader–Willi, and Angelman syndromes, Albright hereditary osteodystrophy (AHO) and pseudohypoparathyroidism Ia (PHP-Ia) and PHP-Ib, and transient neonatal diabetes.
DNA methylation patterns are globally disrupted in cancer, with genome-wide hypomethylation and gene-specific hypermethylation events occurring simultaneously in the same cell.
Carcinogenes affecting OCU - DNA METHYLATION
Arsen, TPA, DMBA
DNA methylation inhibitors:
methylation inhibitors: Zebularine, (often - p16 demethylation; The drug caused a complete depletion of extractable DNA methyltransferase 1 (DNMT1) and partial depletion of DNMT3a and DNMT3b3. )
5-Aza-CdR, (5-Aza-2'-deoxycytidine = decitabine), ( often – p15/INK4B demethylation, p16 dmethylation)
5-Aza-CR (5-azacytidine ),
procainamide, teapolyphenol -egallocatechin-3-gallate(EGCG) ((from C.B.Yoo, J.C.Cheng, P.A.Jones. Zebularine:a new drug for epigenetic therapy. 2004, Biochemical society transactions))
One Carbon Units metabolism inhibitors
dihydroxypropyladenine (DHPA) -dihydroxypropyladenine (DHPA), the competitive inhibitor of cellular S-adenosylhomocysteine hydrolase
Ethionine (Ethi), the hepatocarcinogenic antimetabolite of methionine
a general inhibitor of transmethylation reactions, L-ethionine
nitrous oxide (inactivating methionine synthase)
methotrexate (inhibiting dihydrofolate reductase)
The other OCU affectors:
methyl defficient diet
Met defficient diet
Choline defficient diet
Choline Methionine Folate defficient diet
caseine(?)

---

Kia Ora Nick,

what are your last results?

Kest from Vilnius

---

Hello Kestitus!

How have you been?

Well we have a PhD student working on the project and she has found a correlation between a polymorphism found within the MTHFR gene in cell lines that correlate to earlier quiescence compared to another polymorphism when the cells are grown in low folate medium. The next stage is to of course test the methylation status.

more later I hope!

Nick

---

Still lots of problems, but life is life. You helped to survive last year - thanks. I must think what is the shortest way to evaluate efficiently the role of all possible DNA methylation system disturbances, becouse the amount of possible reason seems to be too great even for single institution when researching. Perhaps, one must start researching methylases inhibition, then Met cycle disturbances inhibitting single enzymes, then Folate system, and then complex of reasons studying all the One Carbon Units metabolism influencies. Some results were obtained during the last 50 years, but there was no systematical research program. I've seen information about DNA methylation society. They must do this and coordinate all the research.

I see, Your friends You help moderating are great experimentators, but sometimes some theoretical abstract brain storm is needed.
Glad to add some ideas.

---

Kestitus,

we as a scientific community need more theorists that look at the "bigger picture". I am a victim of reductionist biology and think in really simple terms.

Your thoughts and ideas have provoked many discussions within our institution and are valued.

An emerging idea is the correlation between genotype and epigenotype, whereby a specific DNA sequence variant is positively associated with whether the surrounding DNA is methylated or not, now this can be for many reasons such as local DNA structure and nuclear peripheral locations.

As with everything in science, the pursuit to answer a question will always lead to more question being asked. It's an exciting time in biology.

Nick

---

Well,

I would like to make some remarks on recent article published in EPIGENETICS recently -
The authors have studied the same problem we discussed last year - Folate cycle and Met cycle interactions.

Long-Range Allosteric Interactions between the Folate and Methionine Cycles Stabilize DNA Methylation Reaction Rate
H. Frederik Nijhout, Michael C. Reed, David F. Anderson, Jonathan C. Mattingly, S. Jill James and Cornelia M. Ulrich

volume 1 | issue 2
april/may/june 2006
Pages: 81 - 87

http://www.landesbioscience.com/journals/e...cs/article/2677

(erratum is at http://www.landesbioscience.com/journals/e...s/article/3281)


They say - Several metabolites in the folate and methionine cycles influence the activities of distant enzymes involved in one-carbon metabolism. ...

Using both steady-state and fluctuation analyses of a mathematical model of methionine metabolism, we investigate the biochemical basis for several of these hypotheses.

We show that the long-range interactions provide remarkable stabilization of the DNA methylation rate in the face of large fluctuations in methionine input. In particular, they enable the system to maintain methylation in the face of low and extremely low protein input. These interactions may therefore have evolved primarily to stabilize DNA methylation under conditions of methionine starvation. In silico experimentation allows us to evaluate the independent effects of various combinations of the long-range interactions, and thereby propose a plausible evolutionary scenario.



