N-Methyl-L-amino acid dehydrogenase from Pseudomonas
putida
A novel member of an unusual NAD(P)-dependent oxidoreductase
superfamily
Hisaaki Mihara
1,
*, Hisashi Muramatsu
1,
*, Ryo Kakutani
1
, Mari Yasuda
2
, Makoto Ueda
2
,
Tatsuo Kurihara
1
and Nobuyoshi Esaki
1
1 Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
2 Yokohama Research Center, Mitsubishi Chemical Corp., Yokohama, Japan
n-Methyl amino acids occur naturally as components
of several depsipeptides, including cyclosporin A [1],
vancomycin, destruxins [2], didemnins [3], dolastatin
[4], and enniatin [5], as well as a calpain inhibitor from
Streptomyces griseus [6] and a pilin from Pseudomonas
aeruginosa [7]. Enniatins consist of alternating residues
of d-2-hydroxyisovalerate and N-methyl-l-valine (or
N-methyl-l-isoleucine) [5]. N-Methylphenylalanine is
the N-terminal amino acid of pilins of most bacterial
pathogens expressing type IV pili [7]. The synthesis of

We found N-methyl-l-amino acid dehydrogenase activity in various bacter-
ial strains, such as Pseudomonas putida and Bacillus alvei, and cloned the
gene from P. putida ATCC12633 into Escherichia coli. The enzyme purified
to homogeneity from recombinant E. coli catalyzed the NADPH-dependent
formation of N-alkyl-l-amino acids from the corresponding a-oxo acids
(e.g. pyruvate, phenylpyruvate, and hydroxypyruvate) and alkylamines (e.g.
methylamine, ethylamine, and propylamine). Ammonia was inert as a sub-
strate, and the enzyme was clearly distinct from conventional NAD(P)-
dependent amino acid dehydrogenases, such as alanine dehydrogenase (EC
1.4.1.1). NADPH was more than 300 times more efficient than NADH as
a hydrogen donor in the enzymatic reductive amination. Primary structure
analysis revealed that the enzyme belongs to a new NAD(P)-dependent
oxidoreductase superfamily, the members of which show no sequence
homology to conventional NAD(P)-dependent amino acid dehydrogenases
and opine dehydrogenases.
Abbreviation
NMAADH, N-methyl-
L-amino acid dehydrogenase.
FEBS Journal 272 (2005) 1117–1123 ª 2005 FEBS 1117
ammonia are formed from methylamine and glutamate
by N-methylglutamate synthase (EC 2.1.1.21) [16].
Another A. aminovorans strain, formerly called Pseudo-
monas MS, produces N-methylalanine from methyl-
amine and pyruvate in the presence of NADPH with
N-methylalanine dehydrogenase (EC 1.4.1.17) [17].
Therefore, in contrast with the N-methyl groups of
depsipeptides and pilins, those of free N-methyl amino
acids are derived from methylamine.
The reaction catalyzed by N-methylalanine dehydro-
genase resembles that of alanine dehydrogenase (EC

completely different from that of conventional amino
acid dehydrogenases and opine dehydrogenases.
Results and Discussion
Identification of a gene encoding N-methyl-L-
amino acid dehydrogenase (NMAADH)
We tested about 100 bacterial strains for the ability to
form N-methylphenylalanine from phenylpyruvate and
methylamine. We found such activity in several strains,
including P. putida, Bacillus alvei, Brevibacterium sac-
chrolyticum, Brevibacterium linens, Agrobacterium
viscosum, Aerobacter aerogenes, P. aeruginosa, and
P. fluorescens. The crude extract from P. putida
ATCC12633 exhibited the highest activity, and this
strain was chosen for further studies.
We cultivated the P. putida strain in 200 L Luria–
Bertani medium and obtained about 2.5 kg wet cells.
NMAADH was purified 150-fold with a yield of
0.066% by purification steps with such chromatogra-
phy columns as SuperQ-Toyopearl, Butyl-Toyopearl,
DEAE-Toyopearl, Green-Sepharose, RESOURCE
PHE, and Blue-Sepharose. SDS ⁄ PAGE analysis and a
TLC-based assay of fractions from the Blue-Sepharose
column revealed that a 36-kDa protein was the enzyme
exhibiting the NMAADH activity. The N-terminal
amino-acid sequence of this protein was determined to
be XAPSTSTVVRVPFTEL. We carried out a blast
search using the unfinished microbial genome database
at TIGR ( with this sequence
and found that it was very similar to that of a putative
protein PP3591 (NCBI database number, AAN69191)

