Isoelectric Points

Wayne R. Baker baker at
Mon Mar 25 22:33:59 EST 1996

:driska at (Stephen P. Driska PhD) wrote:
:>         Wouldn't this be some sort of weighted average?  In other
:> words, if you had a protein with 20 Asp's and 3 Arg's, wouldn't the pI
:> change more by deletion of one of the few Arg's than by deletion of one
:> of the many Asp's?

In bionet.molbio.proteins,  T. Chyau Liang <tliang at> wrote
:The short answer is no.
:Because pI is the average of the two pKa values that link the
:equilibria between the (+1 <==>0) and (0 <==> -1) charged states of the
:protein. In your hypothetical protein the pI would occur when 3 Asp are
:ionized (the other 17 are unionized). 

  That is not possible. Remember that the pKa of an ionizing group gives
the *ratio* of the dissociated and undissociated forms, one of which
will have a charge (depending on if it's a carboxylate, imidazole, etc).
It is inaccurate to say that 3 Asp residues are ionized while the other
17 are not. All ionizing groups within a protein will experience a
degree of pH-dependent ionization (and therefore charge contribution)
governed by the Henderson-Hasselbalch equation. It is inaccurate to say
that 3 Asp residues are ionized while the other 17 are not.

:Therefore, the pI will be the
:average of the pKa values responsible for the ionization of the 3rd and
:4th Asp. 

  The pI is a function of *all* the ionizations in a protein, both
acidic and basic. At a given pH, the net negative charge contribution of
the carboxylates is the number of carboxylates x the net charge per
carboxylate. So it is, in fact, a weighted average. 

:In real protein each Asp will have a slightly different pKa,
:but these pKa don't differ by much. 

  The first statement is true, but the second is not. A survey of the
literature (which, IMHO, is lacking in quantity) shows that pKa values
for Asp range up to 4 pH units. Ranges are slightly less for Glu and
C-term. In general, His, Lys and Arg have less variability although a
few pH units is still seen. 

:Deleting one Asp or Arg will shift the pI by an unappreciable amount in
:this case. 

  That's not how my calculations show it happening. For a protein with
20 Asp (pKa 4.0) and 3 Arg (pKa 12.5), the calculated pI is 3.25. For a
protein with 20 Asp and 2 Arg, the pI is 3.0. For a protein with 19 Asp
and 3 Arg, the pI is 3.27. Clearly, removing an Arg alters the pKa more
than removing a Asp. And intuitively this makes sense: the greater the
number of acidic residues, the lower the pI; the greater the number of
basic residues, the higher the pI. 

Wayne Baker (baker at  	He who has a why to live for
4288 Molecular Biology Building  	can bear almost any how.
Iowa State University            	--Nietzsche

More information about the Proteins mailing list