Active membrane transport and receptor proteins from bacteria.

A general strategy for the expression of bacterial membrane transport and receptor genes in Escherichia coli is described. Expression is amplified so that the encoded proteins comprise 5-35% of E. coli inner membrane protein. Depending upon their topology, proteins are produced with RGSH6 or a Strep tag at the C-terminus. These enable purification in mg quantities for crystallization and NMR studies. Examples of one nutrient uptake and one multidrug extrusion protein from Helicobacter pylori are described. This strategy is successful for membrane proteins from H. pylori, E. coli, Enterococcus faecalis, Bacillus subtilis, Staphylococcus aureus, Microbacterium liquefaciens, Brucella abortus, Brucella melitensis, Campylobacter jejuni, Neisseria meningitides, Streptomyces coelicolor and Rhodobacter sphaeroides.

Weak substrate binding to transport proteins studied by NMR.

The weak binding of sugar substrates fails to induce any quantifiable physical changes in the L-fucose-H+ symport protein, FucP, from Escherichia coli, and this protein lacks any strongly binding ligands for competitive binding assays. Access to substrate binding behavior is however possible using NMR methods which rely on substrate immobiliza-tion for detection. Cross-polarization from proton to carbon spins could detect the portion of 13C-labeled substrate associated with 0.2 micromol of the functional transport system overexpressed in the native membranes. The detected substrate was shown to be in the FucP binding site because its signal was diminished by the unlabeled substrates L-fucose and L-galactose but was unaffected by a three- to fivefold molar excess of the non-transportable stereoisomer D-fucose. FucP appeared to bind both anomers of its substrates equally well. An NMR method, designed to measure the rate of substrate exchange, could show that substrate exchanged slowly with the carrier center (>10(-1) s), although its dynamics are not necessarily coupled strongly to this site within the protein. Relaxation measurements support this view that fluctuations in the interaction with substrate would be confined to the binding site in this transport system.

The post-translational processing and intracellular sorting of carboxypeptidase H in the islets of Langerhans.

The post-translational processing and intracellular sorting of the proinsulin-converting enzyme carboxypeptidase H (CPH) was studied in isolated rat islets of Langerhans. Pulse-chase-radiolabelling experiments using sequence-specific antisera showed that CPH was synthesized initially as a 57-kDa glycoprotein which was processed to a 54-kDa mature form by proteolytic processing at the N-terminus. Processing of the CPH precursor occurred rapidly (t(1/2) = 30) after an initial delay of 15-30 min and the enzyme was secreted in parallel with the insulin-related peptides in response to glucose-stimulation within 1 h after radiolabelling. This indicated that the proteins were packaged into nascent secretory granules at approximately the same rate following synthesis. Conversion of proinsulin and the 57-kDa form was inhibited markedly by chase incubation of islets at 20 degrees C, indicating that maturation of both proteins occurs in a post-Golgi compartment. Affinity purification of the enzyme from insulinoma subcellular fractions showed that the 57-kDa form was associated with endoplasmic reticulum or Golgi elements, and the 54-kDa form was present in secretory granules. Structural analysis showed that the granule form of the enzyme had an N-terminal amino acid sequence beginning at residue 42 of rat CPH, thereby implicating cleavage of the precursor after the fourth Arg in a site containing five consecutive Arg residues. These findings indicate that post-translational processing of CPH is mediated by an endoprotease which cleaves at sites containing multiple basic amino acid residues upon segregation of the enzyme to the secretory granules.

Kinetics and thermodynamics of the binding of forskolin to the galactose-H+ transport protein, GalP, of Escherichia coli.

