PEDS Advance Access originally published online on April 19, 2008
Protein Engineering Design and Selection 2008 21(6):387-393; doi:10.1093/protein/gzn014
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Improving activity and stability of cutinase towards the anionic detergent AOT by complete saturation mutagenesis
1 Institut für Molekulare Enzymtechnologie, Heinrich-Heine-Universität Düsseldorf, Forschungszentrum Jülich, D-52426 Jülich, Germany 2IBB—Institute for Biotechnology and Bioengineering, Centre for Biological and Chemical Engineering, Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal 3evocatal GmbH, Merowingerplatz la, D-40225 Düsseldorf, Germany
4 To whom correspondence should be addressed. joaquim.cabral{at}ist.utl.pt (J.M.S.C.); vbrissos{at}itqb.unl.pt (V.B.)
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Cutinase is an enzyme suitable for detergent applications as well as for organic synthesis in non-aqueous solvents. However, its inactivation in the presence of anionic surfactants is a problem which we have addressed by creating a complete saturation library. For this, the cutinase gene from Fusarium solani pisi was mutated to incorporate all 19 possible amino acid exchanges at each of the 214 amino acid positions. The resulting library was screened for active variants with improved stability in the presence of the anionic surfactant dioctyl sulfosuccinate sodium salt (AOT). Twenty-four sites in cutinase were discovered where amino acid replacements resulted in a 2–11-fold stability increase as compared to the wild-type enzyme.
Keywords: complete saturation mutagenesis/cutinase/stability/surfactant
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Cutinases are named for their ability to degrade cutin polymers, but additionally, they display a rather broad substrate selectivity, ranging from short-chain p-nitrophenylesters to long-chain triglycerides (Kolattukudy, 1984
; Lauwereys, 1991). Cutinase from the phytopathogenic fungus Fusarium solani pisi is an example of a small carboxylic ester hydrolase that shares functional properties of lipases and esterases.
The stability of enzymes is affected by many factors, such as temperature, pH, oxidative stress, solvents, binding of metal ions or co-factors, and the presence of surfactants. Surfactants are major constituents of household detergents where they are mixed with enzymes and therefore surfactant stability is extremely important from an industrial point of view. As cutinase is an efficient catalyst both in solution and at water–lipid interfaces, it is potentially suitable for lipid stain removal applications in the detergent industry (Flipsen et al., 1998). However, its implementation has been hampered by its inactivation in the presence of anionic surfactants.
The problem of detergent inactivation is also relevant for other applications of enzymes in surfactant-containing systems, e.g. in non-aqueous solvents where cutinase, similar to lipases, has proved to be an efficient biocatalyst for organic synthesis. Among the various systems developed, reversed micellar technology has revealed to be a successful methodology for different reactions ranging from hydrolysis of triglycerides (Melo et al., 1995) to esterification (Sebastiao et al., 1993; Cunnah et al., 1996; PintoSousa et al., 1996) and transesterification (Carvalho et al., 1997, 1999). Water-in-oil microemulsions are sustainable due to the amphipathic nature of the surfactant, an indispensable component of these systems. Selection of the proper surfactant is a compromise between the ability of formation of the microemulsion and toxic or inhibitory effects on the enzyme. The anionic surfactant dioctyl sulfosuccinate sodium salt (AOT) is the most commonly used due to its ability to solubilise in water (Maitra, 1984). However, it exhibits inhibitory effects on cutinase resulting in a relative lack of stability of these systems and low enzyme half-lives (Sebastiao et al., 1993), where enzyme inactivation has been recognised as a key problem (Melo et al., 1996).
Anionic detergents presumably bind to an enzyme because the negative head group of the surfactant recognises positive charges on the enzymes’ surface through electrostatic interactions and, cooperatively, the apolar tail recognises a bordering hydrophobic crevice (Juffer et al., 1996; Flipsen et al., 1999). A strategy for improving the resistance to anionic surfactants therefore involves disrupting the binding sites for anionic surfactants to the enzyme. In principle, this can be achieved in two ways: (i) reduction of the electrostatic interaction between the anionic surfactant and the enzyme by replacing one or more positively charged residues located close to a hydrophobic patch or by masking the positive charge by introducing a negative charge nearby (Egmond, 1996); (ii) replacing amino acid residues located in a hydrophobic patch by less hydrophobic ones or by larger ones filling up the crevice.
