One of the most effective ways to deliver cytotoxic drugs is by conjugation, where a relatively small molecular weight drug is covalently linked to usually a protein, which recognizes a specific cellular receptor that is expressed de novo or over-expressed by the targeted cells. In fact, some of the most useful proteins are the monoclonal antibodies (mAbs), which can be selected to recognize a single specific epitope on the cell surface. Yet, proteins like transferrin, hemoglobin, albumin and others are also being used to prepare these conjugates. Because of the recognition site on a protein, the essential pharmacophores on the drug responsible for its pharmacological activity, preparation of a conjugate needs to address preserving the binding capacity of the protein and most likely the release of the drug once it has entered the cell so that it can find its therapeutic target.
Hence, the design of a drug conjugate would require identifying functional groups in the protein that are far from its binding site, so binding of the drug does sterically hinders that site. If the protein is produced by recombinant DNA technology, the appropriate groups like thiol group can be introduced in the protein amino acid sequence; otherwise, there will be a need to introduce new functional groups by selective modification of certain groups in the protein. For instance, thiolation of amino groups using either N-acetyl homocysteine thiolactone (SATA) (Fig. 1) or S-acetylmercaptosuccinic anhydride (SAMSA) is a common approach modifying proteins.
Fig.1. Modification of amino groups with SATA
The success of the modifications would depend on their selectivity and if the reactive groups are far from the protein’s binding site. In some cases, if the ligand is a carbohydrate or other non-peptide compound, it would be possible to saturate the binding site before modification to protect the active site. Once the modification has taken place, the ligand can be removed from the binding site by extensive dialysis or similar procedures. An alternative method would be the reaction of the terminal amino group of a protein with an aldehyde carrying bifunctional linker. This method takes advantage of the lower pKa of the terminal α-amino group (pKa 7.6-8) than that of the lysine ε-amino groups (pKa ~ 9.4). Thus by selecting the pH of the reaction, the aldehyde group will form an imine with the terminal α-amino group, which may be reduced to a stable secondary amine with sodium cyanoborohydride. Evidently, this method would be useful only in proteins where the terminal amino group is not present at the binding site. While modifications of the carboxyl groups are possible, its selectivity would be less than the above-described methods and may lead to cross-linking with neighboring proteins; unless there is evidence that no carboxyl groups are present at the binding site.
In the case of conjugates with mAbs, one of the most used methods is to reduce one of the two disulfide bonds linking together the two IgG heavy chains to form two thiol groups. Under these conditions, the formed thiol groups are present in the Fc region and away from the IgG recognition sites formed by the heavy and light chains. These two thiols can them react with a maleimide of an iodoacetyl group to form a stable thioether; using this approach it is possible to introduce one or two drug molecules per IgG molecule. A complete reduction of the two disulfide bonds may deliver two half-IgG molecules, which can be linked to a drug molecule. However, after reduction of both disulfide bonds, in some cases the two half IgGs would not separate, which may interfere with the conjugate formation. While the reduction of the disulfide bond is quite effective for IgGs, it is doubtful that it may be used with other proteins, a disulfide bond usually stabilizes the three-dimensional structure of a protein.
Because the conjugated drug usually needs to interact with a cellular target after intracellular delivery, it would need to be released from the protein to which it is linked. To achieve this goal, the drug is covalently linked to the protein via a cleavable linker, i.e. a chemical structure that upon exposure to an enzyme, low pH, excess of reducing thiol groups, will split leaving the drug free and able to interact with its target. There are several types of cleavable linkers, and some of the most frequently used are those having a hydrazone group, which is cleaved at acid pH, < 5, in the endolysosome compartment, after uptake of the conjugate by endocytosis (Fig, 2). Another type of cleavable linker is the peptide pro-drug linker (self immolatory) having a Val-Cit-PAB-PNP sequence. This type of linker has two cleavage sites; site “a” specific for the enzyme cathepsin B cleavage, and site “b” that is the (1,6)-fragmentation site of the self-immolative link. To protect the drug, the self-immolatory spacer is located between the site of enzymatic cleavage and the drug; subsequent to the enzymatic cleavage, there is an extensive chemical rearrangement, a 1,6 elimination, that results in the drug release. Self-immolatory linkers are cleaved at the lysosome by proteases, after uptake of the conjugate by endocytosis. (Fig. 3). Although some cleavable linkers containing disulfide bonds have been described, they are not too stable and may be reduced in circulation by compounds like cysteine, glutathione, and other thiols.
Fig. 2. Cyclohexyl-aryl hydrazone cleavable linker
Fig. 3. Peptide pro-drug linker (self immolatory): Mal-Val-Cit-PAB-PNP
The other crucial component of these protein-conjugates, is the drug; which usually is linked by one of its functional groups that is not an essential for its pharmacological activity. Special care should be taken to avoid any significant structural changes that may alter the interactions of the drug with its target. Because the activity would be dependent on the amount of drug, it is important to determine the composition of the conjugate and to have a relatively homogeneous conjugate population.
Reference
Miguel Castro and Dante Marciani; Ebr Bioconjugates January 2012, 44-50. www.samedanltd.com
Ducry, L., Stump, B.; Antibody-drug conjugates: linking cytotoxic payloads to monoclonal antibodies. Bioconjug Chem. 2010 Jan;21(1):5-13. doi: 10.1021/bc9002019.
Roger L. Lundblad; Chemical Reagents for Protein Modification, Fourth Edition Fourth Edition, 2014. CRC Press.
-.-