Roux-en-Y gastric bypass increases systemic but not portal bile acid concentrations by decreasing hepatic bile acid uptake in minipigs

Roux-en-Y gastric bypass (RYGB) surgery is widely used in the management of morbid obesity. RYGB improves metabolism independently of weight loss by still unknown mechanisms. Bile acids (BAs) are good candidates to explain this benefit, since they regulate metabolic homeostasis and their systemic concentrations increase upon RYGB. Here we analyzed the mechanisms underlying the increase in systemic BA concentrations after RYGB and the role of the liver therein. To this aim, we used the Göttingen-like minipig, a human-size mammalian model, which allows continuous sampling and simultaneous analysis of pre-hepatic portal and systemic venous blood. BA concentrations and pool composition were measured in portal blood, containing intestinal reabsorbed BAs and compared to systemic blood during a standardized meal test before and after RYGB. Systemic total BA concentrations increased after RYGB, due to an increase in conjugated BAs. Interestingly, the ratio of portal:systemic conjugated BAs decreased after RYGB, indicating a role for the liver in systemic BA concentrations changes. In line, hepatic expression of BA transporter genes decreased after RYGB. Our results show that the increase in systemic BAs after surgery is due to decreased selective hepatic recapture. Thus, alterations in hepatic function contribute to the increase in systemic BAs after RYGB.

Influence of Roux-en-Y gastric bypass on plasma bile acid profiles: a comparative study between rats, pigs and humans

BACKGROUND

Roux-en-Y gastric bypass (RYGBP) is the most widely used bariatric surgery procedure, which induces profound metabolic and physiological effects, such as substantial improvements in obesity, type 2 diabetes and their comorbidities. Increasing evidence identifies bile acids (BAs) as signaling molecules that contribute to the metabolic improvement after RYGBP. However, how and to what extent BAs mediate the metabolic effects of RYGBP still remains unclear and requires mechanism of action studies using preclinical models. In this study, we compared plasma BA profiles before and after RYGBP in two animal models, rats and pigs, with humans to evaluate their translational potential.

METHODS

Plasma BAs were profiled in rats, pigs and humans by liquid chromatography coupled with tandem mass spectrometry before and after RYGBP.

RESULTS

RYGBP increased baseline plasma total BA concentrations in humans and in the two animal models to a similar extent (∼3-fold increase), despite differences in presurgery BA levels and profiles between the models. However, qualitatively, RYGBP differently affected individual plasma BA species, with similar increases in some free species (cholic acid (CA), chenodeoxycholic acid (CDCA) and deoxycholic acid (DCA)), different increases in glyco-conjugated species depending on the model and globally no increase in tauro-conjugated species whatever the model.

CONCLUSIONS

The tested animal models share similar quantitative RYGBP-induced increases in peripheral blood BAs as humans, which render them useful for mechanistic studies. However, they also present qualitative differences in BA profiles, which may result in different signaling responses. Such differences need to be taken into account when translating results to humans.

Endosomal proteolysis of internalised [ArgA0]-human insulin at neutral pH generates the mature insulin peptide in rat liver in vivo

AIMS/HYPOTHESIS

A proteolysis study of human monoarginyl-insulin ([Arg(A0)]-HI) and diarginyl-insulin ([Arg(B31)-Arg(B32)]-HI) within hepatic endosomes was undertaken to determine whether the endosomal compartment represents a physiological site for the removal of Arg residues and conversion of Arg-extended insulins into fully processed human insulin.

METHODS

The metabolic fate of arginyl-insulins has been studied using the in situ rat liver model system following ligand administration to rats and cell-free hepatic endosomes.

