In a previous paper we presented preliminary experiments aimed at the
preparation of gel particles with the property to recognize selectivel
y some particular protein (hemoglobin, cytochrome C, transferrin) [1].
Using the same method we show in this article that human growth hormo
ne, ribonuclease and myoglobin from horse can also be adsorbed specifi
cally, indicating that the method may be universal or at least applica
ble to a great number of proteins. A gel with specific adsorption of t
hree model proteins was synthesized in order to demonstrate that the b
eds can be employed to remove (traces of) several proteins contaminati
ng a sample (''negative purification''). The degree of selective recog
nition is high, to judge from the fact that myoglobin from horse, but
not that from whale, was adsorbed onto a column designed to bind speci
fically the former protein. This selectivity is noteworthy since these
two proteins have similar amino acid sequences and 3-D structures. Th
e method for the synthesis of the specific gels involves polymerizatio
n of appropriate monomers (for instance acrylamide and its derivatives
) in the presence of the protein to be adsorbed specifically, granulat
ion of the gel formed, packing a column with the gel particles, washin
g the column to remove the protein and finally application of the samp
le for selective adsorption of the protein. The approach resembles tha
t used for entrapment (immobilization) of proteins for affinity chroma
tography and that for molecular imprinting, with the distinct differen
ce that the monomer composition is quite different and thereby the bin
ding mechanism. This mechanism is discussed, for instance, in terms of
(I) a new classification system for chromatographic beds based on the
number of bonds between the solute and the matrix and the strength of
each bond and (2) ''non-specific bonds'' (these bonds are often harmf
ul in conventional chromatography, but we have used them to advantage)
. In this classification system the selective recognition is character
ized by a large number of weak bonds. Therefore, so-called functional
monomers are not used for the preparation of the gels because they oft
en are charged and; accordingly, give rise to strong electrostatic int
eractions, i.e. the beds behave to some extent as ion-exchangers. In m
ost experiments we have used a polyacrylamide gel with large pores to
facilitate diffusion of proteins into and out of the gel granules. Whe
n used in chromatography these soft gels (which can be used repeatedly
) allow only rather low flow rates. This problem can be overcome by a
new approach to prepare the granules. Potential applications of the se
lective beds are discussed, as well as future improvements.