Bioluminescence in the limpet-like snail, latia neritoides

Article abstract of DOI:10.1246/bcsj.78.1197

Latia neritoides is a small limpet-like snail that produces a bright green bioluminescence; it is found only in New Zealand streams. The light-emitting system is unique. Although Latia bioluminescence has been studied since 1880, its mechanism is unclear. Shimomura and Johnson clarified the elements of the mechanism, including the structures of luciferin and luciferase, in 1968. However, neither the emitter nor the mechanism of the excited state of luciferin has been determined. We studied molecular mechanisms to clarify the characteristics of luciferin and luciferase and to produce a new application for this system.

Full text of DOI:10.1246/bcsj.78.1197

ꢀ 2005 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 78, 1197–1205 (2005) 1197
Accounts
Bioluminescence in the Limpet-Like Snail, Latia neritoides
ꢀ;1
Yoshihiro Ohmiya, Satoshi Kojima,1;2 Mitsuhiro Nakamura,1;2 and Haruki Niwa2
1Research Institute for Cell Engineering, National Institute of Advanced Industrial Science and Technology,
Ikeda, Osaka 563-8577
2Department of Applied Physics and Chemistry, University of Electro-Communications, Chofu, Tokyo 182-8585
Received November 30, 2004; E-mail: y-ohmiya@aist.go.jp
Latia neritoides is a small limpet-like snail that produces a bright green bioluminescence; it is found only in New
Zealand streams. The light-emitting system is unique. Although Latia bioluminescence has been studied since 1880, its
mechanism is unclear. Shimomura and Johnson clarified the elements of the mechanism, including the structures of lu-
ciferin and luciferase, in 1968. However, neither the emitter nor the mechanism of the excited state of luciferin has been
determined. We studied molecular mechanisms to clarify the characteristics of luciferin and luciferase and to produce a
new application for this system.
Bioluminescence, the production of light without heat by
living organisms, has elicited the interest of scientists for a
long time. The biological functions of bioluminescence are
thought to involve defence against predators, assistance in pre-
dation, communication in reproduction, and metabolic and
biochemical pathway by-product formation in organisms. Bio-
luminescence reveals a diversity of organisms from bacteria to
fish, because the luciferins (substrates) of the various phyloge-
netically distant systems on unrelated chemical classes and the
luciferases (enzyme) based on different gene family.1 So, for
example, a biologist researches the behavior, ecology, and
taxonomy of bioluminescent organisms to reveal their diversi-
ty. A chemist determines the luciferin structure and analyses
the chemical reaction to produce new chemiluminescent mate-
rials. A biochemist explores a new luciferase from organisms
to understand the metabolic pathway and its mechanism, and
a bioengineer develops a biotechnological tool by using infor-
mation from earlier bioluminescence research. Research on
many bioluminescent organisms has gradually revealed vari-
ous aspects of this mysterious feature.2–4 However, there is
one bioluminescent system that has not been sufficiently clari-
fied to date. Here, we account for the molecular mechanism of
bioluminescence of the limpet-like snail, Latia.
1. Chemical Mechanism of Bioluminescence
Bioluminescence is cold light that is characterized by a wide
range of colors:6 it is blue in jellyfishes, dinoflagellates, and
ostracods; blue-green in bacteria and the limpet-like snail, Lat-
ia; and green-red in fireflies and railroad worms (Table 1). The
light emitted by bioluminescent substrates is identical to that
emitted by chemiluminescent substrates when the energy level
of the light emitter molecules falls from the excited state to
the ground state.5 The variety of colors in bioluminescence
is attributed to differences in the structures of luciferase and
Table 1. Chemistry and Color of Bioluminescence in Different Organisms
Luminous organisms
Bacteria
(Photobacterium, Vibrio)
Dinoflagellates
(Lingulodinium, Pyrocystis)
Cnidarians
(Aequorea, Renilla)
Molluscs
(Latia)
Luciferin, Cofactor
FMNH2, RCHO
Luciferase/kDa
80
Emission max./nm
495–500
Tetrapyrrole, Hþ
Coelenterazine, Ca2þ
Enol formate
130
25
475–483
460–490
536
180
70
Crustacea
Imidazopyrazinone
465
(Vargula; Cypridina)
Insects
(Photinus; Photuris)
Benzothiazoline, Mg2þ, ATP
60
550–620
Published on the web July 1, 2005; DOI 10.1246/bcsj.78.1197