the deduced amino-acid sequence of the enzyme, except
that the initial Met was removed. The initial Met was
also missing in the enzyme purified from P. putida
ATCC12633. The purified enzyme gave a single band
with a molecular mass of 37 kDa on SDS ⁄ PAGE
(Fig. 1). The molecular mass of the native enzyme was
found to be 74 kDa by gel filtration. Therefore, the
enzyme probably consists of two identical subunits.
Substrate specificity and effects of various
compounds on the enzymatic activity
The substrate specificity of the enzyme in NADPH-
dependent reductive amination was examined with
various a-oxo acids and amines. Pyruvate was the best
substrate of the various a-oxo acids tested (Table 1).
The enzyme also acted on a-oxohexanoate, phenylpyru-
vate, a-oxobutyrate, fluoropyruvate, a-oxovalerate,
a-oxoisocaproate, a-oxo-octanoate, and hydroxypyru-
vate. However, branched-chain a-oxo acids, such as
a-oxoisovalerate and a-oxo-b-methylvalerate, were
inert. The best substrate was methylamine, but the
enzyme also showed activity towards ethylamine (4.4%,
relative to methylamine), 2-chloroethylamine (0.74%),
2-bromoethylamine (0.27%), n-propylamine (0.16%)
and dimethylamine (0.12%). Weak activities (0.06–
0.03%, relative to methylamine) were found with
ethylenediamine, hydroxylamine, isopropylamine, n-
butylamine, n-amylamine, n-hexylamine, 1,6-diamino-
hexane, and spermidine. Interestingly, the enzyme was
unable to use ammonia as a substrate and was distinct
from alanine dehydrogenase [18] and N-methylalanine

both reductive amination of phenylpyruvate and oxi-
dative deamination of N-methyl-l-alanine. It was sta-
ble between pH 6.0 and 10.0 and showed maximum
activity at 35 °C. However, it was unstable at this tem-
perature and lost 30% of its original activity after
a 30-min incubation in 20 mm Tris ⁄ HCl buffer at
pH 7.0. The enzyme was stable at temperatures below
30 °C for at least 30 min and used both NAD
+
and
NADH as coenzymes. However, the specific activity of
the enzyme with NADPH (42 UÆmg
)1
) was more than
300 times higher than with NADH (0.13 UÆmg
)1
)in
reductive amination of pyruvate. Oxidative deamina-
tion of N-methyl-l-alanine with NADP
+
(0.36 UÆ
mg
)1
) was markedly lower than reductive amination of
pyruvate with NADPH.
Fig. 1. Overexpression and purification of recombinant NMAADH.
Protein samples from various stages of the purification were sub-
jected to SDS ⁄ PAGE and stained with Coomassie Brilliant Blue.
Lane 1, marker proteins (sizes in kDa are shown); lane 2, crude
extract; lane 3, after the Green-Sepharose CL-4B chromatography;

with various concentrations of NADPH and fixed con-
centrations of pyruvate. These indicate a sequential
mechanism for the NMAADH reactions. In fact, the
data obtained gave a good global fit to the equation of
an ordered Ter Bi mechanism.
To test the kinetic mechanism further, we performed
inhibition studies with both products of the reaction,
NADP
+
and N-methyl-l-alanine. At a fixed subsaturat-
ing concentration of N-methyl-l-alanine (100 mm), the
inhibition by NADPH was competitive linearly with
respect to NADP
+
. With N-methyl-l-alanine as the
variable substrate, at a fixed subsaturating concentra-
tion of NADP
+
(0.2 mm), NADPH was a linear mixed-
type inhibitor. Pyruvate behaved as a noncompetitive
inhibitor with regard to NADP
+
and N-methyl-l-alan-
ine. Methylamine behaved as a mixed-type inhibitor
with respect to NADP
+
and N -methyl-l-alanine. These
results indicate that NADPH-dependent N -methyl-l-
alanine formation proceeds in an ordered sequential Ter
Bi mechanism, in which NADPH, pyruvate, and meth-

[33], ureidoglycolate dehydrogenase [34], and 2,3-
diketo-l-gulonate reductase [35]. The action on a-oxo
acids as substrate is common to all these enzymes, and
NMAADH is a new addition to this superfamily.
Conclusion
We have identified an enzyme catalyzing the NADPH-
dependent formation of N -methyl-l-phenylalanine from
phenylpyruvate and methylamine. The enzyme is
unique in that it does not act on ammonia at all and
shows broad specificity for various a-oxo acids. Accord-
ingly, we named the enzyme N-methyl-l-amino acid
dehydrogenase to distinguish it from the previously
reported N-methylalanine dehydrogenase [17]. More-
over, this study shows that the enzyme belongs to a new
NAD(P)-dependent oxidoreductase family and is struc-
turally distinct from conventional NAD(P)-dependent
amino acid dehydrogenases and opine dehydrogenases.
Experimental procedures
Materials
N-Methyl-l-phenylalanine and d-amino acid oxidase from
porcine kidney were purchased from Sigma (St Louis, MO,
USA). RESOURCE PHE, Superose 12, Sepharose CL-4B,
and molecular-mass marker proteins were obtained from
Amersham Pharmacia Biotech (Uppsala, Sweden). SuperQ-
Toyopearl, DEAE-Toyopearl, and Butyl-Toyopearl were
from Tosoh (Tokyo, Japan). Green-Sepharose was prepared
as described previously [36]. NADH, NADPH, and mole-
cular-mass marker proteins for gel filtration were from
Oriental Yeast (Tokyo, Japan). Restriction enzymes and
kits for genetic manipulation were from Takara Shuzo

