Functional chararacterization of the enzymes TabB and TabD involved in tabtoxin biosynthesis by Pseudomonas syringae

Article abstract of DOI:10.1016/j.bbrc.2018.01.028

Pseudomonas syringae pv. tabaci ATCC 11528 produces tabtoxin, a β-lactam-containing dipeptide phytotoxin. Tabtoxinine-β-lactam (TβL), one of tabtoxin's constituent amino acids, structurally mimics lysine, and many of the proteins encoded by the tabtoxin biosynthetic gene cluster are homologs of lysine biosynthetic enzymes, suggesting that the tabtoxin and lysine biosynthetic routes parallel one another. We cloned and expressed TabB and TabD, predicted homologs of tetrahydrodipicolinate (THDPA)-N-acyltransferase and N-acyl-THDPA aminotransferase, respectively, to determine their activities in vitro. We confirmed that TabB succinylates THDPA and that TabD is a PLP-dependent aminotransferase that utilizes glutamate as an amine donor. Surprisingly, we also found that though TabD could utilize the TabB product N-succinyl-THDPA as a substrate, THDPA itself was also recognized. These observations reveal that TabB functionally duplicates DapD, the THDPA-N-succinyltransferase involved in lysine biosynthesis, and reinforce the close relationship between the metabolic logics underpinning the respective biosynthetic pathways.

Full text of DOI:10.1016/j.bbrc.2018.01.028

Biochemical and Biophysical Research Communications 496 (2018) 212e217
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications
journal homepage: www.elsevier.com/locate/ybbrc
Functional chararacterization of the enzymes TabB and TabD involved
in tabtoxin biosynthesis by Pseudomonas syringae
Margot E. Manning, Eli J. Danson, Christopher T. Calderone*
Department of Chemistry, Carleton College, 1 North College Street, Northfield, MN 55057, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Pseudomonas syringae pv. tabaci ATCC 11528 produces tabtoxin, a b-lactam-containing dipeptide phy-
totoxin. Tabtoxinine-b-lactam (TbL), one of tabtoxin's constituent amino acids, structurally mimics lysine,
Received 11 December 2017
Accepted 4 January 2018
Available online 4 January 2018
and many of the proteins encoded by the tabtoxin biosynthetic gene cluster are homologs of lysine
biosynthetic enzymes, suggesting that the tabtoxin and lysine biosynthetic routes parallel one another.
We cloned and expressed TabB and TabD, predicted homologs of tetrahydrodipicolinate (THDPA)-N-
acyltransferase and N-acyl-THDPA aminotransferase, respectively, to determine their activities in vitro.
We confirmed that TabB succinylates THDPA and that TabD is a PLP-dependent aminotransferase that
utilizes glutamate as an amine donor. Surprisingly, we also found that though TabD could utilize the TabB
product N-succinyl-THDPA as a substrate, THDPA itself was also recognized. These observations reveal
that TabB functionally duplicates DapD, the THDPA-N-succinyltransferase involved in lysine biosynthesis,
and reinforce the close relationship between the metabolic logics underpinning the respective biosyn-
thetic pathways.
Keywords:
Natural product biosynthesis
Tabtoxin
Enzyme function
© 2018 Elsevier Inc. All rights reserved.
1. Introduction
lysine biosynthesis proceeds from THDPA via N-acylation (either
succinylation catalyzed by DapD in the succinylase branch of the
Tabtoxin (1) is a phytotoxin produced by various strains of
Pseudomonas syringae pv. tabaci, the causative agent of wildfire
disease of tobacco [1]. Tabtoxin consists of a dipeptide of threonine
pathway, or acetylation catalyzed by DapH in the acetylase branch
of the pathway, mainly observed in Bacillus); DapC-catalyzed
transamination to form N-acyl-
L,L-diaminopimelate (DAP); deacy-
with the monobactam amino acid tabtoxinine-
Cleavage of the peptide bond by the zinc peptidase TabP releases
L [2], an inhibitor of glutamine synthetase, which in turn leads to
buildup of ammonia and the chlorosis typical of wildfire disease
[3,4]. Very recently, a mechanism for T L inhibition of glutamine
synthetase was reported to involve phosphorylation of the -hy-
droxyl to form a tight-binding inhibitor of the enzyme; this
mechanism is surprising given the presence of the -lactam that
b
-lactam (T
b
L).
lation to form -DAP; epimerization to form meso-DAP; and finally
L,L
LysA-catalyzed decarboxylation to form lysine [7]. (Fig. 1a).
