Crystal structure of maltose phosphorylase from Lactobacillus brevis: unexpected evolutionary relationship with glucoamylases

BACKGROUND

Maltose phosphorylase (MP) is a dimeric enzyme that catalyzes the conversion of maltose and inorganic phosphate into beta-D-glucose-1-phosphate and glucose without requiring any cofactors, such as pyridoxal phosphate. The enzyme is part of operons that are involved in maltose/malto-oligosaccharide metabolism. Maltose phosphorylases have been classified in family 65 of the glycoside hydrolases. No structure is available for any member of this family.

RESULTS

We report here the 2.15 A resolution crystal structure of the MP from Lactobacillus brevis in complex with the cosubstrate phosphate. This represents the first structure of a disaccharide phosphorylase. The structure consists of an N-terminal complex beta sandwich domain, a helical linker, an (alpha/alpha)6 barrel catalytic domain, and a C-terminal beta sheet domain. The (alpha/alpha)6 barrel has an unexpected strong structural and functional analogy with the catalytic domain of glucoamylase from Aspergillus awamori. The only conserved glutamate of MP (Glu487) superposes onto the catalytic residue Glu179 of glucoamylase and likely represents the general acid catalyst. The phosphate ion is bound in a pocket facing the carboxylate of Glu487 and is ideally positioned for nucleophilic attack of the anomeric carbon atom. This site is occupied by the catalytic base carboxylate in glucoamylase.

CONCLUSIONS

These observations strongly suggest that maltose phosphorylase has evolved from glucoamylase. MP has probably conserved one carboxylate group for acid catalysis and has exchanged the catalytic base for a phosphate binding pocket. The relative positions of the acid catalytic group and the bound phosphate are compatible with a direct-attack mechanism of a glycosidic bond by phosphate, in accordance with inversion of configuration at the anomeric carbon as observed for this enzyme.

Use of microbial transglutaminase for the enzymatic biotinylation of antibodies

Nowadays many reagents are available for the biotinylation of proteins. As most of them bind to amino groups of the protein the degree of labelling differs from batch to batch and the possibility exists that the biological activity of the target protein may be affected by the labelling procedure. In the present study we have investigated an enzymatic approach to biotinylation using microbial transglutaminase (MTGase) from Streptoverticillium mobaraense. The proposed method is particularly suitable when only a few biotin molecules need to be attached to the target proteins. The enzyme catalyses the acyl transfer reaction between gamma-carboxyamide groups and various primary amines. This was exploited for biotinylation using two amino-modified biotin derivatives, biotinamido-5-pentylamin (BIAPA) and biotinoyl-1,8-diamino-3, 6-dioxaoctane (BIDADOO) as acyl acceptors and a monoclonal IgG against the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) as the acyl donor. Kinetic studies revealed that the MTGase-mediated reaction proceeds with low velocity and is almost complete after 34 h. Conjugation ratios ranging from 1.1 to 1.9 biotins per IgG were found by mass spectrometry. To investigate the influence of antibody conjugation on antigen binding a competitive ELISA for the determination of 2,4-D employing MTGase-biotinoylated IgGs was developed. In this assay lower limits of detection of 0.3 and 1.0 microg/l of 2,4-D were achieved with BIDADOO- and BIAPA-modified antibodies, respectively.

Molecular dynamics of microbial lipases as determined from their intrinsic tryptophan fluorescence

We have studied the intrinsic tryptophan fluorescence of the lipases from Chromobacterium viscosum (CVL), Pseudomonas species (PSL), and Rhizopus oryzae (ROL) in aqueous buffer, zwitterionic detergent micelles, and isopropanol-water mixtures. It was the purpose of this study to obtain information about biophysical properties of the respective enzymes under conditions that modulate enzyme activities and stereoselectivities to a significant extent. According to their decay-associated emission spectra, CVL tryptophans are located in the hydrophobic interior of the protein. In contrast, the PSL and ROL tryptophans are probably confined to the core and the surface of the lipase. From the tryptophan lifetime distributions it can be concluded that the conformation of CVL is not much affected by detergent or organic solvent (isopropanol). Accordingly, CVL is enzymatically active in these systems and most active in the presence of isopropanol. In contrast, ROL and PSL show high conformational mobility, depending on the solvent, because their lifetime distributions are very different in the presence and absence of detergent or isopropanol. Time-resolved anisotropy studies provided evidence that the lipases exhibit very high internal molecular flexibility. This peculiar feature of lipases is perhaps the key to the great differences in activity and stereoselectivity observed in different reaction media. Furthermore, information about self-association of the lipases in different solvents could be obtained. PSL, but not CVL and ROL, forms aggregates in water. Lipase aggregation can be reversed by the addition of detergent or isopropanol, which competes for the hydrophobic surface domains of this protein. This dissociation could efficiently contribute to the increase in lipase activity in the presence of a detergent or isopropanol.