1198 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
ACCOUNTS
luciferase
Luciferin + O2
(P)*
P + hν
(1)
Scheme 1.
luciferin. Bioluminescent systems are not evolutionarily con-
served; gene codings for luciferase proteins are not homolo-
gous and those for luciferins also differ.3 It is important to note
that all luciferases are oxygenases. They incorporate oxygen
into a product molecule (P) in an electronically excited state
ꢀ
(P) via a dioxetane intermediate; the process is sufficiently
energetic to result in the emission of a photon (Scheme 1).
The luminescence maximum corresponds to the fluorescence
emission maximum of oxyluciferin in the excited state. In gen-
eral, bioluminescence reactions occur when luciferin is oxi-
dized by molecular oxygen and the reaction is catalyzed by lu-
ciferase. Table 1 shows the structures of luciferin and its oxi-
dized product, oxyluciferin, for bacteria, dinoflagellates, Renil-
la, ostracods, and the firefly. Renilla luciferin (coelenterazine)
is structurally quite similar to ostracod Vargula (Cypridina) lu-
ciferin. It contains the imidazopyrazinone skeleton that is sup-
posedly biosynthesized from three amino acids. In the light-
emitting reaction, the luminescence maxima of firefly, Renilla,
and ostracod bioluminescences are 540–630, 450–470, and
450–460 nm, respectively, which correspond to the fluores-
cence emission maxima of their oxyluciferins.3 On the other
hand, the molecular weights of luciferases are widely distribut-
ed (25–180 kDa) and their structures, which originate from the
various phylogenetically distant systems, belong to different
super families, viz. the calcium-binding protein, aequorin, in
jellyfish and the adenylation enzyme in fireflies and railroad
worms. Bioluminescence reactions are triggered by various
cofactors, viz. the calcium ion in Aequorea jellyfish, the proton
(Hþ) in dinoflagellates, and ATP in fireflies and railroad
worms.
The biological function of bioluminescence in fireflies is
thought to be communication for mating, and their color differ-
ences are adapted to their visual systems.6 Luciferin is convert-
ed to luciferyl adenylate, which is then oxidized by molecular
oxygen and catalyzed by luciferase to yield light, oxyluciferin,
CO2, and AMP (Fig. 1). Various explanations for the chemical
origin of color differences have been given. The first is based
on differences in the ionic structure of the excited oxyluciferin;
the first excited monoanion state deprotonates to form a dian-
ion that emits yellow light (Fig. 1a).5 A second explanation in-
volves the polarity of the oxyluciferin binding site,7 and a third
contends that the conformation of oxyluciferin is so rigid that
the first-excited singlet state induces a structural restriction on
luciferase that changes the energy of electronic state transi-
tions (Fig. 1b).8 Luciferases have been cloned and analyzed
from many fireflies and beetles. Luciferases consist of 543–
550 amino acid residues and belong, in a molecular evolution-
ary sense, to the adenylate enzyme family.3,9 The three-dimen-
sional structure of Photinus pyralis luciferase has been deter-
mined, and an activation site and reaction mechanism have
been proposed.10,11 As a result of the acquisition of knowledge
about the chemical reaction and the structure of luciferin and
luciferase, the firefly bioluminescence system is now used
widely as a reporter of gene expression and as a cellular mark-
er.12 Recently, we developed a tricolor reporter in vitro assay
Fig. 1. Chemical mechanism of the bioluminescence reac-
tion of firefly luciferin catalyzed by firefly luciferase. (a)
Luciferin is converted into an adenylate in the presence
of ATP, which is oxygenated in the presence of oxygen
forming a peroxide intermediate by splitting off AMP.
When the energy levels of the excited states fall to the
ground states, a photon of light (ꢀmax ¼ 550{560 nm)
and red light (ꢀmax ¼ 615 nm) is emitted. (b) According
to an alternative mechanism proposed by McCapra, the
conformation of oxyluciferin is so rigid that the first-excit-
ed singlet state suffers a structural restriction of luciferases
that changes the energy of electronic state transitions.
system in which the expression of three genes can monitored
simultaneously. This is accomplished by splitting the emis-
sions from green, orange, and red beetle luciferases with a
long-pass filter.13 Thus, bioluminescence is a unique system
that has great potential for new applications.
Most bioluminescence is derived from the final product of
the luciferin–luciferase reaction. The luminescence reaction
in bacteria is slightly different from other reactions, as it re-
quires luciferase, molecular oxygen, reduced flavin (FMNH2),
and a long-chain aldehyde. Initially, the luciferase forms a
complex with FMNH2, which is subsequently oxidized to its
peroxide. The complex of luciferase and peroxy FMNH2 reacts
with the long-chain aldehyde, and it is then broken down to
4a-hydroxy flavin with the emission of light and the production
of acid. Finally, the hydroxy flavin is converted to FMN with
the loss of H2O. It is thought that the light emitter in bacterial
luminescence is the luciferase-bound 4a-hydroxy flavin
(Fig. 2a).4
One exception to these general reactions is found in dinoflag-
ellates; their oxyluciferin has no potential for fluorescence;14
the luminescence maximum (ꢀmax ¼ 475 nm) corresponds to
the fluorescence emission maximum of luciferin (Fig. 2b).
On the other hand, fluorescence is not found in Latia luciferin
or oxyluciferin. Shimomura and Johnson called for clarifica-
tion of the bioluminescence system of Latia but, after 30 years,
the basis of this emission remains unclear.15–18 We have at-
tempted to understand the bioluminescence mechanism by an-
alysing the relationships between the function and the structure
of luciferin. We also purified luciferase to identify the emis-
sion species and to develop new applications for future use.
2. Biology of Latia
The limpet-like snail L. neritoides lives in swift-running
streams throughout the North Island of New Zealand and is
the only bioluminescent creature known to live in fresh wa-
ter.18–21 Suter was the first to describe specimens and was sur-