The gene cluster responsible for tabtoxin biosynthesis was re-
ported several years ago [8]. (Fig. 1b) Among the predicted proteins
Tb
b
encoded by the cluster are a
clavaminic acid synthetase (TblC), presumably involved in incor-
poration of the -lactam ring in T L. Recently, Wencewicz and
Walsh identified TblF as the ligase that forms the T L-Thr peptide
bond, and Ttr (a protein encoded separately from the gene cluster)
as a T L acetylase that confers resistance upon the producing or-
b-lactam synthetase (TblS) and a
g
b
b
b
b
often serves as a hydrolytically labile warhead in other molecules
[5].
b
T
b
L and lysine are highly structurally homologous, and in accord
with these structural similarities, early feeding studies showed that
L is derived from tetrahydrodipicolinic acid (THDPA), a precursor
ganism [2]. Notably, the tabtoxin biosynthetic gene cluster also
encodes several homologs of lysine biosynthetic enzymes,
including TabB (homologous to the THDPA acyltransferases DapD/
DapH), TabD (homologous to the acyl-THDPA aminotransferase
DapC), and TabA (homologous to the meso-DAP decarboxylase
LysA).
Tb
to lysine produced from aspartate semialdehyde and pyruvate by
sequential action of the enzymes DapA and DapB [6]. In bacteria,
Despite the apparent parallels with the well-studied lysine
biosynthetic pathway, many of the details of T
remain mysterious. As a first step in our efforts to decipher the logic
bL biosynthesis
* Corresponding author.
E-mail address: ccalderone@carleton.edu (C.T. Calderone).
https://doi.org/10.1016/j.bbrc.2018.01.028
0006-291X/© 2018 Elsevier Inc. All rights reserved.

M.E. Manning et al. / Biochemical and Biophysical Research Communications 496 (2018) 212e217
213
Fig. 1. (A) Biosynthesis of lysine in bacteria. After THDPA biosynthesis catalyzed by DapA and DapB, THDPA is either succinylated by DapD (succinylase branch) or acetylated by
DapH (acetylase branch). Homologs to lysine biosynthetic proteins encoded by the tabtoxin biosynthetic gene cluster are shown at each step. (B) Tabtoxin biosynthetic cluster with
functions predicted by sequence homology.
of this pathway, we cloned and expressed TabB and TabD and
characterized their activities. Our results reveal that TabB acts as a
THDPA-N-succinyltransferase in vitro. We also identified and
characterized a second, functionally redundant THDPA-N-succi-
nyltransferase presumed to be the DapD involved in primary
metabolism. Comparison with the lysine biosynthetic pathway
would suggest that the DapC homolog TabD converts the TabB
proteins of interest were transformed into BL21(DE3) cells. 20 mL of
LB broth containing 50 g/mL kanamycin was inoculated with a
single colony and shaken overnight at 37 ꢀC. The culture was
diluted into 2 L LB broth containing 50 g/mL kanamycin and
grown at 37 ꢀC with shaking until the OD600 reached 0.5. The cul-
tures were cooled to 15 ꢀC and 400
M IPTG was added with
shaking continued overnight. The following morning, cells were
harvested by centrifugation at 4 ꢀC (10 min, 3000g) and frozen
at ꢁ80 ꢀC until purification.
m
m
m
product N-succinyl-THDPA into N-succinyl-L,L-DAP; surprisingly,
though TabD could catalyze this transformation efficiently, in our
hands it also utilized THDPA as a substrate, preventing us from
unequivocally determining its role in tabtoxin biosynthesis.
Regardless, these observations provide the first hints as to the early
stages of tabtoxin production and will serve as a starting point for
future exploration of this heretofore poorly understood pathway.
Protein purification. Frozen cell pellets were resuspended in
25 mL lysis buffer (25 mM Tris, 500 mM NaCl, pH 8.0), and lysed by
two passages through a French press. Lysates were subjected to
centrifugation (15000g, 30 min), and the clarified lysate was incu-
bated with 1 mL Ni-NTA agarose resin (Qiagen) at 4 ꢀC for 1 h. The
resin was washed with 10 mL lysis buffer containing 5 mM imid-
azole and subsequently subjected to step-gradient elution with
lysis buffer containing 25e200 mM imidazole (1 mL per fraction).
