Ion-exchange chromatography (IEC)

Consider the  situation where a single  anion type,  say  CH₃COO-, and a number of cation  types,  say Na+, K+ and  Protein+,  exist  together  in  an aqueous solution,  where the  aqueous component  will additionally contribute OH- and H+ ions.  These  ions will establish a dynamic equilibrium, comprised of a number of sub-equilibria, e.g.

The overall  equilibrium condition  will be determined by the  values of  the  respective dissociation constants:-

The overall equilibrium state is a result of-

•  intrinsic  affinities  between  ions  (expressed  in  the  dissociation

constants), and,

•  competition  between  ions  (a  function  of the  relative  concentrations of the  ions).

To  develop  a chromatography  system, the  CH₃COO-  ion, in the  form of a  carboxymethyl  group  (-CH₂COO-),  could be covalently  attached  to the stationary phase and the dissociated cations allowed to move with  the mobile phase: net electrical neutrality  being maintained,  however.  With an  immobilized  anion,  this  would constitute  a  cation  exchange  system Conversely,  immobilization  of  a  cation  would  constitute  an  anion exchange system.

If the  above  ions  were  applied  to  the  chromatography  system as  a sample  “plug” and the  system  was subsequently eluted with a buffer,  say lithium  citrate,  the  differential  affinities  of the  ions  for  the  stationary anion  would  be  manifest as  differential  rates  of  migration  through  the column.  The  ions  would migrate  at  relative  rates  proportional  to  their respective dissociation  constants.  Their  manifest  affinity,  and thus  their absolute rates of migration, would depend upon the competition that  they encountered from the buffer cation, Li⁺ in this example.  With increasing Li⁺ concentration, the sample cations  would face  increasing competition in  their  association  with  the  immobilized  anion  and  so  would  be increasingly dissociated, resulting in an increase in their rate  of migration through the column.  In the  case of proteins,  the  dissociation  constant  is affected by pH, so elution can also be effected by a change in pH.

1 . Ion-exchange “resins”

The term “resin” comes from early  polystyrene-based ion-exchangers which had a translucent  yellow appearance,  like the  resin exudates from pine trees.  The term has stuck, although modern ion-exchangers used  for protein separations are generally opaque and white.

All  ion-exchange  resins  are  comprised  of  a  matrix  to  which  are attached  ionic substituent groups.  For  low  pressure  chromatography  of proteins, the matrix is often comprised of a hydrophilic biopolymer, such as cellulose, Sephadex™, or agarose.  These  materials  cannot  withstand high pressures and for  medium  to  high  pressure  liquid chromatography, the trend is towards silica-based resins, or synthetics such as Trisacryl ™.

Cellulose is a polymer  of ß-D-glucose units,  linked  with  bonds. It  is relatively  inexpensive  and provides good flow  properties,  but  large interstitial spaces lead to relatively poor resolution.  Sephadex consists  of dextran  chains,  comprised  of  linked  dextrose  (glucose)  residues, cross-linked with epichlorhydrin.  The name “Sephadex” is a contraction of the words  “separating”,  “Pharmacia” and  “dextran”.  It is sold  in  the  form  of dry  xerogels  which  absorb  water  and  swell  into hydrated spherical  particles.  Substituted Sephadex ion-exchangers  give good  resolution  but  they  are  subject  to  marked  volume  changes  with changes in buffer ionic strength.  This is a disadvantage as it is  difficult to apply  an  accurate  salt  gradient  to  a  shrinking  gel,  and  it  may  become necessary to re-pack the column after only a few runs.

Agarose  is the  neutral  polysaccharide  component  of agar,  an  extract of  kelp,  which  is  a  type  of  seaweed.  It  is  a  linear  polysaccharide composed  of  alternating  residues  of  D-galactose  and  3,6-anhydro-L galactose, linked by →3  and→  4 bonds.

Figure 1.  The structure of agarose.

Agarose is freely soluble in water at 100°C and upon  cooling  forms  an exceptionally  strong,  so-called macroreticular gel, with large pores  (Fig. 2).

Figure 2.  Comparison of micro- and macroreticular gels.

Sephadex  (Fig.  2A)  is  an  example  of  a  microreticular  gel.  In  a macroreticular gel (Fig. 2B), e.g. agarose, the gel fibres align into bundles resulting  in  a  much  stronger  gel  and  a  larger pore  size  at  a  given  gel concentration.  The  sketch  in  Fig. 2  is  a  2-D  representation,  but gels actually form 3-D labyrinths.

The  macroreticular  structure  of  agarose  makes  it  very  suitable as  a matrix  for  ion-exchangers  as  proteins  have  easy  access  to  the  gel interior,  so that  in  effect  the  gel has a very  large surface area  to  which ionic substituent groups may be attached.  The macroreticular structure is also mechanically  strong,  so that  substituted  agarose  ion-exchangers  do not  shrink  or  swell  with  changes  in  buffer  ionic  strength.  The  gel structure of agarose is maintained by non-covalent bonds and agarose gels cannot be dried and reconstituted.  They are  consequently  supplied in the form  of  a  slurry.  They  also  cannot  be  boiled  or  autoclaved,  as  they simply melt to a sol at high temperatures.

