Investigation of a substrate-specifying residue within Papaver somniferum and Catharanthus roseus aromatic amino acid decarboxylases

Article abstract of DOI:10.1016/j.phytochem.2014.07.007

Plant aromatic amino acid decarboxylases (AAADs) catalyze the decarboxylation of aromatic amino acids with either benzene or indole rings. Because the substrate selectivity of AAADs is intimately related to their physiological functions, primary sequence data and their differentiation could provide significant physiological insights. However, due to general high sequence identity, plant AAAD substrate specificities have been difficult to identify through primary sequence comparison. In this study, bioinformatic approaches were utilized to identify several active site residues within plant AAAD enzymes that may impact substrate specificity. Next a Papaver somniferum tyrosine decarboxylase (TyDC) was selected as a model to verify our putative substrate-dictating residues through mutation. Results indicated that mutagenesis of serine 372 to glycine enables the P. somniferum TyDC to use 5-hydroxytryptophan as a substrate, and reduces the enzyme activity toward 3,4-dihydroxy-L-phenylalanine (dopa). Additionally, the reverse mutation in a Catharanthus roseus tryptophan decarboxylase (TDC) enables the mutant enzyme to utilize tyrosine and dopa as substrates with a reduced affinity toward tryptophan. Molecular modeling and molecular docking of the P. somniferum TyDC and the C. roseus TDC enzymes provided a structural basis to explain alterations in substrate specificity. Identification of an active site residue that impacts substrate selectivity produces a primary sequence identifier that may help differentiate the indolic and phenolic substrate specificities of individual plant AAADs.

Full text of DOI:10.1016/j.phytochem.2014.07.007

Phytochemistry 106 (2014) 37–43
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Investigation of a substrate-specifying residue
within Papaver somniferum and Catharanthus roseus
aromatic amino acid decarboxylases
⇑
Michael P. Torrens-Spence, Michael Lazear, Renee von Guggenberg, Haizhen Ding, Jianyong Li
Department of Biochemistry, Virginia Tech, Blacksburg, VA, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Plant aromatic amino acid decarboxylases (AAADs) catalyze the decarboxylation of aromatic amino acids
with either benzene or indole rings. Because the substrate selectivity of AAADs is intimately related to
their physiological functions, primary sequence data and their differentiation could provide significant
physiological insights. However, due to general high sequence identity, plant AAAD substrate specificities
have been difficult to identify through primary sequence comparison. In this study, bioinformatic
approaches were utilized to identify several active site residues within plant AAAD enzymes that may
impact substrate specificity. Next a Papaver somniferum tyrosine decarboxylase (TyDC) was selected as
a model to verify our putative substrate-dictating residues through mutation. Results indicated that
mutagenesis of serine 372 to glycine enables the P. somniferum TyDC to use 5-hydroxytryptophan as a
substrate, and reduces the enzyme activity toward 3,4-dihydroxy-L-phenylalanine (dopa). Additionally,
the reverse mutation in a Catharanthus roseus tryptophan decarboxylase (TDC) enables the mutant
enzyme to utilize tyrosine and dopa as substrates with a reduced affinity toward tryptophan. Molecular
modeling and molecular docking of the P. somniferum TyDC and the C. roseus TDC enzymes provided a
structural basis to explain alterations in substrate specificity. Identification of an active site residue that
impacts substrate selectivity produces a primary sequence identifier that may help differentiate the ind-
olic and phenolic substrate specificities of individual plant AAADs.
Received 10 March 2014
Received in revised form 2 June 2014
Available online 5 August 2014
Keywords:
Papaver somniferum
Papaveraceae
Catharanthus roseus
Apocynaceae
Aromatic amino acid decarboxylase
Tyrosine decarboxylases
Tryptophan decarboxylases
Ó 2014 Elsevier Ltd. All rights reserved.
1
. Introduction
Aromatic amino acid decarboxylases (AAADs) are a group of
decarboxylases (TDCs), tyrosine decarboxylases (TyDCs) and aro-
matic acetaldehyde synthases (AASs), respectively. TDCs catalyze
decarboxylation of tryptophan and 5-hydroxytryptophan (Noé
et al., 1984; De Luca et al., 1988; Lopez-Meyer and Nessler, 1997;
Yamazaki et al., 2003; Kang et al., 2007; Park et al., 2009), whereas
TyDCs engender decarboxylation of tyrosine and dopa (Facchini
and De Luca, 1995; Lehmann and Pollmann, 2009; Torrens-
Spence et al., 2012, 2013) and AASs the decarboxylation-oxidative
deamination of dopa, tyrosine and phenylalanine (Kaminaga et al.,
2006; Gutensohn et al., 2011; Torrens-Spence et al., 2012),
respectively.