Protein-containing fractions were identified by SDS-PAGE,
collected, and dialyzed twice against 2 L of 25 mM Tris, 50 mM
NaCl, 10% glycerol, pH 8.0. Dialyzed protein was flash-frozen in
liquid nitrogen and stored at ꢁ80 ꢀC. Protein concentrations were
determined by Bradford assay.
2. Materials and methods
Cloning of tabB, tabD, and dapD from P. syringae pv. tabaci ATCC
11528. Genomic DNA from P. syringae pv. tabaci ATCC 11528 was
isolated using a DNeasy Blood and Tissue Kit (Qiagen) according to
the manufacturer's protocol. Sequences encoding tabB, dapD, and
tabD were amplified using Phusion DNA polymerase (New England
Biolabs) and the following primers: tabB 50- ggaattcca-
tatgagtaaccgtgcattgattgaagaagc-30 and 50- gatcctcgagtcggct-
gagcgctcggaccagatcg-3’; tabD 5’- agctcatatgaactccgcaggcaatgctttcg-
3’ and 5’- agctctcgagagtctgcgcagcctcaaggc-3’; dapD 5’- ggaattcca-
tatgtccactaccctgttcagcctgg-3’ and 5’- gatcctcgaggttatgagcgtg-
cagtgcttcgttc-3’. The resulting amplicons were digested with NdeI
and XhoI and ligated into similarly digested pET-29b using T4 DNA
ligase. Plasmids from kanamycin-resistant transformants were
confirmed to contain appropriate inserts by diagnostic digestion
with NdeI and XhoI and DNA sequencing (Dana-Farber Cancer
Institute Molecular Biology Core Facility, Boston, MA).
Preparation of succinyl-CoA and acetyl-CoA. A solution of coen-
zyme A (22
brought to approximately 7.5 by addition of 100
solution was placed on ice. To this solution was added an ice-cold
solution of succinic or acetic anhydride (35 mol) in 500 L aceto-
mmol) was prepared in 1 mL 0.5 M KHCO3. The pH was
mL 1 M HCl, and the
m
m
nitrile. The mixture was incubated on ice for 15 min. Reaction
progress was monitored by Ellman assay; if necessary, additional
anhydride solution was added. Upon reaction completion, 3 mL of
6 M HCl was added to lower the pH to approximately 6. Acetonitrile
was removed in vacuo, and acyl coenzyme A was purified by
Supelco LC-18 column with elution in 1% trifluoroacetic acid and
50% acetonitrile. Acetonitrile was removed in vacuo, followed by
lyophilization. The product was dissolved in water to 10 mM and
Protein overexpression. Ddh was expressed and purified as pre-
viously reported [9]. Sequence-confirmed plasmids encoding

214
M.E. Manning et al. / Biochemical and Biophysical Research Communications 496 (2018) 212e217
stored in aliquots at ꢁ80 ꢀC.
Assay of TabB. THDPA was generated in situ by reaction of meso-
diaminopimelic acid (meso-DAP) with Ddh. Specifically, 1 mM
meso-DAP was combined with 1 mM NADPþ and 5
mM Ddh in
100 mM HEPES pH 7.5 at 25 ꢀC. The reaction was allowed to proceed
to completion as judged by A340. Various concentrations of THDPA
and acyl-CoA were combined in 100 mM HEPES pH 7.5 with 5
TabB in the presence of 200
M DTNB at 25 ꢀC. Reaction progress
was monitored by coenzyme A formation as determined via A412
mM
m
.
Assay of TabD. THDPA and N-succinyl-THDPA were generated as
described above; N-acetyl-THDPA was generated analogously using
the THDPA-N-acetyltransferase DapH from Bacillus megaterium
(data not shown). The in situ-generated THDPA derivative (100
was combined with glutamate or alanine (250 M), coenzyme A
(300 M), -ketoglutarate dehydrogenase or
M), thio-NADþ (300
M), and TabD (5 M)
mM)
m
m
m
a
pyruvate dehydrogenase (0.0625 U), PLP (20
m
m
in 100 mM HEPES pH 7.5 at 25 ꢀC, and reaction progress was
measured by assaying formation of thio-NADH at 398 nm.
