- Email: [email protected]
Synthesis and evaluation of fluorogenic triglycerides as lipase assay substrates
Chemistry and Physics of Lipids 198 (2016) 72–79
Contents lists available at ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
Synthesis and evaluation of fluorogenic triglycerides as lipase assay substrates Rokhsana J. Andersen, Jesper Brask* Novozymes A/S, Krogshøjvej 36, 2880 Bagsværd, Denmark
A R T I C L E I N F O
Article history: Received 26 April 2016 Received in revised form 23 May 2016 Accepted 25 May 2016 Available online 27 May 2016 Keywords: Lipase assay Triglyceride synthesis Fluorogenic substrate FRET EnzChek
A B S T R A C T
Three racemic fluorogenic triglycerides are synthesized and evaluated as lipase assay substrates. The presented synthesis route goes through a key triglyceride intermediate which can be chemoselectively functionalized with a wide range of different probes. Hence the substrate can be tailor-made for a specific assay, or focus can be on low cost in larger scale for applications in high-throughput screening (HTS) assays. In the specific examples, TG-ED, TG-FD and TG-F2 are assembled with the Edans-Dabcyl or the fluorescein-Dabcyl FRET pair, or relying on fluorescein self-quenching, respectively. Proof-of-concept assays allowed determination of 1st order kinetic parameters (kcat/KM) of 460 s1 M1, 59 s1 M1 and 346 s1 M1, respectively, for the three substrates. Commercially available EnzChek lipase substrate provided 204 s1 M1. Substrate concentration was identified as a critical parameter, with measured reaction rates decreasing at higher concentrations when intermolecular quenching becomes significant. ã 2016 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Lipases are interfacially activated enzymes that catalyze the hydrolysis of lipids on the water/oil interphase (Singh and Mukhopadhyay, 2012; Svendsen, 2000). Microbial lipases are produced in large scale by heterologous expression and applied in the detergent industry, but also for baking, dairy, biodiesel, biocatalysis, etc (Hasan et al., 2006). Development of efficient assays for characterization and high-throughput screening of lipases and their protein engineered variants is therefore of significant interest. Many methods have been reported to follow lipase reactions (Stoytcheva et al., 2012; Hasan et al., 2009), typically based on the liberation of free fatty acids (titration of pH or indicator based). For a high-throughput screening (HTS) setup it is often required that lipase activity generates a color or – preferred – a fluorescent response (Schmidt and Bornscheuer, 2005). As there are limitations to the indicator-based assays, synthetic chromogenic or fluorogenic substrates, such as simple alkyl esters of 4-nitrophenyl (pNP) or 4-methylumbelliferone (4MU), have found wide use (Stoytcheva et al., 2012). The price for their convenience is however that these structures are only very remotely related to natural lipids.
* Corresponding author. E-mail address: [email protected] (J. Brask). http://dx.doi.org/10.1016/j.chemphyslip.2016.05.007 0009-3084/ã 2016 Elsevier Ireland Ltd. All rights reserved.
In this paper we describe the synthesis of three fluorogenic triglycerides (TG-ED, TG-FD and TG-F2, Fig. 1) and provide initial data for their use in lipase assays. The concept of having both a fluorophore and a quencher (or more general a donor and an acceptor) in the same substrate, positioned in such way that enzymatic action (hydrolysis) will remove the quencher and allow the product to become fluorescent, is well-known. Often referred to as FRET (Förster resonance energy transfer) substrates, these have been reported for most classes of hydrolytic enzymes, including proteases (Wang et al., 1990; Thompson et al., 2000), amylases (Oka et al., 2012), DNAses (Ghosh et al., 1994) and phospholipases (Wichmann et al., 2007). FRET is the radiationless energy transfer between two adjacent molecules. The efficiency of energy transfer is related to the sixth power of the ratio of the distance R between donor and acceptor and the Förster radius R0 (Mueller et al., 2013). The Förster radius R0 is the distance between donor and acceptor at which the energy transfer is half-maximal and is usually in the range of 1.5–6 nm. Also a number of FRET-type lipase substrates have been published in the literature. Duque and coworkers synthesized a triglyceride analogue with a pyrene fluorophore and a trinitrophenylamino quencher (Duque et al., 1996), whereas Mitnaul and coworkers have suggested to use self-quenching Bodipy-functionalized triglyceride and phosphatidylcholine substrates (Mitnaul et al., 2007). Further, Yang and coworkers synthesized and evaluated a number of non-glyceride FRET-based lipase substrates (Yang et al., 2006).
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
73
Fig. 1. Structures of the three lipase substrates, compared with commercially available EnzChek lipase substrate.
A number of FRET substrates are commercially available from Invitrogen/ThermoFisher under the brand EnzChek1. There is also a commercial EnzChek lipase substrate (cat. no. E33955), featuring a Bodipy-Dabcyl fluorophore-quencher pair, and an ether bond in the 3-position, presumably to avoid unproductive lipase hydrolysis (Fig. 1) (Basu et al., 2011). The advantages of in-house synthesis over a commercial substrate are related to both price and flexibility. The EnzChek lipase substrate is very costly (approx. 300 USD for 100 mg), severely constraining its use in a HTS setup. Synthesis of the triglyceride substrates presented here easily yields 10–1000 mg per batch, and can be upscaled further. In addition, the chemistry is highly flexible since the presented methodology can be readily modified to synthesize substrates with other fatty acids or fluorophores/quenchers. Hence, it can be imagined that the generic approach of using chemoselective reactions to derivatize a functionalized triglyceride scaffold can be further expanded to include not only other fluorophores and quenchers, but also other chromophores, nanoparticles, affinity ligands, etc. Another interesting difference between the commercial EnzCheck lipase substrate and TG-ED and TG-FD is the position of the fluorophore. In the EnzChek substrate, a typical 1,3-specific lipase will hydrolyse the lauric acid carrying the BODIPY fluorophore from the 1-position of the glycerol backbone (Fig. 1). This is in contrast to TG-ED and TG-FD in which the fluorophore is on the 2-position and a 1,3-specific lipase will therefore produce a fluorescent mono/diglyceride structure. Whereas this may be irrelevant for solution-based assays, the difference could potentially be used for imaging purposes to follow the fate of the hydrolysis products in various biological and technical contexts. 2. Experimental 2.1. General procedures All chemicals were purchased from Sigma-Aldrich, except for Edans C2 maleimide which was from AnaSpec (Fremont, CA, USA) (cat. no. AS-81432) and the EnzChek lipase substrate green (cat. no.
E33955) from Thermo Fisher Scientific. Flash chromatography was carried out with pre-packed silica gel cartridges from Biotage. Analytical TLC was performed on Merck silica gel 60 F254 plates. Spots were detected by a UV lamp at 254 nm or 366 nm or by staining with aqueous H2SO4 or KMnO4 followed by heating with a hot air gun. 1H and 13C NMR data were acquired at 300 K on a Bruker Avance III HD 400 MHz instrument equipped with a SmartProbeTM. MALDI-MS data were acquired on a Bruker Ultraflex Extreme, using a Super-DHB matrix with 0.1% TFA. Fluorescence measurements (for kinetic assay data) were performed at room temperature (25 C) with a SpectraMax M2 from Molecular Devices. 2.2. Synthesis of FA-NHMmt 2 and FA-SMmt 4 2.2.1. FA-NHMmt 2 12-Aminododecanoic acid (FA-NH2, 1) (1.08 g, 5 mmol) was suspended in CHCl3-MeCN (5:1, 18 mL) and chloro trimethylsilane (TMS-Cl, 0.63 mL, 1 eq) was added. The mixture was heated to reflux (65 C) under N2 for 2 h. Remains a suspension. After cooling to RT, triethylamine (1.39 mL, 2 eq) was added. Then addition of 4-methoxytrityl chloride (Mmt-Cl), (1.54 g, 1 eq) dissolved in CHCl3 (10 mL). The turbid orange solution was stirred at RT ON. MeOH (1.0 mL, 5 eq) was then added. The orange solution slowly turned yellow. TLC (heptane-EtOAc, 3:1) confirmed full conversion of Mmt-Cl (Rf 0.40) to product (Rf 0.15). Evaporated to an oil, of which 1.7 g is purified by flash chromatography, eluting with heptaneEtOAc (3:1, later 3:2). The product containing fractions were identified by TLC, pooled and evaporated to yield 1.4 g (57%) of yellow viscous oil. 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 4H, Mmt), 7.40 (d, 2H, Mmt), 7.29 (t, 4H, Mmt), 7.20 (t, 2H, Mmt), 6.83 (d, 2H, Mmt CH next to –OMe), 3.81 (s, 3H, –OMe), 2.36 (t, 2H, –CH2COOH), 2.16 (t, 2H, –NHCH2–), 1.68–1.60 (m, 2H, –CH2CH2COOH), 1.53–1.46 (m, 2H, –NHCH2CH2–). 13 C NMR (CDCl3, selected signals in ppm): 113.0 (Mmt CH next to –OMe), 55.2 (–OMe), 43.7 (–NHCH2–), 34.1 (–CH2COOH), 30.6 (–NHCH2CH2–), 24.8 (–CH2CH2COOH).
74
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
2.2.2. FA-SMmt 4 16-Mercaptohexadecanoic acid (FA-SH, 3) (366 mg, 1.27 mmol) and Mmt-Cl (392 mg, 1 eq) was dissolved in DCM-DMF (1:1, 6 mL). Diisopropylethylamine (DIPEA, 0.5 mL, 2.3 eq) was added to the mixture. Stirred at RT for 2 h. TLC (1:1 heptane-EtOAc) then showed full conversion. Rf (Mmt-Cl) = 0.70; Rf (4) = 0.56. The reaction mixture was evaporated, redissolved in CHCl3 and purified by flash chromatography (eluting initially with hexane-EtOAc 4:1, later 1:1). Product containing fractions were identified by TLC, pooled and evaporated to dryness. Yield 458 mg (64%) of a white wax. 1 H NMR (CDCl3, selected signals in ppm): 7.40 (d, 4H, Mmt), 7.31 (d, 2H, Mmt), 7.25 (t, 4H, Mmt), 7.18 (t, 2H, Mmt), 6.80 (d, 2H, Mmt CH next to –OMe), 3.77 (s, 3H, –OMe), 2.34 (t, 2H, –CH2COOH), 2.14 (t, 2H, –CH2S–), 1.66–1.58 (m, 2H, –CH2CH2COOH). 13 C NMR (CDCl3, selected signals in ppm): 180.4 (–COOH), 113.1 (Mmt CH next to –OMe), 65.9 (Mmt quaternary C), 55.2 (–OMe), 34.1 (–CH2COOH), 32.1 (–SCH2–), 24.7 (–CH2CH2COOH).