Interactions of fluorescent triacylglycerol analogs covalently bound to the active site of a lipase from Rhizopus oryzae

Fluorescent triacylglycerol analogs were synthesized as covalent inhibitors of lipase activity. The respective 1(3), 2-O-dialkylglycero-3(1)-alkyl-phosphonic acid p-nitrophenyl esters contain a fluorescent pyrenealkyl chain and a long-chain alkyl residue bound to the sn-2 and sn-1(3) positions of glycerol, respectively. The phosphonic acid p-nitrophenyl ester bond is susceptible to nucleophilic substitution by the active serine residue in the catalytic triad of a lipase, leading to inactivation of the enzyme. The fluorescent dialkylglycerophosphonates contain two chiral centers, the sn-2 carbon of glycerol and the phosphorus atom. The (1-O-hexadecyl-2-O-pyrenedecyl-sn-glycero)-O-(p-nitrophenyl)-n-hex yl- phosphonate, first peak during HPLC separation and the (3-O-hexadecyl-2-O-pyrenedecyl-sn-glycero)-O-(p-nitrophenyl)-n-hex yl- phosphonate, second peak during HPLC separation were found to be potent lipase inhibitors. After incubation of an equimolar amount of these isomers with lipase from Rhizopus oryzae complete inactivation was observed. Stable conjugates containing a 1 : 1 molar ratio of lipid to protein were formed. The spatial proximity of the fluorescently labeled sn-2 alkyl chain of the inhibitor and tryptophan residues of the lipase was assessed by fluorescence resonance energy transfer. The extent of tryptophan fluorescence quenching and the concomitant increase in pyrene fluorescence upon excitation of lipase tryptophans was found to be similar for the above-mentioned isomers. Thus, the (labeled) sn-2 alkyl chains of a triacylglycerol analog are likely to interact with the same binding site of the R. oryzae lipase, irrespective of their steric configuration. However, it was shown that the extent of resonance energy transfer is strongly influenced by the reaction medium, indicating conformational changes of the lipase in different environments.

In situ antigen immobilization for stable organic-phase immunoelectrodes

A new method based on enzymatic single-step in situ synthesis of hapten-carrier conjugates on electrodes is described yielding stable, reproducible, and reusable organic-phase immunoelectrodes (OPIEs). The electrodes developed were tailored for analyte detection in organic solvents and allow for the analysis of soil extracts without further sample processing and cleanup. Catalyzed by transglutaminase from a variant of Streptoverticillium mobaraense, the reaction proceeds in aqueous solution with and without addition of organic media in only 1.5 hours. In this study, the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) was chosen as model compound and chemically amino-functionalized prior to its enzymatic immobilization. The high reproducibility of the immobilization procedure allowed for batch calibration of the immunoelectrodes. Moreover, pure methanol or treatment with diluted sulfuric acid used for regeneration studies did not disturb the hapten layer. The OPIE consists of screen-printed carbon electrodes, monoclonal anti-2,4-D antibodies, and the immunochemical recognition reaction and was optimized with regard to a high stability in organic media. For electrochemical detection, horseradish peroxidase was used as enzyme label together with H2O2 as substrate and hexacyanoferrate (II)/(III) as mediator. The OPIE showed high stability upon storage over 93 days. Response times of 17 s (t95) were found to be advantageous compared to those of other biosensors. Including the immunochemical reactions, the complete assay takes 30 min. A calibration curve for 2,4-D in 30% methanol/buffer obtained with 70 electrodes within 4 weeks revealed a detection limit of 9 mg/L, a sensitivity of 1.3 nA L mg-1 cm-2, and a repeatability of 6.8%. Although we calculated a lowered repeatability for reused electrodes of 13.4% and a slightly decreased sensitivity of 0.9 nA L mg-1 cm-2, multiple-used OPIEs could also be applied for calibration.

