Crosslinking

 Crosslinking refers to linking separate chemical functionalities covalently. Crosslinking is usually achieved through use of a bifunctional crosslinking reagent, specifically, a compound having two reactive groups that react chemically with the two functional groups that become crosslinked. In this case, only two functional groups are usually crosslinked, but, with other types of reagents, higher order crosslinking is possible. Crosslinking can be either intramolecular (between separate functional groups on the same molecule) or intermolecular (between functional groups located on separate molecules). In the latter case, it is usually between two macromolecules or between a macromolecule and a relatively small molecule. Crosslinking is usually used to detect groups that are in much greater proximity than would be expected from just their bulk concentrations, due to the structures and associations of the molecules of which they are part. In biochemical applications, the chemical functions that can be crosslinked may be components of all classes of macromolecules (proteins, nucleic acids, carbohydrates, and lipids). For simplicity, this article focuses on crosslinking involving the functional groups of proteins, predominantly amino, thiol, imidazole and carboxyl groups (ie, the nucleophilic side chains of certain amino acids); however, the concepts are the same for the other classes of macromolecules. It should be noted that crosslinking also occurs naturally, resulting from oxidation, radiation, or the action of certain enzymes.

A distinction is sometimes made between the terms crosslinking and conjugation (or bioconjugation). In this case, crosslinking is used to denote the covalent linkage of only molecules that are associated naturally, such as the subunits of an oligomeric protein or a hormone and its receptor. Conjugation, in contrast, refers to the covalent linkage of molecules that lack affinity for one another. The resulting covalent complexes, referred to as conjugates, have a multitude of uses, especially in biotechnology. The distinction between crosslinking and conjugation is maintained in this article.

1. Uses of crosslinking

 On the basis of the variety of the crosslinking reagents that are available and of the molecules that can be crosslinked, plus the types of experimental questions that can be answered, crosslinking is among the most versatile of biochemical techniques. In the case of proteins, for example, some information about their three-dimensional structure can be gained through identifying specific amino acid residues that are readily crosslinked intramolecularly (1). Crosslinking can also be used as a general conformational probe to detect structural changes in a protein, such as might be induced by ligand binding (2). Inasmuch as crosslinking can hinder conformational transitions and the dissociation of oligomers, it can also be used to stabilize the tertiary or quaternary structures of proteins. Thus, it can “lock” a protein into a particular functional state (3) or stabilize an oligomeric protein for subsequent physical studies, such as electron microscopy. A careful analysis of the mass and composition of crosslinked complexes allows a minimal subunit stoichiometry to be determined; this is especially useful for insoluble oligomers, such as those found in membranes (4). Information concerning the symmetry of oligomeric proteins can also be obtained from crosslinking (5, 6). One of the most common uses of crosslinking is to identify neighboring proteins in a complex (7), and this can be extended to include determination of the specific regions of each that are crosslinked (8). By varying the length of the crosslinker, one can also obtain a limit for the maximum distance that can separate neighboring proteins (6). In addition to giving both relative and absolute structural information, crosslinking can be used for general screening. For example, one can identify the receptor for a particular ligand, such as a hormone (9).

It must be emphasized that for all the above applications, interpretations can be made only for those crosslinked complexes that are generated. The absence of crosslinking does not mean that two molecules do not interact, only that they are not crosslinked under the specific set of experimental conditions chosen, which includes the crosslinking reagent used. Because crosslinking requires the proper positioning of those functional groups having the appropriate chemical reactivities, one cannot expect all interacting molecules to be crosslinked by any given reagent.

 2. Crosslinking Procedures

Crosslinking, like chemical modification of proteins, is an empirical procedure, and both processes are influenced by the same variables, namely temperature, pH, concentrations of reactants, and time. In crosslinking, however, the influence of the latter two variables is more complex. Because nucleophilic side chains of amino acids are considerably more reactive when deprotonated, the pH of the reaction is important in controlling the extent of crosslinking and the selectivity for particular side chains. Regarding concentrations, the amounts of both the crosslinker and the proteins can influence crosslinking, with low concentrations of protein and crosslinker generally favoring intramolecular crosslinking. Adding the crosslinker over time, instead of as a single addition, favors a greater extent of crosslinking. Differences in selectivity of the reactive groups of bifunctional reagents, especially heterobifunctional crosslinkers, for protein functional groups can be maximized using a two-step reaction. In this procedure, one protein is first derivatized with the crosslinking reagent and then purified to remove excess free reagent before being exposed to the second protein. This two-step procedure, which allows for each of the two reactions to be performed under different conditions, is particularly useful for enzyme immobilization and conjugation, when the two molecules do not normally associate with each other.

 Following crosslinking, the covalent complexes formed are commonly identified as new electrophoresis bands in sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), with altered molecular weights, although care must be taken, because intramolecular crosslinking can alter the migration of individual polypeptides with this technique. Alternatively, in simple systems, covalently linked complexes can be identified as components with increased Stokes radii in size exclusion chromatography. If either the crosslinker or the crosslinked protein has identifiable physical or chemical characteristics, these can aid in the identification of complexes. Once crosslinked complexes have been identified, complete analysis of their composition and stoichiometry can still be challenging, especially when multiple proteins are present. One useful method for identifying the proteins composing a complex is to use a cleavable crosslinker so that the original components can be regenerated from the fractionated complex (7). Another useful method is to perform Western blots on the fractionated complex using antibodies against potential components (2) . Once the components have been identified, their stoichiometry in the complex is deduced from its overall mass. The ultimate step in analysis is to determine directly the actual amino acid residues that are crosslinked. Using conventional protein sequencing by Edman Degradation, the sequences of the two crosslinked peptides are obtained simultaneously, with the crosslinked residues represented by gaps in the sequences; however, identification of crosslinked peptides by mass spectrometry is becoming increasingly common.

References

1. F. C. Hartman and F. Wold (1967) Biochemistry 6, 2439–2448. 

2. O. W. Nadeau, D. B. Sacks, and G. M. Carlson (1997) J. Biol. Chem. 272, 26296–26201. 

3. R. E. Benesch and S. Kwong (1991) J. Prot. Chem. 10, 503–510. 

4. E. Heymann and R. Montlein (1980) Biochem. Biophys. Res. Commun. 95, 577–582. 

5. F. Hucho, H. Mullner, and H. Sund (1975) Eur. J. Biochem. 59, 79–87.

6. J. Hajdu, S. R. Wyss, and H. Aebi (1977) Eur. J. Biochem. 80, 199–207. 

7. D. Dey, D. E. Bochkariov, G. G. Jokhadze, and R. R. Traut (1998) J. Biol. Chem. 273, 1670–1676.

8. V. Rossi, C. Gaboriaud, M. Lacroix, J. Ulrich, J. C. Fontecilla-Camps, J. Gagnon, and G. J. Arlaud (1995) Biochemistry 34, 7311–7321. 

9. T. H. Ji (1977) J. Biol. Chem. 252, 1566–1570.