1062
K. Mizuno et al. / Biochemical and Biophysical Research Communications 452 (2014) 1060–1066
(TaKaRa). DNA-free total RNA (5
used for first strand cDNA synthesis with a 20
l
g) from each tissue sample was
L reaction volume.
l) contained 0.1 l of first strand
derivatives (nicotinic acid, methylnicotinate, nicotinamide, N-
methylnicotinamide, nicotinic acid hydrazide and 6-methylnicoti-
nicacid), and three purine/pyrimidine derivatives (uracil,
guanosine and adenine). The recombinant enzyme then only had
methyltransferase activity that produced trigonelline from nico-
tinic acid used as a substrate. As shown in the HPLC profile in
Fig. 3A, the product had the same retention time as trigonelline.
The fractions were collected and concentrated, and then the con-
densed product was applied to mass spectrometer. Since the prod-
uct had the same molecular weight and mass spectrum as
trigonelline, it was determined that the product was trigonelline
(Fig. 3B). It was concluded that, CTgS1 and CTgS2 are trigonelline
synthases that can specifically catalyze N-methylation of nicotinic
acid. The optimum temperature for both CTgS1 and CTgS2 was
25 °C. At this temperature, the optimum pH was 7.5. CTgS1 and
CTgS2 displayed Michaelis–Menten kinetics toward nicotinic acid
l
The PCR reaction mixture (20
l
l
cDNA from the above reaction mixture, 2.5 units of Ex Taq
(TaKaRa), 2 mM MgCl2, 0.25 mM each of deoxyribonucleotide tri-
phosphate and 5 lM of gene-specific primers. The gene-specific
primers used were: 50-CCCATTCCCCAGAATACA-30 and 50-GCTGCAT
TCGTCGCAAA-30 for CTgS1, and 50-ATTCCACAGGTTTGCGAC-30 and
50-ATCCTTTCCCCGTATCAG-30 for CTgS2. A set of specific primers
(50-GCTTTCAACACCTTCTTCAG-30 and 50-GCTGCTCAGGGTGGAA-
GAG-30) for
a-tubulin (accession number AF363630) was used for
the control reaction. For PCR, we used a PTC-200 thermal cycler
(Bio-Rad Laboratories, Inc., CA, USA) with the following program:
94 °C for 1 min, 50 °C for 30 s, and 72 °C for 15 s. The amplification
was done for 25 cycles. The amplicons corresponding to CTgS1,
CTgS2, and tubulin were 139, 167, and 133 bp in length, respec-
tively. The reaction products were visualized by UV light on 1.2%
agarose gels stained with ethidium bromide. The intensity of fluo-
rescence was quantitated with a Macintosh computer using ImageJ
with Km values of 121 and 184
23.4 nmol/min/mg of protein, respectively. The Km values of CTgS1
and CTgS2, with SAM as a substrate were 68 and 120 M, respec-
lM, and Vmax values of 5.20 and
l
tively. The specific activity of recombinant CTgS1 and CTgS2 were
0.43 and 1.3 nkat/mg proteins with nicotinic acid, respectively.
2.6. Analytical procedure
3.3. Gene expression of CTgS1 and CTgS2
Protein concentrations were measured using methods previ-
ously described by Bradford [28]. Nucleotide sequencing was
carried out using an ABI PRISMÒ 3100 genetic analyzer (Life Tech-
nologies Corporation, CA, USA) and was conducted with assistance
from the Life Research Support Center in Akita Prefectural Univer-
sity. Nucleotide and protein sequences were analyzed by computer
using MacVector (MacVector, Inc., NC, USA).
