The Effect of Zinc in Carbon Concentrating Mechanisms in Phaeodactylum tricornutum

Photosynthesis is crucial for life but is a slow process because the CO2 concentration near the principal carbon-assimilation enzyme RuBisCO is extremely low. Very few plants and algae perform a carbon-concentrating mechanism (CCM) to overcome the insufficiency, which are classified into biophysical and biochemical (C4) mechanism. The enzyme CA catalyzes the reversible dehydration of HCO3- to CO2 in biophysical CCMs and its active site contains a Zn2+. In this study, we hypothesized that Zn2+ availability can impact CCMs and therefore investigated the effect of Zn2+ availability on photosynthetic metabolism in a unicellular marine diatom Phaeodactylum tricornutum. P. tricornutum has a sequenced genome and can conduct both biophysical and C4 CCMs. We observed that Zn2+ has a significant effect on cell growth rate but no significant interference on intracellular metabolism, suggesting no essential compensation of C4 CCMs for biophysical CCMs even at low CA activity anticipated at low Zn2+ concentration.

The Role of PME-1 in Cancer: Therapeutic Implication

Protein phosphatase 2A (PP2A) plays a major role in maintaining cellular signaling homeostasis in human cells by reversibly affecting the phosphorylation of a variety of proteins. Protein phosphatase methylesterase-1 (PME-1) negatively regulates PP2A activity by reversible demethylation and active site binding. Thus far, it is known that overexpression of PME-1 in human gliomas contributes to ERK pathway signaling, cell proliferation, and malignant progression. Whether PME-1-mediated PP2A inhibition promotes therapy resistance in gliomas is unknown. Specific PP2A targets regulated by PME-1 in cancers also remain elusive. Additionally, whether oncogenic function of PME-1 can be generalized to various human cancers needs to be investigated. This study demonstrated that PME-1 expression promotes kinase inhibitor resistance in glioblastoma (GBM). PME-1 silencing sensitized GBM cells to a group of clinically used indolocarbazole multikinase inhibitors (MKIs). To facilitate the quantitative evaluation of MKIs by cancer-cell specific colony formation assay, Image-J software-plugin ‘ColonyArea’ was developed. PME-1-silencing was found to reactivate specific PP2A complexes and affect PP2A-target histone deacetylase HDAC4 activity. The HDAC4 inhibition induced synthetic lethality with MKIs similar to PME-1 depletion. However, synthetic lethality by both approaches required co-expression of a pro-apoptotic protein BAD. In gliomas, PME-1 and HDAC4 expression was associated with malignant progression. Using tumor PME-1, HDAC4 and BAD expression based stratification signatures this study defined patient subgroups that are likely to respond to MKI alone or in combination with HDAC4 inhibitor therapies. In contrast to the oncogenic role of PME-1 in certain cancer types, this study established that colorectal cancer (CRC) patients with high tumor PME-1 expression display favorable prognosis. Interestingly, PME-1 regulated survival signaling did not operate in CRC cells. Summarily, this study potentiates the candidacy of PME-1 as a therapy target in gliomas, but argues against generalization of these findings to other cancers, especially CRC.

Hepsine et matriptase activent l’hémagglutinine des virus influenza A et B et leur inhibition représente une nouvelle stratégie thérapeutique n’entraînant pas le développement de résistance