However, mutagenesis in the region around the active site of cutinase for the purpose of improving stability could affect its activity, which can be enhanced by increasing the hydrophobicity at the surface adjacent to the lipid contact zone (Egmond, 1994). The wild-type enzyme contains a ring of positively charged arginine residues, namely R17, R78, R88, R96, R156, R158 and R196, around this hydrophobic surface in order to prevent aggregation of cutinase. This positive charge is also of importance for the proper orientation of the enzyme to a lipid interface (Juffer et al., 1996; Flipsen et al., 1999). As a consequence, changing these surface residues of cutinase for the purpose of improving compatibility with anionic surfactants would probably result in enzyme inactivation.
The molecular structure and dynamics of cutinase are known in detail after extensive X-ray crystallography and NMR studies (Martinez et al., 1992, 1993; Prompers et al., 1997). Taken together with interest from industry for cutinase applications in detergent systems, this enzyme represents an ideal model system to study the molecular basis of enzyme detergent stability.
In the present study, we have created a complete saturation library of cutinase and developed a high throughput screening method to identify cutinase variants which show increased stability in the presence of AOT while retaining enzymatic activity. Complete saturation mutagenesis theoretically allows all single site mutants to be sampled and it was successfully applied to create enantioselective nitrilases (Gray et al., 2001; DeSantis et al., 2003). More recently, complete saturation mutagenesis proved to be faster and less resource intensive than DNA shuffling in the quest to create β-fucosidase from β-galactosidase (Parikh and Matsumura, 2005).
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Bacterial strains and growth conditions
Plasmids were constructed and transformed into E. coli strain XL1-blue (endA1 gyrA96 (nalR) thi-1 recA1 relA1 lac glnV44 F’[::Tn10 proAB+ lacIq 
(lacZ)M15] hsdR17(rK– mK+)). Escherichia coli cells were grown overnight in Luria–Bertani (LB) medium in glass tubes at 37°C and in the presence of appropriate amounts of ampicillin (100 µg/ml). Overexpression of recombinant cutinase (cut) (a gift from Corvas Intl.) and its variants was performed with strain E. coli WK-6 ((lac-pro) galE strA F’[lacIq ZM15 proAB+]). Protein purification was carried out as described in Lauwereys (1991).
The buffer salt tris-(hydroxymethyl)-aminomethane and the solvent acetonitrile were purchased from Merck in p.a. quality. The surfactant dioctyl sulfosuccinate sodium salt (AOT, 99% purity) and the substrate p-nitrophenyl butyrate (p-NPB, 98% purity) were obtained from Sigma.
Enzyme activity was determined using the substrate p-nitropheylbutyrate (70 mM stock solution in acetonitrile). The final concentration of acetonitrile in each reaction was 
1.0% (v:v). Activity assays were performed in 20 mM Tris–HCl, pH 8.0 (assay buffer) at 30°C, and were initiated by addition of 3 nM of pure enzyme. Reaction rates were determined by monitoring spectrophotometrically the release of p-nitrophenol at 400 nm (
= 15 400 M–1 cm–1) using a Hitachi U-2000 spectrophotometer. Enzyme activity linearly increased with the amount of enzyme within the range used in these assays. The average standard deviation for triplicate determinations was 5%.
Cutinase inactivation by AOT was determined by following the inactivation kinetics of 300 nM of pure cutinase in assay buffer at 40°C containing 0.5 mM of surfactant. At various times, 15 µl aliquots were removed and diluted 100-fold in assay buffer containing 0.5 mM of surfactant. The residual activity of the samples was determined by adding 0.7 mM p-NPB and measuring the hydrolysis at 400 nm over 1 min period at 30°C.
The stability plot was fitted with the Henley and Sadana model (Henley and Sadana, 1985; Sadana, 1991).
General DNA techniques and plasmid constructions
Plasmid DNA was isolated using the plasmid purification midi-kit from Qiagen. The purification of PCR products or DNA fragments from agarose gels was performed using the NucleoSpin Extract kit from Macherey-Nagel. Recombinant DNA techniques were performed essentially as described by Sambrook et al. (1989). Transformation of E. coli was done by high-voltage electroporation (Dower et al., 1988) using an E. coli Transformation Apparatus (BioRad). Restriction digestion reactions and ligations were performed with enzymes from Fermentas (St. Leon-Rot, Germany) under conditions recommended by the manufacturer.