Y. Ohmiya et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1199
prised to find ‘‘all the animals highly phosphorescent with a
violet light’’ and commented that ‘‘this was intensified by a
touch with a needle. The secreted mucus was also phosphores-
cent for some time’’.22 Later, Bowden mixed hot water (70 ꢁC)
with an extract of crushed Latia and reported ꢁthe luciferin–
luciferase reaction on mixing a hot water (70 C) extract of
crushed Latia (after cooling) with cold water containing
once-luminescent mucus.23 We collected specimens of Latia
during 1997–1999 from a running stream at Mt. Pirongia,
New Zealand (Figs. 3a, b). These Latia have rather small oval
seed-like shells and a body mass of 0.03–0.05 g. They are
black and are not easily noticed because they cling to the un-
derside of stones in the riverbed (Fig. 3c). A sample of about
100 g was obtained by three people over several days. We con-
firmed that Latia ejects a bright green luminescent mucus
(Fig. 3d) that is similar to that of sea-fireflies after chemical,
mechanical, or electrical stimulation, but we did not observe
the phosphorescence that Suter reported.
3. Chemistry of Latia Luciferin
After Bowden,23 Shimomura and Johnson isolated Latia lu-
ciferin, determined its structure, and proposed a chemical reac-
tion (Scheme 2).18 Latia luciferin is a colorless, non-fluores-
cent liquid and is a highly hydrophobic, non-polar solvent-
soluble compound. Latia luciferin 1 emits a yellowish green
light (ꢀmax ¼ 536 nm) in the presence of Latia luciferase,
the purple protein, and molecular oxygen, decomposing itself
into compound 2 as an oxyluciferin, formic acid, and carbon
dioxide. Interestingly, the luciferin and oxyluciferin in Latia
are not fluorescent and have no absorption in the visible range
of the spectrum, which indicates that both the mechanism and
the emitter of Latia bioluminescence are different from those
of other bioluminescent organisms. This luciferin was subse-
quently synthesized by Fracheboud et al. in 196924 and by
Kishi et al. in 1970.25
Fig. 2. Chemical mechanism of the bioluminescence reac-
tion in bacteria and dinoflagellates. (a) A complex of bac-
terial luciferase and FMNH2 is oxidized to peroxide; the
complex of luciferase and peroxy FMNH2 reacts with
long-chain aldehyde, and is then broken down to 4a-hy-
droxy flavin with light emission. (b) The conversion of di-
noflagellate luciferin to oxyluciferin is triggered by a pro-
ton and catalyzed by luciferase, and then produces light
(ꢀmax ¼ 475 nm).
To understand the molecular function of Latia luciferin, we
prepared it and its analogues by a slight modification of the
Scheme 2.
(a)
(b)
Mt. Pirongia
(c)
(d)
Fig. 3. (a) Location of the collecting point for Latia neritoides in the North Island of New Zealand. (b) View of the collecting point
for Latia neritoides near Mt. Pirongia. (c) Living condition of Latia neritoides on the river. (d) Bioluminescence of Latia neri-
toides.