3. Results
3.1. Expression and characterization of TabB
A comparison of the proteins encoded by the tabtoxin biosyn-
thetic cluster with proteins involved in lysine biosynthesis led us to
target TabB, the predicted THDPA-N-acyltransferase homolog, for
our initial studies. We hypothesized that the lack of DapA and DapB
homologs encoded by the tabtoxin cluster meant that THDPA could
Fig. 2. (A) SDS-PAGE of DapD (37.3 kD) and TabB (30.6 kD). (B) Relative activities of
TabB and DapD toward succinylation of THDPA vs. acetylation of THDPA. (C) Kinetic
parameters of TabB and DapD. We were unable to saturate the enzymes with acetyl-
CoA and therefore could not determine kcat or KM values with respect to acetylation.
(D) Genomic context of the predicted dapD homolog in P. syringae pv. tabaci ATCC
11528.
be diverted from primary metabolism for T
in analogy with its homolog DapD, TabB would catalyze a step very
early in the pathway converting THDPA to T L. Previous work had
also noted that DapB is necessary for tabtoxin biosynthesis, sug-
gesting that THDPA could be a direct precursor toT L in accord with
bL biosynthesis, and that
b
reductive amination of THDPA to meso-DAP; we performed the
reaction in the reverse direction to convert meso-DAP to THDPA in
situ [9]. In the event, when we combined THDPA with succinyl-CoA
in the presence of TabB, we saw an increase in the rate of coenzyme
A thiol formation as measured by Ellman assay. In contrast, we saw
much lower activity using acetyl-CoA as the acyl donor. We further
determined kcat and KM for succinyl-CoA and THDPA, and obtained
values similar to those for physiological THDPA-N-succinyl-
transferases (vide infra), though we were unable to saturate TabB
with acetyl-CoA as a substrate (Fig. 2b and c). Based on these ob-
servations we conclude that TabB is a succinyl transferase that acts
on THDPA during tabtoxin biosynthesis.
b
this model [10]. We further reasoned, given the distinct DapH- and
DapD-dependent succinylase and acetylase branches of the lysine
biosynthetic pathway, that TabB and a presumed primary meta-
bolic DapD encoded in the P. syringae genome would differentially
label THDPA, diverting the respective acylated THDPA products
toward TbL or lysine biosynthesis. We were also encouraged in this
hypothesis by an early report by Liu and Shaw that tabtoxin
biosynthesis proceeded via acetylated THDPA intermediates [11],
whereas P. syringae lysine biosynthesis, as is the case of most non-
Bacillus bacteria, was thought to occur via succinylated THDPA in-
termediates [12].
To determine TabB's activity, we cloned and expressed it as a C-
terminal His6-tagged construct in E. coli (Fig. 2a). The protein
expressed solubly, and we were able to isolate useful quantities
using standard Ni2þ-NTA chromatography. We tested several mol-
ecules structurally related to THDPA as potential TabB substrates,
3.2. Identification of a second THDPA-N-succinyltransferase DapD
We were surprised by the observation that TabB was selective
for succinyl-CoA as an acyl donor over acetyl-CoA. Given the
expectation that P. syringae would utilize succinylated in-
termediates for lysine biosynthesis, this suggests that TabB func-
tionally duplicates a putative DapD utilized in lysine biosynthesis.
Indeed, it was conceivable that TabB could serve dual physiological
roles in P. syringae, both as the THDPA-N-succinyltransferase
necessary for tabtoxin biosynthesis as well as serving as the
THDPA-N-acyltransferase necessary for lysine biosynthesis. Dual-
function enzymes have previously been observed in lysine
biosynthesis: in E. coli, DapC catalyzes transamination of N-succi-
nyl-THDPA during biosynthesis of lysine as well as of N-acety-
lornithine during arginine biosynthesis [15]. To probe whether
there was a second THDPA-N-acyltransferase utilized in primary
metabolism, we scanned the P. syringae pv. tabaci genome and
identified a candidate dapD gene near to ORFs corresponding to
DapC and DapE homologs, suggesting it participates in lysine
biosynthesis (Fig. 2d). We cloned and overexpressed this protein
including meso-DAP,
combining these substrates with 500
CoA and 200
M 5,50-dithio-bis-(2-nitrobenzoic acid) (DTNB) in the
L,L-DAP,
L
-aminoadipic acid, and lysine. Upon
mM acetyl-CoA or succinyl-
m
presence of TabB, we saw no increase in absorbance relative to that
observed due to background thioester hydrolysis. Though it was
likely that none of these substrates were the physiological acyl
acceptor for TabB, DapD from P. aeruginosa can utilize L-amino-
adipic acid as substrate, albeit with a high KM [13], and we were
surprised to see no reactivity for compounds that are presumably
structurally related to TabB's physiological substrate.