Common substituent groups are shown in Table 1.  The common weak base anion exchanger group is DEAE- and the  common  weak acid cation exchanger  group  is CM-.  The  pH  range  over  which  these  groups  are ionized is shown in Fig. 3.

Table  1.  Some  common  ion-exchange  substituent  groups.

Figure  3.  Ionization characteristics  of CM-  and  DEAE-  substituent groups.

It can be seen from Fig. 65 that exchangers comprised of weak  acid or weak  base  substituent  groups  are  not  completely  ionized  at  most  pH values  of  interest  and,  perhaps  a  greater  drawback,  the  degree  of ionization changes with pH in the  useful pH range.  Exchangers based on strong  acids or bases,  by  contrast,  are  Completely  ionized  over  a much larger pH range, so their degree of ionization is less subject to change with pH (Fig. 4).

Figure 4.  Ionization  of strong  acid  and  strong  base  ion-exchange  substituent  groups.

2 . Gradient generators

Ion-exchange  chromatography  often  requires  elution  of the  bound proteins by a change in either ionic strength, pH,  or both.  If the  system being separated  is well characterized,  then  appropriate  stepwise  changes can be made,  An  example  is  in  the  ion-exchange  separation  of  amino acids in an amino  acid analyzer.  More  commonly,  in  protein  isolation the  exact characteristics of the  components being  separated  are  unknown and it  is  then  necessary  to  use  a  gradient  generator  to  effect  an  ionic strength or pH gradient.

Gradients  are  commonly  generated  in  one  of  two  ways. A  two- chamber  device,  with  a magnetic  stirrer  stirring  one  chamber,  may  be used,    but  an  even  simpler arrangement is to have two conical flasks with a siphon  arranged between them  (Fig. 5).

Figure  5. A simple gradient generator set-up.

The  starting  solution  is  placed  in  the  right  hand  vessel  and  the finishing solution in the left hand vessel.  A siphon is established between

The vessels by sucking solution up and clamping the  T-piece  side tubing. The rigid tubing can consist of flexible tubing inserted  inside of,  for  e.g., plastic disposable pipettes.

A  more  sophisticated,  but  more  expensive,  method  of generating  a gradient is to use a micro-processor controlled proportioning valve  which draws liquid alternately  from  one  vessel  and  then  the  other,  in  small amounts  which  gradually change  in  proportion  with time.  A  mixer  is placed in the  line,  downstream of the proportioning  valve,  to  change  the small  stepwise  changes  in  buffer  composition  into  a  gradual  and continuous change.  Gradient generators  based  on  proportioning  valves are  common  components  of  complete  chromatography  systems, commercially available from a number of manufacturers.

3 . Choosing the pH

One  of the  decisions  that  has  to  be  made  before  conducting  ion- exchange chromatography is what pH to use.  The  selection of pH and  of the  type  of ion-exchanger  to  use may  be  facilitated  by  establishing  the so-called titration  curves of the  proteins  in the  mixture  to  be separated. An electrophoretic titration curve can be determined by establishing a pH gradient  in  a  gel    and  conducting  an  electrophoretic separation  at right-angles to the pH gradient (Fig. 6).

Figure 6.  Electrophoretic  titration curves  of proteins.

In  Fig. 6,  pH  “a”  is the  pH  of  maximal  charge  difference  between proteins  A  and  B.  At  this  relatively  low  pH,  both  proteins  have  a positive charge and a cathodic migration.  This  information  suggests that cation-exchange  chromatography,  conducted  at  pH  “a”,  would  most likely effect the best separation between  proteins  A and B.  By  contrast, at pH “d” the proteins have no apparent charge  difference.  This implies that anion-exchange chromatography would be less successful at resolving A and B, especially at pH “d”.  Points “b” and “c” represent the pI values of proteins B and A, respectively.  The difference in these pI values gives an  indication  of the  separation  which may  be expected  from  isoelectric focusing  or Chromatofocusing.

4 . An ion-exchange chromatography run

The  choice  of  the  type  of  exchanger,  cation-  or  anion-exchanger, may be arbitrary,  in the  absence of any  knowledge of the  characteristics of the  protein  of interest,  e.g.  its  pI  and/or  its  pH  stability  range.  As  a first  approach,  it  is generally  best to  choose  a pH,  within  the  protein’s stability  range, where it will adsorb to  the  ion-exchanger.  Usually, this means that an anion-exchanger should be  used at  pH values above the  pI of the  protein  and a cation-exchanger  at  a pI  below the  pI.  (It  must be realized,  however,  that  the  pI  refers  to  an  overall  property  of  the protein, whereas binding to  an  ion-exchanger  is a function  of the  charge on one surface of the  protein.  It  is therefore  possible to  have  a protein bind to an anion-exchanger at a pH below its pI  or to  a cation-exchanger at  a pH above  its  pI.  It  becomes  a matter  of  experimentally  exploring the behaviour of each new unknown protein.)

