Bioluminescence activity of Latia luciferin analogs

Article abstract of DOI:10.1016/S0040-4039(00)00639-0

Latia luciferin analogs were synthesized and their bioluminescence activities were measured. The Latia luciferase was found to recognize strictly the 2,6,6-trimethylcyclohexene ring moiety in the luciferin structure. While the enol ether analogs exhibited no bioluminescence activity, the corresponding enol acetate analog possessed 60% activity compared to natural luciferin having an enol formate structure, implying that the initial step of the light producing reaction is an enzymatic hydrolysis to yield the corresponding enolate anion. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00639-0

Tetrahedron Letters 41 (2000) 4409±4413
Bioluminescence activity of Latia luciferin analogs
Satoshi Kojima,a Shojiro Maki,a Takashi Hirano,a Yoshihiro Ohmiyab and
Haruki Niwaa,*
aDepartment of Applied Physics and Chemistry, University of Electro-communications, Chofu, Tokyo 182-8585, Japan
bDepartment of Biochemistry, Faculty of Education, Shizuoka University, Shizuoka 422-8529, Japan
Received 6 March 2000; accepted 14 April 2000
Abstract
Latia luciferin analogs were synthesized and their bioluminescence activities were measured. The Latia
luciferase was found to recognize strictly the 2,6,6-trimethylcyclohexene ring moiety in the luciferin structure.
While the enol ether analogs exhibited no bioluminescence activity, the corresponding enol acetate analog
possessed 60% activity compared to natural luciferin having an enol formate structure, implying that the
initial step of the light producing reaction is an enzymatic hydrolysis to yield the corresponding enolate
anion. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Latia neritoides; luciferin; luciferase; bioluminescence; enol formate; enol acetate.
The limpet-like snail Latia neritoides is only found in clear and shallow streams of North Island
in New Zealand, and is the only luminous animal which lives in fresh water. On mechanical
stimulation, the snails secrete mucus, from which greenish light (536 nm) is emitted.1 The Latia
bioluminescence system exhibits a typical luciferin±luciferase reaction.2,3 Shimomura established
the structure of Latia luciferin (1) having a characteristic, enol formate functionality, and proposed
the Latia bioluminescence reaction as shown in Scheme 1.2 4 Thus, the Latia bioluminescence
requires four components: the luciferin (1), the luciferase (178 000 Da), a cofactor so-called
`purple protein (38 000 Da)' (red-¯uorescent protein), and molecular oxygen. During the bio-
luminescence reaction the Latia luciferin (1) was oxidized into Latia oxyluciferin (2), CO2, and
formic acid with an emitting the green light. Shimomura suggested that light emission should
arise from a ¯avin compound covalently bound to the luciferase because none of 1, 2 and `purple
protein' have green ¯uorescent properties.3 However, the precise bio-oxidation mechanism of 1
and the detailed molecular bases for the light-emitting process are still unclear. In this paper, we wish
to report the bioluminescence activity of the Latia luciferin analogs and the substrate speci®city
of Latia luciferase.
* Corresponding author. Fax +81-424-86-1966; e-mail: niwa@pc.uec.ac.jp
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00639-0

4410
Scheme 1. Bioluminescence reaction of Latia neritoides
Scheme 2 shows the preparation of Latia luciferin (1) and the analogs 6±9, 12 and 13 studied.8
Latia luciferin (1) was prepared by the slightly modi®ed procedure reported by Kishi.5
Thus, selective hydrogenation of b-ionone (3) with (PPh3)3RhCl6 followed by Horner±Emmons
ole®nation of the resulting ketone 2 (Latia oxy-luciferin) a€orded a,b-unsaturated esters E-4
(overall 76%) and Z-4 (overall 9%) after separation. Reduction of E-4 with DIBAH followed by the
oxidation of the resulting allylic alcohol with MnO2 gave the desired a,b-unsaturated aldehyde E-6
(overall 95%). Baeyer±Villiger oxidation of E-6 was achieved by using 60% hydrogen peroxide in
t-amylalcohol in the presence of selenium dioxide to give Latia luciferin (1) (45%). The corre-
sponding geometrical isomer Z-1 was prepared via Z-4 by a similar procedure (overall 24%).
Scheme 2. Preparation of Latia luciferin (1) and its analogs. Reagents and conditions: (a) (PPh3)3RhCl, H2, benzene, rt,
14 h, 100%; (b) (EtO)2P(O)CH2CO2Et, LiHMDS, THF, re¯ux, 5 h, 76% for E-4, 9% for Z-4 after separation; (c)
DIBAH, toluene, 0ꢀC, 100% for E-5, 100% for Z-5; (d) MnO2, CH2Cl2, rt, 2 h, 95% for E-6, 95% for Z-6; (e) SeO2,
60% H2O2, t-amylalcohol, 0ꢀC, 45% for 1, 25% for Z-1; (f) MeOCH2P+(Ph3)Cl^, LiHMDS, THF, 0ꢀC, 4 h, 95%
(E:Z=1:1); (g) p-TsOH, 3:1 acetone:H2O, re¯ux, 4 h, 100%; (h) p-TsOH, isopropenylacetate, re¯ux, 12 h, 42% for E-9,
21% for Z-9