Differences in substrate selectivity and activity among plant
TyDCs, TDCs and AASs enable individual enzymes to generate spe-
cific products with unique physiological functions. For example,
plant TDCs are required for synthesis of monoterpenoid indole
alkaloids that comprise a diverse group of hundreds of pharmaco-
logically active compounds (Meijer et al., 1993; Berlin et al., 1994;
Facchini et al., 2000), TyDCs are known to function in several
different metabolic pathways including the biosynthesis of
simple alkaloids, complex benzylisoquinoline alkaloids, and
N-hydroxycinnamic acid amides (Leete and Marion, 1953; Ellis,
economically important and phylogenetically diverse enzymes
that are categorically lined through their pyridoxal 5-phosphate
(
PLP) dependence and sequence homology. This family of enzymes
has been studied extensively in mammals where a single enzyme,
,4-dihydroxy-L-phenylalanine (dopa) decarboxylase (DDC), cata-
3
lyzes decarboxylation of both phenolic and indolic amino acids to
generate their corresponding aromatic amines. Mammalian DDC is
responsible for the decarboxylation of dopa and 5-hydroxytrypto-
phan to yield the neurotransmitters dopamine and serotonin,
respectively (Srinivasan and Awapara, 1978; Zhu and Juorio,
1
995). Unlike the single mammalian DDC enzyme, plant AAADs
have evolved with variations in activity and substrate specificity.
The resulting paralogs are formally identified as tryptophan
⇑
Corresponding author. Address: Department of Biochemistry, Engel Hall 204,
Virginia Tech, Blacksburg, VA 24061, United States. Tel.: +1 540 321 5779; fax: +1
5
40 231 9070.
E-mail address: lij@vt.edu (J. Li).
http://dx.doi.org/10.1016/j.phytochem.2014.07.007
031-9422/Ó 2014 Elsevier Ltd. All rights reserved.
0

3
8
M.P. Torrens-Spence et al. / Phytochemistry 106 (2014) 37–43
1
983; Marques and Brodelius, 1988; Trezzini et al., 1993; Facchini
et al., 2000) and AAS enzymes catalyze the production of volatile
flower scents, floral attractants, and defensive phenolic acetalde-
hyde secondary metabolites (Kaminaga et al., 2006; Gutensohn
et al., 2011; Torrens-Spence et al., 2012). It is clear that the phys-
iological functions of plant AAADs are closely related to their
respective activities and substrate specificities; however, due to
the subtlety of the enzymatic divergence of plant AAADs, it has
historically been difficult to predict the function of any given plant
AAAD paralog through sequence comparison. Consequently, being
able to distinguish the activity and substrate specificity of individ-
ual plant AAADs is of practical significance.
In previous work, it was determined that a single active site res-
idue is capable of dictating the activity of plant TyDCs, TDCs and
AASs without altering substrate selectivity. Specifically, the pres-
ence of a tyrosine in an active site catalytic loop dictates decarbox-
ylation chemistry while a phenylalanine substitution at the same
location dictates aldehyde synthase chemistry (Torrens-Spence
et al., 2013). The alteration in enzymatic activity due to a single
amino acid substitution suggests that the functional variations in
plant AAADs may be dictated by a small number of residues. Such
subtle variations in plant AAADs likely explain the difficulty in dif-
ferentiating plant TyDCs, TDCs and AASs through sequences
analysis.
In this study, to further distinguish plant AAADs an attempt to
identify specific structural identifiers capable of differentiating
between the substrate specificity of TDCs and TyDCs was made.
To do so, bioinformatic analyses of characterized plant AAAD
enzymes were performed with predicted several target residues
that might potentially impact substrate specificity in plant AAADs
and these residues assessed through site-directed mutations using
a Papaver somniferum TyDC and a Catharanthus roseus TDC. Results
of the biochemical analyses, in conjunction with molecular model-
ing, provide insights into a residue that impacts plant AAAD indole
and benzene substrate selectivity.
Fig. 1. Dendrogram of recombinantly characterized plant TyDC and TDC sequences.
TDC 2 and C. annuum TDC 2. The sequence and phylogenetic ambi-
guity of TyDC and TDC enzymes produces significant challenges in
plant AAAD gene annotation.
Multiple sequence alignments of characterized TyDC and TDC
sequences were used to identify possible substrate specifying res-
idues. Analysis of residues conserved within 9 or more of the 12
characterized TDC and TyDC sequences yielded twenty-two puta-
tive substrate-specifying residues. These conserved residues were
highlighted in the P. somniferum TyDC 9 sequence (subsequently
referred to as P. somniferum TyDC) (Torrens-Spence et al., 2013)
(Fig. 2). Next, the previously characterized P. somniferum TyDC
was modeled and a carbidopa external aldimine superimposed into
the active site. Residues within four angstroms of the carbidopa
external aldimine were highlighted to produce an additional 18
residues potentially involved in substrate recognition (Fig. 2).
