Re: Alpha Radiation (fwd)

Robert J. Bradbury (
Tue, 3 Aug 1999 03:28:52 -0700 (PDT)

> John Clark <> wrote:

> In case you hadn't noticed Alpern is talking about cancer, "oncogenic transformation"
> just sounds better in a big expensive scientific book.
> >Only a *very* energetic particle has enough energy to interact with the DNA
> >backbone and produce a double strand break.
> Nonsense. In the first place the backbone is not important, it's a break in the rungs
> of the DNA ladder that's important because that's where all the digital information
> is stored.

John, this simply isn't true. I would suggest that you go to PubMed, and lookup a REVIEW article on "poly-ADP-ribose polymerase", "double-strand-breaks" and "apoptosis" (or some combination thereof).

There are many *very* distinct pathways of mutation, repair and cell death. I believe I may have glossed over the pathways involved with base mutations vs. DSB. Producing base mutations involves breaking a H-C bond on a base [in most cases] (low energy). Producing a double strand break requires breaking the bonds on *both* DNA backbones (higher energy). To produce a DSB you have to deposit enough energy in a region covered by the distance between the DNA strands to break both of them in a time that is less than the time required for single-strand break repair to occur. Since DSB require breaking two bonds in an volume that is much bigger than that of a single base it is more difficult than a single base mutation.

If you have enough double strand breaks you activate poly-ADP-ribose polymerase which consumes NAD(P)H producing a "sugar-matrix" (if my recollection serves me correctly) that tries to physically "lock" the DNA in place so the double strand breaks don't flop around the entire nucleus in such a way that makes them difficult to rejoin. However if you deplete all of the NAD(P)H in the cell, you have eliminated all of the reducing equivalents which makes oxidative defense difficult. The cells seem to have a threshold of tolerance for DSB. If you exceed that level (~5 in mammalian cells), then the cell activates the apoptosis (cell death) program (through the p53 pathway I believe).

DSB are very important. There is a specific kind of cancer (Burkitt's Lymphoma I believe) that is the result of two different chomosomes getting broken and put back together improperly so an improper upstream promoter sequence activates a gene that promotes cancer. Since the more DSB you have, the greater the probability of mis-joining chromosomes is, the best bet is to "fall on your sword" when that situation occurs.

Now, all of this is *much* different from the production of hydroxyl radicals that attack the DNA bases, that *may* or *may* not be copied incorrectly. Current thinking is that there are 3-4 different error repair pathways that use 4+ different DNA polymerases in mammalian cells that each have different abilities to copy DNA with/without various mutations accurately or inaccurately. Evolution has presumably tried to evolve the relative activations and probabilities for these paths to serve its own goals (tradeoffs between repair and reproduction and a mutation rate sufficient to drive evolution).

There are 5-8 different human genetic diseases (with large numbers of variants [complementation groups] known) involved in defective DNA repair pathways. These pathways involve at least 30-50 genes. These genes don't exist simply because nature is showing off. They exist because there are many types of damage that can occur and they have to be responded to in different ways at different times. Repair is an appropriate response if the damage isn't too severe. Cell death is an appropriate response if you think the damage might lead to cancer. Copying DNA (with mutations) is an appropriate response if the cell is "essential".

For example, there is a specific mutation - a thymine dimer - that occurs when UV radiation bonds breaks the T-A bonds across the DNA strands and joins two adjacent T-T on the same backbone to each other. There is a specific enzyme DNA photolyase that utilizes low energy light to correct this error. There is a specific DNA polymerase (DNA Pol zeta) that conveniently enough appears to be able to copy through T-T dimers correctly inserting the A's on the opposite strand (while other DNA polymerases halt or mis-copy). Alpha particles or X-rays *don't* produce thymine dimers (relative to the numbers that UV radiation does).

The main point of your comments that I am stuck on is your claim that alpha particles (of various energies) produce mutations that must kill or mutate cells. Given the different energies and the multitude of materials that can be "hit" and the various response pathways that can be invoked, I think that statement is a gross oversimplification.

> In the second place, the chemical bonds in the rungs are much weaker than those in
> the backbone, although even the strongest chemical bonds are of only a few electron
> volts .
> I'm talking about average run of the mill radiation particles and they have thousands,
> often millions, and sometimes many billions of electron volts of energy.

Ok, now we are talking hard numbers. So with all this excess energy, *most* of it will be transfered to the water molecules (since they are the most abundant). Those produce hydroxyl radicals (by knocking off a H atom). The question then becomes what is the probability of interaction of the hydroxyl radical with (a) a DNA base, and (b) the atoms in the DNA backbone. I can't answer the question right now (since I've got to leave for the Bioastronomy conf. in a few hours), but if you are really interested, remind me in a week or so and I'll try to tease the numbers out of my sources.

Since the "data" is that high-LET radiation is more likely to produce mutations than DSB (and cell death), I would presume that the hydroxyl radicals have a higher probability of attacking a base than the backbone. Low-LET radiation may have completely different interaction probabilities.

> Hey, don't tell me how smart you are, show me

I can only try.
> I will too, should that unlikely event ever happen.
> You're welcome to try but you'll have to do better than quoting long passages from a
> book you don't understand.
I understand the physics in the books reasonably well, and the biological systems involved better than anyone but a handful of people. To really answer this question properly requires a simulation at the atomic level. There are programs that do this (at LLNL & LANL), but I doubt we can have access to them :-).