Résumé: Chaque année, les épidémies saisonnières d’influenza causent de 3 à 5 millions de cas sévères de maladie, entraînant entre 250 000 et 500 000 décès mondialement. Seulement deux classes d’antiviraux sont actuellement commercialisées pour traiter cette infection respiratoire : les inhibiteurs de la neuraminidase, tels que l’oseltamivir (Tamiflu) et les inhibiteurs du canal ionique M2 (adamantanes). Toutefois, leur utilisation est limitée par l’apparition rapide de résistance virale. Il est donc d’un grand intérêt de développer de nouvelles stratégies thérapeutiques pour le traitement de l’influenza. Le virus influenza dépend de l’activation de sa protéine de surface hémagglutinine (HA) pour être infectieux. L’activation a lieu par clivage protéolytique au sein d’une séquence d’acides aminés conservée. Ce clivage doit être effectué par une enzyme de l’hôte, étant donné que le génome du virus ne code pour aucune protéase. Pour les virus infectant l’humain, plusieurs études ont montré le potentiel de protéases à sérine transmembranaires de type II (TTSP) à promouvoir la réplication virale : TMPRSS2, TMPRSS4, HAT, MSPL, Desc1 et matriptase, identifiée récemment par notre équipe (Beaulieu, Gravel et al., 2013), activent l’HA des virus influenza A (principalement H1N1 et H3N2). Toutefois, il existe peu d’information sur le clivage de l’HA des virus influenza B, et seulement TMPRSS2 et HAT ont été identifiées comme étant capables d’activer ce type de virus. Les travaux de ce projet de maîtrise visaient à identifier d’autres TTSP pouvant activer l’HA de l’influenza B. L’efficacité de clivage par la matriptase, hepsine, HAT et Desc1 a été étudiée et comparée entre ces TTSP. Ces quatre protéases s’avèrent capables de cliver l’HA de l’influenza B in vitro. Cependant, seul le clivage par matriptase, hepsine et HAT promeut la réplication virale. De plus, ces TTSP peuvent aussi supporter la réplication de virus influenza A. Ainsi, l’utilisation d’un inhibiteur de TTSP, développé en collaboration avec notre laboratoire, permet de bloquer significativement la réplication virale dans les cellules épithéliales bronchiques humaines Calu-3. Cet inhibiteur se lie de façon covalente et lentement réversible au site actif de la TTSP par un mécanisme slow tight-binding. Puisque cet inhibiteur cible une composante de la cellule hôte, et non une protéine virale, il n’entraîne pas le développement de résistance après 15 passages des virus en présence de l’inhibiteur dans les cellules Calu-3. L’inhibition des TTSP activatrices d’HA dans le système respiratoire humain représente donc une nouvelle stratégie thérapeutique pouvant mener au développement d’antiviraux efficaces contre l’influenza.

Understanding immobilised enzymes by NMR spectroscopy

Enzyme immobilisation is the conversion of a soluble enzyme molecule into a solid particle form. This allows the recovery of the enzyme catalyst for its re-use and avoids protein contamination of the product streams. A better understanding of immobilised enzymes is necessary for their rational development. A more rational design can help enormously in the applicability of these systems in different areas, from biosensors to chemical industry. Immobilised enzymes are challenging systems to study and very little information is given by conventional biochemical analysis such as catalytic activity and amount of protein. Here, solid-state NMR has been applied as the main technique to study these systems and evaluate them more precisely. Various approaches are presented for a better understanding of immobilised enzymes, which is the aim of this thesis. Firstly, the requirements of a model system of study will be discussed. The selected systems will be comprehensibly characterised by a variety of techniques but mainly by solid-state NMR. The chosen system will essentially be the enzyme α-chymotrypsin covalently immobilised on two functionalised inorganic supports – epoxide silica and epoxide alumina – and an organic support – Eupergit®. The study of interactions of immobilised enzymes with other species is vital for understanding the macromolecular function and for predicting and engineering protein behaviour. The study of water, ions and inhibitors interacting with various immobilised enzyme systems is covered here. The interactions of water and sodium ions were studied by 17O and 23Na multiple-quantum techniques, respectively. Various pore sizes of the supports were studied for the immobilised enzyme in the presence of labelled water and sodium cations. Finally, interactions between two fluorinated inhibitors and the active site of the enzyme will be explored using 19F NMR, offering a unique approach to evaluate catalytic behaviour. These interactions will be explored by solution-state NMR firstly, then by solid-state NMR. NMR has the potential to give information about the state of the protein in the solid support, but the precise molecular interpretation is a difficult task.

Glutathione Transferases as Detoxification Agents: Structural and Functional Studies