Cross-referencing these active site proximal residues with the con-
served TyDC/TDC residues reduced the hypothetical substrate
specifying residues down to 5 individual amino acids (Fig. 2). These
residues are represented as serine 101, cysteine 170, asparagine
2
. Results
2.1. Primary sequence investigation of recombinant characterized
TyDC and TDC sequences
To begin the investigation of residues that might impact sub-
strate specificity of plant AAAD enzymes, a primary sequence eval-
uation of essentially all the characterized plant TyDC and TDC
sequences was performed (Supplemental Table S1). Pairwise align-
ments of these characterized plant AAAD enzymes indicated that
inter TDC–TyDC class identities are often greater than that of intra
TDC or TyDC classes. For example, despite differences in substrate
specificity, the Capsicum annuum TDC 2 (Park et al., 2009) shares
greater homology with the Arabidopsis thaliana TyDC (Lehmann
and Pollmann, 2009) (66% identity) than it does with the Camptot-
heca acuminata TDC 2 (Lopez-Meyer and Nessler, 1997) (56% iden-
tity). The sequence ambiguity of plant AAADs can be further
illustrated in a dendrogram of characterized TyDC and TDC
sequences (Fig 1). As this figure illustrates, TyDC and TDC
sequences did not appear to exclusively cluster according to their
substrate preference. Despite maintaining identical substrate pro-
files, the characterized TDC sequences appeared to cluster into
two groups. Interestingly, TDC sequences from both Oryza sativa
and C. annuum appear in both clusters. This suggests that this clus-
tering is not due to evolutionary divergence of plant species but
rather the evolutionary divergence of individual AAAD sequences.
In addition to the delocalization of the TDC genes, the dendrogram
suggest variable clustering of characterized TyDC sequences. P.
somniferum TyDC 7, P. somniferum TyDC 9 and Thalictrum flavum
TyDC cluster together while A. thaliana TyDC clusters with O. sativa
3
18, alanine 319 and serine 372 within the P. somniferum TyDC
sequence.
2.2. Generation and investigation of putative substrate specifying P.
somniferum TyDC mutants
Before evaluating the substrate specifying roles of these select
five residues, recombinant P. somniferum TyDC was first evaluated
to determine the substrate range of the wild type TyDC. The
recombinant wild type TyDC was expressed and purified to homo-
geneity. Activity assay results showed that it efficiently catalyzed
the decarboxylation of dopa and tyrosine (at 2 mM substrate con-
centration, the enzyme showed 5910 ± 200 nmol/min/mg protein
to dopa and 5350 ± 200 nmol/min/mg protein to tyrosine), but dis-
played no activity towards phenylalanine or indolic substrates,
including 5-hydroxytryptophan and tryptophan. Next, using the
P. somniferum TyDC as a model, site directed mutations of S101A,
C170A, N318S, A319P and S372G were made based on conserved
active site proximal residues.
To test perturbations in substrate specificity, the P. somniferum
TyDC S101A, C170S, N318S, A319P, and S372G mutant enzymes
were cloned, expressed, and purified to homogeneity. The activi-
ties of the mutant enzymes were tested using dopa, tyrosine,

M.P. Torrens-Spence et al. / Phytochemistry 106 (2014) 37–43
39
Fig. 2. Putative substrate specifying residues form P. Somniferum TyDC 9. Green residues are mostly conserved (residues are conserved between substrate specificities for 8 or
more of the 11 characterized TDC and TyDC sequences). Red residues are within 4 Å of the carbidopa external aldimine from the TyDC 9 homology model. Yellow residues are
amino acids that appear as both green and red residues. Residues in bold and italic indicate amino acids within 4 Å of the carbidopa external aldimine from the beta chain of
the homodimer. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
phenylalanine, tryptophan and 5-hydroxytryptophan. Results
indicated that the S101A, C170S, N318S and A319P mutants had
no obvious alterations in indolic or phenolic substrate specificity.
However, the S372G mutant displayed a broader substrate profile.
The S372G mutant retained activity towards dopa and tyrosine,
had no measurable activity towards tryptophan (Supplementary
Fig. 1A–D), but displayed activity towards 5-hydroxytryptophan
of the S372G and wild type P. somniferum TyDC enzymes was per-
formed using the preferred substrate (dopa) and the putative ind-
olic substrate (5-hydroxytryptophan). The calculated kinetic
parameters indicated that the mutation reduced the enzyme’s
ability to catalyze decarboxylation of dopa, while enabling a
new but weak 5-hydroxytryptophan decarboxylase activity
(Table 1). No kinetic data for 5-hydroxytryptophan was obtain-
able for the wild type TyDC due to the lack of measurable product
formation. Comparative sequence analysis of residues homolo-
gous to the P. somniferum TyDC 372 residue revealed that serine
was stringently conserved in all experimentally verified TyDC
(Fig 3A–D). Initial biochemical testing suggested that the P. som-
niferum TyDC serine 372 residue might play a role in differentiat-
ing phenolic substrates from indolic substrates. To measure
alterations in substrate selectivity, a full kinetic characterization
A
B
C
D
E
F
G
H
Fig. 3. HPLC-EC detection of products from P. somniferum TYDC S372G and C. roseus TDC G370S mutants. The Y-axis represents the output in microamps, the x-axis represents
retention time. Chromatogram (A) illustrates lack of product formed from P. somniferum TyDC wild type and 5-hydroxytryptophan after 20 min of incubation.