RESULTS

While the kinetics of insulin receptor endocytosis after the administration of arginyl-insulins were similar to those observed using human insulin, a more prolonged concentration of endosomal insulin receptor was observed in response to [Arg(A0)]-HI. [Arg(A0)]-HI induced a marked increase in the phosphotyrosine content of endosomal insulin receptor, coinciding with a more sustained endosomal association of growth factor receptor-bound protein 14 (GRB14), and a higher and prolonged activation of mitogen-activated protein kinase pathways. At acidic pH, the endosomal cathepsin D rapidly degraded insulin peptides with similar binding affinity, and generated comparable intermediates for both arginyl-insulins without affecting amino and carboxyl arginyl-peptide bonds. At neutral pH, hepatic endosomes fully processed [Arg(A0)]-HI into mature human insulin while no conversion was observed with [Arg(B31)-Arg(B32)]-HI. The neutral endosomal Arg-convertase was sensitive to bestatin, immunologically distinct from insulin-degrading enzyme, nardilysin or furin, and was potentially related to aminopeptidase-B-type enzyme.

CONCLUSIONS/INTERPRETATION

The data describe a unique processing pathway for the endosomal proteolysis of [Arg(A0)]-HI which involves the removal of Arg(A0) and subsequent generation of mature human insulin through an uncovered neutral Arg-aminopeptidase activity. The endosomal conversion of [Arg(A0)]-HI into human insulin might extend the insulin receptor signalling at this locus.

Identification of insulin domains important for binding to and degradation by endosomal acidic insulinase

The endosomal compartment of hepatic parenchymal cells contains an acidic endopeptidase, endosomal acidic insulinase (EAI), which hydrolyzes internalized insulin at a limited number of sites. Although the positions of these cleavages are partially known, the residues of insulin important in its binding to and proteolysis by EAI have not been defined. To this end, we have studied the degradation over time of native human insulin and three insulin-analog peptides using a soluble endosomal extract from rat liver parenchyma followed by purification of the products by HPLC and determination of their structure by mass spectrometry. We found variable rates of ligand processing, i.e. high ([Asp(B10)]- and [Glu(A13),Glu(B10)]-insulin), moderate (insulin) and low (the H2-analog). On the basis of IC(50) values, competition studies revealed that human and mutant insulins display nearly equivalent affinity for the EAI. Proteolysis of human and mutant insulins by EAI resulted in eight cleavages in the B-chain which occurred in the central region (Glu(B13)-Leu(B17)) and at the C-terminus (Arg(B22)-Thr(B27)), the latter region comprising the initial cleavages at Phe(B24)-Phe(B25) (major pathway) and Phe(B25)-Tyr(B26) (minor pathway) bonds. Except for the [Glu(A13),Glu(B10)]-insulin mutant, only one cleavage on the A-chain was observed at residues Gln(A15)-Leu(A16). Analysis of the nine cleavage sites showed a preference for hydrophobic and aromatic amino acid residues on both the carboxyl and amino sides of a cleaved peptide bond. Using the B-chain alone as a substrate resulted in a 30-fold increase in affinity for EAI and a 6-fold increase in the rate of hydrolysis compared with native insulin. A similar role for the C-terminal region of the B-chain of insulin in the high-affinity recognition of EAI was supported by the use of the corresponding B(22)-B(30) peptide, which displayed an increase in EAI affinity similar to the entire B-chain vs. wild-type insulin. Thus, we have identified a highly specific molecular interaction of insulin with EAI at the aromatic locus Phe(B24)-Phe(B25)-Tyr(B26). Analytical subfractionation of a postmitochondrial supernatant fraction showed that a pulse of internalized [(125)I]Tyr(A14)-H2-analog, a protease-resistant insulin analog, undergoes a greater lysosomal transfer and lesser degradation than [(125)I]Tyr(A14)-insulin, confirming that endosomal sorting is regulated directly or indirectly by endosomal proteolysis.

