To understand how the protein and chromophore components of a light-sensing protein interact to create a light cycle, we performed time-resolved spectroscopy on site-directed mutants of photoactive yellow protein (PYP). Recently determined crystallographic structures of PYP in the ground and colorless I2 states allowed us to design mutants and to study their photosensing properties at the atomic level. We developed a system for rapid mutagenesis and heterologous bacterial expression for PYP apoprotein and generated holoprotein through formation of a covalent thioester linkage with the p-hydroxycinnamic acid chromophore as found in the native protein. Glu46, replaced by Gln, is buried in the active site and hydrogen bonds to the chromophore's phenolate oxygen in the ground state. The Glu46Gln mutation shifted the ground state absorption maximum from 446 to 462 nm, indicating that the color of PYP can be fine-tuned by the alteration of hydrogen bonds. Arg52, which separates the active site from solvent in the ground state, was substituted by Ala. The smaller red shift (to 452 nm) of the Arg52Ala mutant suggests that electrostatic interactions with Arg52 are not important for charge stabilization on the chromophore. Both mutations cause interesting changes in light cycle kinetics. The most dramatic effect is a 700-fold increase in the rate of recovery to the ground state of Glu46Gln PYP in response to a change in pH from pH 5 to 10 (pKa = 8). Prompted by this large effect, we conducted a careful reexamination of pH effects on the wild-type PYP light cycle. The rate of color loss decreased about 3-fold with increasing pH from pH 5 to 10. The rate of recovery to the colored ground state showed a bell-shaped pH dependence, controlled by two pKa values (6.4 and 9.4). The maximum recovery rate at pH 7.9 is about 16 times faster than at pH 5. The effect of pH on Arg52Ala is like that on wild type except for faster loss of color and slower recovery. These kinetic effects of the mutations and the changes with pH demonstrate that both phases in PYP's light cycle are actively controlled by the protein component.

Bovine glucagon, a polypeptide of 29 amino acids, was immunogenic in guinea pigs. The immunologic determinants of glucagon were investigated using isolated tryptic peptides of the hormone. Antibodies from virtually all of more than two dozen animals had specificity primarily for the amino-terminal heptadecapeptide (NM) and showed little or no binding with the carboxy-terminal undeca- and dodecapeptides (C). The smallest synthetic peptide of a series initiated at residue 16 which measurably bound antibody comprised residues 5–16 of glucagon. In cellular immune assays, both NM and C elicited delayed cutaneous reactions and inhibited the migration of peritoneal cells from immune animals. However, only intact glucagon and its C fragment stimulated lymphoid cells to synthesize DNA. While glucagon was somewhat more active than C, the addition of NM to C did not enhance its transforming activity. The smallest synthetic carboxy-terminal peptide with discernible transforming activity comprised residues 19–29 of glucagon. In both native and synthetic C peptide preparations, the undecapeptide was generally more active than the dodecapeptide, although cells from different animals gave different response patterns. The difference between the two is the presence of arginine at the amino-terminus of the peptide chain. Thus, the recognition specificity of populations of antigen-reactive cells from different animals displays a variation which is at least superficially analogous to that of populations of antibody molecules. In limited experiments using NM and C peptides as immunogens, neither gave rise to delayed hypersensitivity or to glucagon-binding circulating antibody, following a regimen which invariably provoked these responses when glucagon itself served as the immunogen. These results indicate that glucagon was cleaved by trypsin along functional lines into two parts, one of which housed the major antigenic determinant and the other of which carried the major immunogenic determinant, and they are highly compatible with a two-cell mechanism of immune induction. An apparent dissociation between the capacity to provoke delayed hypersensitivity reactions and to transform antigen-reactive cells in culture was observed.

Based on the original observation that L-tyrosine-azobenzene-p-arsonate (RAT) I induced specific delayed hypersensitivity in guinea pigs (1), we have shown that this small molecule could serve as a carrier for a macromolecular hapten (2). Conjugates of RAT and poly-3'-n-glutamic acid (PGA) raised humoral antibody specific for the PGA determinant and cellular immunity directed against the RAT determinant in guinea pigs. In addition, prior immunization with RAT potentiated the anti-hapten response (2). These observations prompted the conclusion that RAT, although having a molecular weight of only 409, is a true immunogenThe R A T -P G A conjugates had, on the average, 9-12 R A T groups per molecule of PGA. In the present communication, we report the preparation and use of better defined, small bifunctional antigen molecules composed either of one R A T carrier moiety and one dinitrophenyl (DNP) haptenic group or of two R A T moieties, separated b y spacers of varying size. These molecules permitted an exploration of the spatial relationships between carrier and hapten required for a humoral antibody response, as well as the question of "self-help." The latter refers to the ability of a carrier determinant to cooperate in the humoral response to a second identical determinant. In addition, the structural features of R A T which are essential for immunogenicity have been investigated.

No abstract