Æ7H
2
O
(0.1 gÆL
)1
) (pH 6.9). The cells were harvested by centrifu-
gation, resuspended in 0.1 mL reaction mixture containing
5mm calcium phenylpyruvate and 1 m methylam-
ine ⁄ H
2
SO
4
(pH 8.9), and incubated for 3 h at 30 °C. The
formation of N-methylphenylalanine in the reaction mix-
ture was analyzed by TLC on a silica gel 60 plate (Merck,
Darmstadt, Germany) or by HPLC with an Ultron
ES-PhCD column (Shinwa Kako, Kyoto, Japan). The sol-
vent system for TLC was ethyl acetate ⁄ ethanol ⁄ acetic
acid ⁄ water (5 : 2 : 1 : 1, by vol.), and N-methylphenyl-
alanine was visualized with a coloring reagent containing
0.2% ninhydrin, 0.5% acetic acid, and 95% butan-1-ol.
HPLC analysis was performed with a solvent containing
16 mm KH
2
PO
4
, 20% acetonitrile, and 0.04% phosphoric
acid at a flow rate of 0.85 mLÆmin
)1
at 40 °C, and eluates

Cycler 480 (Wellesley, MA, USA) in a 50-lL reaction mix-
ture containing 1· LA Taq buffer (Takara Shuzo), 2.5 mm
MgCl
2
, 0.4 mm dNTP, 0.2 mm each primer (5¢-GGAAT
TCCATATGTCCGCACCTTCCACCAGCACCG-3¢ and
5¢-GGGAAGCTTTCAGCCAAGCAGCTCTTTCAGG-3¢),
2.5 U LA Taq DNA polymerase, and 115 ng genomic
DNA from P. putida ATCC12633: preincubation at 94 °C
for 1 min and then 30 cycles between 98 °C for 20 s and
68 °C for 3 min and finally at 72 °C for 10 min. The PCR
product was digested with NdeI and HindIII and ligated
into pET21a(+) previously digested with the same restric-
tion enzymes. The resultant plasmid, pDPKA, was intro-
duced into E. coli BL21(DE3) to provide us with
recombinant DpkA.
Purification of recombinant NMAADH
E. coli BL21(DE3) carrying pDPKA was cultivated in
Luria–Bertani medium containing 100 lgÆL
)1
ampicillin at
37 °C for 14 h. The culture was supplemented with 1 mm
isopropyl b-d-thiogalactopyranoside and grown for 3 h.
The wet cells (3.3 g) obtained by centrifugation were sus-
pended in 28 mL of our standard buffer: a 20 mm Tris ⁄ HCl
buffer (pH 7.0) containing 1 mm phenylmethanesulfonyl
fluoride. The crude extract obtained by sonication was loa-
ded on to a Green-Sepharose CL-4B column (100 mL)
equilibrated with the standard buffer. The enzyme was elut-
ed with a linear gradient of 0–1 m NaCl in the buffer. The

ia
and
K
ib
are the dissociation constants for the enzyme–NADPH
complex and the enzyme–pyruvate complex, respectively.
Product inhibition studies were performed with various
concentrations of either NADP
+
or N-methyl-l-alanine as
one substrate and a fixed saturating concentration of the
other. The data were fitted to Eqns (2), (3) and (4), which
describe the competitive, uncompetitive, and noncompeti-
tive inhibition patterns, respectively. P is the concentration
of the product, K
is
is the inhibition constant from the slope
H. Mihara et al. N-Methyl-L-amino acid dehydrogenase from P. putida
FEBS Journal 272 (2005) 1117–1123 ª 2005 FEBS 1121
term, and K
ii
is the inhibition constant from the intercept
term.
v ¼ VABC=ðK
ia
K
ib
K
c
þ K