The binding of the transport inhibitor, forskolin, to the galactose-H+ symporter, GalP, of Escherichia coli was evaluated by equilibrium and time-resolved fluorescence measurements. A quench in protein fluorescence of 8-12% was observed upon the binding of forskolin. The overall dissociation constant (Kd) for forskolin determined by fluorescence titration ranged between 1.2 and 2.2 microM, which is similar to that reported from equilibrium dialysis measurements of the binding of [3H]forskolin (Kd = 0.9-1.4 microM). The kinetics of forskolin binding were measured by stopped-flow fluorescence methods. The protein fluorescence was quenched in a biphasic manner; the faster of these two rates was dependent on the concentration of forskolin and was interpreted as the initial binding step from which both the association (kon) and dissociation (koff) rate constants were determined. The association and dissociation rate constants were 5.4-6.2 microM-1.s-1 and 5.1-11.5 s-1 respectively, and the Kd was calculated to be 1.5 microM. The binding of forskolin was inhibited by D-galactose, but not by L-galactose, and displacement by sugar provided an additional method to calculate the dissociation rate constant for forskolin (koff = 12.4-13.0 s-1). The rate of the slow change in protein fluorescence (3-5 s-1) was independent of the forskolin concentration, indicating an isomerization of the transporter between different conformations, possibly outward- and inward-facing forms. These kinetic parameters were determined at a series of temperatures, so that the thermodynamics of forskolin binding and transporter re-orientation could be analysed. The binding process was entropically driven (delta S = 83.7 J.K-1.mol-1; delta H = 8.25 kJ.mol-1), similar to that for cytochalasin B, which is also an inhibitor of GalP. Measurements of the binding of [3H]forskolin by equilibrium dialysis revealed competitive displacement of bound forskolin by cytochalasin B, possibly suggesting that the sugar, forskolin and cytochalasin B binding sites are overlapping; the Kds for forskolin and cytochalasin B were calculated to be 0.85 microM and 4.77 microM respectively, and the concentration of binding sites was 10.2 nmol.mg-1.

NMR observation of substrate in the binding site of an active sugar-H+ symport protein in native membranes.

NMR methods have been adopted to observe directly the characteristics of substrate binding to the galactose-H+ symport protein GalP, in its native environment, the inner membranes of Escherichia coli. Sedimented inner-membrane vesicles containing the GalP protein, overexpressed to levels above 50% of total protein, were analyzed by 13C magic-angle spinning NMR, when in their normal "fluid" state and with incorporated D-[1-13C]glucose. Using conditions of cross-polarization intended to discriminate bound substrate alone, it was possible to detect as little as 250 nmol of substrate added to the membranes containing about 0.5 mumol (approximately 26 mg) of GalP protein. Such high measuring sensitivity was possible from the fluid membranes by virtue of their motional contributions to rapid relaxation recovery of the observed nuclei and due to a high-resolution response that approached the static field inhomogeneity in these experiments. This good spectral resolution showed that the native state of the membranes presents a substrate binding environment with high structural homogeneity. Inhibitors of the GalP protein, cytochalasin B and forskolin, which are specific, and D-galactose, but not L-galactose, prevent or suppress detection of the 13C-labeled glucose substrate, confirming that the observed signal was due to specific interactions with the GalP protein. This specific substrate binding exhibits a preference for the beta-anomer of D-glucose and substrate translocation is determined to be slow, on the 10(-2) s time scale. The work describes a straightforward NMR approach, which achieves high sensitivity, selectivity, and resolution for nuclei associated with complex membrane proteins and which may be combined with other NMR methodologies to yield additional structural information on the binding site for the current transport system without isolating it from its native membrane environment.

The post-translational processing of chromogranin A in the pancreatic islet: involvement of the eukaryote subtilisin PC2.

The post-translational processing of chromogranin A (CGA) and the nature of the enzyme(s) involved were investigated in rat pancreatic islet and insulinoma tissue. Pulse-chase radiolabelling experiments using sequence-specific antisera showed that the 98 kDa (determined by SDS/PAGE) precursor was processed to an N-terminal 21 kDa peptide, a C-terminal 14 kDa peptide and a 45 kDa centrally located peptide with a rapid time course (t1/2 approx. 30 min) after an initial delay of 30-60 min. The 45 kDa peptide was, in turn, converted partially into a 5 kDa peptide with pancreastatin immunoreactivity and a 3 kDa peptide with WE-14 immunoreactivity over a longer time period. Incubation of bovine CGA with rat insulinoma secretory-granule lysate produced peptides of 18, 16 and 40 kDa via intermediates of 65 and 55 kDa. N-terminal sequence analysis indicated that cleavage occurred at the conserved paired basic sites Lys114-Arg115 and Lys330-Arg331, suggesting that cleavage of the equivalent sites (Lys129-Arg130 and Lys357-Arg358) in the rat molecule produced the initial post-translational products observed in intact pancreatic beta-cells. The enzyme activity responsible for the cleavage of bovine CGA co-chromatographed on DEAE-cellulose with the type-2 proinsulin endopeptidase and with PC2 immunoreactivity. The type-1 enzyme (PC1/3) appeared inactive towards CGA. The requirement for Ca2+ ions and an acidic pH for conversion was consistent with the involvement of a member of the eukaryote subtilisin family, and the composition of the released peptides in pulse-chase and secretion studies suggested that conversion occurred in the secretory-granule compartment. The overall catalytic rate as well as the relative susceptibilities of the Lys114-Arg115 and Lys330-Arg331 sites to cleavage were affected by pH, suggesting that the ionic environment of the processing compartment may play a role in the differential processing of CGA which is evident in various neuroendocrine cells.