The plasmids pMac5–8-cutinase (pMac-cut), pUC-cut and pUC-cut/2 were used as templates in PCR reaction. The construction of the plasmid pMac-cut has been described elsewhere (Lauwereys, 1991).
The cutinase gene was cloned into pUC19 vector (Stratagene) using EcoRI and XbaI to obtain plasmid pUC-cut. This plasmid was transformed into E. coli strain XL1-blue, isolated and characterised by restriction analysis and DNA sequencing.
pUC-cut/2 was constructed from pUC19-cut. Initially, the DNA fragment coding for amino acids residues 1–88 of cutinase (cut) was removed by restriction digestion with EcoRI and Acc65I. The overhangs were filled in by Klenow DNA polymerase treatment and the plasmid was cyclised. This plasmid was transformed into E. coli strain XL1-blue, isolated and characterised by restriction analysis and DNA sequencing.
pMac-
Sm/Sp was constructed by cloning the Sm/Sp cassette from plasmid pHRP316 into the EcoRI and XbaI sites of pMac5–8. This plasmid was transformed into E. coli strain XL1-blue and transformants selected on plates containing streptomycin–spectinomycin (50 µg/ml).
pMac-cut-Sm/Sp was constructed from pMac-cut. The DNA coding for amino acids residues 126–214 of cutinase (cut) was removed by restriction digestion with NotI and SalI and subsequent insertion of a Sm/Sp cassette from plasmid pHRP316 into the same sites. This plasmid was transformed into E. coli strain XL1-blue and transformants selected on plates containing streptomycin–spectinomycin (50 µg/ml).
Complete saturation mutagenesis
Saturation mutagenesis was performed using a megaprimer PCR mutagenesis method (Barettino et al., 1994). Thirty-two-fold degenerate oligonucleotides (N, N, S, where N is an equal mixture of all four nucleotides and S is an equal mixture of G or C) were used for each codon in the gene so that all possible amino acids would be encoded by the resulting codons. Statistically, this level of degeneracy will cover all 20 amino acids. As it is difficult to amplify megaprimers with sizes up to 800 bp (Brons-Poulsen et al., 2001; Barik, 2002), the cutinase gene was divided into the 5'-part covering amino acids 1–111 and the 3'-part covering amino acids 112–214 (Fig. 1). In the first PCR reaction, the megaprimer harbouring the desired point mutation was amplified using a mutagenesis primer carrying a randomised codon at the desired position and one of the upstream primers: pmacup, when the mutagenesis primer contained the mutation between amino acid residues 1–111 of cutinase, or cutup, when the mutagenesis primer contained the mutation between amino acid residues 112–214 of cutinase. The product from this PCR reaction was isolated by agarose gel electrophoresis and the band containing the megaprimer was cut and purified. In the second PCR reaction, the mutated product was amplified using the megaprimer together with the pmacup primer or the cutup primer and an additional downstream primer pucdown. The sequence of the primer used for the first PCR reaction must be absent in the vector used as the template in the second PCR reaction to guarantee a very low frequency of wild-type false positive clones.
|
The PCR reaction was performed using a Mastercycler Gradient (Eppendorf), with the following optimised protocols. PCR1: amplification of DNA fragments was performed in a 50 µl reaction mix containing 1x Pfu buffer, 1 ng plasmid pMac-cut as template, 25 pmol of each primer, 0.2 mM dNTP’s and 2.5 U Pfu (Stratagene) polymerase. A negative control was performed omitting the template. PCR conditions for synthesis of mutant megaprimers were as follows: 1x (3 min 98°C); 35x (1 min 95°C; 2 min 58°C; 1 min 72°C) and 1x (7 min 72°C); PCR2: reactions were carried out in a total volume of 50 µl with 1x Taq buffer with (NH4)2SO4, 1.5 mM MgCl2, 1 ng plasmid pUC-cut (used when the megaprimer contained the mutation between amino acid residues 1–111 of cutinase) or pUC-cut/2 (used when the megaprimer contained the mutation between amino acid residues 112–214 of cutinase) as template, 25 pmol of megaprimer and both upstream and downstream primers, 0.2 mM dNTP’s and 2.5 U Taq (Fermentas) polymerase. Two negative controls were performed, one omitting the template and a second omitting the megaprimer. PCR conditions for synthesis of the final mutated product were as follows: 1x (3 min 98°C); 35x (1 min 95°C; 2 min 65°C; 1 min 72°C) and 1x (7 min 72°C).