Glutathione transferases (GSTs) are a diverse family of enzymes that catalyze the glutathione-dependent detoxification of toxic compounds. GSTs are responsible for the conjugation of the tripeptide glutathione (GSH) to a wide range of electrophilic substrates. These include industrial pollutants, drugs, genotoxic carcinogen metabolites, antibiotics, insecticides and herbicides. In light of applications in biomedicine and biotechnology as cellular detoxification agents, detailed structural and functional studies of GSTs are required. Plant tau class GSTs play crucial catalytic and non-catalytic roles in cellular xenobiotic detoxification process in agronomically important crops. The abundant existence of GSTs in Glycine max and their ability to provide resistance to abiotic and biotic stresses such as herbicide tolerance is of great interest in agriculture because they provide effective and suitable tools for selective weed control. Structural and catalytic studies on tau class GST isoenzymes from Glycine max (GmGSTU10-10, GmGSTU chimeric clone 14 (Sh14), and GmGSTU2-2) were performed. Crystal structures of GmGSTU10-10 in complex with glutathione sulfenic acid (GSOH) and Sh14 in complex with S-(p-nitrobenzyl)-glutathione (Nb-GSH) were determined by molecular replacement at 1.6 Å and 1.75 Å, respectively. Major structural variations that affect substrate recognition and catalytic mechanism were revealed in the upper part of helix H4 and helix H9 of GmGSTU10-10. Structural analysis of Sh14 showed that the Trp114Cys point mutation is responsible for the enhanced catalytic activity of the enzyme. Furthermore, two salt bridges that trigger an allosteric effect between the H-sites were identified at the dimer interface between Glu66 and Lys104. The 3D structure of GmGSTU2-2 was predicted using homology modeling. Structural and phylogenetic analysis suggested GmGSTU2-2 shares residues that are crucial for the catalytic activity of other tau class GSTs–Phe10, Trp11, Ser13, Arg20, Tyr30, Leu37, Lys40, Lys53, Ile54, Glu66 and Ser67. This indicates that the catalytic and ligand binding site in GmGSTU2-2 are well-conserved. Nevertheless, at the ligandin binding site a significant variation was observed. Tyr32 is replaced by Ser32 in GmGSTU2-2 and thismay affect the ligand recognition and binding properties of GmGSTU2-2. Moreover, docking studies revealed important amino acid residues in the hydrophobic binding site that can affect the substrate specificity of the enzyme. Phe10, Pro12, Phe15, Leu37, Phe107, Trp114, Trp163, Phe208, Ile212, and Phe216 could form the hydrophobic ligand binding site and bind fluorodifen. Additionally, side chains of Arg111 and Lys215 could stabilize the binding through hydrogen bonds with the –NO2 groups of fluorodifen. GST gene family from the pathogenic soil bacterium Agrobacterium tumefaciens C58 was characterized and eight GST-like proteins in A. tumefaciens (AtuGSTs) were identified. Phylogenetic analysis revealed that four members of AtuGSTs belong to a previously recognized bacterial beta GST class and one member to theta class. Nevertheless, three AtuGSTs do not belong to any previously known GST classes. The 3D structures of AtuGSTs were predicted using homology modeling. Comparative structural and sequence analysis of the AtuGSTs showed local sequence and structural characteristics between different GST isoenzymes and classes. Interactions at the G-site are conserved, however, significant variations were seen at the active site and the H5b helix at the C-terminal domain. H5b contributes to the formation of the hydrophobic ligand binding site and is responsible for recognition of the electrophilic moiety of the xenobiotic. It is noted that the position of H5b varies among models, thus providing different specificities. Moreover, AtuGSTs appear to form functional dimers through diverse modes. AtuGST1, AtuGST3, AtuGST4 and AtuGST8 use hydrophobic ‘lock–and–key’-like motifs whereas the dimer interface of AtuGST2, AtuGST5, AtuGST6 and AtuGST7 is dominated by polar interactions. These results suggested that AtuGSTs could be involved in a broad range of biological functions including stress tolerance and detoxification of toxic compounds.

Structural Aspects of the Distinct Biochemical Properties of Glutaredoxin 1 and Glutaredoxin 2 from Saccharomyces cerevisiae

Glutaredoxins (Grxs) are small (9-12 kDa) heat-stable proteins that are ubiquitously distributed. In Saccharomyces cerevisiae, seven Grx enzymes have been identified. Two of them (yGrx1 and yGrx2) are dithiolic, possessing a conserved Cys-Pro-Tyr-Cys motif. Here, we show that yGrx2 has a specific activity 15 times higher than that of yGrx1, although these two oxidoreductases share 64% identity and 85% similarity with respect to their amino acid sequences. Further characterization of the enzymatic activities through two-substrate kinetics analysis revealed that yGrx2 possesses a lower Km for glutathione and a higher turnover than yGrx1. To better comprehend these biochemical differences, the pK(a) of the N-terminal active-site cysteines (Cys27) of these two proteins and of the yGrx2-C30S mutant were determined. Since the pK(a) values of the yGrx1 and yGix2 Cys27 residues are very similar, these parameters cannot account for the difference observed between their specific activities. Therefore, crystal structures of yGrx2 in the oxidized form and with a glutathionyl mixed disulfide were determined at resolutions of 2.05 and 1.91 angstrom, respectively. Comparisons of yGrx2 structures with the recently determined structures of yGrx1 provided insights into their remarkable functional divergence. We hypothesize that the substitutions of Ser23 and Gln52 in yGrx1 by Ala23 and Glu52 in yGrx2 modify the capability of the active-site C-terminal cysteine to attack the mixed disulfide between the N-terminal active-site cysteine and the glutathione molecule. Mutagenesis studies supported this hypothesis. The observed structural and functional differences between yGrx1 and yGrx2 may reflect variations in substrate specificity. (C) 2008 Elsevier Ltd. All rights reserved.