The post-translational processing and intracellular sorting of PC2 in the islets of Langerhans.

Proinsulin conversion in the insulin secretory granule is mediated by two sequence-specific endoproteases related to the Kex2 homologues, PC2 and PC3 (Bennett, D. L., Bailyes, E. M., Nielsen, E., Guest, P. C., Rutherford, N. G., Arden, S. D., and Hutton, J. C. (1992) J. Biol. Chem. 267, 15229-15236; Bailyes, E. M., Bennett, D. L., and Hutton, J. C. (1992) Enzyme, in press). Radiolabeling studies using isolated rat islets showed that PC2 was synthesized initially as a 76-kDa glycoprotein which was converted by limited proteolysis to the mature 64-66-kDa form. Conversion was initiated approximately 1 h after synthesis and proceeded via intermediates of 71, 68, and 66 kDa with a t1/2 of 140 min. Release of only the 66- and 64-66-kDa radiolabeled forms of PC2 was induced by glucose and then only at times more than 2 h following synthesis. Proinsulin conversion, by contrast, was more rapid (delay = 30 min, t1/2 = 60 min), and release commenced as soon as 1 h after synthesis with the secreted material being comprised of the precursor, intermediate, and mature forms of insulin. Ultrastructural analysis of islet beta cells showed that PC2 was concentrated in secretory granules. Subcellular fractionation combined with immunoblot analysis showed that insulinoma secretory granules contained only the mature 64-66-kDa form of PC2, whereas fractions enriched in Golgi and endoplasmic reticulum contained a mixture of the 76- and 66-kDa forms of the enzyme. These results indicate that post-translational proteolysis of PC2 is initiated before sorting into the regulated pathway of secretion and that the relative proportions of proinsulin and PC2 packaged into secretory granules will change with physiological conditions.

Identification of the type 2 proinsulin processing endopeptidase as PC2, a member of the eukaryote subtilisin family.

Enzymological studies have implicated two Ca(2+)-dependent endopeptidases in the conversion of proinsulin to insulin; a type 1 activity which cleaves on the C-terminal side of Arg31-Arg32 and a type 2 activity which cleaves C-terminally to Lys64-Arg65 in the proinsulin sequence. The possibility that these enzymes are related to the recently discovered family of mammalian subtilisin-like gene products (furin, PC2, and PC3) and the yeast propheromone-converting enzyme (KEX-2), was investigated. Degenerate oligonucleotide primers flanking the putative catalytic domain within this gene family were used in a polymerase chain reaction to amplify related sequences from rat insulinoma cDNA. One major product of 700 base pairs was obtained which was greater than 99% identical to the corresponding rat PC2 sequence. This cDNA was subcloned into the bacterial expression vector pGEX-3X to generate a recombinant protein for antibody production. Western blot analysis showed the immunoreactivity was prominent in neuroendocrine tissues as a 65-kDa protein. It was concentrated in secretory granule-enriched fractions of insulinoma tissue, where it was present as a readily solubilized monomeric protein. Deglycosylation studies using endoglycosidase H and N-glycanase showed that the 65-kDa protein was comprised of approximately 9% carbohydrate, consistent with the presence of three consensus sequences for N-linked glycosylation in rat PC2. The immunoreactivity co-eluted with the type 2 proinsulin endopeptidase on gel filtration and ion-exchange chromatography and the antisera specifically immunoprecipitated type 2 activity from insulin granule extracts. N-terminal sequence analysis of the immunoreactive protein gave two sequences which corresponded to residues 109-112 and 112-119 of rat PC2. This indicated that posttranslational processing of PC2 itself occurs C-terminally to basic amino acids to produce the mature enzyme. It is concluded that PC2 is the type 2 endopeptidase involved in proinsulin conversion. Localization of PC2 immunoreactivity to other tissues of the diffuse neuroendocrine system suggests that the type 2 endopeptidase also functions in the processing of precursor forms of other prohormones and polypeptide neurotransmitters.