Chromatograms (B) illustrate the formation of 5-hydroxytryptamine generated from 5-hydroxytryptophan and the P. somniferum TyDC S372G mutant after 10 min of
incubation. Chromatograms (C) illustrate the formation of 5-hydroxytryptamine generated from 5-hydroxytryptophan and the P. somniferum TyDC S372G after 20 min of
incubation. Chromatogram (D) shows the peak generated from the injection of 150 pmol of 5-hydroxytryptamine standard. Chromatogram (E) illustrates the lack of product
formed from the C. roseus TDC wild type and the dopa reaction mixture after 20 min of incubation. Chromatogram (F) illustrate the formation of dopamine generated from
incubating dopa and the C. roseus TDC G370S mutant for 10 min. Chromatograms (G) illustrate the formation of dopamine generated from dopa and the C. roseus TDC G370S
mutant after 20 min of incubation. Chromatogram (H) shows the peak generated from the injection of 150 pmol of dopamine standard.

4
0
M.P. Torrens-Spence et al. / Phytochemistry 106 (2014) 37–43
Table 1
Kinetic parameters of Papaver somniferum wild type and mutant TyDC9 enzymes.
Substrate
Enzyme
kcat (1/s)
Error (1/s)
K
m
(mM)
Error (mM)
kcat/K
m
(1/(s mM))
Error (1/(s mM))
Dopa
Dopa
PsTyDC (WT)
PsTyDC (S372G)
PsTyDC (WT)
PsTyDC (S372G)
CrTDC (WT)
8.02
1.47
ND
0.26
2.54
1.37
ND
0.46
0.30
ND
0.03
0.10
0.11
ND
0.75
2.10
ND
3.95
0.12
0.21
ND
0.10
0.20
ND
0.45
0.02
0.05
ND
10.97
0.72
ND
0.07
21.92
7.05
ND
0.66
0.21
ND
0.01
4.49
2.20
ND
5
5
-Hydroxytryptophan
-Hydroxytryptophan
Tryptophan
Tryptophan
Dopa
CrTDC (G370S)
CrTDC (WT)
Dopa
CrTDC (G370S)
0.37
0.04
3.40
0.21
0.11
0.01
Values represent mean SE (n = 3).
ND, not determined.
a carbidopa external aldimine was superimposed into the active
site of the models. Model analysis suggests likely involvement of
the serine 372 (TyDC sequence) and glycine 370 (TDC sequence)
in substrate recognition. In the ligand bound models, this residue
was located in the back of the active site approximately three to
0
five angstroms away from the 3 hydroxyl of the carbidopa external
aldimine (Fig 5). Comparison of the wild type and mutant models
indicated that the mutations altered the size of the active site
pocket by approximately one and a half angstroms. For example,
within the P. somniferum TyDC model, the serine 372 glycine muta-
tion increases the distance between the 3 prime hydroxyl group of
the carbidopa ligand and the 372 residue from 3.6 Å to 5.0 Å. Con-
versely, the reverse mutation in the C. roseus TDC enzyme
decreases the distance from 5.0 Å to 3.6 Å. The theoretical altera-
tions in active site volume might enable the accommodation of
specific sizes of substrates, whereas the structurally larger serine
residue may reduce the size of the active site to accommodate phe-
nolic substrates while the structurally smaller glycine residue may
increase the size of the active site to accommodate larger indolic
compounds. Biochemical characterization of the substrate-impact-
ing residue, in coordination with the ligand bound homology
model analyses, helps to illuminate the functional roles of these
active site residues.
Fig. 4. Sequence alignment of a key residue within the characterized TDC and TyDC
sequences. Within the TDC sequences, the residue homologous to the P. somniferum
TYDC serine 372 is highlighted in red. Within the TyDC sequences, the residue
homologous to the P. somniferum TYDC serine 372 is highlighted in yellow. (For
interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)
enzymes, while glycine was conserved in all experimentally
verified TDC enzymes (Fig 4).
2.3. Generation and biochemical investigation of the C. roseus G 370S
TDC mutant
3
. Discussion
Based on the findings from the mutation of the S372 TyDC res-
idue, a similar study was conducted to broaden the substrate spec-
ificity of a TDC to include phenolic substrates. A C. roseus TDC was
selected as the candidate for this study and a TDC G370S mutant
was expressed, purified and evaluated for altered substrate speci-
ficity. This TDC, like other verified TDCs, contains the conserved
glycine 370 (equivalent to serine 372 of the P. somniferum TyDC)
Previous work identified a single residue responsible for differ-
entiation of AAAD and AAS activity. Within plant AAADs and AASs,
the presence of the active site tyrosine residue appears to dictate
(Fig 4). An electrochemical HPLC assay determined that, unlike
the wild type enzyme, the C. roseus TDC G370S mutant was able
to catalyze decarboxylation of dopa (Fig 3E–H) and tyrosine (Sup-
plementary Fig. 2A–D). The C. roseus TDC G370S enzyme displayed
no activity towards phenylalanine (Supplementary Fig. 1E–G). Full
kinetic characterization of the G370S and wild type C. roseus TDCs
showed significant alterations in the enzyme’s substrate profile.