We subsequently decided to test THDPA as an acyl acceptor for
the TabB-catalyzed reaction. THDPA is not commercially available,
though a chemical synthesis has been reported [14]. In our hands,
we were unable to obtain pure THDPA using the reported protocol
and so turned to in situ generation of THDPA from meso-DAP
catalyzed by the enzyme Ddh. Physiologically, Ddh catalyzes the

M.E. Manning et al. / Biochemical and Biophysical Research Communications 496 (2018) 212e217
215
and observed efficient THDPA-N-succinyltransferase activity.
Strikingly, DapD and TabB displayed similar kinetic parameters
(Fig. 2c). Thus, it appears that P. syringae has two distinct, but
functionally redundant THDPA-N-succinyltransferases encoded by
dapD and tabB.
perform a full kinetic characterization of TabD, we observed that
TabD was able to convert THDPA analogs to the corresponding DAP
product, in accord with its homology to DapC (Fig. 3b). TabD
showed no activity in the presence of alanine, demonstrating that it
is specific for glutamate as the amine donor, consistent with char-
acterized DapC isoforms [17]. However, among the THDPA de-
rivatives we tested, THDPA itself was preferred by a factor of
approximately 2.5ꢂ, with the two N-acyl-THDPA substrates being
processed at similar rates.
3.3. Expression and characterization of TabD
The next step in lysine biosynthesis after formation of N-acyl-
THDPA is PLP-dependent transamination catalyzed by DapC. Given
the presence of tabD encoding a DapC homolog in the tabtoxin
biosynthetic cluster, we set out to clone and overexpress TabD to
determine whether it could accept N-succinyl-THDPA as a substrate
for transamination to generate N-succinyl-DAP. However, when we
isolated His6-tagged TabD we saw no evidence of PLP binding based
on absorbance at 420 nm, in contrast to typical PLP-dependent
enzymes which form stable (nanomolar to low micromolar KD)
lysine-PLP imine complexes [16]. To confirm that the overexpressed
4. Discussion
Tabtoxin is marked by its structural similarity to lysine as well as
by the number of homologs to lysine metabolic enzymes encoded
within its biosynthetic cluster in addition to several tabtoxin-
specific enzymes including a predicted
b-lactam synthase and
clavaminic acid synthase. It is not a priori apparent how the lysine-
related reactions and tabtoxin-specific reactions are sequenced
with respect to one another during the biosynthesis of tabtoxin. We
took the known succinylase/acetylase pathways of lysine biosyn-
thesis as the starting point of our explorations of tabtoxin
biosynthesis.
TabD was well-folded and active, we combined 5
mM TabD with
20 M PLP and observed formation of a bright yellow solution, and
m
a difference spectrum of the PLP þ TabD mixture and an equivalent
solution of PLP lacking TabD showed a clear absorbance at 420 nm
typical of a lysine-PLP imine (Fig. 3a). Ultrafiltration of the
PLP þ TabD mixture removed the absorbance, demonstrating that
the TabD-PLP imine is unstable.