With  the  resin  chosen  and  the  column  packed,  it  is  necessary  to equilibrate  the  column  with  several  column  volumes  (“colvols”)  of starting  buffer,  a  buffer  of  low  ionic  strength  and  of  a  pH  which  will promote  binding  of  the  protein  of  interest  to  the  resin.  The  sample protein  mixture  must  contain  a  low  salt  concentration,  achieved  by equilibrating it with the  starting  buffer;  either  by dialysis,  ultrafiltration, molecular exclusion  chromatography  or,  following  TPP,  by  simply  re- dissolving the precipitate in the starting buffer.

The sample solution is applied to the  column and chased with at  least 2 colvols  of starting  buffer in  order to  elute  the  unbound  fraction.  The A₂₈₀  may  be  monitored  during  this  process  and  elution  with  starting buffer  stopped  once  the  A₂₈₀  returns to  the  baseline.  At  this  point  a buffer  gradient  may  be  applied.  As  a  first  approach,  a  gradient  of increasing  ionic  strength  is the  best  choice,  and  is  applicable  to  both cation- and anion-exchange.

The  resolution  of peaks  is a function  of the  steepness  of the  eluting gradient. A shallow gradient gives better resolution, but takes  more  time, so  a trade-off must be  made.  With  an  unknown  system,  the  best  first approach is to use a steep gradient, as this gives a quick assessment of the number of peaks to  be expected,  and the  separation  can  subsequently be optimized.  A suitably steep gradient for a first approach  is  M NaCl in starting buffer, in 3 colvols.  For subsequent optimization  the  gradient limits  or the  number  of colvols  can  be  altered,  to  change  the  gradient slope.

It must be  appreciated that  the  gradient  is applied to  the  inlet  side of the column, whereas monitoring of the effluent is done on the outlet side. There  is  thus  a  colvol  difference  between  the  influent  and  effluent streams,  Consequently,  after  application  of the  gradient,  it  is necessary to  elute  with  at  least  one  colvol  of  finishing  buffer to  ensure  that  the whole gradient itself is eluted and that all peaks  which would be eluted by the  gradient are,  in fact,  washed from  the  column.  An  example  of  the reporting of an ion-exchange chromatography run is presented in Fig.  7.

Figure  7.  An example  of ion-exchange chromatography :  Purification of cathepsin S,  from a TPP fraction from bovine spleen, by  chromatography on S-Sepharose. Column, 2.5  an i.d.  x 17  cm;  Starting  buffer, 20  inM Na-acetate containing  1 mM EDTA and 0.02% NaN3, pH 5.0; Gradient, 0------300  in M NaCl in 600 ml.  (Note that the column is eluted with  1  column volume of finishing  buffer  after the  gradient  is  applied,  in  order  to elute  proteins  which  may  have  been  displaced  by  the  buffer but  not  yet  eluted  from the column.)  The  thick  line  indicates  cathepsin  S activity  and  the  thin  line  is  A₂₈₀.  (The apparent  activity  in  the  break-through  peak  reflects  the  fact  that  the  assay  is  not absolutely  specific  for cathepsin  S).

S-Sepharose is a cation  exchanger.  If an  anion  exchanger were used, elution  with an  ionic  strength  gradient  could  be  effected  in  exactly  the same  way. A  different  strategy  is  needed  for  cation  and  anion exchangers, however, if a pH gradient is used to  elute  the  bound proteins. In  the  case  of  a  cation  exchanger,  a  gradient  of  rising  pH  is  required, whereas with an anion exchanger a gradient of descending pH is required. After  completion  of the  chromatography  run,  with  either  a cation  or anion exchanger, firmly bound  sample  components  may  be eluted with a high  salt  concentration,  such  as  1 M  NaCl.  In  the  case  of  cation exchangers, high salt concentration may be combined with a high pH, and with  anion  exchangers  the  high  salt  may  be  combined  with  a  low pH. The extreme pH values used must, of course, be within the stability limits of the ion-exchanger resin.  Finally,  the  column  is re-equilibrated with several  colvols of starting buffer, in preparation for the next run.

Ion-exchange  chromatography  thus  requires  cycling  through  large changes  in  ionic  strength  and  possibly  also  of  pH.  As  mentioned previously  it is an  advantage  if the  resin  can  withstand the required ionic strength and pH changes, without shrinking  or swelling, as  this  makes  it  possible  to  cycle  through  the  changes  in  buffer composition without having to repack the column.

References

-Dennison, C. (2002). A guide to protein isolation . School of Molecular mid Cellular Biosciences, University of Natal . Kluwer Academic Publishers new york, Boston, Dordrecht, London, Moscow .