Comparison of the kinetic values between the wild type TDC and
the mutant TDC illustrated that the glycine 370 serine substitution
reduces binding affinity and the rate of the enzyme towards tryp-
tophan while enabling the enzyme to utilize dopa as a substrate
(
Table 1).
Fig. 5. Homology model active site analysis of the P. somniferum TyDC wild type,
the P. somniferum TyDC S372G, the C. roseus TDC wild type and the C. roseus TDC
G370S mutant. Ligands and models were generated using the PDB model 1JS3. The
blue structure represents the external aldimine of carbidopa and PLP. The magenta
residue represents the substrate impacting amino acid. Model (A) illustrates the
distance between the ligand and P. somniferum TyDC wild type 372 residue. Model
2
.4. Molecular modeling of the substrate-impacting residue
Homology models of P. somniferum TyDC wild type, P. somnife-
(
B) illustrates the distance between the ligand and C. roseus TDC wild type 370
rum TyDC S372G, C. roseus TDC wild type and C. roseus TDC G370S
were generated and analyzed in an effort to investigate the struc-
ture–function relationship of the specificity-impacting residue. To
reveal the possible interactions of the enzyme with its substrates,
residue. Model (C) illustrates the distance between the ligand and P. somniferum
TyDC mutant residue 372. Model (D) illustrates the distance between the ligand and
C. roseus TDC mutant residue 370. (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)

M.P. Torrens-Spence et al. / Phytochemistry 106 (2014) 37–43
41
decarboxylation activity, while a phenylalanine substitution seems
tobe responsible for decarboxylation-oxidative deamination activ-
ity (Torrens-Spence et al., 2013). The resulting research indicates
that biochemical and biophysical characteristics of individual
active site amino acids may promote/enhance specific catalytic
reactions; thereby dictating their reaction chemistry. Conse-
quently, it is reasonable to suggest that the molecular mechanism
responsible for differentiating plant AAADs may be dictated by a
select few amino acids.
TDCs and TyDCs have different physiological functions; there-
fore being able to distinguish substrate selectivity without lab-
intensive experimental verification is highly useful. Despite their
high sequence identity, it is well known that plant TyDCs do not
use indolic amino acids as substrate and TDCs have no any activity
to amino acids with benzene rings. The rigorous substrate selectiv-
ity of both TDCs and TyDCs has been stressed in many publications
residue (implicating indolic substrate specificity) and the presence
of the homologous Y350 residue (implicating decarboxylation
activity and not aldehyde synthase activity) (Torrens-Spence
et al., 2013), it can be speculated that the annotated Fragaria vesca
TyDC, Medicago truncatula TyDC, Setaria italica TyDC and
Theobroma cacao TyDC sequences are all likely TDCs. Such primary
sequence annotation illustrates the practicality of substrate speci-
ficity and activity differentiation within plant type II PLP decarbox-
ylases. The study herein is, therefore, one step towards a better
classification of plant TyDCs and TDCs.
4
. Concluding remarks
The functional evolution of AAAD enzymes within different
branches of the phylogenetic tree reflects the physiological needs
of different species. In plants, AAADs are an integral part of the
pathways for the biosynthesis of numerous chemical compounds
responsible for mitigating interactions with their biotic and abiotic
environments (Leete and Marion, 1953; Ellis, 1983; Meijer et al.,
(De Luca et al., 1988; Facchini and De Luca, 1995; Facchini et al.,
2
000). Moreover, sequence analysis and hydropathy plots suggest
that the unique substrate specificity of plant TyDCs and TDCs are
the result of relatively minor amino acid substitutions (De Luca
et al., 1988; Facchini and De Luca, 1995; Facchini et al., 2000).
In this study, attempts were made to identify key residues capa-
ble of differentiating the phenolic selective TyDCs from indolic
selective TDCs. Extensive comparative analyses of characterized
TyDC and TDC sequences enabled us to select several potential tar-
get residues. Site-directed mutagenesis of these residues within
two model enzymes provided evidence to suggest that the P. som-
niferum TyDC 372 residue and the C. roseus TDC 370 residue collec-
tively impact substrate selectivity. The new found activity of the P.
somniferium S372G TyDC towards 5-hydroxytryptophan and the C.
roseus G370S TDC towards dopa appears to be significant. The
kinetic values of these mutant enzymes and their new substrates
are comparable to some characterized AAAD enzymes. For exam-
ple, the kcat values of the two mutant enzymes towards their
new substrates are larger than the kcat values of C. annuum TDC
1
2
993; Trezzini et al., 1993; Berlin et al., 1994; Facchini et al.,
000). For example, P. somniferum contains approximately 15 TyDC
sequences that play roles in production of benzylisoquinoline alka-
loids, such as the pharmacologically active morphine, codeine and
papaverine (Facchini and De Luca, 1995; Facchini et al., 2000).
However, despite variable functions, plant TyDCs and TDCs do
not have easily recognizable motifs in primary sequences.
