Site-directed alterations to the geometry of the aspartate transcarbamoylase zinc domain: selective alteration to regulation by heterotropic ligands, isoelectric point, and stability in urea.

Structural aspects requisite for allosteric function in the regulatory chain of aspartate transcarbamoylase were explored by site-specific amino acid insertion or substitution within the zinc domain in the region of contact between the catalytic and regulatory chains. Amino acid substitution at two positions yielded enzymes which retained a maximum velocity similar to that of the wild-type enzyme but responded differently from the native enzyme in the presence of regulatory nucleoside triphosphates. A change of zinc coordinate amino acid C109 to histidine and a change of E119 to aspartic acid resulted in enzymes which demonstrated synergistic inhibition by CTP and UTP but not inhibition by CTP in either phosphate buffer or a morpholino-based tri-partate (TP) buffer at pH 7. At pH 8.3, where there is a higher proportion of T-state conformers in the native enzyme, the mutants diverged from their similar kinetic behavior. C109H remained an enzyme which was not inhibited by CTP but was still inhibited by CTP+UTP. E119D was inhibited by both CTP and CTP+UTP. Activation of the mutants by ATP was found to vary either with pH or with phosphate as a buffer component. C109H was activated by ATP in phosphate, while in TP at either pH 7 or 8.3 its activation by ATP was diminished or absent. E119D was activated by ATP in phosphate at pH 7 or in TP at pH 8.3, but not in TP at pH 7. In TP at pH 7, where neither mutant was activated by ATP, the S values and Hill coefficients of the unliganded mutant enzymes resembled those of the ATP-liganded wild-type enzyme. While neither mutation would be predicted to alter the net charge of the holoenzyme, differences in the isoelectric point of the mutants were observed if phosphate was present. This result suggests that the isoelectric point of aspartate transcarbamoylase is conformationally dependent and that the mutants exist in an altered conformation. In addition, the stabilities of both mutant holoenzymes were reduced substantially from those of the wild-type enzyme in 4 M urea. C109H was more stable at pH 8.25 in a Tris buffer; E119D was more stable at pH 7 in the phosphate buffer. Potential effects of these mutations on the active site chemistry and geometry are discussed.

Separation and determination of alpha-amino acids by boroxazolidone formation.

Reaction of an alpha-amino acid (alpha-AA) with 1,1-diphenylborinic acid (DPBA) leads to the formation of a kinetically stable adduct at pH 2-5 in which both the alpha-amino and the alpha-carboxyl groups are bound to boron forming a cyclic mixed anhydride termed a boroxazolidone. In this adduct, the greater than N:B bond is coordinate, involving the free electron pair of nitrogen, thereby satisfying the octet rule for the second electron shell of boron (Group IIIA). Consequently, the alpha-amino function of the boroxazolidone can be primary, secondary, or tertiary, as demonstrated by boroxazolidone formation with glycine, N-methylglycine, and N,N-dimethylglycine. On reaction with DPBA, the alpha-AA moiety of N-terminal gamma-glutamyl peptides is also derivatized as demonstrated by the formation of a glutathione boroxazolidone. The 1,1-diphenylboroxazolidone adducts of alpha-AA may be separated by reversed-phase (RP)-HPLC (AA-DPBA/RP-HPLC) enabling the derivatization procedure to be used as a precolumn reaction for alpha-AA analysis. Under the conditions we describe here, DPBA is not stably reactive with the epsilon-amino group of lysine. Furthermore, it does not complex with amide bonds of the peptide backbone or to any side chains of the common amino acids. Reaction of an alpha-AA mixture with DPBA, followed by RP-HPLC (AA-DPBA/RP-HPLC) is then a simple method by which to analyze alpha-AA in a mixture with peptides and amines. Precolumn reaction with DPBA may be used to separate peptides from alpha-AA and from those peptides which contain an alpha-AA moiety. Unreacted peptides are bound only weakly to the HPLC column and thus are separated from reacted alpha-amino acids which are retained as 1,1-diphenylboroxazolidones until their selective elution. This method is particularly suited for the analysis of alpha-amino acids that are derived from post-translational modification of protein side chains.

Biologic activity of a C2-derived peptide. Demonstration of a specific interaction with guinea pig lung tissues.

A synthetic peptide derived from the carboxy terminus of C2b has been investigated for its ability to induce the contraction of guinea pig lung parenchymal strips. This peptide is known to enhance vascular permeability in guinea pig and human skin, and to induce contraction of estrous rat uterus. This C2 peptide (C2 207-223) is active from 5 x 10(-5) M to 5 x 10(-4) M and is not tachyphylactic to itself. No cross-activity between C2 207-223 and C5a or C3a could be demonstrated. C2 207-223 is not inhibited by antihistamines or cyclooxygenase inhibitors. These data indicate that the peptide exerts its action via a mechanism distinct from those of the C3a and C5a anaphylatoxins.

Angioedema induced by a peptide derived from complement component C2.