The PCR-amplified cut mutant genes were cloned into vector pMac-Sm/Sp (when the mutated product was amplified using the pmacup primer) or pMac-cut-Sm/Sp (when the mutated product was amplified using the cutup primer) using EcoRI and XbaI or Acc65I and XbaI restriction sites, respectively. The pool of mutated DNA was transformed into E. coli WK6.
Pre-screening for enzymatic activity
The recombinant library was pre-screened for activity on tributyrin agar plates (Kok et al., 1993). Release of free fatty acids due to the hydrolytic activity of cutinase yields halos around the colonies of active variants.
High throughput screening for activity/stability
In order to screen mutants with enhanced activity/stability, transformants were inoculated on LB agar plates containing 100 µg/ml ampicillin and incubated at 37°C overnight. Colonies appearing after transformation were picked with a toothpick and cultured overnight with shaking at 37°C in 96 deep well microtiter plates filled with 200 µl of LB medium aseptically supplemented with ampicillin 100 µg/ml and sealed. Two 96-plates were screened per codon and every plate contained a column of cutinase wild-type control. The absence of contaminating cutinase activity in the pMac5–8 WK6 expression system was established by assaying the empty vector control under the same reaction conditions. After overnight growing, 800 µl of TB medium (12 g/l tryptone, 24 g/l yeast extract, 5 ml/l glycerol, 3.81 g/l KH2PO4, 12.51 g/l K2HPO4) supplemented with ampicillin 100 µg/ml was dispensed and after 4 h of shaking at 25°C (OD580 1) cutinase expression was induced by adding 20 µl of 5 mM isopropyl-β-D-thiogalactoside (IPTG). The induced cultures were incubated for an additional 20 h and cells were separated from the culture medium by centrifugation at 3220 g for 30 min.
A high-throughput robot assisted screening, using an automated Tecan-pipetting-workstation, was developed to identify cutinase mutants with improved stability in the presence of surfactants. A 5 µl aliquot from each culture supernatant was transferred to individual microtiter plates in duplicate. Afterwards, one plate was filled with 195 µl of 20 mM Tris–HCl buffer, pH 8.0 and the other with 195 µl of 20 mM Tris–HCl buffer, pH 8.0 with 0.5 mM AOT. Both plates were then incubated for 30 min at 40°C. Aliquots of 5 µl each were subsequently transferred to individual microtiter plates in duplicate. Afterwards, 195 µl of substrate (p-nitrophenyl-palmitate) solution (Winkler and Stuckmann, 1979) was dispensed, and the kinetic assay was followed using a plate reader at OD400. Reactions were typically monitored for 6-data points at 60 s intervals. Rates of cutinase activity in the presence and absence of AOT were compared in order to measure surfactant stability. Clones exceeding 15% surfactant stability relative to the wild-type were chosen and plated onto LB plates. This collection of potential positive variants was then re-screened under the same conditions for confirmation purposes. Those clones passing the secondary assay were re-grown, and their plasmids were isolated and sequenced.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Surfactant stability of the parental cutinase
In the present study, we have used a complete saturation mutagenesis approach to search cutinase for amino acids contributing to increased stability in the presence of the anionic surfactant AOT. Activation of wild-type cutinase was not observed upon incubation with AOT concentrations of 0.3–1 mM. However, cutinase was rapidly inactivated upon incubation with AOT within 1 h at 40°C (Fig. 2) with a half life of 25 min for 300 nM pure cutinase in 0.5 mM AOT at pH 8.