Critical role of tyrosine 277 in the ligand-binding and transactivating properties of retinoic acid receptor alpha

Retinoic acid receptors specifically bind all-trans-retinoic acid (RA) and function as RA-inducible transcriptional regulatory factors. Binding of RA to RARalpha, beta, and gamma is sensitive to nitration with tetranitromethane, a tyrosine-specific modifying reagent. To identify tyrosine residue(s) that are important for RA binding, we carried out chemical modification experiments with purified RARalpha ligand-binding domain (RARalpha-LBD) subjected to partial acid hydrolysis and selective proteolysis. The chemically modified peptides containing each of the three Tyr residues present in the RARalpha-LBD sequence were then analyzed and identified by high-performance liquid chromatography coupled to electrospray ionization mass spectrometry (HPLC/ESI-MS). We found that RA binding to RARalpha-LBD protected Tyr(277)-containing peptides from nitration. Protection of Tyr(277) could result either from direct masking by the bound ligand or from ligand-induced changes in receptor conformation and tyrosine accessibility. The role of Tyr residues was further documented by site directed mutagenesis using three site-specific RARalpha mutants: Y208A, Y277A, and Y362A. The affinity for RA of these mutant receptors was in the range of that of the wild-type protein, except for the Y277A receptor mutant, which displays a 15-20-fold reduction in affinity and transactivation activity for RA. Whereas mutation of Tyr(277) into alanine had a variable effect on different agonists and antagonists binding, it caused a dramatic decrease of retinoid-dependent transactivation activity. This later effect was also observed with mutation of Tyr(277) into phenylalanine. It is unlikely that major conformational changes are responsible for the lower affinity of RA binding and RA-dependent transactivation since these mutants displayed wild-type dimerization and DNA-binding activities. Limited proteolysis revealed that upon ligand binding, the Y277A mutant induced a conformational change slightly different from that obtained with the wild-type protein. These data could suggest that Tyr(277) play a critical role in the ligand-induced conformational changes required for the activation of RARalpha.

Characterization of specificity of subtilisin Carlsberg towards peptide T by high-performance liquid chromatography and electrospray mass spectrometry

Peptide T has a sequence (Ala-Ser-Thr-Thr-Thr-Asn-Tyr-Thr) belonging to HIV envelope that is involved in the interaction with CD(4) receptor of T lymphocytes. Research of protease activities towards this peptide is very relevant for AIDS therapy. Characterization of specificity of subtilisin Carlsberg towards this very hydrophilic peptide is proposed by using high-performance liquid chromatography and mass spectrometry. Peptide T was totally hydrolysed by the protease after 24 h. Separation of hydrophilic fragments was perfected with an hydrophilic stationary phase and a reversed acetonitrile gradient. Peptide masses were determined by ion spray mass spectrometry. Four primary and one secondary hydrolysis products were found, corresponding to cleavage at the carboxylic side of threonine. Specifities of subtilisin Carlsberg towards the Segments 19 to 26 of bovine pancreatic ribonuclease A, an homologous fragment of peptide T, and peptide T were compared.

Characterization of a multi-lipopeptides mixture used as an HIV-1 vaccine candidate

A multi-component vaccine has been defined, which contains six different synthetic 24- to 32-amino acid lipopeptides derived from the sequence of HIV-1 proteins. The physicochemical properties of the lipopeptide components were compatible with multi-dimensional analysis, using RP-HPLC, Edman sequencing, electrospray mass spectrometry, and 2D-NMR. Detailed analysis of the impurity profiles led to the detection and evaluation of the relative proportions of most by-products: several contaminants resulted from the formation of acetylated fragments, transpeptidation reactions with succinimide or piperidide formation, or methionine and/or tryptophan mono-oxidations. The first batch to be produced underwent extensive pharmacotoxicological testings to confirm its safety; this vaccine candidate has now been used in phase I clinical trials. Despite the complexity of such multi-lipopeptide vaccines, our findings suggest the possibility of preparing a clear and precise assignment of by-products to toxicologically qualified impurities in the eventuality of a future production of several successive batches.