1200 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
ACCOUNTS
between function and the structure of luciferin and suggest the
molecular mechanism. Latia luciferin benzoate analogues27
and luciferin analogues having a methyl-substituted phenyl
group instead of the natural 2,6,6-trimethylhexene ring28 were
prepared. Enol acetate E-7 and its regioisomer Z-7 showed
66% and 44% of the activity of 1, respectively, indicating that
the formyl (–CHO) group is not functionally essential. For the
series of benzoate analogues, the luminescence peak intensities
of E-8 to E-12 were 4.2–0.79%, but their total light activities
for 2 h were 85–16% compared with 1, suggesting that
these analogues have a potential bioluminescence reaction
(Table 2). The reason for the delay in the bioluminescence re-
action is unclear, although it may be caused by factors such as
their bulky structure, electronic state, and cleavage rate of the
enol ester moiety, and the charge of the modified moiety
(Fig. 5). Thus, Latia luciferase has an esterase activity and
generates the enolate anion 13 (Fig. 12) from enol esters in
the environment of the enzyme. On the other hand, the ana-
logues E-14, E-15, and E-16, having methyl-substituted phenyl
groups instead of the natural 2,6,6-trimethylcyclohexene ring
system, showed luminescent activities, although they were less
potent than that of the authentic Latia luciferin. The biolumi-
nescence spectra of E-14, E-15, and E-16 were identical to
those of the authentic Latia luciferin, indicating that the
2,6,6-trimethylcyclohexene moiety of luciferin does not
affect the bioluminescence spectra (Fig. 6). These results indi-
cate that the number and the position of the methyl group(s) on
the phenyl ring are important for substrate recognition in the
Latia luciferase. In addition, the 2,6-dimethyl group on the
phenyl ring may be in a good position, as with natural lucifer-
in. Interestingly, no delayed emission in bioluminescence was
observed for the luciferin analogues 14–16, although they ex-
hibited a lower emission activity. From these results, it is de-
duced that luminescence can be controlled by modification of
the enol ester and 2,6,6-trimethylcyclohexene moieties of lu-
Fig. 4. Preparation of Latia luciferin (1) and its isomer (Z-
1). Reagents and conditions: (a) Bu3SnH, toluene, 95%;
(b) (EtO)2P(O)CH2CO2Et, LiHMDS, THF, reflux, 5 h,
76% for E-4, 9% for Z-4 after separation; (c) DIBAH, tol-
uene, 0 ꢁC, 100% for E-5, 100% for Z-5; (d) MnO2,
CH2Cl2, rt, 2 h, 95% for Eꢁ-6, 95% for Z-6; (e) SeO2,
60% H2O2, t-amylalcohol, 0 C, 45% for 1, 25% for Z-1.
procedure reported by Kishi (Fig. 4).26 Selective hydrogena-
tion of ꢁ-ionone (3) was followed by Horner–Emmons olefina-
tion of the resulting ketone (2) (Latia oxyluciferin). The proce-
dure afforded ꢂ,ꢁ-unsaturated esters E-4 and Z-4 after separa-
tion. Reduction of E-4 with DIBAH, followed by the oxidation
of the resulting allylic alcohol with MnO2, gave the desired
ꢂ,ꢁ-unsaturated aldehyde E-6. Baeyer–Villiger oxidation of
E-6 was achieved by using 60% hydrogen peroxide in t-pentyl
alcohol in the presence of selenium(IV) oxide to give Latia lu-
ciferin (1). The corresponding geometrical isomer Z-1 was pre-
pared via Z-4 by a similar procedure. Using a crude Latia lu-
ciferase solution, the Z-1 regioisomer of Latia luciferin (1) had
60% of the activity of 1, indicating that the recognition site of
the enol ester moiety in luciferase has a certain degree of flex-
ibility. On the other hand, ꢂ,ꢁ-unsaturated aldehyde E-6,
which is supposed to be a biosynthetic precursor of 1, and
Z-6 had no bioluminescence activity. The bioluminescence
properties of Latia luciferin analogues clarify the relationship
Table 2. Bioluminescence Activities of Latia Luciferin and its Analogues
O
R
O
NO
2
CN
OMe
NMe
H
CH
3
1
E-7
E-8
85
E-9
71
E-10
69
E-11
28
E-12
16
Total Light %
100
66
Peak Intensity %
100
N.D.
4.2
3.2
2.9
1.1
0.79
OCHO
R
1
E-14
32
E-15
E-16
E-17
E-18

Total Light %
100
Peak Intensity %
100


7.7
3.0
1.4

7.2
0.54

Y. Ohmiya et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1201
E-8
E-11
E-9
E-10
1
E-7
E-12
Fig. 5. Kinetics of the bioluminescence reaction of Latia
luciferase and Latia luciferin and its analogues.
Fig. 7. Analysis of Latia luciferase by SDS-PAGE (12.5%)
at various steps of the purification procedure. All samples
were adjusted to the same bioluminescent activity (2 ꢂ
105 RLU). Lane 1, marker proteins; Lane 2, 33–60%
(NH4)2SO4 fractionation; Lane 3, HiPrep Sephacryl S-
200; Lane 4, HiTrap Con A; Lane 5, Mono Q; and Lane
6, Superdex 200.
E-14
HiPrep Sephacryl S-200 was slightly yellow. This fraction had
two fluorescence peaks (575 and 670 nm), and one of these
(575 nm) could correspond to that of the purple protein.17
The fluorescent elements did not bind to the Con A column
during the next purification step for affinity chromatography.
The active fraction did not show the fluorescent peaks at 575
nm or 670 nm and the specific activity of luciferase was
increased, indicating that the enzyme and the purple protein
are separable. These fractions were subjected to ion exchange
FPLC using a Mono Q column. The elution pattern of the ac-
tive fractions was widely distributed between samples contain-
ing 0.3 and 0.4 M NaCl, showing that the enzyme is heteroge-
neous. After the active fractions had been concentrated, the en-
zyme was purified to a single band on SDS-PAGE by silver
staining using size-exclusion FPLC with two Superdex 200
columns (Fig. 7). The reaction between synthetic luciferin
E-7
1
Fig. 6. Bioluminescence in vitro spectrum of purified Latia
luciferase and synthetic Latia luciferin and its analogues.
Bioluminescence spectra were obtained using an
AB1850 spectrofluorometer (ATTO) equipped with a
back-illumination-type attachment LN/CCD-512TKB
(Princeton Instruments, CA). Width of the slit was 0.5
mm, and spectral resolution was 20 nm. The biolumines-
cent spectrum was collected for 10 s.
and the purified enzyme produced a greenish light (ꢀmax
¼
ciferin. The bioluminescence spectra of all luciferin analogues
were the same as those of Latia luciferin, which strongly indi-
cates that the light emitter cannot be luciferin or its analogues
and is contained in Latia luciferase.
536 nm), which is similar to the spectrum of luciferin and
the crude extract in vitro, and therefore reconfirmed that the
purified enzyme was Latia luciferase (Fig. 8). The relative mo-
lecular mass of Latia luciferase on a Superdex 200 size-exclu-
sion column was about 180 kDa. Figure 7 and Figure 9a show
SDS-PAGE patterns for each fraction from the ammonium
sulfate precipitate to the purified enzyme, suggesting that the
molecular mass of luciferase is 31 kDa. Furthermore, during
MALDI-TOF analysis, the observed molecular mass was
31.6 kDa (Fig. 9b). These results suggest that Latia luciferase
is a homo-oligomeric protein with 31.6 kDa subunits and a ho-
mohexamer of ca. 180 kDa corresponding to Shimomura’s pu-
rification. On the other hand, the luciferase was completely
bound to the Con A column without methyl ꢂ-D-glucopyrano-
side. When the luciferase was treated with PNGase F, the mo-
lecular mass of the resultant protein was 29–30 kDa on SDS-
PAGE (Fig. 9a) and 29.1 kDa on MALDI-TOF (Fig. 9c), sug-
gesting that Latia luciferase is an N-linked glycoprotein. The
molecular mass of the carbohydrate chain was estimated to
4. Biochemistry of Latia Luciferase
Shimomura and Johnson identified two proteins that partic-
ipate in the light-emitting reaction and purified a 173 kDa pro-
tein considered to be luciferase (EC 1.14.99.21) and a purple
protein of 39 kDa as cofactor.17,18 Their purification procedure
included DEAE-cellulose, Sephadex G-200, and Sephadex G-
25 column chromatography. The purified luciferase is biolumi-
nescent, with a molecular mass of 173 kDa as estimated by ul-
tracentrifugation. We also purified Latia luciferase to clarify
the structure. The enzyme responsible for the secretion of lu-
ciferase was predicted to be a glycoprotein. The enzyme was
isolated from the extract in five steps including affinity chro-
matography for oligosaccharide chains. The active fraction
that was first isolated by size-exclusion chromatography using