|
Pre-screening for enzymatic activity
The percentage of lipolytically active variants tested in a pre-screening on tributyrin agar plates is shown in Fig. 3. Substitutions at 33% of the amino acid positions resulted in inactive variants. As expected, most of them have solvent-inaccessible side-chains. The catalytic residues Ser-120, Asp-175 and His-188 are exceptions. It was observed that the variants of the oxyanion hole residues Ser-42 and Gln-121 and the variants at positions Gly-82, Asn-84 and Ala-85 that are located in the putative lid helix (residues 81–85) were mostly inactive. Asn-84 participates indirectly in the oxyanion hole by maintaining Ser-42 O
in a suitable orientation and probably the properties of the other residues of the helical lid are connected with the integrity of the oxyanion hole. Also, a high percentage of inactive variants were found at the positions of four cysteines as expected because the two disulphide bridges, Cys31–Cys109 and Cys171–Cys178 play an important role for the structural integrity of this enzyme. Among the other positions most sensitive to substitution, five, Gly-117, Gly-118, Gly-148, Asn-152 and Gly-174, are small residues lining the active site. Probably the presence of bulky side chains at these positions interferes with substrate binding.
|
Surfactant stability of single site variants
A complete saturation library of the parental cutinase gene was constructed using the megaprimer method resulting in 214 pools of mutagenised genes, with each pool randomised at one of the amino acid positions. A total of 41 216 clones were picked and high throughput screens were conducted. To correctly interpret the data and reduce the probability of selecting false positives, variability among individual assays was minimised during each step of the process. In order to validate the screening, a set of control experiments were performed. Multiple colonies of the wild-type cutinase were grown and processed in 96-well microplates, in conditions identical to the mutant’s cultivation and assaying. Measuring the screening variability is important to determine a lower limit above which the observed improvement can be considered significant. A good way to estimate the variability or dispersion of data is by determining the coefficient of variance (CV) (Salazar and Sun, 2003).
One way to overcome some of the asynchrony between the wells is to perform a pre-incubation and to extend the growth period so that all the cultures eventually reach the maximum level allowed by the medium (Lamsa et al., 2004). The growth curve was followed together with the enzyme activity and no decline was observed in these measurements during the incubation and a CV of <10% was achieved.
Another important source of error is poor design of the activity assay (Salazar and Sun, 2003). Assay conditions were adjusted for quantitative measurements, i.e. initial rate and saturating substrate concentration for the wild-type, to avoid very active variants depleting the substrate in short times. The lowest experimental CV achieved was nearly 15%, mostly due to the surfactants and the substrate emulsion used in the activity assay. Besides that, as temperature or time increases, the variability associated with this type of measurement also generally increases (Lamsa et al., 2004).
Those clones more stable in the presence of AOT than the wild-type controls after incubation were recovered and the respective genes were sequenced to determine the amino acid substitutions conferring increased stability.
Complete saturation mutagenesis offers a tremendous advantage over other less comprehensive mutagenic techniques such as error prone PCR, since all possible amino acids can be accessed for their contribution to a particular property. Here, we have scanned the primary structure of cutinase from F. solani pisi revealing an AOT stability map of this enzyme (Fig. 4A).
|
A total of 24 single-site mutations at 18 different sites were identified as being more stable in the presence of AOT. Figure 4B shows a ribbon plot of the wild-type cutinase indicating the location of these amino acids. All mutations were located at or near the enzyme surface. In general, surface residues are far more tolerant for substitutions than core residues (Rennell et al., 1991
). Of the 24 mutations found, only one (R17S) was the result of a single base change codon; 12 were double changes and the remaining 11 three-base changes (all the position in the respective codon were changed). These mutations would never have been captured by a standard mutagenesis protocol such as error prone PCR, which accesses at random one-base change per codon. These mutations increase the stability of the wild-type enzyme in the presence of AOT between 2- and 11-fold (Table I) and substitution S54D was found to be the most effective regarding the improvement of the anionic surfactant AOT tolerance and maintaining enzymatic activity without surfactant.
|
Structural domains and amino acids substitutions involved in AOT stability Figure 4B allows one to distinguish five different regions that are important in the stabilisation of cutinase in the presence of AOT: (i) the loop near the N-terminus; (ii) the binding loop 40–52; (iii) the binding loop 171–191; (iv) the loop 151–166 and (v) the
-helices 51–63 and 191–211 and the loop 64–66. In the loop near the N-terminus, we found three variants that were more stable in the presence of AOT: the elimination of a positively charged amino acid (R17S) or the introduction of a negatively charged amino acid (T18D), potentially with electrostatic interaction with the lipid contact zone (aa40–aa52), may reduce the electrostatic interaction between AOT and the enzyme or the introduction of an amino acid with a larger side chain (G26P) may reduce the interaction between the anionic surfactant molecules and the enzyme.