Mechanistic Dissection of the Cop9 Signalosome’s Deneddylation Activity on Cullin-RING Ligases

We set out to understand the precise mechanisms that regulate the activation and deactivation of Cullin-RING Ligases (CRLs). While a great deal of work has already gone into identifying the players involved in these pathways and the cellular consequences associated with the loss of each, the biochemical mechanisms regulating these steps have remained elusive. In this work we sought to gain a better understanding of the mechanisms behind these steps by teasing apart specific their biochemical reactions. By measuring the individual microscopic rate constants of the reactions we have shed light on both the proper sequence of events in the regulation of CRLs as well as how they are in fact controlled.

Prior to this work, it was believed that CSN deactivated CRLs by binding them and enzymatically removing the activating post-translation modification Nedd8. It was believed that CSN could not bind to CRLs while they were active due to the steric hindrance by the CRL substrates, and that they would remain bound to deneddylated CRLs as a sequestering agent until a new substrate could displace it. We now have some insight that substrates themselves cannot inhibit CSN very well, but that the active ubiquitination by an E2 enzyme precludes CSN binding and activity. When the substrate for a CRL becomes depleted, CSN then binds to the CRL in a low affinity, low activity conformation. This triggers a conformational change that pulls the autoinhibitory Ins-1 loop away from the active site in the catalytic subunit Csn5, resulting in a large increase in affinity and cleavage of the isopeptide bond between CRLs and Nedd8. Upon dissociation of Nedd8, CSN rapidly returns to the low affinity state and dissociates from the CRL, allowing it reenter its activation cycle.

Inhibition of Neurotoxic Secretory Phospholipases A(2) Enzymatic, Edematogenic, and Myotoxic Activities by Harpalycin 2, an Isoflavone Isolated from Harpalyce brasiliana Benth

Secretory phospholipases A(2) (sPLA(2)) exert proinflammatory actions through lipid mediators. These enzymes have been found to be elevated in many inflammatory disorders such as rheumatoid arthritis, sepsis, and atherosclerosis. The aim of this study was to evaluate the effect of harpalycin 2 (Har2), an isoflavone isolated from Harpalyce brasiliana Benth., in the enzymatic, edematogenic, and myotoxic activities of sPLA2 from Bothrops pirajai, Crotalus durissus terrificus, Apis mellifera, and Naja naja venoms. Har2 inhibits all sPLA(2) tested. PrTX-III (B. pirajai venom) was inhibited at about 58.7%, Cdt F15 (C. d. terrificus venom) at 78.8%, Apis (from bee venom) at 87.7%, and Naja (N. naja venom) at 88.1%. Edema induced by exogenous sPLA(2) administration performed in mice paws showed significant inhibition by Har2 at the initial step. In addition, Har2 also inhibited the myotoxic activity of these sPLA(2)s. In order to understand how Har2 interacts with these enzymes, docking calculations were made, indicating that the residues His48 and Asp49 in the active site of these enzymes interacted powerfully with Har2 through hydrogen bonds. These data pointed to a possible anti-inflammatory activity of Har2 through sPLA(2) inhibition.