1202 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
ACCOUNTS
(a)
(a)
(b)
(c)
(b)
Fig. 8. (a) Bioluminescence in vivo spectrum of Latia. (b)
Bioluminescence in vitro spectrum of synthetic Latia lu-
ciferin and purified Latia luciferase. Bioluminescence
spectra were measured using an AB1850 spectrofluoro-
meter (ATTO) equipped with a back-illumination-type at-
tachment LN/CCD-512TKB (Princeton Instruments, CA).
The width of the slit was 0.5 mm, and spectral resolution
was 20 nm. The bioluminescent spectrum was collected
for 10 s.
Fig. 9. The change of the molecular mass by the enzymatic
digestion using endoglycosidase F (PNGase F). (a) Analy-
sis by SDS-PAGE (12.5%). Lane 1, marker proteins; Lane
2, purified Latia luciferase; Lane 3, deglycosylated Latia
luciferase. (b) Matrix-assisted laser desorption/ionization
(MALDI) mass spectra for purified Latia luciferase and
(c) deglycosylated Latia luciferase. An Applied Biosystem
mass spectrometer was used to obtain MALDI mass spec-
tra for determining molecular mass in a linear time of
flight (TOF) mode. The nitrogen laser was set to deliver
337 nm wavelength pulses (5 ns) to the sample. Sinapinic
acid was used as the matrix.
Table 3. Purification of Luciferase from Latia neritoides
Total activity
/RLU
Protein
/mg
Specific activity
/RLU mgꢄ1
Yield
/%
Fold
/Times
Latia (39 g)
ꢃ
1. (NH4)2SO4 prep.
2. Sephacryl S200
3. Con A
4. Mono Q
5. Superdex 200
3:95 ꢂ 108
2:63 ꢂ 108
2:30 ꢂ 108
2:19 ꢂ 108
1:87 ꢂ 108
117
3:38 ꢂ 106
5:04 ꢂ 106
3:49 ꢂ 107
1:10 ꢂ 107
3:59 ꢂ 108
100
67
58
55
1
1.5
10.3
32.5
105
52.2
6.60
2.00
0.52
47
be about 2 kDa. In this procedure, for example, approximately
520 mg of the purified luciferase was obtained from 39 g (wet
weight) of the Latia sample, and its activity was about 105-
fold greater than that of the fraction from the 33–60% ammo-
nium sulfate precipitation. This indicates that the purple pro-
tein could not have contributed to the apparent biolumines-
cence (Table 3). Shimomura, Johnson, and co-workers report-
ed that, for a fixed amount of luciferin and purple protein, the
rate of the light emitting reaction is essentially of the first order
in luciferase, with the total amount of emitted light being
independent of the luciferase concentration.17 In contrast, the
kinetics of light emission are complex for a fixed amount of
luciferase and luciferin and a varying amount of purple pro-
tein, with the time lag being attributed to a slow interaction be-
tween purple protein, luciferin, and oxygen. The fluorescence
properties of purple protein do not overlap with Latia biolumi-
nescence. These results suggest that purple protein is not
essential but may have some influence on the reaction.
Table 4. Enzymatic Properties of Latia Luciferase
Processing Type
Specific activity/RLU mgꢄ1
N-linked glycoprotein
3:59 ꢂ 108
ꢃ
Molecular mass
Gel filtration
SDS-PAGE
MALDI-TOF
Isoelectric point
Km
180 kDa
31 kDa
31.6 kDa
5.4–6.0
8.7 mM
around pH 7.2. The Km value of luciferase for luciferin was
about 8.7 mM when calculated with a linear Lineweaver–Burk
plot, indicating that the reaction between the purified enzyme
and luciferin effectively produced a greenish light, and that the
binding of luciferin to luciferase was highly specific. The ac-
tivity of the active form of Latia luciferaꢁse was maintained
above 90% when incubated for 12 h at 37 C, whereas its ac-
tivity when incubated at 50 ꢁC for 10 min was only about 50%.
This partially inactivated luciferase was reanalyzed using size-
exclusion FPLC (Fig. 10). The elution pattern showed two
We summarize the enzymatic properties of purified Latia
luciferase in Table 4. The specific activity of our purified lu-
ciferase was 3:59 ꢂ 108 RLU/mg and its optimum pH was