With TabD in hand, we examined its ability to catalyze trans-
amination of several THDPA structural variants in the presence of
exogenous PLP. We generated THDPA and N-succinyl-THDPA in situ
via reaction with Ddh or with Ddh and TabB respectively, as
described above. Given the homology with DapC and the existence
of DapC isoforms from the acetylase branch of lysine biosynthesis
that presumably utilize N-acetyl-THDPA as a substrate, we also
generated N-acetyl-THDPA using Ddh and DapH from Bacillus
megaterium for testing (data not shown). Among characterized
DapC homologs, glutamate is the preferred amine donor, but we
also tested the common amine donor alanine [15,17]. Our assay
Of the homologs of lysine biosynthetic enzymes, TabB corre-
sponds to the earliest step in the pathway represented in the tab-
toxin biosynthetic cluster. TabB is homologous to THDPA-N-
acyltransferases; its lysine biosynthetic counterparts succinylate
(DapD) or acetylate (DapH) THDPA. In vitro, we observed selective
succinylation of THDPA catalyzed by TabB with efficiencies ex-
pected of authentic THDPA-N-succinyltransferases. Comparing the
sequence of TabB with that of a succinyl-specific DapD homolog
from an unknown organism originally reported as Mycobacterium
bovis [19] (PDB: 1KGQ; 67% similar, 51% identical to TabB) revealed
that all DapD residues making direct contact with THDPA and the
succinyl moiety of the succinyl-CoA are conserved in TabB (Fig. 4).
Thus, though it is impossible to discount the possibility that an
unidentified enzyme encoded in the tabtoxin biosynthetic cluster
converts THDPA to an unknown product that is the physiological
substrate of TabB, we believe that our observations compellingly
demonstrate that TabB is a THDPA-N-succinyltransferase.
measured formation of
vate (from alanine) by a coupled oxidation catalyzed by
a
-ketoglutarate (from glutamate) or pyru-
-keto-
a
glutarate or pyruvate dehydrogenase with concomitant formation
of thio-NADH (lmax ¼ 398 nm) [18]. Though we were unable to
Complicating this interpretation of the role of TabB is the
Fig. 3. (A) SDS-PAGE of TabD (42.2 kD). (B) Spectrophotometric assay for PLP binding by TabD. Spectra were taken of 5
TabD þ 20 M PLP at pH 7.5. The difference spectrum between the TabD þ PLP and PLP only spectra show an absorbance at 420 nm typical of a covalently-bound PLP. (C) Relative
activities of TabD toward N-succinyl-THDPA, N-acetyl-THDPA, and THDPA. The no substrate experiment lacked an amine acceptor. Under these conditions, TabD generated thio-
NADH utilizing THDPA as a substrate at a rate of 0.45
mol minꢁ1 mgꢁ1
mM TabD, 20
mM PLP (lmax ¼ 388 nm), and 5
mM
m
m
.

216
M.E. Manning et al. / Biochemical and Biophysical Research Communications 496 (2018) 212e217
convert N-succinyl-THDPA to N-succinyl-DAP with an activity only
approximately 2.5-fold less than THDPA itself, maintaining the
possibility that TabD is a physiological N-succinyl-THDPA amino-
transferase. Unfortunately, preliminary studies aimed at obtaining
other proteins in the tabtoxin cluster led to insoluble expression;
this insolubility prevents us at present from examining later steps
in the tabtoxin biosynthesis to gain insight into the specific roles
and sequencing of the TabB- and TabD-catalyzed reactions.
Finally, we observed similar activity of TabD toward both N-
succinyl- and N-acetyl-THDPA. Preliminary studies in our labora-
tory on other DapC homologs drawn from both the succinylase and
acetylase branches of lysine biosynthesis suggest that this is typical
of DapC homologs from succinylase-utilizing organisms (data not
shown). This lack of selectivity between succinylated and acety-
lated substrates is perhaps not surprising given the presumed
absence of potentially competitive acetylated substrates in vivo.
Simplistically, one can speculate that the succinyl-binding pocket is
large enough to accommodate either the succinylated or acetylated
substrate. However, this model fails to account for the apparent
preference for non-acylated THDPA. It is of course also possible that
another enzyme encoded in the tabtoxin cluster converts the TabB
product N-succinyl-THDPA to an unknown intermediate that is the
physiological substrate of TabD, with the activity toward THDPA
and N-acyl-THDPA being in vitro artifacts.
In conclusion, these results confirm the homology-based as-
signments of TabB as a THDPA-N-succinyltransferase and TabD as a
glutamate- and PLP-dependent aminotransferase. Based on the
catalytic efficiency of TabB and comparison with bona fide THDPA-
N-succinyltransferase sequences, we believe that the observed
succinylation of THDPA represents TabB's physiological activity.
However, though our results confirm that TabD is able to catalyze
transamination of THDPA derivatives, we are unable to say with
confidence what its specific role in tabtoxin biosynthesis might be.
