Construction, purification and biological activities of recombinant human interleukin-2 analogs.

Fourteen human interleukin-2 (IL-2) analogs have been cloned and expressed in E. coli, starting from a chemically synthesized gene for human IL-2 optimized for expression in E. coli. These analogs were purified to greater than 95% purity as determined by SDS-PAGE, and were measured for biological activity in a 3H-thymidine incorporation assay using an IL-2 dependent murine T-cell line (CTLL). One analog was made which eliminated the N-terminal 23 amino acids from the protein by replacing one restriction endonuclease fragment with another. This analog, which begins at an internal methionine, had no detectable CTLL activity. Thirteen analogs were constructed using oligonucleotide site-directed mutagenesis. Four of these analogs were truncated at various residues near the C-terminus (residues 106, 116, 121 and 126). These analogs had at least 500-fold lower CTLL activities than the natural recombinant IL-2. The remaining nine analogs had substitutions at 1, 2, or all 3 of the three cysteine residues in the protein (residues 58, 105 and 125). Substituting an alanine, asparagine, aspartic acid, or serine at residue 125 resulted in highly active molecules with CTLL activities similar to that of the natural recombinant IL-2. The analogs with alanine and serine substitutions at residue 125 actually had slightly higher CTLL activities than the natural recombinant IL-2. Substituting alanine for cysteine at position 125 and serine for cysteine at either position 58 or 105 yielded analogs with about 150-fold lower CTLL activities than natural recombinant IL-2. Substituting an alanine for the cysteine at position 125 and serines for cysteines at both positions 58 and 105 resulted in an analog with 30-fold lower CTLL activity than the natural recombinant IL-2. The ten analogs with less than 1.0% of the CTLL activity of natural recombinant IL-2 were tested for competition with the natural recombinant IL-2 by mixing a 10-to 100- fold excess of the analog with the natural recombinant IL-2 and assaying the mixture in the CTLL assay. None of these analog mixtures resulted in a lower activity than mixing the natural recombinant IL-2 with buffer alone, implying that none of these analogs effectively competes with the natural recombinant IL-2 for binding to IL-2 receptors during incubation with the CTLL cells. If reduced binding does occur, it may be the direct cause of their lower activities.