During the screening, six variants were identified in four positions belonging to the substrate binding regions (40–52, 171–191) that were more stable in the presence of AOT: the introduction of an amino acid with a larger side chain (G41A) may fill up the crevice of the substrate binding thereby reducing the capability of surfactant molecules to penetrate via the cavity into the hydrophobic core of the enzyme or the introduction of a negatively charged amino acid (T45D) may decrease the sensitivity against the AOT; the introduction of an amino acid with a larger side chain (T179N/E, A185T) may reduce the interaction between the surfactant and the enzyme or the introduction of a cysteine (T179C) next to another cysteine (Cys178) that forms a disulphide bond Cys171–Cys178 near the active site may introduce some structural modification that stabilises the enzyme against the AOT. Variant T179E was the less sensitive to AOT (11.5-fold increase) although it almost loses the activity without surfactant. This substitution represents an example of a high improvement in one property with a concomitant loss of another property.
Molecular dynamics simulations by Creveld et al. (1998) showed that the large loop encompassing residues 151–166 did not create any large hydrophobic patch upon generation of unfolding motions. These authors identified this region as a weak spot in the molecular architecture and thus, projected a minor sensitivity towards surfactants. However, in this study, several mutants were identified which are more stable in the presence of AOT. The observed amino acid substitutions include (i) residues with a positive charge (K151M, R156K/N and R158L); (ii) the introduction of a less hydrophobic residue (L153Q/A); (iii) the substitution of an amino acid with a negative charge (A164E) shielding the positive charge of R158 and resulting in a putative additional hydrogen-bond between carboxylate of E164 and R158. These results suggest that the residues of this loop are involved in the destabilisation of the enzyme in the presence of anionic surfactants.
A large hydrophobic crevice, which was identified by molecular dynamic computer simulations, is thought to be the most important weak region for unfolding by anionics (Creveld et al., 1998). This crevice is created when helices, here encompassing residues 191–211 and 51–63, move apart in the course of unfolding. A ‘kink motion’, i.e. a movement of the -helix between residues 51–55 and a conformational change in the loop encompassing residues 64–66 characterises this motion. In this region, we observed that substitutions at Ser-54, -57, -61, Lys-65 and Arg-196 show variants with a decrease in the sensitivity against AOT: the replacement of the serines present in the -helix 51–63 by a negatively charged amino acid (S54D/E, S57D and S61D) or the elimination of a positively charged amino acid (K65P and R196A/E) may decrease the sensitivity against the AOT. R196A does not have a significant decrease in activity (15%) when compared with the variant R196E (80%). This can be due to the fact that alanine is a residue with a high helix propensity and consequently stabilises the -helix where this residue is located (Matthews et al., 1987; Zhang et al., 1991). Of the variants that retained catalytic activity similar to native cutinase, S54D had the most reduced sensitivity to AOT. The introduction of a negative charge may repel AOT molecules, but besides that there is a probability that an additional hydrogen bond is formed between carboxylate D54 and main chain carbonyl A195 which prevents the separation of helices encompassing residues 191–211 and 51–63, thus avoiding the formation of the large hydrophobic crevice.
Amino acid substitutions conferring AOT sensitivity About 27% of the amino acids substitutions resulted in inactive variants in the presence of AOT. This happens with most of the negatively charged residues. Probably, their elimination will increase the interactions between the anionic surfactant and the positively charged residues, decreasing the stability of cutinase.