Regardless, our results are further evidence for the close relation-
ship between lysine and tabtoxin biosynthesis.
Fig. 4. Binding site for THDPA and succinyl moiety taken from DapD from an unknown
organism with L-aminoadipic acid used as a THDPA analog (PDB: 1KGQ) [19]. Corre-
sponding amino acids in TabB identified by sequence alignment are shown in bold
below each label.
observation that the P. syringae genome encodes a second THDPA-
N-succinyltransferase that we assigned as the THDPA-N-succinyl-
transferase DapD utilized in lysine biosynthesis based on its
genomic context and confirmation of its activity in vitro. It is sur-
prising that the P. syringae genome encodes two THDPA-N-succi-
nyltransferases; however, this does explain the observation that
dapD and tabB are separately dispensible for both lysine and tab-
toxin biosynthesis [11]. Though we did not explicitly probe the
in vivo roles of these two apparently functionally redundant en-
zymes in these experiments, we speculate that the tabtoxin gene
cluster including TabB was acquired in toto by horizontal gene
transfer. Preliminary analysis of the P. syringae DapD and TabB
suggest that they are from distinct phylogenetic clades, supporting
this hypothesis [20].
We also explored the activity of the DapC homolog TabD in vitro.
In lysine biosynthesis, DapC catalyzes transamination of the DapD/
DapH product N-acyl-THDPA to generate N-acyl-L,L-DAP in a PLP-
Acknowledgements
dependent manner using glutamate as the amine donor. Typical
PLP-dependent aminotransferases exist in a resting state in which
the PLP is covalently bound to a conserved active-site lysine via
imine linkage with high affinity. However, TabD did not copurify
with PLP as is typical of PLP-dependent aminotransferases, though
we were able to reconstitute the holo-enzyme in the presence of
exogenous excess PLP. In accord with its observed sequence ho-
mology, the PLP-bound TabD was able to convert THDPA derivatives
to the corresponding DAP derivatives using glutamate as the amine
donor. Interestingly, we observed that TabD actually slightly prefers
THDPA as a substrate instead of the TabB product N-succinyl-
THDPA as we had expected. The preference of TabD for THDPA is
difficult to reconcile with our model of tabtoxin biosynthesis
directly paralleling the early stages of lysine biosynthesis. We were
unable to perform a full kinetic characterization of TabD; it is
possible that DapD/TabB are much more efficient than TabD,
thereby out-competing TabD for THDPA. Alternatively, one can
imagine that TabD acts before TabB in the tabtoxin biosynthesis,
with TabB's physiological substrate being some diamine related to
DAP. We do not find this explanation compelling; in our hands,
despite the reported promiscuity of TabB homologs toward their
acyl acceptor substrates [13], we did not see any detectable activity
toward any of the several diamine substrates tested. Further, as
described above, the complementarity of the active sites of TabB
and the bona fide THDPA N-succinyltransferase DapD suggests
THDPA is the physiological substrate of TabB. However, despite its
preference for THDPA, we emphasize that in vitro TabD was able to
ꢀ
We thank Andre O. Hudson (Rochester Institute of Technology)
for providing the Ddh expression construct, and Carleton College
for funding.
Transparency document
Transparency document related to this article can be found
online at https://doi.org/10.1016/j.bbrc.2018.01.028.
References
[1] W.W. Stewart, Isolation and proof of structure of wildfire toxin, Nature 229
(1971) 174e178.
[2] T.A. Wencewicz, C.T. Walsh, Pseudomonas syringae self-protection from
tabtoxinine-beta-lactam by ligase TblF and acetylase Ttr, Biochemistry 51
(2012) 7712e7725.
[3] S.L. Sinden, R.D. Durbin, Glutamine synthetase inhibition: possible mode of
action of wildfire toxin from Pseudomonas tabaci, Nature 219 (1968)
379e380.
[4] M.D. Thomas, P.J. Langston-Unkefer, T.F. Uchytil, R.D. Durbin, Inhibition of
glutamine synthetase from pea by tabtoxinine-beta-lactam, Plant Physiol. 71
(1983) 912e915.
[5] C.T. Walsh, T.A. Wencewicz, Antibiotics: Challenges, Mechanisms, Opportu-
nities, ASM Press, United State of America, 2016.