Although it is sometimes possible to classify the substrate specific-
ity of a given AAAD through extensive sequence comparison, plant
AAAD differentiation remains quite speculative. The findings
regarding the S372G and G370S mutants indicate that the phenolic
and indolic substrates in plant AAADs could be primarily dictated
by this single active site amino acid. The stringent conservation
for serine and glycine in verified plant TyDC and TDC, respectively,
support this consideration (Fig 4). Future research is needed to
fully reveal the active site conformations and residues invariably
conserved amongst distinct plant AAAD classes. Upon completion,
AAAD fingerprints should enable the proper annotation of AAAD
and AAS genes without expensive and laborious enzyme expres-
sion and characterization.
1
and C. annuum TDC 2 [Park et al., 2009] towards their preferred
substrate tryptophan. Moreover, the substrate specificity of the
mutants towards their new substrates is comparable to those of
an investigated orphan TDC enzyme (K
m
3.0 mM towards trypto-
phan) (Gibson et al., 1972). Although both mutations enabled
expanded substrate profiles, the substitutions did not lead to loss
of their original activities. However, it is reasonable to suggest that
this active site residue is a key amino acid responsible for impact-
ing substrate specificity. This hypothesis is supported by the gly-
cine conservation in verified TDCs and the serine conservation in
verified TyDCs (Fig. 4).
Subsequent analysis of the active site conformation of the C.
roseus TDC and P. somniferum TyDC ligand bound homology models
provides some structural basis for alterations in substrate profiles.
It can be proposed that the mutation of the P. somniferum 372 res-
idue from serine to glycine enables broader substrate selectivity by
increasing the active site to accommodate structurally large sub-
strates like 5-hydroxytryptophan, while the C. roseus G370S muta-
tion reduces the size of the active site to accommodate the
structurally smaller dopa. Molecular interaction measurements
support this hypothesis. Homology models of the two mutants
illustrate an alteration of active site pocket of 1.4 Å. The altered
volume of the active site could enable structurally different com-
pounds to enter the active site and form the external aldimine with
the PLP cofactor.
5
. Experimental
5.1. Reagents
Tyrosine, tyramine, tryptophan, tryptamine, 5-hydroxytrypto-
phan, 5-hydroxytryptamine, dopa, dopamine, phenylalanine,
phenylethylamine, pyridoxal 5-phosphate (PLP), phthaldialdehyde,
HCO H, and CH CN were purchased from Sigma (St. Louis, MO).
2 3
The IMPACT-CN protein expression system was purchased from
New England Biolabs (Ipswich, MA).
5.2. Preparation of recombinant proteins
P. somniferum and C. roseus RNA extraction, cDNA production,
vector cloning and wild type protein expression were conducted
as previously described (Torrens-Spence et al., 2013). Primer pairs
were synthesized and used for amplification and SapI mutagene-
sis of the expression mutants (Supplemental Table S2). The
resulting PCR products were ligated together into IMPACT-CN
bacterial expression plasmids. DNA sequencing was utilized to
verify the sequence and frame of each mutant cDNA insert. Trans-
formed bacterial colonies, expressing the wild type and mutant
proteins, were selected and used for large-scale expression of
individual recombinant proteins. Bacterial cells were cultured at
37 °C. After induction with 0.15 mM IPTG, the cells were cultured
A brief Genbank evaluation of this 372 residue AAAD has been
performed in an effort to annotate non-characterized plant AAAD
sequences using the results from this study. An analysis of the NCBI
plants (taxid: 3193) database enables tentative identification of
several sequences annotated as TyDCs that are likely TDCs. For
example, based on the conservation of the homologous G372

4
2
M.P. Torrens-Spence et al. / Phytochemistry 106 (2014) 37–43
at 15 °C for 24 h. The soluble fusion proteins were applied to a
column packed with chitin beads and subsequently hydrolyzed
under reducing conditions. The affinity purification resulted in
the isolation of each individual recombinant protein at about
6. Bioinformatic analysis sequences
All of the following accession numbers are from the NCBI
Genbank
8
5% purity. Further purifications of recombinant proteins were
C. acuminata TDC
[AAB39709], C. annuum TDC 1 [ACN62127], C. annuum TDC 2
[ACN62126], O. pumila TDC [BAC41515], O. sativa TDC
[AK069031], O. sativa TDC 2 [AK103253], C. roseus TDC [P17770],
P. somniferum TYDC [AAC61842], P. somniferum TyDC
[AAC61843], T. flavum TyDC [AAG60665], and A. thaliana TyDC
[NP_001078461], P. crispum TyDC [Q06086], S. lycopersicum AAAD
1A [NP_001233845], S. lycopersicum AAAD 1B [NP_001233852], S.
lycopersicum AAAD 2 [NP_001233859], P. somniferum TyDC 1
1 [AAB39708], C. acuminata TDC 2
achieved by Mono-Q and gel filtration chromatography. Purified
recombinant enzymes were concentrated to approximately
1
PLP using a Centricon YM-50 concentrator (Millipore). Purity of
the recombinant proteins was evaluated by SDS–PAGE. The con-
centrations of the purified recombinant proteins were determined
by a Bio-Rad protein assay kit (Hercules, CA) using bovine serum
albumin as a standard.