2.3.3. TG-(SH)-NH2 9 TG-(SMmt)-NHMmt 8 (86 mg, 0.063 mmol) was dissolved in DCM (3 mL). The mixture was cooled on ice and TFA (60 mL, 12 eq) and triethylsilane (30 mL, 3 eq) was added. The clear yellow solution was stirred 20 min at 0 C and 30 min at RT. TLC (heptaneEtOAc 3:1) of the now colorless solution showed full conversion with Rf (product) = 0. The mixture was concentrated to dryness and then purified by flash chromatography, eluting initially with heptane-EtOAc 1:2, then shifting to an EtOAc-MeOH gradient (8:1–8:2). Product containing fraction were identified by TLC, pooled and evaporated to yield 54 mg (92%, assuming TFA-salt). TLC (DCM-MeOH 9:1) showed pure product with Rf = 0.67. 1 H NMR (CDCl3, selected signals in ppm): 5.30–5.25 (m, 1H, H2), 4.30 (dd, 2H, H-1/3), 4.16 (dd, 2H, H-1/3), 2.92 (t, 2H, –CH2NH2), 2.54 (q, 2H, –CH2SH), 2.36–2.30 (m, 3 2H, –CH2COO–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 68.9 (C-2), 62.1 (C-1/3), 40.0 (–CH2NH2), 34.0 (–CH2COO–), 24.6 (–CH2SH), 14.1 (–CH3).
2.3. Synthesis of TG-ED 11 and TG-FD 12 2.3.1. DG-NHMmt 6 1-Stearoyl-rac-glycerol 5 (4.00 g, 11.16 mmol), N-(3-dimethylaminopropyl)-N0 -ethylcarbodiimide hydrochloride (EDAC, 6.30 g, 2.5 eq) and 4-(dimethylamino)pyridine (DMAP, 297 mg, 0.2 eq) was added to a solution of FA-NHMmt 2 (6.80 g, 1.2 eq) in THF-DCM (1:1, 100 mL) cooled to 0 C on ice. After stirring for 1 h, the reaction was continued for 2 h at RT. TLC (hexane-EtOAc 3:1) then showed full conversion. Rf (6) = 0.30 and Rf (7) = 0.60. Purification by flash chromatography, eluting with a heptane-EtOAc (4:1). Yield 3.72 g (40%). 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 4H, Mmt), 7.40 (d, 2H, Mmt), 7.28 (t, 4H, Mmt), 7.19 (t, 2H, Mmt), 6.83 (d, 2H, Mmt CH next to –OMe), 4.23–4.09 (m, glycerol backbone, 5H), 3.80 (s, 3H, –OMe), 2.37 (t, 2 2H, –CH2COO–), 2.13 (t, 2H, –CH2NH–), 1.67–1.61 (m, 4H, –CH2CH2COO–), 1.55–1.45 (m, 2H, –CH2CH2NH–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 173.9 (–COO–), 113.0 (Mmt CH next to –OMe), 70.3 (Mmt quaternary C), 68.4 (C-2), 65.1 (C-1/3), 55.2 (–OMe), 43.6 (–CH2NH–), 34.1 (–CH2COO–), 24.9 (–CH2CH2COO–), 22.7 (–CH2CH3), 14.1 (–CH3). 2.3.2. TG-(SMmt)-NHMmt 8 DG-NHMmt 6 (200 mg, 0.24 mmol), EDAC (119 mg, 2.5 eq), DMAP (6 mg, 0.2 eq) and FA-SMmt 4 (136 mg, 1 eq) were dissolved in DCM-THF (3:2, 5 mL). The solution was stirred under N2 at 0 C for 1 h and then allowed to warm to RT. TLC (heptane-EtOAc 3:1) after 18 h showed almost full conversion with Rf (8) = 0.5. The reaction mixture was evaporated in vacuo and purified by flash chromatography, eluting with heptane-EtOAc 7:1. This yielded 279 mg (87%) pure target product. 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 4H, Mmt), 7.44–7.38 (m, 6H, Mmt), 7.35–7.16 (m, 14H, Mmt), 6.83 (d, 2 2H, Mmt CH next to –OMe), 5.31–5.26 (m, 1H, H-2), 4.31 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.81 (s, 3H, –OMe), 3.80 (s, 3H, –OMe), 2.33 (t, 3 2H, –CH2COO–), 2.18–2.11 (m, 2 + 2H, –CH2NH– + –CH2S–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 173.3 (–COO–), 172.9 (–COO–), 113.1 (Mmt CH next to –OMe), 113.0 (Mmt CH next to – OMe), 70.3 (Mmt quaternary C), 68.9 (C-2), 65.9 (Mmt quaternary C), 62.1 (C-1/3), 55.2 (–OMe), 55.1 (–OMe), 43.6 (–CH2NH–), 34.2 (–CH2COO–), 34.1 (–CH2COO–), 24.9 (–CH2CH2CO–), 24.8 (–CH2CH2CO–), 22.7 (–CH2CH3), 14.1 (–CH3).
2.3.4. TG-(SH)-NHDabcyl 10 TG-(SH)-NH2 9 (103 mg, 0.110 mmol assuming TFA-salt) was dissolved in DMF and added Dabcyl-NHS (53 mg, 1.1 eq) and triethylamine (83 mL, 5 eq). The mixture was stirred for 2 h at 40 C. TLC (heptane-EtOAc 3:2) then showed product 10 with Rf = 0.51, an unidentified byproduct with Rf = 0.41, and excess Dabcyl-NHS with Rf = 0.21. The mixture is purified by flash chromatography, eluting with a gradient of heptane-EtOAc, 5:2–4:2. Evaporation of fractions yielded 60 mg (51%) of the desired product 10 as a red solid material. 1 H NMR (CDCl3, selected signals in ppm): 7.94–7.86 (m, 6H, dabcyl), 6.78 (d, 2H, dabcyl), 6.21 (t, 1H, –CH2NHCO–), 5.32–5.25 (m, 1H, H-2), 4.31 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.49 (q, 2H, – CH2NHCO–), 3.13 (s, 6H, –NMe2 dabcyl), 2.54 (q, 2H, –CH2SH), 2.37– 2.29 (m, 3 2H, –CH2COO–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 68.9 (C-2), 62.8 (C-1/ 3), 40.3 (–NMe2 dabcyl), 40.2 (–CH2–NHCO–), 34.0 (–CH2COO–), 24.6 (–CH2SH), 14.0 (–CH3). 2.3.5. TG-ED 11 TG-(SH)-NHDabcyl 10 (56 mg, 0.052 mmol) was suspended in DMF-EtOAc 3:1 (2 mL). EDANS C2 maleimide (18 mg, 1 eq) and triethylamine (25 mL, 3 eq) was added. The mixture was stirred 1 h at RT and then 1 h at 30 C. TLC (DCM-MeOH 20:3) then showed TG-ED 11 with Rf = 0.31. The mixture was purified by flash chromatography eluting with a gradient going from 3% to 10% MeOH in DCM. The product containing fraction were pooled and evaporated to yield 55 mg (74%) of the desired TG-ED 11 as an orange-red solid. 1 H NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 8.17 (dd, 2H, edans), 7.92-7.79 (m, 1 + 6H, edans + dabcyl), 7.44–7.34 (m, 2H, edans), 6.74 (d, 2H, dabcyl), 6.71 (t, 1H, –CH2NHCO–), 6.57 (d, 1H, edans), 5.26–5.20 (m, 1H, H-2), 4.27 (dd, 2H, H-1/3), 4.12 (dd, 2H, H-1/3), 4.00-3.90 (m, 2H, –CH2CH2NH-edans), 3.72 (dd, 1H, –S– CH– maleimide), 3.48–3.33 (m, 2 + 2H, –CH2NHCO– + –CH2–NHedans), 3.08 (s, 6H, –NMe2 dabcyl), 2.53 (dd, 2H, –CH2S–), 2.29 (t, 3 2H, –CH2COO–), 0.85 (t, 3H, –CH3). 13 C NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 68.9 (C2), 62.1 (C-1/3), 45.9 (–S–CHCH2– maleimide), 42.9 (–CH2–NHedans), 40.0 (–CH2–NHCO–), 40.0 (–NMe2 dabcyl), 39.3 (–S–CH– maleimide), 38.4 (–CH2CH2NH-edans), 36.2 (–CH2S–), 33.9 (–CH2COO–), 14.0 (–CH3). MALDI-MS (exact mass = 1422.86 Da): 1423.82 [M + H]+, 1445.76 [M + Na]+, 1461.75 [M + K]+.
