Characterization of an epoxide hydrolase from the Florida red tide dinoflagellate, Karenia brevis Author Sun, Pengfei Department of Chemistry and Biochemistry, Florida International University, Miami, FL 33199, USA Author Leeson, Cristian Author Zhi, Xiaoduo Author Leng, Fenfei Author Pierce, Richard H. Author Henry, Michael S. Author Rein, Kathleen S. text Phytochemistry 2016 2016-02-29 122 11 21 http://dx.doi.org/10.1016/j.phytochem.2015.11.002 journal article 285216 10.1016/j.phytochem.2015.11.002 18ecad6c-abce-4590-9b84-edd0e9b30119 1873-3700 PMC4724521 26626160 10485359 2.3. Characterization of K. brevis EH Fluorescent probe sets 3b / 5b and 3a / 5a were used for initial characterization of the expressed K. brevis EH ( Fig. 3C ). The EH was optimally active at pH 6.8–7.1, with half-maximal activity at pH 6.5 and 8.3. Consistent with the selectivity observed in the K. brevis CFE, the purified EH showed a preference for epoxide 5b over 5a . The ratio of initial rates ( V o ) for substrates 5a : 5b was 1:4. Epoxide hydrolases of the a,b- hydrolase type share significant homologies with other hydrolases, including esterases. As done with the CFE, EH activity was distinguished from esterase activity using the substrates 3a and 5a . No activity was observed using the esterase probes indicating that the expressed protein was indeed an EH. Substrate and enantioselectivity of the K. brevis EH was assessed by kinetic studies. Kinetic parameters of the enzyme using substrates 5a and 5b were performed using the sensitive fluorescent assay. Other substrates having no fluorescent reporter group were evaluated by GC–MS analysis of the reaction mixtures using 1- naphthol as an internal standard. The data shown in Table 1 indicates that the K. brevis EH has a strong preference for terminal epoxides over internal epoxides (Entries 3–9 vs. Entries 1–2). Aliphatic substituents are preferred over aromatic substituents by approximately 5:1 (Entries 7–9 vs. Entries 3–6) and the enzyme shows only a slight preference for R over S enantiomers (Entries 3 and 4). The reaction rates using 1,2-disubstituted epoxides as substrates are too slow to be determined using the GC–MS method. However, disubstituted epoxides proved to be inhibitors of the K. brevis EH only when an aromatic substituent is present (Entries 10–12 vs. Entries 13–14) and cis -disubstituted epoxides are more potent inhibitors than those that are trans -disubstituted (Entries 10–12). A noteworthy distinction between the monensin ( 2 ) pathway and the proposed brevetoxin ( 1 ) pathway is the site of nucleophilic attack on the epoxide substrate ( Scheme 1 ). In the case of non-ladder polyethers, the majority of epoxide ring opening reactions proceed via attack at the proximal carbon ( exo-tet cyclization) while polyether ladder formation requires attack at the distal carbon ( endo-tet cyclization). Both the empirical Baldwin’s rules ( Baldwin, 1976 ) and molecular orbital calculations ( Gruber et al., 1999 ) indicate that the endo-tet cyclization pathway is kinetically disfavored. The exception among the bacterial polyketides is lasalocid. In this polyether, the pyran ring arises from the endo-tet cyclization of an intermediate epoxide. This reaction is catalyzed by the EH Lsd19. Lsd19 is a bifunctional enzyme consisting of two domains (A and B). Each domain is responsible for the cyclization of one of the rings with Lsd19B catalyzing the endo-tet cyclization of the terminal ring ( Minami et al., 2011 ). Fig. 3. (A) Alignment of putative K. brevis EH 1 with Aspergillus niger EH. Catalytic residues are indicated by .. (B) SDS PAGE of expressed protein after purification and dialysis. Lanes M : Precision Plus Protein Unstained Standards (Biorad 161-0363) lane 1: lysate; lanes 2–3: column flow-through; lanes 4–5: column washes; lane 6: elution. C. EH ( 5b ) and esterase ( 3b ) assay of purified K. brevis EH. Table 1 Kinetic parameters of the K. brevis EH with various substrates and inhibitors.

Entry	Substrate	K m (LM)	k (s –1) cat	k (s –1)/ K (M) cat m	IC50 (mM)
1	Probe 5a (R = Ph)	13.6	7.81E–4	57.4	–
2	Probe 5b (R = Et)	161	4.46E–3	27.7	–
3	( R ) styrene oxide	6240	9.01E–1	144	–
4	( S ) styrene oxide	6430	8.32E–1	129	–
5	( R/S ) styrene oxide	6570	8.58E–1	131	–
6	a- Methylstyrene oxide	5310	3.31E–1	62.3	–
7	( R ) 1-butane oxide	25,800	6.15	238	–
8	( R / S ) 1-butane oxide	22,200	5.36	241	–
9	1,2,7,8-Diepoxyoctane	4,800	7.12	1483	–
10	cis -b- Methylstyrene oxide	–	–	–	3.05
11	trans -b- Methylstyrene oxide-	–	–	–	14.28
12	b- Dimethylstyrene oxide	–	–	–	6.99
13	cis -2,3-Epoxybutane	–	–	–	–
14	2,3-Dimethyl-2,3-epoxybutane	–	–	–	–

Scheme 4. Regioisomeric cyclizations of trans -4,5-epoxy-hexanol ( 10 ). The ability of the K. brevis EH to perform cyclization was also evaluated. The substrate trans -4,5-epoxy-hexanol 9 was prepared as previously described ( Coxon et al., 1973 ).This substrate may undergo endo-tet cyclization to produce pyran 10 , or exo-tet cyclization to produce furan 11 ( Scheme 4 ).The substrate was incubated with the expressed EH in reaction buffer and the reaction was monitored by GC–MS. Both possible products were identified by GC–MS, by comparison with authentic samples however, neither the percent conversion nor the product ratio differed from that of the buffer alone.