1
0 mg/ml protein in 25 mM HEPES (pH 7), containing 0.04 mM
9
7
[
[
P54768], P. somniferum TyDC 2 [P54769], Fragaria vesca TyDC
XP_004292248], Medicago truncatula TyDC [XP_003625397],
5
.3. Homology modeling and ligand superimposition
Setaria italica TyDC [XP_004956078] and Theobroma cacao TyDC
EOX99271].
[
Modeller (Šali and Blundell, 1993) was utilized to produce ten P.
somniferum TyDC and C. roseus TDC homology models for the sub-
strate analog bound protein data bank models 1JS3 (Burkhard
et al., 2001). The model from each PDB template with the optimal
molpdf and DOPE score were selected for further evaluation.
Mutant homology models were generated in the same manner
from the optimal wild type models. Pymol was utilized to align
the PLP coenzyme and substrate analog from 1JS3 (Burkhard
et al., 2001) onto the corresponding homology models and to visu-
alize the active site residues. The residues proximal to the ligand
from the homology models were then compared with their homol-
ogous residues from characterized TyDC and TDC enzymes to iden-
tify potential substrate specifying residues.
Acknowledgement
This study was supported through Virginia Tech Biochemistry
College of Agricultural and Life Sciences funding.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.phytochem.2014.
0
7.007.
References
5.4. Kinetic analysis
Berlin, J., Ruegenhagen, C., Kuzovkina, I.N., Fecker, L.F., Sasse, F., 1994. Are tissue
cultures of Peganum harmala a useful model system for studying how to
manipulate the formation of secondary metabolites? Plant Cell Tissue Organ
Cult. 38, 289–297.
Burkhard, P., Dominica, P., Borri-Voltattorni, C., Jansonius, J.N., Malashkevich, V.N.,
2001. Structural insight into Parkinson’s disease treatment from drug-inhibited
DOPA decarboxylase. Nat. Struct. Biol. 8 (11), 963–967.
De Luca, V., Marineau, C., Brisson, N., 1988. Molecular cloning and analysis of cDNA
encoding a plant tryptophan decarboxylase: comparison with animal dopa
decarboxylase. PNAS 86 (8), 2582–2586.
Ellis, B.E., 1983. Production of hydroxyphenylethanol glycosides in suspension-
cultures of Syringa vulgaris. Phytochemistry 22, 1941–1943.
The specific activity of the wild type P. somniferum TyDC
towards tyrosine and dopa was analyzed in 50
ll phosphate buf-
fer (pH 7.5) containing 2 mM substrate with enzyme (2
lg) for an
incubation time of 2.5 min. This analysis determined that the wild
type enzyme possessed typical TyDC activity. The S372G mutant
displayed both typical TyDC activity and novel decarboxylation
activity towards 5-HTP. The kinetic parameters of the P. somnife-
rum TyDC wild type, P. somniferum S372G mutant, C. roseus TDC
wild type and C. roseus TDC G370S mutant enzymes were subse-
Facchini, P.J., De Luca, V., 1995. Expression in Echerichia coli and partial
characterization of two tyrosine/dopa decarboxylases from opium poppy.
Phytochemistry 38, 1119–1126.
quently analyzed. Reaction mixtures of 50
recombinant protein (25 g for P. somniferum TyDC 5-hydroxytry-
ptophan, 3 g for the P. somniferum TyDC dopa, 25 g for C. roseus
TDC dopa and 3 g for the C. roseus TDC tryptophan) and varying
concentration of substrate (0.05–20 mM of 5-hydroxytryptophan,
.005–5 mM of tryptophan and 0.05–15 mM of dopa) were pre-
pared in 50 mM phosphate buffer (pH 7.5) and incubated at
5 °C. An equal volume of 0.8 M HCO H was added to each reac-
ll assays containing
l
Facchini, P.J., Huber-Allanach, K.L., Tari, L.W., 2000. Plant aromatic L-amino acid
l
l
decarboxylases: evolution, biochemistry, regulation, and metabolic engineering
applications. Phytochemistry 54, 121–138.
Gibson, R.A., Barret, G., Wightman, F., 1972. Biosynthesis and metabolism of indol-
l
3yl-acetic acid: III. Partial purification and properties of a tryptamine-forming
L-
0
tryptophan decarboxylase from tomato shoots. J. Exp. Bot. 23 (3), 775–789.
Gutensohn, M., Klempien, A., Kaminaga, Y., Nagegowda, D.A., Negre-Zakharov, F.,
Huh, J.H., Luo, H., Weizbauer, R., Mengiste, T., Tholl, D., Dudareva, N., 2011. Role
of aromatic aldehyde synthase in wounding/herbivory response and flower
scent production in different Arabidopsis ecotypes. Plant J. 66, 591–602.