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
2.3.6. TG-FD 12 TG-(SH)-NHDabcyl 10 (55 mg, 0.052 mmol) in DMF-DCM (1:1, 2 mL) was added N-(5-fluoresceinyl)maleimide (25 mg, 1.1 eq) and triethylamine (25 mL, 3 eq). The mixture was stirred at 30 C for 2 h. TLC (DCM-MeOH 9:1) then showed essentially full conversion with product Rf = 0.16. Purification by flash chromatography using a gradient from 7% MeOH to 15% MeOH in DCM gave pure TG-FD 12 as an orange solid. Yield 37 mg (47%). 1 H NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 7.99– 7.96 (m, 2H, fluorescein), 7.88–7.83 (m, 6H, dabcyl), 7.60 (dd, 1H, fluorescein), 7.26 (d, 1H, fluorescein), 6.75 (d, 2H, dabcyl), 6.70– 6.61 (m, 3 + 1H fluorescein + –CH2NHCO–), 6.53 (dd, 2H, fluorescein), 5.28–5.21 (m, 1H, H-2), 4.28 (dd, 2H, H-1/3), 4.13 (dd, 2H, H-1/3), 3.92 (dd, 1H, –S–CH– maleimide), 3.46–3.33 (m, 2 + 2H, – CH2NHCO– + –S–CHCH2– maleimide), 3.10 (s, 2 3H, –NMe2 dabcyl), 2.72 (dd, 2H, –CH2S–), 2.30 (t, 3 2H, –CH2COO–), 0.86 (t, 3H, –CH3). 13 C NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 68.9 (C2), 61.8 (C-1/3), 40.5 (–NMe2 dabcyl), 40.2 (–CH2–NHCO–), 38.9 (–S–CH– maleimide), 36.1 (–CH2S–), 35.9 (–S–CHCH2– maleimide), 34.0 (–CH2COO–), 13.9 (–CH3). MALDI-MS (exact mass = 1503.86 Da): 1504.95 [M + H]+, 1526.94 [M + Na]+, 1542.91 [M + K]+. 2.4. Synthesis of TG-F2 14 2.4.1. TG-(NHMmt)-NHMmt 7 TG-(NHMmt)-NHMmt 7 was isolated as a byproduct from the synthesis of DG-NHMmt 6 described above. Yield 1.29 g (9%). 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 8H, Mmt), 7.40 (d, 2H, Mmt), 7.31–7.25 (m, 10H, Mmt), 7.22–7.16 (m, 4H, Mmt), 6.83 (d, 2 2H, Mmt CH next to –OMe), 5.32–5.24 (m, 1H, H-2), 4.32 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.81 (s, 2 3H, –OMe), 2.33 (t, 3 2H, –CH2COO–), 2.13 (t, 2 2H, –CH2NH–), 1.68–1.58 (m, 3 2H, –CH2CH2COO–), 1.52–1.42 (m, 2 2H, –NHCH2CH2–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 113.1 (Mmt CH next to –OMe), 69.0 (C-2), 62.0 (C-1/3), 55.2 (–OMe), 43.5 (–CH2NH–), 33.7 (–CH2COO–), 30.8 (–CH2CH2NH–), 24.8 (–CH2CH2COO–), 14.0 (–CH3). 2.4.2. TG-(NH2)-NH2 13 TG-(NHMmt)-NHMmt 7 (200 mg, 0.154 mmol) was dissolved in DCM (6 mL) at 0 C. Then added TFA (0.6 mL) and triethylsilane (0.3 mL, 12 eq). The dark yellow solution was stirred for 30 min at 0 C and then 2 h at RT. TLC (heptane-EtOAc 3:1) then showed full deprotection with Rf (7) = 0.42 and Rf (13) = 0. The reaction mixture was evaporated and used in the next step without purification.
75
The mixture was stirred at 40 C for 3 h, after which TLC (DCMMeOH 9:1) showed good conversion with Rf (14) = 0.46. Purification by flash chromatography, eluting with a gradient going from 5% to 50% MeOH in DCM, resulted in 44 mg (20%) pure TG-F2 as an orange solid, as well as a larger fraction of product needing further purification. 1 H NMR (CD3OD, selected signals in ppm): 8.50–6.50 (18H, fluorescein signals), 5.32–5.23 (m, 1H, H-2), 4.37 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.44 (dt, 2H, –CH2NHCO–), 3.31 (dt, 2H, –CH2NHCO–), 2.33 (t, 3 2H, –CH2COO–), 0.90 (t, 3H, –CH3). 13 C NMR (CD3OD, selected signals in ppm): 69.4 (C-2), 61.9 (C-1/ 3), 39.6 (–CH2–NHCO–), 33.5 (–CH2COO–), 13.3 (–CH3). MALDI-MS (exact mass = 1468.76 Da): 1469.81 [M + H]+, 1491.81 [M + Na]+, 1507.79 [M + K]+. 2.5. Assays Substrate stock solutions (5–10 mM) were made in DMSO. A buffer was made consisting of 50 mM Tris, 2 mM CaCl2, 0.1 mg/mL SB3-14 surfactant, 1 mg/mL BSA (FFA-free), and adjusted to pH 8.0 with HCl. “Assay substrate solutions” were then made by rapidly stirring the buffer while slowly adding the DMSO-stock solution until a final concentration of 800 mM substrate. The assay was performed in black 96-well plates using 40 mL buffer, 50 mL diluted substrate (800 mM to 6.25 mM in a 2-fold dilution series), and 10 mL lipase (Lipolase 100 L diluted 50 x with water). The buffer/substrate mixtures were equilibrated at room temperature for 1 h before the final addition of lipase. Then fluorescence was recorded every 30 s for a total of 15 min. The spectrometer was set for excitation at 350 nm and emission at 460 nm (for TG-ED), or excitation at 492 nm and emission at 517 nm (for TG-FD and TG-F2). When the commercial EnzChek lipase substrate was used in the assay, excitation and emission wavelengths were adjusted to 482 and 515 nm, respectively. From the reaction profiles showing relative fluorescence units (RFU) vs time, the initial velocity v0 was determined by linear regression over the first 5 min. Standard curves, showing RFU vs concentration of fully hydrolysed substrate was constructed from the same plates as used for the kinetic measurements after several days incubation at room temperature. Plotting v0 against substrate concentration allowed determination of vmax/KM as the slope of the initial linear part (up to 25 mM substrate). Dividing with the lipase in-assay concentration (1.3 mM) gave the kcat/KM. The kinetic measurements were done in triplicates with results reported as the average plus/minus the standard deviation. 3. Results and discussion 3.1. Synthesis
2.4.3. TG-F2 14 The crude TG-(NH2)-NH2 13 (0.154 mmol) was dissolved in DMF (3 mL) to which 5-carboxyfluorescein NHS-ester (100 mg, 1.4 eq) and triethylamine (245 mL, 10 eq) was added. Using 1.4 eq 5-carboxyfluorescein was a mistake – it should have been 2 eq.
The lipase substrates were designed to resemble natural triglycerides as much as possible. To allow for a flexible approach, the triglyceride skeleton was first assembled, to which fluorophores and quenchers were added using chemoselective
Scheme 1. Protection of functionalized fatty acids.
76
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
Scheme 2. Synthesis of DG-NHMmt 6 and TG-(NHMmt)-NHMmt 7.
Scheme 3. Synthesis of TG-(SH)-NHDabcyl 10.
reactions. Hence, a key intermediate in the synthesis is the functionalized triglyceride TG-(SH)-NH2 9 (Scheme 3), displaying the two handles, an amine and a thiol. A range of different fluorophores and quenchers are commercially available as NHSesters and maleimides, allowing for orthogonal couplings to this scaffold. TG-(SH)-NH2 9 was, as illustrated in Schemes 1–3, assembled from commercially available 1-stearoyl-rac-glycerol 5 and Mmtprotected amino- (FA-NHMmt 2) and mercapto fatty acids (FASMmt 4). The initial coupling of FA-NHMmt 2 yielded a mixture of di- and triglycerides, but with the desired diglyceride (DG-NHMmt 6) as the main product. The byproduct, TG-(NHMmt)-NHMmt 7, was later used for synthesis of TG-F2 (Scheme 5). It was initially attempted to couple unprotected FA-SH 3 to diglyceride 6. However, this resulted in a complex mixture. It was therefore decided to protect the thiol functionality like the amino group in FA-NH2 with a Mmt protecting group. The 4-methoxytrityl (Mmt) was chosen over the standard trityl protection group since it is significantly more acid labile, which
could be important when deprotecting the triglyceride product. The conditions chosen for protecting FA-SH 3 was inspired by a procedure described by Mourtas and coworkers (Mourtas et al., 2001), whereas protection of FA-NH2 1 followed a procedure by Barlos and coworkers for trityl protection of amino acids, in which the carboxylic acid is temporarily protected as a TMS-ester (Barlos et al., 1982). The EDAC conditions chosen for the coupling reactions to form the di- and triglycerides was inspired by a procedure described by Whitten and coworkers (Whitten et al., 2012). Mild acidic conditions (1–2% TFA in DCM) using triethylsilane as scavenger for the Mmt cation resulted in fast and clean deprotection to key intermediate 9. Using similar deprotecting conditions on diglyceride 6 had resulted in 10–20% acyl migration (estimated by NMR, data not shown). This issue was avoided with the chosen strategy in which the triglyceride is assembled before deprotection. As for the choice of FRET-pairs, Edans-Dabcyl (TG-ED) and Fluorescein-Dabcyl (TG-FD) were chosen since the probes are available and relatively inexpensive in synthetic quantities and
Scheme 4. Synthesis of TG-ED 11 and TG-FD 12.
