Preliminary thermodynamic studies of a high affinity monoclonal anti-fluorescyl antibody (4-4-20) indicate that its active site is a shallow hydrophobic pocket, which binds fluorescein primarily through entropy. Moreover, high resolution diffraction data (2.5 Angstrom) is available for the liganded antigen binding fragment (Fab) of this antibody, which crystallizes in 16% polyethyleneglycol. Liganded Fab fragments also crystallize in a less polar solvent system (46.7% 2-methyl-2,4-pentanediol). The affinity of the intact IgG molecule is 300-fold lower in this solvent. Correlated crystal and solution studies of both crystal systems are planned. With this information we hope to explain the molecular basis of antigen binding to this antibody. Crystallization trials are currently in progress with four other monoclonal anti-fluorescyl antibodies which are idiotypically related to 4-4-20, but exhibit affinities which vary over a 1000-fold range. Solution studies are planned to determine whether the entropy predominance and active site characteristics of 4-4-20 are common to the other clones as well. Idiotypic determinants are generally thought to be located at or near the active site. Crystal studies should clarify the nature of idiotypic determinants and whether idiotypic reagents are valid tools for identifying related active sites. High resolution diffraction data (2.0 Angstrom) is available for the unliganded Fab fragment of a monoclonal antibody which binds single stranded DNA (BV04-01). In solution, the protein exhibited a base specificity for pyrimidines, with greater affinity for thymine than uracil. This antibody is of clinical importance because it was isolated from a mouse with an autoimmune syndrome similar to systemic lupus erythematosus. We have obtained a low-resolution (6 Angstrom) structure and plan to extend this solution to higher resolution. We also plan to perfuse oligonucleotides into crystals to determine the molecular basis of the observed pyrimidine specificity.