Y. Ohmiya et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1203
luciferase fluoresced visibly in an alkaline or KCN solution.18
The spectrum of this fluorescence is very similar to that of Lat-
ia bioluminescence and to the fluorescence of flavin adenine
dinucleotide (FAD). This result indicates that Latia luciferase
might be bound to a flavin group, which may be triggered by
the bioluminescent reaction and accepted by the excited ener-
gy of oxyluciferin, although its fluorescence was quenched un-
der the reduced state. We also detected the maximum fluores-
cent wavelength at about 535 nm but it was very weak, similar
to the fluorescence of flavin (Fig. 11b), which supports the pos-
sibility that Latia luciferase has a tightly bound flavin group
that constitutes the light emitter.18 However, the FAD as a
chromophore, or light-mediating species could not be identi-
fied. In other systems, accessory fluorescent proteins, such as
GFP in coelenterates, receive the excitation energy of the oxy-
luciferin via energy transfer and emit it with their own fluores-
cence spectra. The intermediate state of luciferin has not been
determined. In other cases such as in the bacterial reaction, the
emitter is a reaction intermediate. Sometimes, as in the case of
dinoflagellates, although the precise chemical identity of the
emitter is unknown, it is accepted that it is a polypyrrole deriv-
ative of luciferin. Furthermore, although Pholas and Diptera
luciferin has not been identified, the luciferases have been
identified,31,32 and Pholas luciferase has been cloned.31 In con-
trast, there is, until now, no strong candidate for the role of
emitter in the case of Latia.
The quantum yieldꢁof luciferin is very low, 0.003 per lu-
ciferin oxidized at 25 C, but as much as 8 per purple protein
molecule, which apparently recycles, indicating that the reac-
tion efficiency is very low compared with the other biolumi-
nescent reaction.17 In general, the quantum yield of biolumi-
nescence is very high: for example, 0.88 for firefly, 0.23 for ae-
quorin, and 0.31 for ostracod.33–35 The reason for the low effi-
ciency in Latia bioluminescence is not clear but may relate to
non-fluorescent properties of luciferin and oxyluciferin or to
energy transfer to the bounded chromophore from excited oxy-
luciferin.
Fig. 10. Size-exclusion FPLC of partially denatured lu-
ciferase. The protein was detected by its absorbance at
280 nm (solid line) and bioluminescence activity ( ). In-
set: analysis of two peaks by SDS-PAGE (12.5%); Lane 1,
marker proteins; Lane 2, 0.5 mg of peak 1; and Lane 3, 0.5
mg of peak 2.
peaks on Superdex 200: the upper peak fraction at around 180
kDa had bioluminescence activity, whereas a new peak frac-
tion at around 30 kDa had no activity. Interestingly, each peak
on SDS-PAGE analysis showed the same band at 31 kDa,
which corresponds to the monomeric component of luciferase.
These results suggest that the oligomeric form could be active,
whereas dissociation of the enzyme causes inactivation of the
bioluminescence reaction. However, it is unclear whether the
monomeric component of the inactive peak had been modified.
The well-known bacterial luciferase is a hetero-dimeric protein
consisting of ꢂ and ꢁ subunits,29 and Oplophorus luciferase is
a hetero-oligomeric enzyme, consisting of 35 kDa and 19 kDa
peptides.30 In this series of luciferases, Latia luciferase is the
first example of one in which the homo-oligomeric form is es-
sential for the active enzyme.
5. Light Emitter of Latia Bioluminescence
The reaction between the synthetic Latia luciferin and the
purified Latia luciferase produces a greenish light that coin-
cides with the natural bioluminescence spectrum (Fig. 8). In
this case, it does not require other proteins or cofactors, indi-
cating that the luciferase contains a chromophore as the light
emitter. The enzyme is colorless and does not have maximal
absorption in the visible range (Fig. 11a). Shimomura and
Johnson reported that the normally colorless, non-fluorescent
6. Remarks: Remaining Mystery
The proposed mechanism of Latia bioluminescence is sum-
marized in Fig. 12. In the luciferin–luciferase reaction, Latia
luciferin probably converts to the excited state of oxyluciferin
via the dioxetanone structure; then its excited energy is trans-
ferred to the bounded chromophore on Latia luciferase, and
bioluminescence is produced. The excited state of oxyluciferin
(a)
(b)
Fig. 11. (a) UV–vis absorption spectrum of Latia luciferase. (b) Fluorescent spectrum of Latia luciferase.