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
they are both well-described donor-quencher pairs. Both Edans and Fluorescein have emission spectra nicely overlapping the absorption spectrum of the Dabcyl dark quencher. Fluorescein has in contrast to Edans overlapping excitation and emission spectra, meaning it has self-quenching properties (homo-FRET) (Tramier et al., 2003). This is the basis for the design of TG-F2. The probes are attached to the triglyceride scaffold via chemoselective reactions, i.e. coupling of NHS-esters to the primary amine and addition of maleimides to the thiol. The couplings can either be performed sequentially with intermediate purification or one-pot (data not reported). Isolated yields (nonoptimized) were generally in the range of 50–80%. TG-ED 11, TG-FD 12 and TG-F2 14 were hence prepared in 40–50 mg scale and isolated as pure orange/yellow solids (Schemes 4 and 5). 3.2. Assays To obtain proof-of-concept for TG-ED, TG-FD and TG-F2 in lipase assays, reaction conditions close to that reported by Basu and coworkers (Basu et al., 2011) for the commercial EnzChek lipase substrate was chosen. The authors report that the use of Zwittergent surfactant is critical, whereas alternatives such as Triton X-100 would lead to complete dequenching of the substrate. It is however not specified exactly which type Zwittergent is used. After a few attempts, we found a buffer system consisting of 50 mM Tris, 2 mM CaCl2, 0.1 mg/mL SB3-14 surfactant (Zwittergent), 1 mg/mL BSA (FFA-free), pH 8.0 to work satisfactorily. The Thermomyces lanuginosus lipase (TLL) was found to be a natural choice for the assay development work. The enzyme is a classical interfacially activated lipase, it is very well studied and furthermore an industry standard (Fernandez-Lafuente, 2010). It is commercialized by Novozymes under the tradename Lipolase1 100 L for the detergent industry. It is also distributed by Sigma (cat. no. L0777). A 500 x dilution (in-assay concentration) of Lipolase 100 L gave suitable reaction rates. The substrates were readily dissolved in DMSO to make 5–10 mM stock solution. Good reproducibility was obtained by making “assay substrate solutions” from rapidly stirring the buffer while slowly adding the DMSO-stock solution until a final concentration of 800 mM substrate. After final substrate dilutions in the microtiter plate, the plate was left to equilibrate 1 h at room temperature in the dark. This step was essential for obtaining flat baselines of blank samples (no enzyme), presumably by giving the substrate time to be distributed in the surfactant micelles. There was no sign of turbidity or precipitation of substrate under any conditions. It appears that the fluorophores significantly increases solubility of the triglycerides. Having a water-soluble
77
Table 1 Spectroscopic and kinetic results. Substrate
TG-ED
Ex/em wavelengths (nm) 350/460 Standard curve slope (RFU/mM) 34 3 1st order kinetics (up to 25 mM) 1 1 kcat/KM (s M ) 460 24
TG-FD
TG-F2
492/517
492/517
181 33
4655 875
59 9
346 15
substrate clearly simplifies assay development. However, the interfacial activation then has to be provided by the surfactant system. Standard curves, showing fluorescence response (RFU) vs substrate/product concentration, were constructed from a substrate dilution series, incubated with lipase to full hydrolysis. This was also used to find optimal excitation and emission wavelength (Table 1). Not surprisingly, standard curves were not linear over the entire substrate range, but only up to approx. 25 mM, after which the curve flattened out, presumably due to intermolecular quenching at higher concentrations. Kinetic studies were constructed initially aiming towards finding kcat and KM values. However, that proved impossible as KM-values for all three substrates were out of the linear range of the standard curve. Indeed, measured initial reaction rates seemed to go through a maximum and appeared very low at high substrate concentrations (Fig. 2). Plotting initial reaction rates versus substrate concentration in the linear range allowed determination of kcat/KM from the slope (1st order reaction kinetics, Table 1). All three substrates showed essentially flat backgrounds (Fig. 2), once the substrate was fully equilibrated in the micelles. Further, all three showed a good window between fully hydrolysed and non-hydrolyzed substrate, up to a factor 15 for TG-FD and 3–4 for TG-ED and TG-F2 at 25 mM concentration. Looking at the RFU scale (of the standard curve or the reaction profile), TG-F2 is clearly much more fluorescent that TG-ED and TG-FD, due to the very efficient Dabcyl-quenching vs the homo-FRET quenched TG-F2. In applications were a high-fluorescent background can allowed, TG-F2 will have the advantage of being prepared in fewer synthetic steps. The 1st order kinetic values for the three substrates do not appear dramatically different, although with TG-FD being more slowly processed by the lipase (Table 1). The substrates were designed with the fluorescent probes at the end of C12 or longer fatty acids, which is expected to be outside the lipase active site. Any differences between the substrates are therefore not likely a direct substrate/enzyme fit effect, but more a question of how the
Scheme 5. Synthesis of TG-F2 14.
78
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
Fig. 2. Raw kinetics data for 25 mM substrate, with and without enzyme (left) and plots of v0 vs substrate concentration (right) for TG-ED (top), TG-FD (middle) and TG-F2 (bottom).
substrate is presented to the enzyme, the orientation or distribution in the micelles. For comparison, the commercial EnzChek lipase substrate was assayed using the same conditions. This resulted in kcat/KM = 204 s1 M1. Also the commercial substrate showed lower measured reaction rates at high substrate concentrations. Again this rate constant is in the same range as those found for the triglyceride structures, indicating that the ether-linked C6-alkyl chain in EnzChek (Fig. 1) does not dramatically effect Lipolase activity. 4. Conclusion Three fluorogenic lipase substrates were synthesized using chemoselective couplings to a racemic triglyceride scaffold. The amino- and thiol-functionalized triglyceride was assembled using standard carbodiimide couplings of Mmt-protected amino- or thiol fatty acids to the glycerol backbone. Following deprotection under mildly acidic conditions, quenchers and fluorophores were added in high-yielding, click-like (Kolb et al., 2001) reactions of NHSesters or maleimides. The resulting Dabcyl-quenched substrates, TG-ED and TG-FD and the fluorescein self-quenched TG-F2, were subsequently evaluated in proof-of-concept assays with the commercially available Thermomyces lanuginosus lipase Lipolase 100L. This showed rewarding lipase responses with flat blanks, yielding up
to a factor of 15 x difference in fluorescence signal between hydrolysed and intact substrate (TG-FD at 25 mM concentration). Varying the substrate concentrations revealed this as a critical parameter, since high substrate concentrations resulted in lower measured reaction rates, presumably due to intermolecular quenching. Conditions for the proof-of-concept assays are largely unoptimized, but inspired by those reported by Basu and coworkers (Basu et al., 2011). It is therefore likely that assay performance can be further improved and/or conditions adapted to mimic a specific biological system or technical enzyme application. In conclusion the results illustrate the applicability of fluorogenic lipase substrates that structurally are closely related to natural triglycerides, and that these are easily prepared in mg and gram scale. Acknowledgements Mr Clive Phipps Walter, Novozymes, is acknowledged for running MALDI-MS analyses of the three substrates. References Barlos, K., Papaioannou, D., Theodoropoulos, D., 1982. Efficient one-pot synthesis of N-trityl amino-acids. J. Org. Chem. 47, 1324–1326.
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79 Basu, D., Manjur, J., Jin, W.J., 2011. Determination of lipoprotein lipase activity using a novel fluorescent lipase assay. J. Lipid Res. 52, 826–832. Duque, M., Graupner, M., Stutz, H., Wicher, I., Zechner, R., Paltauf, F., Hermetter, A., 1996. New fluorogenic triacylglycerol analogs as substrates for the determination and chiral discrimination of lipase activities. J. Lipid Res. 37, 868–876. Fernandez-Lafuente, R., 2010. Lipase from Thermomyces lanuginosus: uses and prospects as an industrial biocatalyst. J. Mol. Catal. B Enzym. 62, 197–212. Ghosh, S.S., Eis, P.S., Blumeyer, K., Fearon, K., Millar, D.P., 1994. Real-time kinetics of restriction-endonuclease cleavage monitored by fluorescence resonance energy-transfer. Nucleic Acids Res. 22, 3155–3159. Hasan, F., Shah, A.A., Hameed, A., 2006. Industrial applications of microbial lipases. Enzyme Microb. Tech. 39, 235–251. Hasan, F., Shah, A.A., Hameed, A., 2009. Methods for detection and characterization of lipases: a comprehensive review. Biotechnol. Adv. 27, 782–798. Kolb, H.C., Finn, M.G., Sharpless, K.B., 2001. Click chemistry: diverse chemical function from a few good reactions. Angew. Chem. Int. Ed. 40, 2004–2021. Mitnaul, L.J., Tian, J., Burton, C., Lam, M.H., Zhu, Y.P., Olson, S.H., Schneeweis, J.E., Zuck, P., Pandit, S., Anderson, M., Maletic, M.M., Waddell, S.T., Wright, S.D., Sparrow, C.P., Lund, E.G., 2007. Fluorogenic substrates for high-throughput measurements of endothelial lipase activity. J. Lipid Res. 48, 472–482. Mourtas, S., Gatos, D., Kalaitzi, V., Katakalou, C., Barlos, K., 2001. S-4-Methoxytrityl mercapto acids: synthesis and application. Tetrahedron Lett. 42, 6965–6967. Mueller, S., Galliardt, H., Schneider, J., Barisas, B.G., Seidel, T., 2013. Quantification of forster resonance energy transfer by monitoring sensitized emission in living plant cells. Front. Plant Sci. 4, 1–20. Oka, H., Koyama, T., Hatano, K., Matsuoka, K., 2012. Synthetic studies of bifluorescence-labeled maltooligosaccharides as substrates for alpha-amylase on
79
the basis of fluorescence resonance energy transfer (FRET). Bioorg. Med. Chem. 20, 435–445. Schmidt, M., Bornscheuer, U.T., 2005. High-throughput assays for lipases and esterases. Biomol. Eng. 22, 51–56. Singh, A.K., Mukhopadhyay, M., 2012. Overview of fungal lipase: a review. Appl. Biochem. Biotech. 166, 486–520. Stoytcheva, M., Montero, G., Zlatev, R., Leon, J.A., Gochev, V., 2012. Analytical methods for lipases activity determination: a review. Curr. Anal. Chem. 8, 400–407. Svendsen, A., 2000. Lipase protein engineering. Biochim. Biophys. Acta 1543, 223–238. Thompson, V.F., Saldana, S., Cong, J.Y., Goll, D.E., 2000. A bodipy fluorescent microplate assay for measuring activity of calpains and other proteases. Anal. Biochem. 279, 170–178. Tramier, M., Piolot, T., Gautier, I., Mignotte, V., Coppey, J., Kemnitz, K., Durieux, C., Coppey-Moisan, M., 2003. Homo-fret versus hetero-fret to probe homodimers in living cells. Method. Enzym. 360, 580–597. Wang, G.T., Matayoshi, E., Huffaker, H.J., Krafft, G.A., 1990. Design and synthesis of new fluorogenic hiv protease substrates based on resonance energy-transfer. Tetrahedron Lett. 31, 6493–6496. Whitten, K.M., Makriyannis, A., Vadivel, S.K., 2012. Application of chemoenzymatic hydrolysis in the synthesis of 2-monoacylglycerols. Tetrahedron 68, 5422–5428. Wichmann, O., Gelb, M.H., Schultz, C., 2007. Probing phospholipase A(2) with fluorescent phospholipid substrates. Chembiochem 8, 1555–1569. Yang, Y.Z., Babiak, P., Reymond, J.L., 2006. Low background fret-substrates for lipases and esterases suitable for high-throughput screening under basic (pH 11) conditions. Org. Biomol. Chem. 4, 1746–1754.