Quantitative analysis of human serum corticosterone by high-performance liquid chromatography coupled to electrospray ionization mass spectrometry

An original method based upon high-performance liquid chromatography coupled to electrospray ionization mass spectrometry has been developed for corticosterone (B) quantification in human serum. After extraction by diethyl ether using triamcinolone (T) as an internal standard, solutes are separated on a C18 microbore column (250 X 1.0 mm, I.D.), using acetonitrile-water-formic acid (40:59.9:0.1, v/v/v) as the mobile phase (flow-rate 40 microl/min). Detection is performed on an API 1 single quadrupole mass spectrometer equipped with a ESI interface and operated in positive ionization mode. Corticosterone quantifications were realized by computing peak area ratios (B/T) of the serum extracts analyzed in SIM mode (m/z 347 and m/z 395 for B and T. respectively), and comparing them with the calibration curve (r=0.998).

Comparative oxidation of loxapine and clozapine by human neutrophils

The clozapine-induced agranulocytosis could be due to the formation of a reactive intermediate formed in polymorphonuclear neutrophils and granulocyte precursors with the myeloperoxidase-hydrogen peroxide system. On the contrary, no case of agranulocytosis has been described for loxapine, an other neuroleptic drug with a very close structural analogy. We have compared the clozapine and loxapine interaction with the oxidative burst and particularly with this enzymatic complex. On the one hand, the assay of the oxidative species demonstrated a different impact for the two neuroleptics. The 50% inhibitory concentration was 92 microM for hydrogen peroxide and 40 microM for hypochlorous acid for loxapine. The loxapine target is located before the myeloperoxidase-hydrogen peroxide system in the oxidative stream, whereas clozapine diverts the chlorination pathway of the enzyme. On the other hand, the in vitro metabolism of drugs by the myeloperoxidase-hydrogen peroxide system has been investigated by mass spectrometry. Loxapine remains inert but clozapine undergoes the oxidation. The glutathione or ascorbate addition in the medium leads to a removal of the oxidation. Glutathione is able to trap the toxic intermediate and could avoid its formation.

DNA-interacting proteins in the spermiogenesis of the mollusc Murex brandaris

Sperm chromatin of Murex brandaris (a neogastropod mollusc) undergoes a series of structural transitions during spermiogenesis. The DNA-interacting proteins responsible for these changes as well as the mature protamines present in the ripe sperm nucleus have been characterized. The results reveal that spermiogenic nuclear proteins are protamine precursors that are subjected to a substantial number of small N-terminal deletions that gradually modify their overall charge. The composition of mature protamines is remarkably simple in turn, promoting an efficient and extremely tight packaging of DNA. The pattern of spermiogenic chromatin condensation in M. brandaris clearly departs from that corresponding to vertebrate chromatin.

On-line HPLC-electrospray ionization mass spectrometry: a pharmacological tool for identifying and studying the stability of Gd3+ complexes used as magnetic resonance imaging contrast agents

The identification of MRI contrast agents (CAg) as gadolinium complexes often used at very low concentrations in Pharmacology was carried out by ESI-MS or HPLC-ESI-MS. Firstly, Omniscan, Dotarem and Magnevist were tested. In these compounds, the Gd3+ ion must be solidly chelated by linear or macrocyclic ligands because of the severe toxicity of the free Gd3+. Spectra were obtained at low voltage, preserving the non-covalent binding integrity of the complexes, and at various higher voltages showing the progressive destruction of the complexes. Secondly, a direct reaction of these drugs with the oxidative human neutrophil production, induced in vitro by Phorbol 12-myristate 13-acetate enhancing the respiratory burst, was investigated. This was done to mimic what happens in the case of inflammatory diseases, or infection, or when people are likely to develop anaphylactoid reactions, as the i.v. injection of CAg causes contact between the complexes and neutrophils in the blood. Analysis by HPLC-ESI-MS coupling did not show any direct reaction between Gd complexes and the chemical compounds in the neutrophil oxidative metabolism, even if uncertainty remains as regards meglumine salt. HPLC-ESI-MS is a good way of visualizing characteristic, Gd isotopic distribution and of following its associations in biological samples.