Insulin secretory granule biogenesis. Co-ordinate regulation of the biosynthesis of the majority of constituent proteins.

Two-dimensional gel-electrophoretic analysis combined with fluorography and densitometric quantification was used to examine the effects of glucose on the biosynthesis of rat pancreatic islet proteins. An increase in the medium glucose concentration from 2.8 to 16.7 mM produced a 10-20 fold stimulation in the synthesis of 10 out of 260 detected islet proteins, as judged by incorporation of [35S]methionine during a 20 min incubation. The synthetic rates of the majority of the remaining proteins were stimulated by 2-4-fold. Greater resolution achieved by pulse-chase labelling and subcellular fractionation showed that, of 32 major proteins localized to insulin secretory granules, the biosynthesis of 25 were stimulated 15-30-fold by glucose. By contrast, only eight of 160 proteins in the soluble fraction showed a response of similar magnitude. It is concluded that there is a major and co-ordinated activation of the biosyntheses of proteins destined for secretory granules, which most likely occurs at the level of translational initiation and signal-recognition-particle-mediated translocation into the endoplasmic reticulum lumen. However, it is clear that not all granule proteins, or the majority of proteins translocated across the endoplasmic reticulum membrane, are affected in an equivalent manner. In addition, the synthesis of a small number of cytosolic proteins may be increased markedly by insulinotropic stimuli.

The tissue distribution of rat chromogranin A-derived peptides: evidence for differential tissue processing from sequence specific antisera.

The distribution of chromogranin A and related peptides in rat tissues was investigated using sequence specific antisera. N- and C-terminal antisera and a presumptive C-terminal rat pancreastatin antiserum immunostained an extensive neuroendocrine cell population throughout the gastro-entero-pancreatic tract, anterior pituitary, thyroid and all adrenomedullary cells. However, mid- to C-terminal antisera immunostained a subpopulation of chromogranin A positive cells. Most notable of these was with the KELTAE antiserum (R635.1) which immunostained discrete clusters of adrenomedullary cells and antiserum A87A which immunostained a subpopulation of cells in the anterior pituitary and throughout the gastrointestinal tract. The present study has demonstrated the widespread occurrence of chromogranin A and related peptides in rat neuroendocrine tissues and provides evidence of tissue and cell specific processing.

Isolation, characterization and localization of a lectin within the vitelline membrane of the hen's egg.

A lectin with an affinity for certain sulphated polysaccharides, such as fucoidin and dextran sulphate, has been isolated from the vitelline membrane of hens' eggs and purified to homogeneity as assessed by two-dimensional gel electrophoresis. Polyclonal and monoclonal antibodies have been raised to the lectin and used in indirect immunofluorescence microscopy to localize the agglutinin in the outer layer of the vitelline membrane, where the lectin persists prior to the breakdown of the vitelline membrane. The quantity of lectin extracted from the two layers of the membrane, which have been separated by the method of Bellairs, Harkness & Harkness (1963), correlated well with the results of immunofluorescence microscopy. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis of the two layers of the membrane indicates that each layer has a distinctive polypeptide composition, the outer layer containing in particular lysozyme and avidin. The evidence obtained in this study indicates that the lectin is not involved in adhesion of the blastoderm to the vitelline membrane; neither is it involved in the expression of the blastoderm nor in maintaining the strength of the membrane. The possible roles in promoting transport of solutes across the membrane as well as providing bactericidal properties to the egg are discussed.