Contents lists available at ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
Synthesis and evaluation of fluorogenic triglycerides as lipase assay substrates Rokhsana J. Andersen, Jesper Brask* Novozymes A/S, Krogshøjvej 36, 2880 Bagsværd, Denmark
A R T I C L E I N F O
Article history: Received 26 April 2016 Received in revised form 23 May 2016 Accepted 25 May 2016 Available online 27 May 2016 Keywords: Lipase assay Triglyceride synthesis Fluorogenic substrate FRET EnzChek
A B S T R A C T
Three racemic fluorogenic triglycerides are synthesized and evaluated as lipase assay substrates. The presented synthesis route goes through a key triglyceride intermediate which can be chemoselectively functionalized with a wide range of different probes. Hence the substrate can be tailor-made for a specific assay, or focus can be on low cost in larger scale for applications in high-throughput screening (HTS) assays. In the specific examples, TG-ED, TG-FD and TG-F2 are assembled with the Edans-Dabcyl or the fluorescein-Dabcyl FRET pair, or relying on fluorescein self-quenching, respectively. Proof-of-concept assays allowed determination of 1st order kinetic parameters (kcat/KM) of 460 s1 M1, 59 s1 M1 and 346 s1 M1, respectively, for the three substrates. Commercially available EnzChek lipase substrate provided 204 s1 M1. Substrate concentration was identified as a critical parameter, with measured reaction rates decreasing at higher concentrations when intermolecular quenching becomes significant. ã 2016 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Lipases are interfacially activated enzymes that catalyze the hydrolysis of lipids on the water/oil interphase (Singh and Mukhopadhyay, 2012; Svendsen, 2000). Microbial lipases are produced in large scale by heterologous expression and applied in the detergent industry, but also for baking, dairy, biodiesel, biocatalysis, etc (Hasan et al., 2006). Development of efficient assays for characterization and high-throughput screening of lipases and their protein engineered variants is therefore of significant interest. Many methods have been reported to follow lipase reactions (Stoytcheva et al., 2012; Hasan et al., 2009), typically based on the liberation of free fatty acids (titration of pH or indicator based). For a high-throughput screening (HTS) setup it is often required that lipase activity generates a color or – preferred – a fluorescent response (Schmidt and Bornscheuer, 2005). As there are limitations to the indicator-based assays, synthetic chromogenic or fluorogenic substrates, such as simple alkyl esters of 4-nitrophenyl (pNP) or 4-methylumbelliferone (4MU), have found wide use (Stoytcheva et al., 2012). The price for their convenience is however that these structures are only very remotely related to natural lipids.
* Corresponding author. E-mail address: [email protected] (J. Brask). http://dx.doi.org/10.1016/j.chemphyslip.2016.05.007 0009-3084/ã 2016 Elsevier Ireland Ltd. All rights reserved.
In this paper we describe the synthesis of three fluorogenic triglycerides (TG-ED, TG-FD and TG-F2, Fig. 1) and provide initial data for their use in lipase assays. The concept of having both a fluorophore and a quencher (or more general a donor and an acceptor) in the same substrate, positioned in such way that enzymatic action (hydrolysis) will remove the quencher and allow the product to become fluorescent, is well-known. Often referred to as FRET (Förster resonance energy transfer) substrates, these have been reported for most classes of hydrolytic enzymes, including proteases (Wang et al., 1990; Thompson et al., 2000), amylases (Oka et al., 2012), DNAses (Ghosh et al., 1994) and phospholipases (Wichmann et al., 2007). FRET is the radiationless energy transfer between two adjacent molecules. The efficiency of energy transfer is related to the sixth power of the ratio of the distance R between donor and acceptor and the Förster radius R0 (Mueller et al., 2013). The Förster radius R0 is the distance between donor and acceptor at which the energy transfer is half-maximal and is usually in the range of 1.5–6 nm. Also a number of FRET-type lipase substrates have been published in the literature. Duque and coworkers synthesized a triglyceride analogue with a pyrene fluorophore and a trinitrophenylamino quencher (Duque et al., 1996), whereas Mitnaul and coworkers have suggested to use self-quenching Bodipy-functionalized triglyceride and phosphatidylcholine substrates (Mitnaul et al., 2007). Further, Yang and coworkers synthesized and evaluated a number of non-glyceride FRET-based lipase substrates (Yang et al., 2006).
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
73
Fig. 1. Structures of the three lipase substrates, compared with commercially available EnzChek lipase substrate.
A number of FRET substrates are commercially available from Invitrogen/ThermoFisher under the brand EnzChek1. There is also a commercial EnzChek lipase substrate (cat. no. E33955), featuring a Bodipy-Dabcyl fluorophore-quencher pair, and an ether bond in the 3-position, presumably to avoid unproductive lipase hydrolysis (Fig. 1) (Basu et al., 2011). The advantages of in-house synthesis over a commercial substrate are related to both price and flexibility. The EnzChek lipase substrate is very costly (approx. 300 USD for 100 mg), severely constraining its use in a HTS setup. Synthesis of the triglyceride substrates presented here easily yields 10–1000 mg per batch, and can be upscaled further. In addition, the chemistry is highly flexible since the presented methodology can be readily modified to synthesize substrates with other fatty acids or fluorophores/quenchers. Hence, it can be imagined that the generic approach of using chemoselective reactions to derivatize a functionalized triglyceride scaffold can be further expanded to include not only other fluorophores and quenchers, but also other chromophores, nanoparticles, affinity ligands, etc. Another interesting difference between the commercial EnzCheck lipase substrate and TG-ED and TG-FD is the position of the fluorophore. In the EnzChek substrate, a typical 1,3-specific lipase will hydrolyse the lauric acid carrying the BODIPY fluorophore from the 1-position of the glycerol backbone (Fig. 1). This is in contrast to TG-ED and TG-FD in which the fluorophore is on the 2-position and a 1,3-specific lipase will therefore produce a fluorescent mono/diglyceride structure. Whereas this may be irrelevant for solution-based assays, the difference could potentially be used for imaging purposes to follow the fate of the hydrolysis products in various biological and technical contexts. 2. Experimental 2.1. General procedures All chemicals were purchased from Sigma-Aldrich, except for Edans C2 maleimide which was from AnaSpec (Fremont, CA, USA) (cat. no. AS-81432) and the EnzChek lipase substrate green (cat. no.
E33955) from Thermo Fisher Scientific. Flash chromatography was carried out with pre-packed silica gel cartridges from Biotage. Analytical TLC was performed on Merck silica gel 60 F254 plates. Spots were detected by a UV lamp at 254 nm or 366 nm or by staining with aqueous H2SO4 or KMnO4 followed by heating with a hot air gun. 1H and 13C NMR data were acquired at 300 K on a Bruker Avance III HD 400 MHz instrument equipped with a SmartProbeTM. MALDI-MS data were acquired on a Bruker Ultraflex Extreme, using a Super-DHB matrix with 0.1% TFA. Fluorescence measurements (for kinetic assay data) were performed at room temperature (25 C) with a SpectraMax M2 from Molecular Devices. 2.2. Synthesis of FA-NHMmt 2 and FA-SMmt 4 2.2.1. FA-NHMmt 2 12-Aminododecanoic acid (FA-NH2, 1) (1.08 g, 5 mmol) was suspended in CHCl3-MeCN (5:1, 18 mL) and chloro trimethylsilane (TMS-Cl, 0.63 mL, 1 eq) was added. The mixture was heated to reflux (65 C) under N2 for 2 h. Remains a suspension. After cooling to RT, triethylamine (1.39 mL, 2 eq) was added. Then addition of 4-methoxytrityl chloride (Mmt-Cl), (1.54 g, 1 eq) dissolved in CHCl3 (10 mL). The turbid orange solution was stirred at RT ON. MeOH (1.0 mL, 5 eq) was then added. The orange solution slowly turned yellow. TLC (heptane-EtOAc, 3:1) confirmed full conversion of Mmt-Cl (Rf 0.40) to product (Rf 0.15). Evaporated to an oil, of which 1.7 g is purified by flash chromatography, eluting with heptaneEtOAc (3:1, later 3:2). The product containing fractions were identified by TLC, pooled and evaporated to yield 1.4 g (57%) of yellow viscous oil. 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 4H, Mmt), 7.40 (d, 2H, Mmt), 7.29 (t, 4H, Mmt), 7.20 (t, 2H, Mmt), 6.83 (d, 2H, Mmt CH next to –OMe), 3.81 (s, 3H, –OMe), 2.36 (t, 2H, –CH2COOH), 2.16 (t, 2H, –NHCH2–), 1.68–1.60 (m, 2H, –CH2CH2COOH), 1.53–1.46 (m, 2H, –NHCH2CH2–). 13 C NMR (CDCl3, selected signals in ppm): 113.0 (Mmt CH next to –OMe), 55.2 (–OMe), 43.7 (–NHCH2–), 34.1 (–CH2COOH), 30.6 (–NHCH2CH2–), 24.8 (–CH2CH2COOH).
74
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
2.2.2. FA-SMmt 4 16-Mercaptohexadecanoic acid (FA-SH, 3) (366 mg, 1.27 mmol) and Mmt-Cl (392 mg, 1 eq) was dissolved in DCM-DMF (1:1, 6 mL). Diisopropylethylamine (DIPEA, 0.5 mL, 2.3 eq) was added to the mixture. Stirred at RT for 2 h. TLC (1:1 heptane-EtOAc) then showed full conversion. Rf (Mmt-Cl) = 0.70; Rf (4) = 0.56. The reaction mixture was evaporated, redissolved in CHCl3 and purified by flash chromatography (eluting initially with hexane-EtOAc 4:1, later 1:1). Product containing fractions were identified by TLC, pooled and evaporated to dryness. Yield 458 mg (64%) of a white wax. 1 H NMR (CDCl3, selected signals in ppm): 7.40 (d, 4H, Mmt), 7.31 (d, 2H, Mmt), 7.25 (t, 4H, Mmt), 7.18 (t, 2H, Mmt), 6.80 (d, 2H, Mmt CH next to –OMe), 3.77 (s, 3H, –OMe), 2.34 (t, 2H, –CH2COOH), 2.14 (t, 2H, –CH2S–), 1.66–1.58 (m, 2H, –CH2CH2COOH). 13 C NMR (CDCl3, selected signals in ppm): 180.4 (–COOH), 113.1 (Mmt CH next to –OMe), 65.9 (Mmt quaternary C), 55.2 (–OMe), 34.1 (–CH2COOH), 32.1 (–SCH2–), 24.7 (–CH2CH2COOH).