Kaminaga, Y., Schnepp, J., Peel, G., Kish, C.M., Ben-Nissan, G., Weiss, D., Orlova, I.,
Lavie, O., Rhodes, D., Wood, K., Porterfield, D.M., Cooper, A.J., Schloss, J.V.,
Pichersky, E., Vainstein, A., Dudareva, N., 2006. Plant phenylacetaldehyde
synthase is a bifunctional homotetrameric enzyme that catalyzes phenylalanine
decarboxylation and oxidation. J. Biol. Chem. 281, 23357–23366.
Kang, S., Kang, K., Lee, K., Back, K., 2007. Characterization of rice tryptophan
decarboxylases and their direct involvement in serotonin biosynthesis in
transgenic rice. Planta 227, 263–272.
Leete, E., Marion, L., 1953. The biogenesis of alkaloids: VI. The formation of
hordenine and N-methyltyramine from tyrosine in barley. Can. J. Chem. 31,
126–128.
2
2
tion mixture (15 min for P. somniferum TyDC 5-hydroxytrypto-
phan assays, 2.5 min for P. somniferum dopa assays, 15 min for
C. roseus TDC dopa assays and 2.5 min for C. roseus tryptophan
assays) with supernatants individually was subjected to HPLC-
electrochemical analysis. Kinetic data points were performed in
triplicate and kinetic values were evaluated by hyperbolic regres-
sion. An isocratic running buffer consisting of 50 mM phosphate
buffer pH 4.3, 25% acetonitrile, and 0.5 mM octyl sulfate was used
for the 5-hydroxytryptophan characterization, whereas an iso-
cratic running buffer consisting of 50 mM phosphate buffer pH
Lehmann, T., Pollmann, T., 2009. Gene expression and characterization of a stress-
induced tyrosine decarboxylase from Arabidopsis thaliana. FEBS Lett. 583, 1895–
4
.3, 15% acetonitrile, and 0.5 mM octyl sulfate was used for the
dopa characterization. The amounts of products in reaction mix-
tures containing different concentrations of substrate were quan-
tified based on a standard curve generated using authentic
standards at identical conditions of HPLC-electrochemical
analysis.
1
900.
Lopez-Meyer, M., Nessler, C.L., 1997. Tryptophan decarboxylase is encoded by two
autonomously regulated genes in Camptotheca acuminata which are
differentially expressed during development and stress. Plant J. 11, 1167–1175.
Marques, I.A., Brodelius, P.E., 1988. Elicitor-induced
L-tyrosine decarboxylase from
plant cell suspension cultures. Plant Physiol. 88, 46–51.

M.P. Torrens-Spence et al. / Phytochemistry 106 (2014) 37–43
43
Meijer, A.H., Verpoorte, R., Hoge, J.H.C., 1993. Regulation of enzymes and genes
involved in terpenoid indole alkaloid biosynthesis in Catharanthus roseus. J.
Plant Res. 3, 145–164.
Noé, W., Mollenschott, C., Berlin, J., 1984. Tryptophan decarboxylase from
Catharanthus roseus cell suspension cultures: purification, molecular and
kinetic data of the homogenous protein. Plant Mol. Biol. 3 (5), 281–288.
Park, S., Kang, K., Lee, K., Choi, D., Kim, Y., Back, K., 2009. Induction of serotonin
biosynthesis is uncoupled from the coordinated induction of tryptophan
biosynthesis in pepper fruits (Capsicum annuum) upon pathogen infection.
Planta 230, 1197–1206.
Šali, A., Blundell, T.L., 1993. Comparative protein modelling by satisfaction of spatial
restraints. J. Mol. Biol. 234, 779–815.
Srinivasan, K., Awapara, J., 1978. Substrate specificity and other properties of DOPA
decarboxylases from guinea pig kidneys. BBA Enzymol. 526, 597–604.
Torrens-Spence, M.P., Gillaspy, G., Zhao, B., Harich, K., White, R.H., Li, J., 2012.
Biochemical evaluation of a parsley tyrosine decarboxylase results in a novel
4-hydroxyphenylacetaldehyde synthase enzyme. Biochem. Biophys. Res.
Commun. 2, 211–216.
Torrens-Spence, M.P., Liu, P., Ding, H., Harich, K., Gillaspy, G., Li, J., 2013.
Biochemical evaluation of the decarboxylation and decarboxylation-
deamination activities of plant aromatic amino acid decarboxylases. J. Biol.
Chem. 288, 2376–2387.
Trezzini, G.F., Horrichs, A., Somssich, I.E., 1993. Isolation of putative defense-related
genes from Arabidopsis thaliana and expression in fungal elicitor-treated cells.
Plant Mol. Biol. 21, 385–389.
Yamazaki, Y., Sudo, H., Yamazaki, M., Aimi, N., Saito, K., 2003. Camptothecin
biosynthetic genes in hairy roots of Ophiorrhiza pumila: cloning,
characterization and differential expression in tissues and by stress
compounds. Plant Cell Physiol. 44, 395–403.
Zhu, M.-Y., Juorio, A.V., 1995. Aromatic
L-amino acid decarboxylase: biological
characterization and functional role. Gen. Pharmacol.: Vasc. Syst. 26, 681–696.