However, it was also possible to observe this situation with the substitutions of several positively charged residues. (i) Arginines that are less solvent accessible, Arg-20, Arg-40, Arg-138 and Arg-166 lead to variants that are mostly inactive as expected. However, the variants that showed activity in the absence of surfactant were mostly inactive in the presence of AOT, with the exception of the variants at Arg-138 that mostly remained unaffected. (ii) Arg-88 is located in one of the binding loops (73–91) and is one of the arginines that constitute the ring of positively charged residues around the hydrophobic surface adjacent to the lipid contact zone. It can be argued that the elimination of this arginine would prevent the binding of the anionic surfactants to the enzyme. Instead, it was observed that the active variants of this residue show a higher sensitivity against the anionic surfactant. (iii) Of the five major helices present in cutinase, the C-terminal helix is the one with the highest content of non-polar residues and simultaneously, it has three residues with positively charged side chains, Lys-206, Arg-208 and Arg-211. We observed that the variants of Arg-208 were mostly inactive and a high percentage of the active variants of these three residues were more sensitive to AOT. It seems that in this case the concentration of positive charges decreases the sensitivity to the surfactant and their elimination will facilitate further association of the detergent or even have a drastic structural effect in the case of Arg-208. A similar effect had been observed in unfolding studies of cellulose Cel45 from Humicola insolens by anionic surfactants (Otzen et al., 1999). In this study, double and triple mutants (Arg) in close sequence proximity bearing positive charges were constructed and a decrease was observed in sensitivity to the anionic surfactant. The authors argued that local concentrations of charges with the same sign disfavours the binding of a ligand with hydrophobic groups, since these will decrease the dielectric constant of the local environment, and thus increase the repulsion among the charged groups.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Complete saturation mutagenesis and subsequent high throughput screening allowed us to identify cutinase variants which showed increased stability in the presence of AOT without decreasing the enzymatic activity. The observed amino acid substitutions would have been impossible to predict a priori with current methods. Their localisation in a model derived from the wild-type cutinase crystal structure revealed hot spots at the cutinase surface; and it is important to note that most of them are remote from the active site or located close to the substrate binding loops. Our results also allowed one to attribute an important role to positively charged amino acids of cutinase in the sensitivity against the model detergent AOT. We believe that the mutants showing substitutions in the large hydrophobic crevice (S54D, S57D, S61D, K65P, R196A), that is thought to be the region more involved in the unfolding by anionics, will be very important to obtain an enzyme less sensitive to AOT.
|
Footnotes |
|---|
Edited by David Ollis
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
V. Brissos gratefully acknowledges PhD grant SFRH/BD/9019/2002 from Fundação para a Ciência e Tecnologia. Additionally, the authors wish to thank Holger Gieren of the Institute of Molecular Enzyme Technology for his help and valuable advice with robotic screening of cutinase variants.
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Barettino D., Feigenbutz M., Valcarcel R., Stunnenberg H.G. Nucleic Acids Res. (1994) 22:541–542.
Barik S. Methods in Molecular Biology—Chen B.-Y., Janes H.W., eds. (2002) NJ: Human Press Inc. 189–196.
Brons-Poulsen J., Nohr J., Larsen L.K. Methods in Molecular Biology—Braman J., ed. (2001) NJ: Humana Press Inc. 71–76.
Carvalho C.M.L., Serralheiro M.L.M., Cabral J.M.S., Aires-Barros M.R. Enzyme Microb. Tech. (1997) 21:117–123.[CrossRef][Web of Science]
Carvalho C.M.L., Cabral J.M.S., Aires-Barros M.R. Enzyme Microb. Tech. (1999) 24:569–576.[CrossRef][Web of Science]
Creveld L.D., Amadei A., van Schaik R.C., Pepermans H.A.M., de Vlieg J., Berendsen H.J.C. Proteins (1998) 33:253–264.[CrossRef][Web of Science][Medline]
Cunnah P.J., AiresBarros M.R., Cabral J.M.S. Biocatal. Biotransform. (1996) 14:125–146.[CrossRef]
DeSantis G., et al. J. Am. Chem. Soc. (2003) 125:11476–11477.[CrossRef][Web of Science][Medline]
Dower W.J., Miller J.F., Ragsdale C.W. Nucleic Acids Res. (1988) 16:6127–6145.
Egmond M.R., Van der Hijden H.T.M., Musters W., Peters H., Verrips C.T., De Vlieg J. Modified Cutinases, DNA, Vector and Host (1994) WO: World Patent. 94–14963.
Egmond M.R., de Vlieg J., Verheij H.M., de Haas G.H. Engineering of/with Lipases—Malcata F.X., ed. (1996) Dordrecht, The Netherlands: Kluwer Academic Publishers. 193–202.