Characterization of metal and nucleotide liganded forms of adenylate kinase by electrospray ionization mass spectrometry

Complexes of adenylate kinase from Escherichia coli, Bacillus subtilis, and Bacillus stearothermophilus with the bisubstrate nucleotide analog P1,P5-di(adenosine 5')-pentaphosphate and with metal ions (Zn2+ and/or Mg2+) were analyzed by electrospray ionization mass spectrometry. P1,P5-di(adenosine 5')-pentaphosphate. adenylate kinase complex was detected in the positive mode at pH as low as 3.8. Binding of nucleotide to adenylate kinase stabilizes the overall structure of the protein and preserves the Zn2+ chelated form of the enzyme from the gram-positive organisms. In this way, it is possible in a single mass spectrometry experiment to screen metal-chelating adenylate kinases, without use of radioactively labeled compounds. Binding of Mg2+ to enzyme via P1,P5-di(adenosine 5')-pentaphosphate was also demonstrated by mass spectrometry. Although no amino acid side chain in adenylate kinase is supposed to interact with Mg2+, Asp93 in porcine muscle cytosolic enzyme, equivalent to Asp84 in the E. coli adenylate kinase, was proposed to stabilize the nucleotide.Mg2+ complex via water molecules.

Evidence that human milk isolated cyclophilin B corresponds to a truncated form

Cyclophilin B (CyPB) is a member of the cyclophilin family (cyclosporin A-binding proteins) with specific N- and C-terminal extensions. In contrast to cyclophilin A, CyPB owns a signal sequence leading to its translocation in the endoplasmic reticulum. CyPB was reported to be present in human blood and milk, suggesting it is secreted. For this purpose, CyPB was purified to homogeneity from human milk and compared to recombinant CyPB expressed in E. coli. Ion spray mass spectrometry revealed that CyPB secreted in human milk exhibits a lower molecular mass than the one expected. Identification of phenylalanine as the C-terminus amino-acid residue of human milk CyPB indicates that the difference in molecular mass may be explained by the absence of the five C-terminal amino-acid residues AIAKE. These results suggest that in the sequence VEKPFAIAKE known to be responsible for retention of CyPB in the endoplasmic reticulum, the sequence AIAKE is more particularly necessary. Our findings raise the possibility that the CyPB may be processed to promote its release. As recombinant CyPB was shown to bind specifically to Jurkat cells, a lymphoblastic T-cell line, we then wanted to investigate the binding of human milk CyPB to these cells. Despite lacking the five C-terminal amino-acid residues, human milk CyPB is able to inhibit the binding of recombinant CyPB to Jurkat T cells.

Squid spermiogenesis: molecular characterization of testis-specific pro-protamines

Cuttlefish spermiogenesis is characterized by a two-step nuclear protein transition: histones-->spermatid-specific protein (protein T)-->sperm protamine (protein Sp). A similar situation can be observed in another Cephalopod species, the squid Loligo pealeii. The protein T from Loligo consists of two structural variants, T1 and T2 (molecular masses: 10788 and 10791 Da respectively), phosphorylated to different degrees (2-6 phosphate groups). The primary structures of these two variants and of the protamine variant Sp2 were established from sequence analysis and mass spectrometric data of the proteins and their fragments. T1 and T2 are closely related proteins of 79 residues. The complete structural identity of the C-terminal domain (residues 22-79) of protein T2 with the sperm protamine Sp2 (molecular mass 8562 Da, 58 residues) strongly suggests that the testis-specific protein T2 is indeed the precursor of the protamine. The transition between the precursor protein T and protein Sp results from a hydrolytic cleavage similar to that found in many proteins that are synthesized as precursors. The processing mechanism involves the specific cleavage of a Gly-Arg bond in the sequence Met/Leu18-Lys-Gly-Gly-Arg-Arg23. Furthermore, the study provides molecular evidence on the taxonomic relationship between Loligo and Sepia.