1204 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
ACCOUNTS
Fig. 12. Proposed chemical mechanism of the bioluminescence reaction of Latia luciferin catalyzed by luciferase. Latia luciferin in
the luciferin–luciferase reaction probably converts to the excited stated of oxyluciferin via the dioxetanone structure; its excited
energy is transferred to the bounded chromophore on Latia luciferase and bioluminescence is produced.
12 K. V. Wood, Y. A. Lam, and W. D. McElroy, J. Biolumin.
Chemilumin., 4, 289 (1989).
does not determine color but is an energy source for producing
light. Therefore, the emitter remains a mystery. We think that
it may be possible to control the bioluminescence reaction, in-
cluding its color and lifetime, with recombinant proteins re-
constructed from flavin derivates or luciferin analogues. This
is one of the reasons that prompted us to reinvestigate the Latia
system. In view of the ease of synthesis of luciferin and
the greenish emission, we also hoped that a recombinant Latia
luciferase might become a useful tool as a reporter of gene
expression.
13 Y. Nakajima, M. Ikeda, T. Kimura, S. Honma, Y. Ohmiya,
and K. Honma, FEBS Lett., 565, 122 (2004).
14 H. Nakamura, Y. Kishi, O. Shimomura, D. Morse, and
J. W. Hastings, J. Am. Chem. Soc., 111, 7607 (1989).
15 O. Shimomura and F. H. Johnson, ‘‘Bioluminescence in
Progress,’’ ed by F. H. Johnson and Y. Haneda, Princeton Univ.
Press, NJ (1966), pp. 391–404.
16 O. Shimomura and F. H. Johnson, Biochemistry, 7, 1734
(1968).
17 O. Shimomura and F. H. Johnson, Biochemistry, 7, 2574
(1968).
18 O. Shimomura and F. H. Johnson, Proc. Natl. Acad. Sci.
U.S.A., 69, 2086 (1972).
19 J. E. Gray, Proc. Zool. Soc. London, 17, 164 (1849).
20 S. Moore and V. B. Meyer-Rochow, Verh. Int. Ver. Theor.
Angew. Limnol., 23, 2189 (1988).
21 V. B. Meyer-Rochow and S. Moore, Int. Revue ges. Hydro-
biol., 73, 21 (1988).
We are pleased to acknowledge our indebtedness to Prof.
V. B. Mayer-Rochow, Oulu University, Finland, Dr. M. R.
Scarsbrook of National Institute of Water and Atmospheric
Research (NIWA), New Zealand, for collecting the specimens
of Latia and Drs. Nobuyoshi Ohba, Shojiro Maki, and Takashi
Hirano for their valuable comments during this work. We also
acknowledge the financial support from the Japan Society for
the Promotion of Science.
22 H. Suter, Trans. N. Z. Inst., 23, 93 (1890).
23 B. J. Bowden, Biol. Bull., 99, 373 (1950).
24 M. G. Fracheboud, O. Shimomura, R. K. Hill, and F. H.
Johnson, Tetrahedron Lett., 10, 3951 (1969).
25 F. Nakatsubo, Y. Kishi, and T. Goto, Tetrahedron Lett., 11,
381 (1970).
26 S. Kojima, S. Maki, T. Hirano, Y. Ohmiya, and H. Niwa,
Tetrahedron Lett., 41, 4409 (2000).
27 M. Nakamura, M. Masaki, S. Maki, R. Matsui, M. Hieda,
M. Mamino, T. Hirano, Y. Ohmiya, and H. Niwa, Tetrahedron
Lett., 45, 2203 (2004).
References
1 E. N. Harvey, ‘‘Bioluminescence,’’ Academic Press Inc.,
New York (1952).
2
3
4
5
L. J. Kricka, Methods Enzymol., 305, 333 (2000).
Y. Ohmiya, Seikagaku, 76, 5 (2004).
J. W. Hastings, Gene, 173, 5 (1996).
T. Wilson and J. W. Hastings, Annu. Rev. Cell Dev. Biol.,
14, 197 (1998).
6
7
M. A. Branham and J. W. Wenzel, Cladistics, 19, 1 (2003).
L. J. White, R. Rapport, H. H. Seliger, and T. A. Hopkins,
28 M. Nakamura, M. Mamino, M. Masaki, S. Maki, R.
Matsui, S. Kojima, T. Hirano, Y. Ohmiya, and H. Niwa, Tetra-
hedron Lett., 46, 53 (2005).
Bioorg. Chem., 1, 92 (1971).
F. McCapra, D. J. Gilfoyle, D. W. Young, N. J. Church,
8
and P. Spencer, ‘‘Bioluminescence and Chemiluminescence:
Fundamental and Applied Aspects,’’ ed by A. Campbell, L. J.
Kricka, and P. E. Stanley, John Wiley and Sons, UK (1994),
pp. 387–391.
29 J. W. Hastings, K. Weber, J. Friedland, A. Eberhard, G. W.
Mitchell, and A. Gunsalus, Biochemistry, 12, 4681 (1969).
30 S. Inouye, K. Watanabe, H. Nakamura, and O. Shimomura,
FEBS Lett., 481, 19 (2000).
9
V. R. Viviani, Cell. Mol. Life. Sci., 59, 1833 (2002).
10 E. Conti, N. P. Franks, and P. Brick, Structure, 4, 287
(1996).
31 S. L. Dunstan, G. B. Sala-Newby, A. B. Fajardo, K. M.
Taylor, and A. K. Campbell, J. Biol. Chem., 275, 9403 (2000).
32 V. R. Viviani, J. W. Hastings, and T. Wilson, Photochem.
Photobiol., 75, 22 (2002).
11 E. Conti, T. Stachelhaus, M. A. Marahiel, and P. Brick,
EMBO J., 16, 4174 (1997).
33 H. H. Seliger and W. D. McElroy, Arch. Biochem. Bio-