Flipsen J.A.C., Appel A.C.M., van der Hijden H.T.W.M., Verrips C.T. Enzyme Microb. Tech. (1998) 23:274–280.[CrossRef][Web of Science]
Flipsen J.A.C., van Schaick M.A., Dijkman R., van der Hijden H.T.W.M., Verheij H.M., Egmond M.R. Chem. Phys. Lipids (1999) 97:181–191.[CrossRef][Web of Science][Medline]
Gray K.A., Richardson T.H., Kretz K., Short J.M., Bartnek F., Knowles R., Kan L., Swanson P.E., Robertson D.E. Adv. Synth. Catal. (2001) 343:607–617.[CrossRef]
Henley J.P., Sadana A. Enzyme Microb. Tech. (1985) 7:50–60.[CrossRef][Web of Science]
Juffer A.H., Argos P., DeVlieg J. J. Comput. Chem. (1996) 17:1783–1803.[Web of Science]
Kok R.G., Christoffels V.M., Vosman B., Hellingwerf K.J. J. Gen. Microbiol. (1993) 139:2329–2342.
Kolattukudy P.E. Lipases—Borgstrom B., Brockman H., eds. (1984) Amsterdam: Elsevier. 471–504.
Lamsa M., Jensen N.B., Krogsgaard S. Enzyme Functionality: Design, Engineering and Screening—Svendsen A., ed. (2004) NJ: Marcel Dekker, Inc. 525–556.
Lauwereys M., de Geus P., de Meutter J., Stanssen P., an Matthyssens G. Lipases, Structure, Mechanism and Genetic Engineering—Alberghina L., Schmid R.D., Verger R., eds. (1991) Weinheim: VCH. 243–251.
Longhi S., Czjzek M., Lamzin V., Nicolas A., Cambillau C. J. Mol. Biol. (1997) 268:779–799.[CrossRef][Web of Science][Medline]
Maitra A. J. Phys. Chem. (1984) 88:5122–5125.[CrossRef][Web of Science]
Martinez C., Degeus P., Lauwereys M., Matthyssens G., Cambillau C. Nature (1992) 356:615–618.[CrossRef][Medline]
Martinez C., Degeus P., Stanssens P., Lauwereys M., Cambillau C. Protein Eng. (1993) 6:157–165.
Matthews B.W., Nicholson H., Becktel W.J. Proc. Natl Acad. Sci. USA (1987) 84:6663–6667.
Melo E.P., Aires-Barros M.R., Cabral J.M. Appl. Biochem. Biotechnol. (1995) 50:45–56.[CrossRef][Web of Science][Medline]
Melo E.P., Costa S.M.B., Cabral J.M.S. Photochem. Photobiol. (1996) 63:169–175.[Medline]
Otzen D.E., Christiansen L., Schulein M. Protein Sci. (1999) 8:1878–1887.[Web of Science][Medline]
Parikh M.R., Matsumura I. J. Mol. Biol. (2005) 352:621–628.[CrossRef][Web of Science][Medline]
Pinto-Sousa A.M., Cabral J.M.S., Aires-Barros M.R. Biotechnol. Lett. (1996) 18:583–586.[CrossRef][Web of Science]
Prompers J.J., Groenewegen A., VanSchaik R.C., Pepermans H.A.M., Hilbers C.W. Protein Sci. (1997) 6:2375–2384.[Web of Science][Medline]
Rennell D., Bouvier S.E., Hardy L.W., Poteete A.R. J. Mol. Biol. (1991) 222:67–87.[CrossRef][Web of Science][Medline]
Sadana A. Biocatalysis: Fundamentals in Enzyme Deactivation Kinetics (1991) Englewwod Cliffs, NY: Prentice Hall.
Salazar O., Sun L. Methods in Molecular Biology—Arnold F.H., Georgious G., eds. (2003) Totowa, NJ: Humana Press Inc. 85–96.
Sambrook J., Fritsch E.F., Maniatis T. Molecular Cloning: A Laboratory Manual (1989) New York: Cold Spring Harbor.
Sebastiao M.J., Cabral J.M.S., Aires-Barros M.R. Biotechnol. Bioeng. (1993) 42:326–332.[Medline]
Winkler U.K., Stuckmann M. J. Bacteriol. (1979) 138:663–670.
Zhang X.J., Baase W.A., Matthews B.W. Biochemistry (1991) 30:2012–2017.[CrossRef][Web of Science][Medline]
Received December 26, 2007; revised February 26, 2008; accepted March 5, 2008.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||