2.3.3. TG-(SH)-NH2 9 TG-(SMmt)-NHMmt 8 (86 mg, 0.063 mmol) was dissolved in DCM (3 mL). The mixture was cooled on ice and TFA (60 mL, 12 eq) and triethylsilane (30 mL, 3 eq) was added. The clear yellow solution was stirred 20 min at 0 C and 30 min at RT. TLC (heptaneEtOAc 3:1) of the now colorless solution showed full conversion with Rf (product) = 0. The mixture was concentrated to dryness and then purified by flash chromatography, eluting initially with heptane-EtOAc 1:2, then shifting to an EtOAc-MeOH gradient (8:1–8:2). Product containing fraction were identified by TLC, pooled and evaporated to yield 54 mg (92%, assuming TFA-salt). TLC (DCM-MeOH 9:1) showed pure product with Rf = 0.67. 1 H NMR (CDCl3, selected signals in ppm): 5.30–5.25 (m, 1H, H2), 4.30 (dd, 2H, H-1/3), 4.16 (dd, 2H, H-1/3), 2.92 (t, 2H, –CH2NH2), 2.54 (q, 2H, –CH2SH), 2.36–2.30 (m, 3 2H, –CH2COO–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 68.9 (C-2), 62.1 (C-1/3), 40.0 (–CH2NH2), 34.0 (–CH2COO–), 24.6 (–CH2SH), 14.1 (–CH3).
2.3. Synthesis of TG-ED 11 and TG-FD 12 2.3.1. DG-NHMmt 6 1-Stearoyl-rac-glycerol 5 (4.00 g, 11.16 mmol), N-(3-dimethylaminopropyl)-N0 -ethylcarbodiimide hydrochloride (EDAC, 6.30 g, 2.5 eq) and 4-(dimethylamino)pyridine (DMAP, 297 mg, 0.2 eq) was added to a solution of FA-NHMmt 2 (6.80 g, 1.2 eq) in THF-DCM (1:1, 100 mL) cooled to 0 C on ice. After stirring for 1 h, the reaction was continued for 2 h at RT. TLC (hexane-EtOAc 3:1) then showed full conversion. Rf (6) = 0.30 and Rf (7) = 0.60. Purification by flash chromatography, eluting with a heptane-EtOAc (4:1). Yield 3.72 g (40%). 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 4H, Mmt), 7.40 (d, 2H, Mmt), 7.28 (t, 4H, Mmt), 7.19 (t, 2H, Mmt), 6.83 (d, 2H, Mmt CH next to –OMe), 4.23–4.09 (m, glycerol backbone, 5H), 3.80 (s, 3H, –OMe), 2.37 (t, 2 2H, –CH2COO–), 2.13 (t, 2H, –CH2NH–), 1.67–1.61 (m, 4H, –CH2CH2COO–), 1.55–1.45 (m, 2H, –CH2CH2NH–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 173.9 (–COO–), 113.0 (Mmt CH next to –OMe), 70.3 (Mmt quaternary C), 68.4 (C-2), 65.1 (C-1/3), 55.2 (–OMe), 43.6 (–CH2NH–), 34.1 (–CH2COO–), 24.9 (–CH2CH2COO–), 22.7 (–CH2CH3), 14.1 (–CH3). 2.3.2. TG-(SMmt)-NHMmt 8 DG-NHMmt 6 (200 mg, 0.24 mmol), EDAC (119 mg, 2.5 eq), DMAP (6 mg, 0.2 eq) and FA-SMmt 4 (136 mg, 1 eq) were dissolved in DCM-THF (3:2, 5 mL). The solution was stirred under N2 at 0 C for 1 h and then allowed to warm to RT. TLC (heptane-EtOAc 3:1) after 18 h showed almost full conversion with Rf (8) = 0.5. The reaction mixture was evaporated in vacuo and purified by flash chromatography, eluting with heptane-EtOAc 7:1. This yielded 279 mg (87%) pure target product. 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 4H, Mmt), 7.44–7.38 (m, 6H, Mmt), 7.35–7.16 (m, 14H, Mmt), 6.83 (d, 2 2H, Mmt CH next to –OMe), 5.31–5.26 (m, 1H, H-2), 4.31 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.81 (s, 3H, –OMe), 3.80 (s, 3H, –OMe), 2.33 (t, 3 2H, –CH2COO–), 2.18–2.11 (m, 2 + 2H, –CH2NH– + –CH2S–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 173.3 (–COO–), 172.9 (–COO–), 113.1 (Mmt CH next to –OMe), 113.0 (Mmt CH next to – OMe), 70.3 (Mmt quaternary C), 68.9 (C-2), 65.9 (Mmt quaternary C), 62.1 (C-1/3), 55.2 (–OMe), 55.1 (–OMe), 43.6 (–CH2NH–), 34.2 (–CH2COO–), 34.1 (–CH2COO–), 24.9 (–CH2CH2CO–), 24.8 (–CH2CH2CO–), 22.7 (–CH2CH3), 14.1 (–CH3).
2.3.4. TG-(SH)-NHDabcyl 10 TG-(SH)-NH2 9 (103 mg, 0.110 mmol assuming TFA-salt) was dissolved in DMF and added Dabcyl-NHS (53 mg, 1.1 eq) and triethylamine (83 mL, 5 eq). The mixture was stirred for 2 h at 40 C. TLC (heptane-EtOAc 3:2) then showed product 10 with Rf = 0.51, an unidentified byproduct with Rf = 0.41, and excess Dabcyl-NHS with Rf = 0.21. The mixture is purified by flash chromatography, eluting with a gradient of heptane-EtOAc, 5:2–4:2. Evaporation of fractions yielded 60 mg (51%) of the desired product 10 as a red solid material. 1 H NMR (CDCl3, selected signals in ppm): 7.94–7.86 (m, 6H, dabcyl), 6.78 (d, 2H, dabcyl), 6.21 (t, 1H, –CH2NHCO–), 5.32–5.25 (m, 1H, H-2), 4.31 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.49 (q, 2H, – CH2NHCO–), 3.13 (s, 6H, –NMe2 dabcyl), 2.54 (q, 2H, –CH2SH), 2.37– 2.29 (m, 3 2H, –CH2COO–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 68.9 (C-2), 62.8 (C-1/ 3), 40.3 (–NMe2 dabcyl), 40.2 (–CH2–NHCO–), 34.0 (–CH2COO–), 24.6 (–CH2SH), 14.0 (–CH3). 2.3.5. TG-ED 11 TG-(SH)-NHDabcyl 10 (56 mg, 0.052 mmol) was suspended in DMF-EtOAc 3:1 (2 mL). EDANS C2 maleimide (18 mg, 1 eq) and triethylamine (25 mL, 3 eq) was added. The mixture was stirred 1 h at RT and then 1 h at 30 C. TLC (DCM-MeOH 20:3) then showed TG-ED 11 with Rf = 0.31. The mixture was purified by flash chromatography eluting with a gradient going from 3% to 10% MeOH in DCM. The product containing fraction were pooled and evaporated to yield 55 mg (74%) of the desired TG-ED 11 as an orange-red solid. 1 H NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 8.17 (dd, 2H, edans), 7.92-7.79 (m, 1 + 6H, edans + dabcyl), 7.44–7.34 (m, 2H, edans), 6.74 (d, 2H, dabcyl), 6.71 (t, 1H, –CH2NHCO–), 6.57 (d, 1H, edans), 5.26–5.20 (m, 1H, H-2), 4.27 (dd, 2H, H-1/3), 4.12 (dd, 2H, H-1/3), 4.00-3.90 (m, 2H, –CH2CH2NH-edans), 3.72 (dd, 1H, –S– CH– maleimide), 3.48–3.33 (m, 2 + 2H, –CH2NHCO– + –CH2–NHedans), 3.08 (s, 6H, –NMe2 dabcyl), 2.53 (dd, 2H, –CH2S–), 2.29 (t, 3 2H, –CH2COO–), 0.85 (t, 3H, –CH3). 13 C NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 68.9 (C2), 62.1 (C-1/3), 45.9 (–S–CHCH2– maleimide), 42.9 (–CH2–NHedans), 40.0 (–CH2–NHCO–), 40.0 (–NMe2 dabcyl), 39.3 (–S–CH– maleimide), 38.4 (–CH2CH2NH-edans), 36.2 (–CH2S–), 33.9 (–CH2COO–), 14.0 (–CH3). MALDI-MS (exact mass = 1422.86 Da): 1423.82 [M + H]+, 1445.76 [M + Na]+, 1461.75 [M + K]+.