Rational design of Rhizopus oryzae lipase with modified stereoselectivity toward triradylglycerols

The binding site of sn-1(3)-regioselective Rhizopus oryzae lipase (ROL) has been engineered to change the stereoselectivity of hydrolysis of triacylglycerol substrates and analogs. Two types of prochiral triradylglycerols were considered: 'flexible' substrates with ether, benzylether or ester groups, and 'rigid' substrates with amide or phenyl groups, respectively, in the sn-2 position. The molecular basis of sn-1(3) stereoselectivity of ROL was investigated by modeling the interactions between substrates and ROL, and the model was confirmed by experimental determination of the stereoselectivity of wild-type and mutated ROL. For the substrates, the following rules were derived: (i) stereopreference of ROL toward triradylglycerols depends on the substrate structure. Substrates with 'flexible' sn-2 substituents are preferably hydrolyzed at sn-1, 'rigid' substrates at sn-3. (ii) Stereopreference of ROL toward triradylglycerols can be predicted by analyzing the geometry of the substrate docked to ROL: if the torsion angle phiO3-C3 of glycerol is more than 150 degrees, the substrate will preferably be hydrolyzed in sn-1, otherwise in sn-3. For ROL, the following rules were derived: (i) residue 258 affects stereoselectivity by steric interactions with the sn-2 substituent rather than polar interactions. To a lower extent, stereoselectivity is influenced by mutations further apart (L254) from residue 258. (ii) With 'rigid' substrates, increasing the size of the binding site (mutations L258A and L258S) shifts stereoselectivity of hydrolysis toward sn-1, decreasing its size (L258F and L258F/L254F) toward sn-3.

Inhibition of lipases from Chromobacterium viscosum and Rhizopus oryzae by tetrahydrolipstatin

Tetrahydrolipstatin is known as an inhibitor for pancreatic lipase but not for microbial lipases. In this paper we demonstrate that in the presence of water-insoluble substrates like tributyrin or olive oil, tetrahydrolipstatin inhibits the lipases of Chromobacterium viscosum and Rhizopus oryzae, although with different potency. In contrast to porcine pancreatic lipase, which forms an irreversible and covalent enzyme-inhibitor complex with tetrahydrolipstatin, the inhibition of the microbial lipases is reversible as the inhibitor can be removed from the enzyme-inhibitor complex by solvent extraction. Moreover, after inhibition of Chromobacterium viscosum lipase tetrahydrolipstatin remains chemically unchanged.

Maltose phosphorylase from Lactobacillus brevis: purification, characterization, and application in a biosensor for ortho-phosphate

With the goal to obtain maltose phosphorylase as a tool to determine ortho-phosphate, the enzyme from Lactobacillus brevis was purified to 98% by an expeditious FPLC-aided procedure which included anion exchange chromatography, gel filtration, and hydroxyapatite chromatography. The native maltose phosphorylase had a molecular mass of 196 kDa and consisted of two 88 kDa subunits. In isoelectric focusing two isoforms with pI values of 4.2 and 4.6 were observed. Maximum enzyme activity was obtained at 36 degrees C and pH 6.5 and was independent of pyridoxal 5'-phosphate. The apparent K(m) values with maltose and phosphate as substrates were 0.9 mmol l-1 and 1.8 mmol l-1, respectively. Maltose phosphorylase could be stored in 10 mM phosphate buffer pH 6.5 at 4 degrees C with a loss of activity of only 7% up to 6 months. The stability of the enzyme at high temperatures was enhanced significantly using additives like phosphate, citrate, and imidazole. The purified maltose phosphorylase was used as key enzyme in a phosphate sensor consisting of maltose phosphorylase and glucose oxidase. A detection limit of 0.1 microM phosphate was observed and the sensor response was linear in the range between 0.5 and 10 microM.

Inhibition of microbial lipases with stereoisomeric triradylglycerol analog phosphonates

1,2(2,3)-Diradylglycero O-(p-nitrophenyl) n-hexylphosphonates were synthesized, with the diradylglycerol moiety being di-O-octylglycerol, 1-O-hexadecyl-2-O-pyrenedecanylglycerol, or 1-O-octyl-2-oleoyl-glycerol, and tested for their ability to inactivate lipases from Chromobacterium viscosum (CVL) and Rhizopus oryzae (ROL). The experimental data indicate the formation of stable, covalent 1:1 enzyme-inhibitor adducts with the di-O-alkylglycero phosphonates. The differences in reactivity of diastereomeric phosphonates with opposite configuration at the glycerol backbone was less expressed with both enzymes tested as compared to the influence of the stereochemistry at the phosphorus. Both lipases exhibited the same preference for the chirality at the phosphorus that was independent from the absolute configuration at the glycerol backbone. However, with CVL and ROL the inhibitors with the active site serine-directed phosphonate linked at position sn-1 of the glycerol moiety reacted significantly faster than the corresponding sn-3 analogs, reflecting the sn-1 stereopreference of the enzymes towards triacylglycerol analogs with a sn-2 O-alkyl substituent. In contrast, the phosphonates based on the 1-O-octyl-2-oleoylglycerol did not significantly inactivate CVL. Unexpectedly, these substances were hydrolyzed in the presence of lipase.