Sequence analysis and structural features of the largest known protamine isolated from the sperm of the archaeogastropod Monodonta turbinata

Protamine of the archaeogastropod mollusc Monodonta turbinata has been isolated and characterized. With a mass of 13,476 Da, it is the largest known protamine. Amino acid sequence of this protamine (106 residues) was established from data provided by automated sequence analysis and mass spectrometry of the protein and of its fragments. The primary structure of the NH2-terminal region exhibits repetitive sequence motifs "Basic-Ser" (mainly R-S) and both central and COOH-terminal regions are composed by arginine clusters. The amino acid sequence of Monodonta turbinata protamine shows structural similarities with other protamines from invertebrates and from birds and mammals.

Application of electrospray and fast atom bombardment mass spectrometry to the identification of post-translational and other chemical modifications of proteins and peptides

Mass spectrometry is a very powerful tool in the identification of chemical modifications of proteins and peptides. Often these modifications cannot be determined by conventional techniques. This report describes the combined use of electrospray ionization mass spectrometry and fast atom bombardment mass spectrometry to complete the primary structure of proteins and peptides. Examples illustrate how mass spectrometry is used to locate sites of phosphorylation, methylation and acetylation, and identify blocking groups and unexpected side reactions such as deamidation or alkylation.

Chondroitin sulphate covalently cross-links the three polypeptide chains of inter-alpha-trypsin inhibitor

Inter-alpha-trypsin inhibitor (ITI) is a tight complex of three different proteins: bikunin and two heavy chains H1 and H2. In order to demonstrate that the three chains are covalently linked by a chondroitin sulphate chain as previously proposed [Enghild, J. J., Salvesen, G., Hefta, S. A., Thogersen, I. B., Rutherford, S. and Pizzo, S. V. (1991) J. Biol. Chem. 266, 747-751], ITI was extensively digested with thermolysin and the glycosaminoglycan-containing fragment was isolated from the digest by ion-exchange chromatography. Its peptide structural determination and mass spectrometry analysis both provide evidence that the different peptide chains constituting ITI are associated by the new cross-link described as the protein-glycosaminoglycan-protein cross-link.

A corrected primary structure for dog-fish Scylliorhinus caniculus protamine Z3

We have redetermined the primary structure for dog-fish protamine using automated amino-acid sequencing associated to mass spectrometry techniques and report, on the basis of these findings, that the previously published amino-acid sequence is incorrect. The correct protamine sequence is 37 amino acids long and differs from the original published sequence by the C-terminal hexapeptide Arg32-Gly-Arg-Arg-Ser-Arg37.

Nuclear transition protein 1 from ram elongating spermatids. Mass spectrometric characterization, primary structure and phosphorylation sites of two variants

The ram transition protein 1 (TP1) is present in spermatid cell nuclei in the nonphosphorylated, monophosphorylated and diphosphorylated forms. Its primary structure was determined by automated Edman degradation of S-carboxamidomethylated protein and of peptides generated by cleavage with thermolysin and endoproteinase Lys-C. The ram TP1 is a small basic protein of 54 residues and structurally very close to other mammalian TP1. The mass spectrometric data obtained from the protein and its fragments reveal that ram TP1 is indeed a mixture (approximately 5:1) of two structural variants (Mr 6346 and 6300). These variants differ only by the nature of the residue at position 27 (Cys in the major variant and Gly in the minor variant). The study of phosphorylation sites has shown that four different serine residues could be phosphorylated in the monophosphorylated TP1, at positions 8, 35, 36 or 39. From previous physical studies, it has been postulated that the Tyr32 surrounded by two highly conserved basic clusters was responsible for the destabilization of chromatin by intercalation of its phenol ring between the bases of double-stranded DNA. The presence of three phosphorylatable serine residues in the very conserved sequence 29-42 is another argument for the involvement of this region in the interaction with DNA.