As CTgS1 and CTgS2 are highly homologous with other coffee
caffeine synthases (CCS1, CTS1, CTS2 and CmXRS1) at the nucleotide
level (over 80% identical), the amount of transcripts from these
genes cannot be detected by Northern blot analysis. The expression
of CTgS1 and CTgS2 was evaluated by semi-quantitative RT-PCR
designed for specific amplification and detection of these tran-
scripts (Fig. 4). Although the transcripts of CTgS1 and CTgS2 were
detected in all organs (flower buds, closed leaves, young leaves,
and mature leaves) (Fig. 4A), and in the developing stages of fruits
(DAF 0 to DAF 30) used in these experiments, the gene expression
in DAF 30 and the mature leaves were the strongest for those genes
in the evaluated organs (Fig. 4B). The transcript accumulation of
CTgS1 was maximum in DAF 5 fruits, and then tended to decrease
slowly. On the other hand, CTgS1 in leaves was expressed in mature
leaves rather than in young leaves. The signals for CTgS2 transcripts
were extremely weak in leaves.
3. Results
3.1. Comparison of coffee trigonelline synthases (CTgS1 and CTgS2)
with other motif B0-methyltransferases
CTgS1 (Accession Number AB054842) and CTgS2 (Accession
Number AB054843) consisted of 1373-bp and 1434-bp sequences,
respectively, and encoded 386 amino acid residues. Their amino
acid sequences were highly identical (over 95% identity) to each
other. CTgS1 shared a high degree of sequence identity (82.3%,
80.8%, and 82.9%) with coffee caffeine synthase (CCS1), coffee theo-
bromine synthase (CTS1), and coffee 7-methylxanthosine synthase
(CmXRS1), respectively (Fig. 2A). CTgS2 had the same values of
sequence identity for the specified caffeine synthetic enzymes.
These values indicated that CTgS1 and CTgS2 are highly homolo-
gous to caffeine synthetic enzymes from coffee. On the other hand,
CTgS1 and CTgS2 also shared approximately 40% identity with tea
caffeine synthase, cacao theobromine synthase, C. breweri salicy-
late methyltransferase [14], and Arabidopsis thaliana jasmonate
methyltransferase [17], and they have three conserved regions as
SAM binding motifs. (Fig. 2A). The phylogenetic tree among the
CCS family and those related enzymes, such as B0-MTs, is indicated
in Fig. 2B. Thus, CTgS1 and CTgS2 are highly homologous with caf-
feine synthetic enzymes from coffee, and enzymes of the motif B0
methyltransferase family [22,23] (SABATH family).
4. Discussion
In this study, we successfully isolated and characterized the
trigonelline synthase genes from coffee. The amino acid sequence
for CTgS1 and CTgS2 are similar to the coffee caffeine synthases,
with the sequence identity between CCS1 and CTgS1 being more
than 82%. It was then assumed that the enzymes could accept xan-
thine derivatives as substrates, so we attempted to measure the
enzymatic activity with those compounds. However, no enzymatic
activity could be detected when the xanthine derivatives were
used as substrates.
It was previously speculated that trigonelline is transported to
seeds after it is produced in other tissues [29]. The levels of trigo-
nelline synthases gene expression clearly increased in fruits from 1
to 2 weeks after flowering (Fig. 4B). In the process of coffee bean
maturation, the pulp and peel develop first, and at this time, the
endosperm and embryo that form the bean are not yet sufficiently
developed. This observation suggested that the trigonelline was
transported into the developing endosperm after being produced
in the pulp. However, to prove that trigonelline was transported,
the trigonelline synthase expression levels in the seed, fruit pulp,
and peel 30 days after flowering needed to be examined.
3.2. Catalytic properties of CTgS1 and CTgS2
The molecular mass for both the recombinant CTgS1 and CTgS2
was 43.2 kDa. The recombinant enzymes were assayed with typical
substrates for B0-MTs (salicylic acid, benzoic acid, jasmonic acid,
indoleacetic acid and gibberellic acid), xanthine derivatives for
substrates of caffeine synthases (xanthosine, xanthine, 7-methyl-
xanthine, theobromine, theophylline and paraxanthine), pyridine
Previous studies have reported trigonelline biosynthesis based
on biochemical analysis and characterization of enzymes purified
from plant tissue. Here, we identified and characterized the gene