A New Multi-Objective Approach for Molecular Docking Based on RMSD and Binding Energy

Ligand-protein docking is an optimization problem based on predicting the position of a ligand with the lowest binding energy in the active site of the receptor. Molecular docking problems are traditionally tackled with single-objective, as well as with multi-objective approaches, to minimize the binding energy. In this paper, we propose a novel multi-objective formulation that considers: the Root Mean Square Deviation (RMSD) difference in the coordinates of ligands and the binding (intermolecular) energy, as two objectives to evaluate the quality of the ligand-protein interactions. To determine the kind of Pareto front approximations that can be obtained, we have selected a set of representative multi-objective algorithms such as NSGA-II, SMPSO, GDE3, and MOEA/D. Their performances have been assessed by applying two main quality indicators intended to measure convergence and diversity of the fronts. In addition, a comparison with LGA, a reference single-objective evolutionary algorithm for molecular docking (AutoDock) is carried out. In general, SMPSO shows the best overall results in terms of energy and RMSD (value lower than 2A for successful docking results). This new multi-objective approach shows an improvement over the ligand-protein docking predictions that could be promising in in silico docking studies to select new anticancer compounds for therapeutic targets that are multidrug resistant.

Harpalycin 2 inhibits the enzymatic and platelet aggregation activities of PrTX-III, a D49 phospholipase A(2) from Bothrops pirajai venom

Background: Harpalycin 2 (HP-2) is an isoflavone isolated from the leaves of Harpalyce brasiliana Benth., a snakeroot found in northeast region of Brazil and used in folk medicine to treat snakebite. Its leaves are said to be anti-inflammatory. Secretory phospholipases A(2) are important toxins found in snake venom and are structurally related to those found in inflammatory conditions in mammals, as in arthritis and atherosclerosis, and for this reason can be valuable tools for searching new anti-phospholipase A(2) drugs.Methods: HP-2 and piratoxin-III (PrTX-III) were purified through chromatographic techniques. The effect of HP-2 in the enzymatic activity of PrTX-III was carried out using 4-nitro-3-octanoyloxy-benzoic acid as the substrate. PrTX-III induced platelet aggregation was inhibited by HP-2 when compared to aristolochic acid and p-bromophenacyl bromide (p-BPB). In an attempt to elucidate how HP-2 interacts with PrTX-III, mass spectrometry, circular dichroism and intrinsic fluorescence analysis were performed. Docking scores of the ligands (HP-2, aristolochic acid and p-BPB) using PrTX-III as target were also calculated.Results: HP-2 inhibited the enzymatic activity of PrTX-III (IC50 11.34 +/- 0.28 mu g/mL) although it did not form a stable chemical complex in the active site, since mass spectrometry measurements showed no difference between native (13,837.34 Da) and HP-2 treated PrTX-III (13,856.12 Da). A structural analysis of PrTX-III after treatment with HP-2 showed a decrease in dimerization and a slight protein unfolding. In the platelet aggregation assay, HP-2 previously incubated with PrTX-III inhibited the aggregation when compared with untreated protein. PrTX-III chemical treated with aristolochic acid and p-BPB, two standard PLA(2) inhibitors, showed low inhibitory effects when compared with the HP-2 treatment. Docking scores corroborated these results, showing higher affinity of HP-2 for the PrTX-III target (PDB code: 1GMZ) than aristolochic acid and p-BPB. HP-2 previous incubated with the platelets inhibits the aggregation induced by untreated PrTX-III as well as arachidonic acid.Conclusion: HP-2 changes the structure of PrTX-III, inhibiting the enzymatic activity of this enzyme. In addition, PrTX-III platelet aggregant activity was inhibited by treatment with HP-2, p-BPB and aristolochic acid, and these results were corroborated by docking scores.

Protein disulfide isomerase as a target for besnoitiosis therapy : molecular characterization and studies of its role in infection and host immune response

Tese de Doutoramento em Ciências Veterinárias na especialidade de Sanidade Animal

Tail gas catalyzed N2O decomposition over Fe-beta zeolite. On the promoting role of framework connected AlO6 sites in the vicinity of Fe by controlled dealumination during exchange