The level of SAM and the velocity of the DNA methyltransferase reaction depend on
the properties of both the methionine and folate cycles. These metabolic cycles in turn
depend on the dietary intake of certain nutrients (methionine, choline, betaine) and vitamins
(B12, folate, B6) and are affected by polymorphisms in the genes for enzymes of the
methionine and folate cycles.3-6 Because defects in folate and methionine metabolism are
associated with a large number of serious disorders (several types of cancer,7-10 cardiovascular
disease,11-13 neural tube defects14 and neurodegenerative diseases15,16), the genes
and enzymes of these cycles have been well studied.

I'm not sure they were studied well enough evaluating all complexity of regualtion and all disturbancies. Where is Ser, Gly input for example?

The authors stressed - We recognize that no mathematical model can capture the full
complexity of a biological system. First, there are substantial
uncertainties and variations in measurements of kinetics because
experimental data come from a diversity of tissues, organisms, and
experimental procedures. Second, many of the substrates and enzymes
participate in other reactions that are not in this system.
Nevertheless, the model has allowed us to verify the effects of
mechanisms proposed in the experimental literature and to quantify
the relative magnitudes of the effects. Perhaps the greatest advantage
of an explicit mathematical model is that it allows us to perform in
silico experiments in which one or more regulations are removed,
experiments that would be difficult or impossible to do in vivo.

The conclusions the authors made are -

The stabilizing effect of the long-range interactions on the DNA
methylation rate can be explained as follows. As methionine input
falls, SAM concentration declines, which has two effects. First, the
decline in SAM reduces the activity of CBS so a larger fraction of
homocysteine is remethylated, which tends to maintain the flux
around the methionine cycle. Second, the decline in SAM releases
the inhibition of MTHFR. This causes the concentration of
5mTHF to rise, which increases the inhibition of GNMT. Thus,
even though the flux from SAM to SAH is lower, the inhibition of
GNMT causes a larger fraction of the flux to be carried by DNMT.
This second mechanism was originally hypothesized by Wagner, Briggs
and Cook.44 The mathematical model shows that each step in this
relatively complicated causal chain is, indeed, correct, and that the
combination of long-range interactions and the presence of the
GNMT reaction are responsible for the stabilization of the DNA
methylation rate.

Correspondingly, we found that folate status
significantly affects methylation rate only at low methionine input
(Fig. 6). Thus, the long-range regulatory mechanisms that we have
described, as well as the connection between the folate cycle and the
methionine cycle may well have evolved to protect methylation rates
against periods of low and very low methionine input. Periods of
protein starvation are common for some human populations today
and must have been common for paleolithic humans, who are
believed to have been primarily hunters and meat eaters.


(Homo sapiens but not neanderthals - fish eaters too).

This article is usefull stressing the role of feedbacks and constructing some scenario of Folate, Methionine cycles interractions and regulation.

The mathematics is some artificial language which allows to speak and speculate about the studied part of reality evaluating not only quolitative but numeral aspects too. If someone uses such tools which allows to simulate real system behavior on computer, this gives the image of scientific research to be complex, authoritative and solid. If model creators were lucky, sometimes such models are really useful, becouse they allow to ivestigate and predict some system behavior. But the problem is that such biological systems behavior is not like the simple velocity dependence on petrol amount in the engine of the car.

I would like to ask the authors - is it possible to simulate classical experimental results using rats and prolonged methionine starvation diet to start liver malignization? If yes (I'm not sure), so do there the math model really fits the time moment when malignisation starts?

But Iwould like to stress again - such deduction models are useful stressing the main feature of system regulation science - the role of feedbacks, but the real system mode they used is the great simplification of OCU metabolism I understand and presented in figure above. If the new elements and more feedbacks are added - the amount of equations is larger and analyse such system using only the pencil is impossible, but computer monsters allow to do this. In the future more realistic models (even using some input stochastics) for evaluating of numerical features will be constructed, I hope but I would like to stress the other approach - empyrical one.

Well, it takes much more time and grants but the experimental measuring SAM dependence on the different Met and Folate inpute would be very useful too. Then using some methods of mathematical statistics (regression analysis or more complex - stochastic system identification science) is possible to obtain some formula describing this dependence. So, those two approaches are somehow usefull, but this is the head ache of mathematicians not the molecular biologists who investigate mainly the structure of the system and events on it. But if you know some boring mathematicians or even some math- or cyber- students having not understanding about the main problems of BIOLOGY tell them this story.

Kestutis

---