Y. Ohmiya et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1205
phys., 88, 136 (1960).
34 O. Shimomura and F. H. Johnson, Nature, 227, 1356
(1970).
35 O. Shimomura and F. H. Johnson, Photochem. Photobiol.,
12, 291 (1970).
Yoshihiro Ohmiya Group leader of Cell Dynamics research Group, Research Institute for Cell
Engineering, National Institute of AIST, was born in Hokkaido in 1960. He received his B.S. de-
gree in 1983 and M.S. degree in 1985 under the direction of Professor Hirotaro Kambe of Gunma
University, and D.S. degree under the direction of Professor Yoichi Kondo of Gunma University.
He became a Postdoctoral researcher for Osaka Bioscience Institute during 1990–1992 and a re-
searcher for JSPS during 1992–1995. After working, he was appointed as an Assistant Professor
of the Department of Chemistry, Faculty of Education, Shizuoka University in 1996. He was ap-
pointed as a Group leader of the National Institute of AIST in 2001. His current research interests
are basic and applied science of bioluminescence systems of various bioluminescent organisms.
Satoshi Kojima was born in 1973 in Aichi. He graduated from the University of Electro-Commu-
nications in 1996 and received his Ph.D. degree from the same university in 2001. He received
the JSPS Research Fellowships for Young Scientists during 1998–2001 and 2001–2004. He cur-
rently researches about dipteran bioluminescence with Prof. J. W. Hastings at Harvard University.
His current research interests include highly efficient light production on bioluminescence reac-
tion, and the development of new biological imaging tools.
Mitsuhiro Nakamura was born in 1974 in Hyogo. He graduated from Kwansei Gakuin University
in 1997 and received his Ph.D. degree from Nagoya University in 2002. He then became a re-
search scientist at Nagoya University from April to May, 2002. He served as a research scientist
at The University of Electro-Communications from June to October, 2002, then served as a re-
search scientist at National Institute of Advanced Industrial Science and Technology from No-
vember, 2002 to September, 2003, and next served as a SVBL researcher at The University of
Electro-Communications from October, 2003 to March, 2004. He serves as a research associate
at RIKEN Harima Institute as of November, 2004. He was selected as the special lecturer of
younger generation at Annual Meeting of The Chemical Society of Japan in 2004. He currently
researches about protein structural–functional analysis using mass spectrometry and synthetic
low-molecular organic substances at RIKEN Harima Institute.
Haruki Niwa Professor of Bioorganic Chemistry, Department of Applied Physics and Chemistry,
University of Electro-Communications, was born in Aichi in 1947. He received his B.S. degree in
1971 and M.S. degree in 1973 under the direction of Professor Yoshimasa Hirata of Nagoya Uni-
versity. After working for seven years at Ono Pharmaceutical Co., Ltd., he was appointed as an
Assistant Professor of Department of Chemistry, Faculty of Science, Nagoya University, in 1980.
While working at the company, he joined Corey’s group for one year and then he received his
Ph.D. degree from Nagoya University in 1979. After he joined Professor Yamada’s group at
Nagoya University, he was promoted to an Associate Professor in 1984. In 1987, he received
an ‘‘Incentive Award in Synthetic Organic Chemistry’’ from The Society of Synthetic Organic
Chemistry, Japan. He was appointed as a Professor of The University of Electro-communications
in 1994. His current research interests are isolation and structure determination of new luciferins
of various bioluminescent organisms, synthesis of luciferins and the related luminous substances,
application of bioluminescence systems for bioimaging, and development of new luminescent
and fluorescent materials based on the molecules of bioluminescence systems.