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
2.3.6. TG-FD 12 TG-(SH)-NHDabcyl 10 (55 mg, 0.052 mmol) in DMF-DCM (1:1, 2 mL) was added N-(5-fluoresceinyl)maleimide (25 mg, 1.1 eq) and triethylamine (25 mL, 3 eq). The mixture was stirred at 30 C for 2 h. TLC (DCM-MeOH 9:1) then showed essentially full conversion with product Rf = 0.16. Purification by flash chromatography using a gradient from 7% MeOH to 15% MeOH in DCM gave pure TG-FD 12 as an orange solid. Yield 37 mg (47%). 1 H NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 7.99– 7.96 (m, 2H, fluorescein), 7.88–7.83 (m, 6H, dabcyl), 7.60 (dd, 1H, fluorescein), 7.26 (d, 1H, fluorescein), 6.75 (d, 2H, dabcyl), 6.70– 6.61 (m, 3 + 1H fluorescein + –CH2NHCO–), 6.53 (dd, 2H, fluorescein), 5.28–5.21 (m, 1H, H-2), 4.28 (dd, 2H, H-1/3), 4.13 (dd, 2H, H-1/3), 3.92 (dd, 1H, –S–CH– maleimide), 3.46–3.33 (m, 2 + 2H, – CH2NHCO– + –S–CHCH2– maleimide), 3.10 (s, 2 3H, –NMe2 dabcyl), 2.72 (dd, 2H, –CH2S–), 2.30 (t, 3 2H, –CH2COO–), 0.86 (t, 3H, –CH3). 13 C NMR (CDCl3 + 5 dr CD3OD, selected signals in ppm): 68.9 (C2), 61.8 (C-1/3), 40.5 (–NMe2 dabcyl), 40.2 (–CH2–NHCO–), 38.9 (–S–CH– maleimide), 36.1 (–CH2S–), 35.9 (–S–CHCH2– maleimide), 34.0 (–CH2COO–), 13.9 (–CH3). MALDI-MS (exact mass = 1503.86 Da): 1504.95 [M + H]+, 1526.94 [M + Na]+, 1542.91 [M + K]+. 2.4. Synthesis of TG-F2 14 2.4.1. TG-(NHMmt)-NHMmt 7 TG-(NHMmt)-NHMmt 7 was isolated as a byproduct from the synthesis of DG-NHMmt 6 described above. Yield 1.29 g (9%). 1 H NMR (CDCl3, selected signals in ppm): 7.49 (d, 8H, Mmt), 7.40 (d, 2H, Mmt), 7.31–7.25 (m, 10H, Mmt), 7.22–7.16 (m, 4H, Mmt), 6.83 (d, 2 2H, Mmt CH next to –OMe), 5.32–5.24 (m, 1H, H-2), 4.32 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.81 (s, 2 3H, –OMe), 2.33 (t, 3 2H, –CH2COO–), 2.13 (t, 2 2H, –CH2NH–), 1.68–1.58 (m, 3 2H, –CH2CH2COO–), 1.52–1.42 (m, 2 2H, –NHCH2CH2–), 0.91 (t, 3H, –CH3). 13 C NMR (CDCl3, selected signals in ppm): 113.1 (Mmt CH next to –OMe), 69.0 (C-2), 62.0 (C-1/3), 55.2 (–OMe), 43.5 (–CH2NH–), 33.7 (–CH2COO–), 30.8 (–CH2CH2NH–), 24.8 (–CH2CH2COO–), 14.0 (–CH3). 2.4.2. TG-(NH2)-NH2 13 TG-(NHMmt)-NHMmt 7 (200 mg, 0.154 mmol) was dissolved in DCM (6 mL) at 0 C. Then added TFA (0.6 mL) and triethylsilane (0.3 mL, 12 eq). The dark yellow solution was stirred for 30 min at 0 C and then 2 h at RT. TLC (heptane-EtOAc 3:1) then showed full deprotection with Rf (7) = 0.42 and Rf (13) = 0. The reaction mixture was evaporated and used in the next step without purification.
75
The mixture was stirred at 40 C for 3 h, after which TLC (DCMMeOH 9:1) showed good conversion with Rf (14) = 0.46. Purification by flash chromatography, eluting with a gradient going from 5% to 50% MeOH in DCM, resulted in 44 mg (20%) pure TG-F2 as an orange solid, as well as a larger fraction of product needing further purification. 1 H NMR (CD3OD, selected signals in ppm): 8.50–6.50 (18H, fluorescein signals), 5.32–5.23 (m, 1H, H-2), 4.37 (dd, 2H, H-1/3), 4.17 (dd, 2H, H-1/3), 3.44 (dt, 2H, –CH2NHCO–), 3.31 (dt, 2H, –CH2NHCO–), 2.33 (t, 3 2H, –CH2COO–), 0.90 (t, 3H, –CH3). 13 C NMR (CD3OD, selected signals in ppm): 69.4 (C-2), 61.9 (C-1/ 3), 39.6 (–CH2–NHCO–), 33.5 (–CH2COO–), 13.3 (–CH3). MALDI-MS (exact mass = 1468.76 Da): 1469.81 [M + H]+, 1491.81 [M + Na]+, 1507.79 [M + K]+. 2.5. Assays Substrate stock solutions (5–10 mM) were made in DMSO. A buffer was made consisting of 50 mM Tris, 2 mM CaCl2, 0.1 mg/mL SB3-14 surfactant, 1 mg/mL BSA (FFA-free), and adjusted to pH 8.0 with HCl. “Assay substrate solutions” were then made by rapidly stirring the buffer while slowly adding the DMSO-stock solution until a final concentration of 800 mM substrate. The assay was performed in black 96-well plates using 40 mL buffer, 50 mL diluted substrate (800 mM to 6.25 mM in a 2-fold dilution series), and 10 mL lipase (Lipolase 100 L diluted 50 x with water). The buffer/substrate mixtures were equilibrated at room temperature for 1 h before the final addition of lipase. Then fluorescence was recorded every 30 s for a total of 15 min. The spectrometer was set for excitation at 350 nm and emission at 460 nm (for TG-ED), or excitation at 492 nm and emission at 517 nm (for TG-FD and TG-F2). When the commercial EnzChek lipase substrate was used in the assay, excitation and emission wavelengths were adjusted to 482 and 515 nm, respectively. From the reaction profiles showing relative fluorescence units (RFU) vs time, the initial velocity v0 was determined by linear regression over the first 5 min. Standard curves, showing RFU vs concentration of fully hydrolysed substrate was constructed from the same plates as used for the kinetic measurements after several days incubation at room temperature. Plotting v0 against substrate concentration allowed determination of vmax/KM as the slope of the initial linear part (up to 25 mM substrate). Dividing with the lipase in-assay concentration (1.3 mM) gave the kcat/KM. The kinetic measurements were done in triplicates with results reported as the average plus/minus the standard deviation. 3. Results and discussion 3.1. Synthesis
2.4.3. TG-F2 14 The crude TG-(NH2)-NH2 13 (0.154 mmol) was dissolved in DMF (3 mL) to which 5-carboxyfluorescein NHS-ester (100 mg, 1.4 eq) and triethylamine (245 mL, 10 eq) was added. Using 1.4 eq 5-carboxyfluorescein was a mistake – it should have been 2 eq.
The lipase substrates were designed to resemble natural triglycerides as much as possible. To allow for a flexible approach, the triglyceride skeleton was first assembled, to which fluorophores and quenchers were added using chemoselective
Scheme 1. Protection of functionalized fatty acids.
76
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
Scheme 2. Synthesis of DG-NHMmt 6 and TG-(NHMmt)-NHMmt 7.
Scheme 3. Synthesis of TG-(SH)-NHDabcyl 10.
reactions. Hence, a key intermediate in the synthesis is the functionalized triglyceride TG-(SH)-NH2 9 (Scheme 3), displaying the two handles, an amine and a thiol. A range of different fluorophores and quenchers are commercially available as NHSesters and maleimides, allowing for orthogonal couplings to this scaffold. TG-(SH)-NH2 9 was, as illustrated in Schemes 1–3, assembled from commercially available 1-stearoyl-rac-glycerol 5 and Mmtprotected amino- (FA-NHMmt 2) and mercapto fatty acids (FASMmt 4). The initial coupling of FA-NHMmt 2 yielded a mixture of di- and triglycerides, but with the desired diglyceride (DG-NHMmt 6) as the main product. The byproduct, TG-(NHMmt)-NHMmt 7, was later used for synthesis of TG-F2 (Scheme 5). It was initially attempted to couple unprotected FA-SH 3 to diglyceride 6. However, this resulted in a complex mixture. It was therefore decided to protect the thiol functionality like the amino group in FA-NH2 with a Mmt protecting group. The 4-methoxytrityl (Mmt) was chosen over the standard trityl protection group since it is significantly more acid labile, which
could be important when deprotecting the triglyceride product. The conditions chosen for protecting FA-SH 3 was inspired by a procedure described by Mourtas and coworkers (Mourtas et al., 2001), whereas protection of FA-NH2 1 followed a procedure by Barlos and coworkers for trityl protection of amino acids, in which the carboxylic acid is temporarily protected as a TMS-ester (Barlos et al., 1982). The EDAC conditions chosen for the coupling reactions to form the di- and triglycerides was inspired by a procedure described by Whitten and coworkers (Whitten et al., 2012). Mild acidic conditions (1–2% TFA in DCM) using triethylsilane as scavenger for the Mmt cation resulted in fast and clean deprotection to key intermediate 9. Using similar deprotecting conditions on diglyceride 6 had resulted in 10–20% acyl migration (estimated by NMR, data not shown). This issue was avoided with the chosen strategy in which the triglyceride is assembled before deprotection. As for the choice of FRET-pairs, Edans-Dabcyl (TG-ED) and Fluorescein-Dabcyl (TG-FD) were chosen since the probes are available and relatively inexpensive in synthetic quantities and
Scheme 4. Synthesis of TG-ED 11 and TG-FD 12.
