There have recently been some discussions about Tm.
<9302082056.AA0279>, from D. Natale, states that the important Tm is during
the hybridization, not the wash. If this is so, then why can I reduce the
embarrassement-quotient of a blot by a wash at a slightly higher temperature
after I have performed a one-day exposure? What is the explantion for the
ability to strip bound probe in the effort to examine the immobilized nucleic
acid by a second probe? Clearly, if no radiolabelled DNA bound to the membrane
during the hybridization step, then nothing which one might attempt during the
wash step will rectify the situation. But if radiolabelled DNA is bound to
targets which are distantly related to unrelated, then the temperature during
the wash step is of paramount importance.
<9302091527.AA08274>, also from D. Natale, states that Tm is the
temperature at which 50% of the DNA molecules are single-stranded. Is this
situation distinguishable from that in which 100% of the DNA molecules are 50%
single-stranded? The data from Breslauer et al. is useful for short DNA.
Unfortunately, data from 20 years ago showed that the Tm is not dependent upon
the length of the duplex for DNA longer than 550 to 650 bp in length. This is
is contrast with DNA shorter than this fuzzy threshold. So the Breslauer data
might not be universally useful for all perfect-match DNA probes. Careful
reading of the Breslauer paper shows that they examined only those sequences
in which all bases are base-paired. The Breslauer data can bring nothing to
bear upon the problem of mismatched probes.
<9302102215.AA24441>, from B. Gross, describes an experiment in which
duplex DNA is heated and the process followed spectrophotometrically. Unless
New Hampshire has different physics than the rest of the world, the result
will be an increase in absorbance of the solution. Duplex DNA is hypochromic
relative to an equimolar nucleotide solution of the same composition. This is
due to the base-stacking of base-paired duplex DNA. As the DNA is heated, some
of the base pairs are disrupted. This results in a increase in the absorbance
of the solution. This is independent of the concentration of base-pairs. Do
not confuse this property with the absolute absorbance of the solution, which
is concentration dependent. This heat-dependent increase in absorbance of the
solution ceases when all of the base-pairs are disrupted. This too is
concentration independent. The temperature at which the increase of absorbance
is 50% of the maximum which could occur is defined as the Tm. It is
concentration independent and represents the temperature at which 50% of the
base-pairs are disrupted. Unless you have a really special spectrophotometer
capable of distinguishing individual polynucleotide chains, there is no way to
know whether 50% of the molecules are 100% dissociated or 100% of the
molecules are 50% dissociated or some other algebraic combination yielding 50%
of the base-pairs disrupted exists. At the Tm there is an equilibrium between
disrupted base-pairs and non-disrupted base-pairs such that the fractional
amount of disrupted base-pairs is constant. At the Tm, reassociation of
disrupted base-pairs is not a second-order reaction. The non-disrupted base-
pairs direct the reassociation of adjacent disrupted base-pairs. These are the
nearest neighbor interactions cataloged in the Breslauer paper. This is a
different experiment from the reassociation of completely dissociated
polynucleotide chains, which is exactly described in the Cot differential
equation. Approximately 20 years ago, workers in the lab of E. Davidson showed
empirically that the rate of reassociation of completely denatured DNA of
moderate complexity occurs most rapidly at Tm - 25oC to 30oC. I have yet to
hear or see an explanation which can be considered quantitatively, in contrast
to qualitatively or descriptively, correct.
<9302120152.AA23099> states that Tm is an alternate form of the
equilibrium constant. Is this quantitatively correct?
<9302120547.AA29021> states that the Tm of DNA denaturation is surely
DNA concentration dependent. The thought experiment of removing those base
pairs having disrupted hydrogen bonds is offered as proof for this assertion.
How is this experiment to be performed? Will this disturb the polynucleotide
nature of the compound being examined? If so, does this not change the
experiment in mid-course? Therefore any understandings arrived at upon
completion of the experiment are not necessarily applicable to the materials
introduced at the start of the experiment (grins not withstanding). The
alternate view, that the Tm of a 1 mg/ml solution of a particular piece of DNA
is different than a 100 ug/ml solution of the same DNA, is ridiculous. The
appropriate analogy is the boiling temperature of a particular chemical
compound. Is the exact value of this property dependent upon the concentration
of the chemical compound? Then we think of hybridization. "Higher single
stranded DNA will favor hybridization."??? I assume that "concentration of"
was omitted between "Higher" and "single stranded". BFD. Simple examination
and comprehension of the Cot differential equation would have led even Dan
Quayle to such an understanding. (Yes miracles are possible in thought
experiments!) Otherwise we should begin including in the hybridization buffer
either marijuana or LSD, depending upon how much hybridization we wish to
achieve. "High temperature favours ssDNA state, and low temperature favours
dsDNA state." Really? DNA longer than 600 bp with 60% GC in 1 M salt will be
almost entirely double stranded at 90oC. (The computed Tm for these conditions
is 106oC) How low does the temperature have to be to allow annealing of the
reverse primer to single-stranded template derived from a -U phagemid? How
much of a molar excess of the primer will "drive" this? On to the orginal
question, a thermodynamic babble about stringency. The assertion that washing
of blots is a kinetic problem requires the immediate question, kinetics to
what equilibrium: i) removal of non-specifically bound probe or ii) conversion
of non-specifically bound probe to specifically bound probe? The assertion
that all the annealing oligonucleotide probe will be washed away at some
temperature below the Tm is an oblique way to tell me that the tube of duplex
DNA fragment (>500 bp, < 1 kb) prepared three years ago and stored at 4oC is
comprised of a significant fraction of single stranded material. The
suggestion that increasing the time of washing will do something to the blot
comparable to increasing the temperature at which the wash is performed
indicates to me that either you are doing your washing for entirely too short
of a time, or that you have no understanding about the nature of the duplex
DNA produced during hybridization. Yes, the process of discriminating between
non-specifically bound probe and specifically bound probe is a thermodynamic
problem. However it is also a statistic and engineering problem. As I noted
above, the Breslauer et al. data only provides the thermodynamic information
for perfect-match situations. Do you have the requisite information for all of
the possible mismatch situations? Assuming you do, you could compute the
conditions to allow dissociation of all (100% since 5% remaining will be a
problem) non-specifically bound probe, but could you devise an apparatus to
achieve the thermal precision your computation demanded. Also, if your
hybridization conditions produced 100 DNA duplexes with perfect matches at the
same time as 100 DNA duplexes with a single mismatch, how many perfect matches
would remain when the desired number of mismatch duplexes had been removed? Do
you think your answer depends upon the location of the mismatch in the duplex
(exact middle versus near one end)?
With respect to the last paragraph. The programs compute only the Tm
simply because that property is all you need to know to use the piece of DNA
in any meaningful and pertinent manner. Despite your infatuation with the
kinetic phenomenon, it is not considered because to do so is to be distracted
by an irrelevancy.
Brian Smith-White
preissj at clvax1.cl.msu.edu