4411
Analogs 7, 8 and 9 were synthesized from ketone 2 (Scheme 2). Wittig reaction of 2 with
methoxymethylidenetriphenylphosphorane gave enol ether 7 as a 1:1 separable mixture of
E:Z-isomers, which, upon acidic hydrolysis, a€orded the corresponding aldehyde 8 (100%).
Treatment of 8 with isopropenyl acetate-p-toluenesulfonic acid7 gave the desired enol acetates E-9
(45%) and Z-9 (21%). By using quite similar sequences, analogs 11 (a racemic, 1:1 mixture of cis
and trans ring-isomers) and 13 were prepared starting with 10 (a racemic, 1:1 mixture of cis and
trans ring-isomers) obtained by hydrogenation of 3 and 12, respectively (Scheme 3).
Scheme 3.
A crude Latia luciferase solution used for the bioluminescence measurement was prepared as
follows: Specimens of frozen Latia were homogenized in 50 mM Tris±HCl bu€er (pH 7.2) at 0ꢀC
and the homogenized mixture was centrifuged at 7000 rpm for 20 min at 4ꢀC. The supernatant
was subjected to fractional ammonium sulfate precipitation. The active precipitates obtained by
33±60% saturation were collected, resuspended in a 50 mM Tris±HCl bu€er (pH 7.2), and
subjected to gel ®ltration using Superdex 200 with the same bu€er, yielding a crude luciferase
solution containing `purple protein'. The protein concentration of the luciferase solution was
determined by Protein Assay Kit1 (Bio Rad) to be 0.14 mg/ml.
The bioluminescent activity was measured by mixing 100 ml of the luciferase solution with 300 ml
of Latia luciferin (1) or the analog solution in water (an 8Â10^3 M EtOH solution was diluted
with deionized H2O to the ®nal concentration of 8Â10^6 M). The photons generated were counted
by a luminometer Lumi Counter 1000 (Niti-on) for 10 min in each measurement. The results are
summarized in Table 1. The cyclohexane analog 11 and the phenyl analog 13 had essentially no
bioluminescence activity. These results suggest that the 2,6,6-trimethylcyclohexene ring moiety
may be recognized strictly by the luciferase, and essential for binding to the active site.
The Z-1 regioisomer of Latia luciferin (1) had 60% activity of 1, as reported by Kishi.5 This
result indicates that the recognition site of the enol ester moiety in the luciferase has a certain
¯exibility.
a,b-Unsaturated aldehyde E-6 is supposed to be a biosynthetic precursor of Latia luciferin (1)
because a certain ¯avin-containing oxygenase is known to catalyze Baeyer±Villiger-type oxidation
of ketones and aldehydes.9 However, both E-6 and Z-6 had no bioluminescence activity, indicating
that Latia luciferase has no such activity. We expected to ®nd that if the bioluminescence reaction
starts with an electron transfer reaction from the substrate to a ¯avin component such as ¯avin
(hydro)peroxide, enol methyl ether analogs E-7 and Z-7, having a more electron donating double
bond, may act as better substrates. However, both analogs were not active substrates. As mentioned
by Shimomura, aldehyde 8, the hydrolyzed product of 1 also has no bioluminescence activity.

4412
Table 1
Relative bioluminescence activity of Latia luciferin analogs
Surprisingly, enol acetate E-9 showed 66% activity compared to that of natural Latia luciferin (1)
having the enol formate functionality. Furthermore, the regioisomer Z-9 also had 44% activity
compared to 1. These results indicate that the formyl (-CHO) group is not essential functionality.
Thus, the Latia bioluminescence system is completely di€erent from that of bacteria. In addition,
the Latia luciferase has an esterase activity and generation of the enolate anion 14 from the enol
esters in the luciferase environment, but not the enol 15, may be an important step for the bio-
luminescence reaction. Further study using substrate analogs is now in progress.
In summary, Latia luciferase was found to recognize the 2,6,6-trimethylcyclohexene moiety
strictly. The enol acetate functionality is also susceptible to the bioluminescence reaction.
Acknowledgements
S.K. thanks the JSPS (Japan Society for the Promotion of Science) Research Fellowships for
Young Scientists. Financial support from the Ministry of Education, Science, Sports and Culture
(Grant-in-Aid for Scienti®c Research Nos. 09041100 and 1068502), the Naito Foundation, and
the Shorai Foundation is gratefully acknowledged. We are deeply indebted to Dr. Mike R.
Scarsbrook (National Institute of Water and Atmospheric Research Ltd., Hamilton, New Zealand)
for his great assistance in collecting biological specimens.

4413
References
1. Johnson, F. H.; Haneda, Y. Bioluminescence in Progress; Princeton University Press: NJ, 1966.
2. Shimomura, O.; Johnson, F. H. Biochemistry 1968, 7, 1734±1738; 2574±2580.
3. Shimomura, O.; Johnson, F. H.; Kohama, Y. Proc. Nat. Acad. Sci. USA 1972, 69, 2086±2089.
4. Fracheboud, M. G.; Shimomura, O.; Hill, R. K.; Johnson, F. H. Tetrahedron Lett. 1969, 3951±3954.
5. Nakatsubo, F.; Kishi, Y.; Goto, T. Tetrahedron Lett. 1970, 381±384.
6. Sims, J. J.; Honwad, V. K.; Selman, L. H. Tetrahedron Lett. 1969, 87±89.
7. Halsall, T. G.; Theobald, D. W.; Walshaw, K. B. J. Chem. Soc. 1964, 1029±1037.
8. Satisfactory IR, 1H NMR, 13C NMR, and high resolution MS spectral data were obtained for all new compounds.
9. Branchaud, B. P.; Walsh, C. T. J. Am. Chem. Soc. 1985, 107, 2143±2160.