MD & MC; Modelling of counterions effects in polyelectrolytes.

 	Hi Folks,
 		All this recent talk about electrostatic charges on
 	protein surfaces and a previous posting about site-specific
 	counter-ion binding around the pectic acid in plant cell walls has
 	brought up a worry that I've had for some time now, I wonder
 	what others think about this.
 		This problem applies to all polyelectolyte modelling
 	but I will use DNA to explain it since it's what I'm familiar
 	with.
 		In most of the modelling of DNA we place sodium
 	counter-ions next to the -ve charged phosphate groups to
 	make the system neutral.
 		Now Debye-Manning theory determined that only around
 	half the phosphates are neutralised at low ionic concentrations
 	and that only when we get to very high salt concentrations do we get
 	( I vaguely remember ) around 0.7 of the phosphates neutralised
 	by sodium ions, the rest sitting in a counter-ion atmosphere
 	some distance away from the DNA.
 		So the dilemna in my mind is that we model DNA with
 	all the phosphates neutralised very locally as if we are modelling
 	it in an extremely concentrated salt solution ( which it would
 	denature in ).
 		Would it not be better to neutralise half the phosphate
 	groups locally with counterions and simulate a charged system ?
 	Statistical mechanics would predict that the DNA would have
 	between 0 and 24 ( for dodecamers ) ions bound onto it with
 	different probabilities for each depending on salt concentration,
 	so why does everyone model it with 24 counter-ions ? When
 	at physiological conditions I believe it is most likely to have around
 	12 directly bound onto it ?
 		This seems to apply to a lot of polyelectrolyte simulations
 	 ... full counter-ion neutralisation seems to be what is modelled,
 	when electrolytic theory predicts that most polyelectrolytes are
 	only partially neutralised by site-specific counter-ions in
 	solution.
 		What do others think ?
 		Maybe I'm being pedantic ( or silly ! ) but the amount
 	of counter-ions included in a simulation extensively effects the
 	structure of the solute.
 		Any better ways to include counterion concentration
 	effects ? I suppose a Debye screening term dependant on
 	concentration could be applied between all charged atoms separated
 	by solvent in the system ... be computational due to the exponential
 	term though ... is there a general concensus of belief on this
 	problem ?
 	Sorry about the large bandwidth.
 	Kevin.
 =============================================================================
     K.Bryson                 email: kb7-0at0-tower.york.ac.uk
     Biophysics Group         Tel  : +44 904 430000 Extn. 2236
     Physics Department       Fax  : +44 904 432214
     University of York
     Heslington              "Molecular modelling of DNA and its
     YORK, UK                    interaction with small molecules."
     YO1 5DD
 =============================================================================