ii
Þ ð4Þ
Analytical size-exclusion chromatography
The protein quaternary structure was analyzed by an
A
¨
KTAexplorer system (Amersham Biosciences, Amersham,
Buckinghamshire, UK) using a YMC-Pack Diol 200 col-
umn (YMC Co, Ltd, Kyoto, Japan). The column was
equilibrated and operated at a flow rate of 1.0 mLÆmin
)1
with a 0.1 m potassium phosphate buffer (pH 7.0) contain-
ing 0.2 m NaCl. The protein standards used were cyto-
chrome c, myokinase, enolase, lactate dehydrogenase, and
glutamate dehydrogenase from Oriental Yeast, Osaka,
Japan.
Other analytical methods
The N-terminal amino-acid sequence of the enzyme was
determined with an automated Shimadzu PPSQ10 protein
sequencer (Kyoto, Japan). The nucleotide sequence of
DNA was determined with an Applied Biosystems 370A
DNA sequencer (Foster City, CA, USA).
Acknowledgements
This work was supported in part by a Grant-in-Aid
for Scientific Research on Priority Areas (B) 13125203
(to NE) from the Ministry of Education, Culture,
Sports, Science, and Technology of Japan, by Grant-
in-Aid for Encouragement of Young Scientists
15780070 (to HM) from the Japan Society for the Pro-
motion of Science, by the National Project on Protein

8 Zocher R, Keller U & Kleinkauf H (1982) Enniatin
synthetase, a novel type of multifunctional enzyme cata-
lyzing depsipeptide synthesis in Fusarium oxysporum.
Biochemistry 21, 43–48.
9 Strom MS, Nunn DN & Lory S (1993) A single bifunc-
tional enzyme, PilD, catalyzes cleavage and N-methyla-
tion of proteins belonging to the type IV pilin family.
Proc Natl Acad Sci USA 90, 2404–2408.
10 Shibata K, Tarui A, Todoroki N, Kawamoto S,
Takahashi S, Kera Y & Yamada R (2001) Occurrence
of N-methyl-l-aspartate in bivalves and its distribution
compared with that of N-methyl-d-aspartate and d,l-
aspartate. Comp Biochem Physiol B Biochem Mol Biol
130, 493–500.
11 Tarui A, Shibata K, Takahashi S, Kera Y, Munegumi
T & Yamada RH (2003) N-methyl-d-glutamate and
N-methyl-l-glutamate in Scapharca broughtonii (Mol-
lusca) and other invertebrates. Comp Biochem Physiol B
Biochem Mol Biol 134, 79–87.
12 Cannon JR & Williams JR (1982) The alkaloids of
gastrolobium-callistachys. Aust J Chem 35, 1497–1500.
13 Shaw WV, Tsai L & Stadtman ER (1966) The enzy-
matic synthesis of N-methylglutamic acid. J Biol Chem
241, 935–945.
14 Kung HF & Wagner C (1970) The enzymatic synthesis
of N-methylalanine. Biochim Biophys Acta 201, 513–516.
15 Urakami T, Araki H, Oyanagi H, Suzuki KI &
Komagata K (1992) Transfer of Pseudomonas-Amino-
vorans (Dendooren Dejong 1926) to Aminobacter
General-Nov as Aminobacter-Aminovorans Comb-Nov

and characterization of the crown gall specific
enzyme nopaline synthase. Biochemistry 18, 3755–
3760.
24 Brunhuber NM & Blanchard JS (1994) The biochemis-
try and enzymology of amino acid dehydrogenases. Crit
Rev Biochem Mol Biol 29, 415–467.
25 Dairi T & Asano Y (1995) Cloning, nucleotide sequen-
cing, and expression of an opine dehydrogenase gene
from Arthrobacter sp. strain 1C. Appl Environ Microbiol
61, 3169–3171.
26 Rossmann MG, Moras D & Olsen KW (1974) Chemical
and biological evolution of nucleotide-binding protein.
Nature 250, 194–199.
27 Muramatsu H, Mihara H, Kakutani R, Yasuda M,
Ueda M, Kurihara T & Esaki N (2004) Enzymatic
synthesis of N-methyl-l-phenylalanine by a novel
enzyme, N-methyl-l-amino acid dehydrogenase, from
Pseudomonas putida. Tetrahedron: Asymmetr 15, 2841–
2843.
28 Smith EL, Austen BM, Blumenthal KM & Nyc JF
(1975) Glutamate dehydrogenase. In The Enzymes, 3rd
edn. (Boyer, P D, ed.), pp. 293–367. Academic Press,
New York.
29 Ohshima T, Misono H & Soda K (1978) Properties of
crystalline leucine dehydrogenase from Bacillus sphaeri-
cus. J Biol Chem 253, 5719–5725.
30 Hummel W, Weiss N & Kula MR (1984) Isolation and
characterization of a bacterium possessing l-phenylala-
nine dehydrogenase-activity. Arch Microbiol 137, 47–52.
31 Asano Y, Yamaguchi K & Kondo K (1989) A new

Holmes M, et al. (2002) Complete genome sequence and
comparative analysis of the metabolically versatile Pseu-
domonas putida KT2440. Environ Microbiol 4, 799–808.
H. Mihara et al. N-Methyl-L-amino acid dehydrogenase from P. putida
FEBS Journal 272 (2005) 1117–1123 ª 2005 FEBS 1123