A novel route to prepare highly active and stable N2O decomposition catalysts is presented, based on Fe-exchanged beta zeolite. The procedure consists of liquid phase Fe(III) exchange at low pH. By varying the pH systematically from 3.5 to 0, using nitric acid during each Fe(III)-exchange procedure, the degree of dealumination was controlled, verified by ICP and NMR. Dealumination changes the presence of neighbouring octahedral Al sites of the Fe sites, improving the performance for this reaction. The so-obtained catalysts exhibit a remarkable enhancement in activity, for an optimal pH of 1. Further optimization by increasing the Fe content is possible. The optimal formulation showed good conversion levels, comparable to a benchmark Fe-ferrierite catalyst. The catalyst stability under tail gas conditions containing NO, O2 and H2O was excellent, without any appreciable activity decay during 70 h time on stream. Based on characterisation and data analysis from ICP, single pulse excitation NMR, MQ MAS NMR, N2 physisorption, TPR(H2) analysis and apparent activation energies, the improved catalytic performance is attributed to an increased concentration of active sites. Temperature programmed reduction experiments reveal significant changes in the Fe(III) reducibility pattern with the presence of two reduction peaks; tentatively attributed to the interaction of the Fe-oxo species with electron withdrawing extraframework AlO6 species, causing a delayed reduction. A low-temperature peak is attributed to Fe-species exchanged on zeolitic AlO4 sites, which are partially charged by the presence of the neighbouring extraframework AlO6 sites. Improved mass transport phenomena due to acid leaching is ruled out. The increased activity is rationalized by an active site model, whose concentration increases by selectively washing out the distorted extraframework AlO6 species under acidic (optimal) conditions, liberating active Fe species.

Conformational states of HIV-1 reverse transcriptase for nucleotide incorporation vs. pyrophosphorolysis – binding of foscarnet

HIV-1 reverse transcriptase (RT) catalytically incorporates individual nucleotides into a viral DNA strand complementing an RNA or DNA template strand; the polymerase active site of RT adopts multiple conformational and structural states while performing this task. The states associated are dNTP binding at the N site, catalytic incorporation of a nucleotide, release of a pyrophosphate, and translocation of the primer 3′-end to the P site. Structural characterization of each of these states may help in understanding the molecular mechanisms of drug activity and resistance and in developing new RT inhibitors. Using a 38-mer DNA template-primer aptamer as the substrate mimic, we crystallized an RT/dsDNA complex that is catalytically active, yet translocation-incompetent in crystals. The ability of RT to perform dNTP binding and incorporation in crystals permitted obtaining a series of structures: (I) RT/DNA (P-site), (II) RT/DNA/AZTTP ternary, (III) RT/AZT-terminated DNA (N-site), and (IV) RT/AZT-terminated DNA (N-site)/foscarnet complexes. The stable N-site complex permitted the binding of foscarnet as a pyrophosphate mimic. The Mg2+ ions dissociated after catalytic addition of AZTMP in the pretranslocated structure III, whereas ions A and B had re-entered the active site to bind foscarnet in structure IV. The binding of foscarnet involves chelation with the Mg2+ (B) ion and interactions with K65 and R72. The analysis of interactions of foscarnet and the recently discovered nucleotide-competing RT inhibitor (NcRTI) α-T-CNP in two different conformational states of the enzyme provides insights for developing new classes of polymerase active site RT inhibitors.

Exploring the role of the α-carboxyphosphonate moiety in the HIV-RT activity of α-carboxy nucleoside phosphonates

As α-carboxy nucleoside phosphonates (α-CNPs) have demonstrated a novel mode of action of HIV-1 reverse transcriptase inhibition, structurally related derivatives were synthesized, namely the malonate 2, the unsaturated and saturated bisphosphonates 3 and 4, respectively and the amide 5. These compounds were evaluated for inhibition of HIV-1 reverse transcriptase in cell-free assays. The importance of the α-carboxy phosphonoacetic acid moiety for achieving reverse transcriptase inhibition, without the need for prior phosphorylation, was confirmed. The malonate derivative 2 was less active by two orders of magnitude than the original α-CNPs, while displaying the same pattern of kinetic behavior; interestingly the activity resides in the “L”-enantiomer of 2, as seen with the earlier series of α-CNPs. A crystal structure with an RT/DNA complex at 2.95 Å resolution revealed the binding of the “L”-enantiomer of 2, at the polymerase active site with a weaker metal ion chelation environment compared to 1a (T-α-CNP) which may explain the lower inhibitory activity of 2.

Fabrication of heterogeneous biocatalyst tethering artificial prosthetic groups to obtain omega-3-fatty acids by selective hydrolysis of fish oils

The active site of lipase from Bacillus thermocathenolatus was selectively modified with allyl and naphthyl chains at different positions. Lipase immobilization and selective tethering of a naphthyl side chain to its position 320 improve both the hydrolysis rate of fish oils and the selectivity towards the eicosapentaenoic acid acyl chains. © The Royal Society of Chemistry 2016.