Crystal structure of a bacterial lipase from Chromobacterium viscosum ATCC 6918 refined at 1.6 angstroms resolution

The crystal structure of a lipase from the bacterium Chromobacterium viscosum ATCC 6918 (CVL) has been determined by isomorphous replacement and refined at 1.6 angstroms resolution to an R-factor of 17.8%. The lipase has the overall topology of an alpha/beta type protein, which was also found for previously determined lipase structures. The catalytic triad of the active center consists of the residues Ser87, Asp263 and His285. These residues are not exposed to the solvent, but a narrow channel connects them with the molecular surface. This conformation is very similar to the previously reported closed conformation of Pseudomonas glumae lipase (PGL), but superposition of the two lipase structures reveals several conformational differences. r.m.s. deviations greater than 2 angstroms are found for the C alpha-atoms of the polypeptide chains from His15 to Asp28, from Leu49 to Ser54 and from Lys128 to Gln158. Compared to the PGL structure in the CVL structure, three alpha-helical fragments are shorter, one beta-strand is longer and an additional antiparallel beta-sheet is found. In contrast to PGL, CVL displays an oxyanion hole, which is stabilized by the amide nitrogen atoms of Leu17 and Gln88, and a cis-peptide bond between Gln291 and Leu292. CVL contains a Ca2+, like the PGL, which is coordinated by four oxygen atoms from the protein and two water molecules.

Hydrolysis and esterification of acylglycerols and analogs in aqueous medium catalyzed by microbial lipases

The stereoselectivity of microbial lipases from Chromobacterium viscosum (CVL) and Rhizopus arrhizus (RAL) towards monoacylglycerols (rac-1(3)-oleoylglycerol and 2-oleoylglycerol), diacylglycerols (1,3-dioleoylglycerol and rac-1,2(2,3)-dioleoylglycerol) and 2-O-ether analogs (rac-1(3)-oleoyl-2-O-hexadecylglycerol and rac-1(3)-octanoyl-2-O-hexadecylglycerol) was determined. The results of the hydrolysis of 2-O-ether analogs confirmed the importance of the substituent at C-2 of acylglycerols in the stereoselective recognition by microbial lipases and also showed that acylation of mono- and diradylglycerols with oleic acid overlaps the hydrolysis reaction in aqueous medium. With the short-chain, water-soluble octanoic acid no significant esterification occurred. Using rac-1,2(2,3)-dioleoylglycerol as a substrate for the hydrolysis with RAL and CVL, the appearance of 1,3-dioleoylglycerol and of 1(3)-monooleoylglycerol was demonstrated. The possibility of chemical vs. enzyme-catalyzed isomerization of 1,2-dioleoylglycerol and of 2-oleoylglycerol is discussed.

Enantiomeric perylene-glycerolipids as fluorogenic substrates for a dual wavelength assay of lipase activity and stereoselectivity

A new type of fluorogenic alkyldiacyl glycerols was synthesized and used as fluorogenic substrates for the analysis of lipase activities and stereoselectivities. These compounds contain perylene as a fluorophore and the trinitrophenylamino (TNP) residue as a quencher. Both substituents are covalently bound to the omega-ends of the sn-2 and sn-1 (3) acyl chains, respectively. Upon glycerolipid hydrolysis, the residues are separated from each other thus allowing determination of lipase activity by the continuous increase in fluorescence intensity which is caused by dequenching. Using enantiomeric pairs of these compounds, we were able to analyze lipase stereoselectivity depending on the reaction medium. Mixtures of enantiomeric fluorogenic alkyldiacyl glycerols, selectively labelled with pyrene or perylene as fluorophores, can be used for a dual-wavelength "stereoassay" of lipases. Since absorption and emission maxima of both labels are clearly separated, hydrolysis of the respective enantiomeric substrates can be determined simultaneously, and the difference in the rates of hydrolysis can be taken as a parameter for the stereopreference of a lipase. Hydrolysis rates measured with perylene-substituted lipids are generally lower than those obtained with the pyrene analogs. Thus, with a mixture of perylene and pyrene-substituted lipids, we observe a higher apparent stereoselectivity of lipases since we measure a combination of stereo- and substrate selectivity. In the presence of albumin, all microbial lipases tested so far exhibit stereopreference for the sn-1 glycerol position. In our assay, the apparent stereoselectivities are highest if in the presence of albumin, the sn-1 position carries pyrene and the sn-3 position is substituted with perylene. The lipase stereoselectivity assay described here requires the simultaneous measurement of the fluorescence intensities at two different wavelengths in a single cuvette and can thus be carried out using existing and cheap instrumentation that was developed for the fluorimetric analysis of Ca+2 concentrations.