Synthetic peptides that correspond to the COOH-terminal portion of C2b enhance vascular permeability in human and guinea pig skin. In human studies, 1 nmol of the most active peptide of 25-amino acid residues produced substantial local edema. A pentapeptide and a heptapeptide corresponding to the COOH-terminal sequence of C2b each induced contraction of estrous rat uterus in the micromole range; a peptide of 25 amino acids from this region induced a like contraction of rat uterus at a concentration 20-fold lower than the smaller peptides. The vascular permeability of guinea pig skin was enhanced by doses of these synthetic peptides in a similar fashion as that observed for the concentration of rat uterus. The induction of localized edema by intradermal injection in both the guinea pig and the human proceeds in the presence of antihistaminic drugs, suggesting that there is a histamine-independent component to the observed increase in vascular permeability. Cleavage of C2 with the enzymic subcomponent of C1, C1s, yields only C2a and C2b, and no small peptides, whereas cleavage of C2 with C1s and plasmin yields a set of small peptides. These plasmin-cleaved peptides are derived from the COOH terminus of C2b, and they induce the contraction of estrous rat uterus.

C1s-induced vascular permeability in C2-deficient guinea pigs.

Normal guinea pigs that have been intradermally injected with C1s exhibit increased vascular permeability at the injection site. Guinea pigs that are genetically deficient in complement component C2 do not exhibit increased vascular permeability when given a similar injection. The C2-deficient guinea pigs respond normally to injections of bradykinin and kallikrein, suggesting that these animals can respond to kinins and have a normal kininogen pathway. When the C2-deficient guinea pigs are given guinea pig C2 before C1s injection, increased vascular permeability is observed. These results demonstrate a definite requirement for complement component C2 in the generation of C1s-induced vascular permeability.

Ultrastructure and composition of bovine conglutinin.

Conglutinin binds in a Ca2+-dependent manner to the carbohydrate portion of zymosan and cell-bound iC3b (complement subcomponent C3b cleaved by Factor I in the presence of factor H) similarly to lectin-like proteins that participate in the clearance of plasma glycoproteins. This carbohydrate-binding protein has been found to include both collagenous and non-collagenous domains. Electron micrographs of bovine conglutinin are presented in which conglutinin appears as a tetramer of four 'lollipop' structures emanating from a central hub. The stem region, linking each head to the central hub, is quite stiff, whereas the hub-stem junction is a flexible hinge. From electron micrographs of a pepsin digest of conglutinin, the linkage region is identified as the collagenous portion of the macromolecule. Conglutinin is a multimer of a single polypeptide chain. From sedimentation equilibria of unreduced as compared with reduced and alkylated conglutinin, there are determined to be three disulphide-linked chains. These data, combined with information on the subunit polypeptide of conglutinin, suggest a model for conglutinin in which four disulphide-linked trimers are associated via the N-termini to form the intact macromolecule as viewed in the electron microscope. The ultrastructure of conglutinin appears ideally suited to its lectin-like function.

Conformation and restricted segmental flexibility of C1, the first component of human complement.

Seventy selected images of chemically crosslinked C1 are analyzed to illustrate structural details of the C1qC1r2C1s2 complex. From inspection of these images, the C1r2C1s2 tetramer can be seen to be located in the region of the C1q arms, cleanly separated from the C1q heads and from at least 90%, if not all, of the C1q stem. From measurements made upon 65 images, the semicone angles formed between the spreading arms and the symmetry axis passing through the stem of C1 may be calculated. Unlike C1q, for which a wide variety of angles is found, the C1 complex appears to possess a restricted range of angular flexibility with an average value of about 50 degrees. The volume inside the cone formed by the spreading arms of C1q is too small to contain the entire C1r2C1s2 tetramer; at least some of the tetramer must lie outside the cone when it is bound to C1q to form C1. From our knowledge of the sizes and structures of its subunits, and from symmetry considerations, a model is proposed for the configuration of the C1 complex in which the middle portion of the C1r2C1s2 tetramer is centrally located among the arms close to the stem of the C1q and with the two protruding ends of the tetramer wrapped around the outside of the cone. Functional implications of this more rigid structure are discussed with relevance to C1q-induced aggregation of latex beads and C1-induced disaggregation.

Ultrastructure of the first component of human complement: electron microscopy of the crosslinked complex.

Electron micrographs are shown of the first component of human complement (C1) which has been crosslinked with a water-soluble carbodiimide to prevent dissociation into its C1q and C1r2C1s2 subunits. Two projections of the crosslinked molecule are seen in the electron micrographs, which are called "top" and "profile." In both views, the C1q heads are visible. From the top, the C1r2C1s2 tetrameric subunits appears to be located centrally on the C1q and folded to form a compact mass obscuring most of the arms and central bundle. In profile, the tetramer appears to be located in the region of the arms between the C1q heads and the central bundle. Both the heads and the rod-like central bundle appear to be free of C1r2C1s2 in these profile projections. Sometimes it is possible to count more than six domains in the region of the C1q heads, as though a portion of the tetramer had unfolded to protrude among the heads.