[6] C.J. Unkefer, R.E. London, R.D. Durbin, T.F. Uchytil, P.J. Langston-Unkefer, The
biosynthesis of tabtoxinine-beta-lactam. Use of specifically 13C-labeled
glucose and 13C NMR spectroscopy to identify its biosynthetic precursors,
J. Biol. Chem. 262 (1987) 4994e4999.
[7] G. Scapin, J.S. Blanchard, Enzymology of bacterial lysine biosynthesis, Adv.
Enzymol. Relat. Area Mol. Biol. 72 (1998) 279e324.
[8] T.G. Kinscherf, D.K. Willis, The biosynthetic gene cluster for the beta-lactam
antibiotic tabtoxin in Pseudomonas syringae, J. Antibiot. 58 (2005) 817e821.

M.E. Manning et al. / Biochemical and Biophysical Research Communications 496 (2018) 212e217
217
[9] A.O. Hudson, A. Klartag, C. Gilvarg, R.C. Dobson, F.G. Marques, T. Leustek, Dual
diaminopimelate biosynthesis pathways in Bacteroides fragilis and Clos-
tridium thermocellum, Biochim. Biophys. Acta 1814 (2011) 1162e1168.
[10] L. Liu, P.D. Shaw, Characterization of dapB, a gene required by Pseudomonas
syringae pv. tabaci BR2.024 for lysine and tabtoxinine-beta-lactam biosyn-
thesis, J. Bacteriol. 179 (1997) 507e513.
[15] R. Ledwidge, J.S. Blanchard, The dual biosynthetic capability of N-acetylorni-
thine aminotransferase in arginine and lysine biosynthesis, Biochemistry 38
(1999) 3019e3024.
[16] A.C. Eliot, J.F. Kirsch, Pyridoxal phosphate enzymes: mechanistic, structural,
and evolutionary considerations, Annu. Rev. Biochem. 73 (2004) 383e415.
[17] S. Weyand, G. Kefala, M.S. Weiss, The three-dimensional structure of N-suc-
cinyldiaminopimelate aminotransferase from Mycobacterium tuberculosis,
J. Mol. Biol. 367 (2007) 825e838.
[11] L. Liu, P.D. Shaw,
A possible role for acetylated intermediates in dia-
minopimelate and tabtoxinine-beta-lactam biosynthesis in Pseudomonas
syringae pv. tabaci BR2.024, J. Bacteriol. 179 (1997) 5922e5927.
[12] S. Weinberger, C. Gilvarg, Bacterial distribution of the use of succinyl and
acetyl blocking groups in diaminopimelic acid biosynthesis, J. Bacteriol. 101
(1970) 323e324.
[18] A.J. McCoy, N.E. Adams, A.O. Hudson, C. Gilvarg, T. Leustek, A.T. Maurelli, L,L-
diaminopimelate aminotransferase,
a trans-kingdom enzyme shared by
Chlamydia and plants for synthesis of diaminopimelate/lysine, Proc. Nat.
Acad. Sci. U. S. A. 103 (2006) 17909e17914.
[13] R. Schnell, W. Oehlmann, T. Sandalova, Y. Braun, C. Huck, M. Maringer,
M. Singh, G. Schneider, Tetrahydrodipicolinate N-succinyltransferase and
dihydrodipicolinate synthase from Pseudomonas aeruginosa: structure anal-
ysis and gene deletion, PLos One 7 (2012), e31133.
[14] E.J.T. Chrystal, L. Couper, D.J. Robins, Synthesis of a key intermediate in the
diaminopimelate pathway to L-lysine - 2,3,4,5-tetrahydropicolinic acid, Tet-
rahedron 51 (1995) 10241e10252.
[19] T.W. Beaman, K.W. Vogel, D.G. Drueckhammer, J.S. Blanchard, S.L. Roderick,
Acyl group specificity at the active site of tetrahydridipicolinate N-succinyl-
transferase, Protein Sci. Publ. Protein Soc. 11 (2002) 974e979.
[20] L. Schuldt, S. Weyand, G. Kefala, M.S. Weiss, The three-dimensional structure
of a mycobacterial DapD provides insights into DapD diversity and reveals
unexpected particulars about the enzymatic mechanism, J. Mol. Biol. 389
(2009) 863e879.