Inversion of lipase stereospecificity for fluorogenic alkyldiacyl glycerols. Effect of substrate solubilization

We synthesized enantiomeric 1-O-alkyl-2,3-diacyl-sn-glycerol and 3-O-alkyl-1,2-diacyl-sn-glycerol containing pyrene as a fluorescent reporter and the trinitrophenylamino residue as a fluorescence quencher; both reporter groups were covalently bound to the omega end of the acyl chains at positions sn-2 and sn-3(1), respectively. The fluorescence of the intact substrate molecules was very low. Chemical or enzymic release of the fatty acyl chains lead to fluorescence dequenching. The rate of lipolysis could be measured from the time-dependent increase in fluorescence intensity. We used the respective substrates for the continuous determination of activity and stereopreference of four different microbial lipases from Chromobacterium viscosum, Candida rugosa, Pseudomonas sp., Rhizopus arrhizus, as well as cutinase from Fusarium solani and lipoprotein lipase from bovine milk. The stereopreference of the lipases depended, in general, on how the substrate was solubilized in the reaction medium. All lipases under investigation preferentially hydrolysed the sn-1 acyl ester bond, if the lipid analog was dispersed in albumin-containing Tris/HCl buffer in the absence of detergent or organic solvent. In mixtures of 1:1 (by vol.) water/ethanol, the enzymes showed higher activity toward the sn-3 acyl ester bond, except for lipoprotein lipase which preferred the sn-1 acyl isomer under all conditions tested. Different stereopreferences were observed with the different lipases if the substrate was solubilized by amphiphiles (micelles of N-dodecyl-N,N-dimethyl-3-ammonio-1-propanesulfonate). C. rugosa lipase and F. solani cutinase showed high stereopreference for the sn-3 acyl ester, whereas Pseudomonas sp. lipase and C. viscosum lipase hydrolysed both enantiomers at similar rates. From spectroscopic studies, it can be inferred that the conformation of the fluorescent lipids is probably similar in water, mixtures of water and organic solvents, and in micelles. The possible effects of reaction conditions on substrate accessibility and enzyme conformation on stereoselectivity of the respective lipases are discussed.

Stereoselectivity of microbial lipases. The substitution at position sn-2 of triacylglycerol analogs influences the stereoselectivity of different microbial lipases

In the present study, the stereoselectivity of purified lipases from Candida rugosa, Chromobacterium viscosum, Pseudomonas species and Rhizopus arrhizus towards triacylglycerols in comparison to various structural analogs were investigated. Different triacylglycerol analogs with distinct polarities at position sn-2 of the glycerol backbone (1,3-diacyl-2-X-glycerol, where 2-X = 2-acyloxy, 2-alkyloxy, 2-deoxy-2-alkyl, or 2-deoxy-2-phenyl) were synthesized. Substrate hydrophobicity and steric requirement was modified by variation of the alkyl and acyl chain length. Hydrolysis of these substrates demonstrated that minor structural variations at C2 of triacylglycerol strongly affect the stereoselectivity of the lipases tested. It was noteworthy that the variation of substrate structure did not only affect the quantity of stereoselectivity expressed as percentage enantiomeric excess, but also resulted in a reversal of stereopreference in some cases. Replacement of the acylester in position 2 of glycerol by a non-ester-linked aliphatic moiety shifted the preference of Chromobacterium viscosum lipase from sn-3 to sn-1. Lipases from Chromobacterium viscosum. Pseudomonas species and Rhizopus arrhizus exhibited sn-3 preference with 2-deoxy-2-phenyl analogs, while towards substrates with a 2-deoxy-2-alkyl moiety sn-1 stereobias was recorded. Candida rugosa lipase was rather insensitive to substrate variations concerning the polarity at position 2 of the glycerol backbone. However, variation of the acyl chain length significantly influenced stereoselectivity of this lipase.

Crystallization and preliminary X-ray analysis of a lipase from Chromobacterium viscosum

Lipase from Chromobacterium viscosum has been purified to homogeneity and crystallized in a form suitable for X-ray diffraction analysis from 10-14% polyethylene glycol 4000 and 10-14% 2-methyl-2,4-pentane diol at pH 6.4 in the presence of 0.25%(w/v) n-octyl-beta-D-glucopyranoside. These crystals belong to space group P2(1)2(1)2 with refined lattice constants a = 41.1 A, b= 156.8, c = 43.6 A, indicating a cell content of one monomer per asymmetric unit of the crystal. The crystals diffract to a resolution of 2.2 A.