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
they are both well-described donor-quencher pairs. Both Edans and Fluorescein have emission spectra nicely overlapping the absorption spectrum of the Dabcyl dark quencher. Fluorescein has in contrast to Edans overlapping excitation and emission spectra, meaning it has self-quenching properties (homo-FRET) (Tramier et al., 2003). This is the basis for the design of TG-F2. The probes are attached to the triglyceride scaffold via chemoselective reactions, i.e. coupling of NHS-esters to the primary amine and addition of maleimides to the thiol. The couplings can either be performed sequentially with intermediate purification or one-pot (data not reported). Isolated yields (nonoptimized) were generally in the range of 50–80%. TG-ED 11, TG-FD 12 and TG-F2 14 were hence prepared in 40–50 mg scale and isolated as pure orange/yellow solids (Schemes 4 and 5). 3.2. Assays To obtain proof-of-concept for TG-ED, TG-FD and TG-F2 in lipase assays, reaction conditions close to that reported by Basu and coworkers (Basu et al., 2011) for the commercial EnzChek lipase substrate was chosen. The authors report that the use of Zwittergent surfactant is critical, whereas alternatives such as Triton X-100 would lead to complete dequenching of the substrate. It is however not specified exactly which type Zwittergent is used. After a few attempts, we found a buffer system consisting of 50 mM Tris, 2 mM CaCl2, 0.1 mg/mL SB3-14 surfactant (Zwittergent), 1 mg/mL BSA (FFA-free), pH 8.0 to work satisfactorily. The Thermomyces lanuginosus lipase (TLL) was found to be a natural choice for the assay development work. The enzyme is a classical interfacially activated lipase, it is very well studied and furthermore an industry standard (Fernandez-Lafuente, 2010). It is commercialized by Novozymes under the tradename Lipolase1 100 L for the detergent industry. It is also distributed by Sigma (cat. no. L0777). A 500 x dilution (in-assay concentration) of Lipolase 100 L gave suitable reaction rates. The substrates were readily dissolved in DMSO to make 5–10 mM stock solution. Good reproducibility was obtained by making “assay substrate solutions” from rapidly stirring the buffer while slowly adding the DMSO-stock solution until a final concentration of 800 mM substrate. After final substrate dilutions in the microtiter plate, the plate was left to equilibrate 1 h at room temperature in the dark. This step was essential for obtaining flat baselines of blank samples (no enzyme), presumably by giving the substrate time to be distributed in the surfactant micelles. There was no sign of turbidity or precipitation of substrate under any conditions. It appears that the fluorophores significantly increases solubility of the triglycerides. Having a water-soluble
77
Table 1 Spectroscopic and kinetic results. Substrate
TG-ED
Ex/em wavelengths (nm) 350/460 Standard curve slope (RFU/mM) 34 3 1st order kinetics (up to 25 mM) 1 1 kcat/KM (s M ) 460 24
TG-FD
TG-F2
492/517
492/517
181 33
4655 875
59 9
346 15
substrate clearly simplifies assay development. However, the interfacial activation then has to be provided by the surfactant system. Standard curves, showing fluorescence response (RFU) vs substrate/product concentration, were constructed from a substrate dilution series, incubated with lipase to full hydrolysis. This was also used to find optimal excitation and emission wavelength (Table 1). Not surprisingly, standard curves were not linear over the entire substrate range, but only up to approx. 25 mM, after which the curve flattened out, presumably due to intermolecular quenching at higher concentrations. Kinetic studies were constructed initially aiming towards finding kcat and KM values. However, that proved impossible as KM-values for all three substrates were out of the linear range of the standard curve. Indeed, measured initial reaction rates seemed to go through a maximum and appeared very low at high substrate concentrations (Fig. 2). Plotting initial reaction rates versus substrate concentration in the linear range allowed determination of kcat/KM from the slope (1st order reaction kinetics, Table 1). All three substrates showed essentially flat backgrounds (Fig. 2), once the substrate was fully equilibrated in the micelles. Further, all three showed a good window between fully hydrolysed and non-hydrolyzed substrate, up to a factor 15 for TG-FD and 3–4 for TG-ED and TG-F2 at 25 mM concentration. Looking at the RFU scale (of the standard curve or the reaction profile), TG-F2 is clearly much more fluorescent that TG-ED and TG-FD, due to the very efficient Dabcyl-quenching vs the homo-FRET quenched TG-F2. In applications were a high-fluorescent background can allowed, TG-F2 will have the advantage of being prepared in fewer synthetic steps. The 1st order kinetic values for the three substrates do not appear dramatically different, although with TG-FD being more slowly processed by the lipase (Table 1). The substrates were designed with the fluorescent probes at the end of C12 or longer fatty acids, which is expected to be outside the lipase active site. Any differences between the substrates are therefore not likely a direct substrate/enzyme fit effect, but more a question of how the
Scheme 5. Synthesis of TG-F2 14.
78
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79
Fig. 2. Raw kinetics data for 25 mM substrate, with and without enzyme (left) and plots of v0 vs substrate concentration (right) for TG-ED (top), TG-FD (middle) and TG-F2 (bottom).
substrate is presented to the enzyme, the orientation or distribution in the micelles. For comparison, the commercial EnzChek lipase substrate was assayed using the same conditions. This resulted in kcat/KM = 204 s1 M1. Also the commercial substrate showed lower measured reaction rates at high substrate concentrations. Again this rate constant is in the same range as those found for the triglyceride structures, indicating that the ether-linked C6-alkyl chain in EnzChek (Fig. 1) does not dramatically effect Lipolase activity. 4. Conclusion Three fluorogenic lipase substrates were synthesized using chemoselective couplings to a racemic triglyceride scaffold. The amino- and thiol-functionalized triglyceride was assembled using standard carbodiimide couplings of Mmt-protected amino- or thiol fatty acids to the glycerol backbone. Following deprotection under mildly acidic conditions, quenchers and fluorophores were added in high-yielding, click-like (Kolb et al., 2001) reactions of NHSesters or maleimides. The resulting Dabcyl-quenched substrates, TG-ED and TG-FD and the fluorescein self-quenched TG-F2, were subsequently evaluated in proof-of-concept assays with the commercially available Thermomyces lanuginosus lipase Lipolase 100L. This showed rewarding lipase responses with flat blanks, yielding up
to a factor of 15 x difference in fluorescence signal between hydrolysed and intact substrate (TG-FD at 25 mM concentration). Varying the substrate concentrations revealed this as a critical parameter, since high substrate concentrations resulted in lower measured reaction rates, presumably due to intermolecular quenching. Conditions for the proof-of-concept assays are largely unoptimized, but inspired by those reported by Basu and coworkers (Basu et al., 2011). It is therefore likely that assay performance can be further improved and/or conditions adapted to mimic a specific biological system or technical enzyme application. In conclusion the results illustrate the applicability of fluorogenic lipase substrates that structurally are closely related to natural triglycerides, and that these are easily prepared in mg and gram scale. Acknowledgements Mr Clive Phipps Walter, Novozymes, is acknowledged for running MALDI-MS analyses of the three substrates. References Barlos, K., Papaioannou, D., Theodoropoulos, D., 1982. Efficient one-pot synthesis of N-trityl amino-acids. J. Org. Chem. 47, 1324–1326.
R.J. Andersen, J. Brask / Chemistry and Physics of Lipids 198 (2016) 72–79 Basu, D., Manjur, J., Jin, W.J., 2011. Determination of lipoprotein lipase activity using a novel fluorescent lipase assay. J. Lipid Res. 52, 826–832. Duque, M., Graupner, M., Stutz, H., Wicher, I., Zechner, R., Paltauf, F., Hermetter, A., 1996. New fluorogenic triacylglycerol analogs as substrates for the determination and chiral discrimination of lipase activities. J. Lipid Res. 37, 868–876. Fernandez-Lafuente, R., 2010. Lipase from Thermomyces lanuginosus: uses and prospects as an industrial biocatalyst. J. Mol. Catal. B Enzym. 62, 197–212. Ghosh, S.S., Eis, P.S., Blumeyer, K., Fearon, K., Millar, D.P., 1994. Real-time kinetics of restriction-endonuclease cleavage monitored by fluorescence resonance energy-transfer. Nucleic Acids Res. 22, 3155–3159. Hasan, F., Shah, A.A., Hameed, A., 2006. Industrial applications of microbial lipases. Enzyme Microb. Tech. 39, 235–251. Hasan, F., Shah, A.A., Hameed, A., 2009. Methods for detection and characterization of lipases: a comprehensive review. Biotechnol. Adv. 27, 782–798. Kolb, H.C., Finn, M.G., Sharpless, K.B., 2001. Click chemistry: diverse chemical function from a few good reactions. Angew. Chem. Int. Ed. 40, 2004–2021. Mitnaul, L.J., Tian, J., Burton, C., Lam, M.H., Zhu, Y.P., Olson, S.H., Schneeweis, J.E., Zuck, P., Pandit, S., Anderson, M., Maletic, M.M., Waddell, S.T., Wright, S.D., Sparrow, C.P., Lund, E.G., 2007. Fluorogenic substrates for high-throughput measurements of endothelial lipase activity. J. Lipid Res. 48, 472–482. Mourtas, S., Gatos, D., Kalaitzi, V., Katakalou, C., Barlos, K., 2001. S-4-Methoxytrityl mercapto acids: synthesis and application. Tetrahedron Lett. 42, 6965–6967. Mueller, S., Galliardt, H., Schneider, J., Barisas, B.G., Seidel, T., 2013. Quantification of forster resonance energy transfer by monitoring sensitized emission in living plant cells. Front. Plant Sci. 4, 1–20. Oka, H., Koyama, T., Hatano, K., Matsuoka, K., 2012. Synthetic studies of bifluorescence-labeled maltooligosaccharides as substrates for alpha-amylase on
79
the basis of fluorescence resonance energy transfer (FRET). Bioorg. Med. Chem. 20, 435–445. Schmidt, M., Bornscheuer, U.T., 2005. High-throughput assays for lipases and esterases. Biomol. Eng. 22, 51–56. Singh, A.K., Mukhopadhyay, M., 2012. Overview of fungal lipase: a review. Appl. Biochem. Biotech. 166, 486–520. Stoytcheva, M., Montero, G., Zlatev, R., Leon, J.A., Gochev, V., 2012. Analytical methods for lipases activity determination: a review. Curr. Anal. Chem. 8, 400–407. Svendsen, A., 2000. Lipase protein engineering. Biochim. Biophys. Acta 1543, 223–238. Thompson, V.F., Saldana, S., Cong, J.Y., Goll, D.E., 2000. A bodipy fluorescent microplate assay for measuring activity of calpains and other proteases. Anal. Biochem. 279, 170–178. Tramier, M., Piolot, T., Gautier, I., Mignotte, V., Coppey, J., Kemnitz, K., Durieux, C., Coppey-Moisan, M., 2003. Homo-fret versus hetero-fret to probe homodimers in living cells. Method. Enzym. 360, 580–597. Wang, G.T., Matayoshi, E., Huffaker, H.J., Krafft, G.A., 1990. Design and synthesis of new fluorogenic hiv protease substrates based on resonance energy-transfer. Tetrahedron Lett. 31, 6493–6496. Whitten, K.M., Makriyannis, A., Vadivel, S.K., 2012. Application of chemoenzymatic hydrolysis in the synthesis of 2-monoacylglycerols. Tetrahedron 68, 5422–5428. Wichmann, O., Gelb, M.H., Schultz, C., 2007. Probing phospholipase A(2) with fluorescent phospholipid substrates. Chembiochem 8, 1555–1569. Yang, Y.Z., Babiak, P., Reymond, J.L., 2006. Low background fret-substrates for lipases and esterases suitable for high-throughput screening under basic (pH 11) conditions. Org. Biomol. Chem. 4, 1746–1754.