Characterization of alkaline phosphatase PhoK from Sphingomonas sp. BSAR-1 for phosphate monoester synthesis and hydrolysis

Article abstract of DOI:10.1016/j.bbapap.2019.140291

The biocatalytic activity of a so far underexploited alkaline phosphatase, PhoK from Sphingomonas sp. BSAR-1, was extensively studied in transphosphorylation and hydrolysis reactions. The use of high-energy phosphate donors and oligophosphates as suitable

Full text of DOI:10.1016/j.bbapap.2019.140291

BBA-ProteinsandProteomics1868(2020)140291
Contents lists available at ScienceDirect
BBA - Proteins and Proteomics
journal homepage: www.elsevier.com/locate/bbapap
Characterization of alkaline phosphatase PhoK from Sphingomonas sp. BSAR-
1 for phosphate monoester synthesis and hydrolysis
Michael Lukeschb, Gábor Tasnádia,b, Klaus Ditrichc, Mélanie Hallb,⁎, Kurt Faberb,⁎
a Austrian Centre of Industrial Biotechnology, c/o, Austria
b Department of Chemistry, Organic & Bioorganic Chemistry, University of Graz, Heinrichstrasse 28, 8010 Graz, Austria
c White Biotechnology Research Biocatalysis, BASF SE, Carl-Bosch-Strasse 38, 67056 Ludwigshafen, Germany
A R T I C L E I N F O
A B S T R A C T
Keywords:
Biotransformation
Enzyme
Phosphatase
Phosphate ester
Transphosphorylation
The biocatalytic activity of a so far underexploited alkaline phosphatase, PhoK from Sphingomonas sp. BSAR-1,
was extensively studied in transphosphorylation and hydrolysis reactions. The use of high-energy phosphate
donors and oligophosphates as suitable phosphate donors was evaluated, as well as the hydrolytic activity on a
variety of phosphate monoesters. While substrates bearing free hydroxy group displayed only moderate re-
activity as acceptors for transphosphorylation by PhoK, strong hydrolytic activity on a broad variety of phos-
phate monoesters under alkaline conditions was observed. Site-directed mutagenesis of selected amino acid
residues in the active site provided valuable insights on their involvement in enzyme catalysis. The key residue
Thr89 so far postulated to engage in enzyme phosphorylation was confirmed to be crucial for catalysis and could
be replaced by serine, albeit with much lower catalytic efficiency.
1. Introduction
like enzymes [10].
PafA and PhoK (~26% protein sequence identity) appear evolu-
Alkaline phosphatases along with nucleotide phosphodiesterases,
co-factor independent phosphoglycerate mutases, phosphonate
monoester hydrolases and aryl sulfatases are metalloenzymes, members
of the large alkaline phosphatase (AP) superfamily that share similar
metal-binding site architecture and protein fold, and are responsible for
a range of catalytic activities [1–3]. Over the past decades, a number of
promiscuous activities across the family members have been observed
[4]. Catalytic promiscuity of enzymes along with particular structural
features reflects evolutionary relationship and provides understanding
of factors that influenced their evolution [5]. This valuable information
might help to guide enzyme engineering and de novo enzyme design
[6]. A particularly interesting case is the evolutionary interconnection
between alkaline phosphatases from E. coli (EcAP) [7], Chryseo-
bacterium meningosepticum (PafA) [8], and Sphingomonas sp. BSAR-1
(PhoK) [9], and the nucleotide pyrophosphatase/phosphodiesterase
(NPP) from Xanthomonas citri [4i], all characterized by a bimetallic
Zn2+-center. EcAP, PafA and PhoK catalyze the hydrolysis of phosphate
monoesters, while NPP is specific for phosphate diesters. These en-
zymes were subject of recent studies shedding light on the evolution
and catalytic diversification of enzymes in the AP superfamily, and
PhoK was recently suggested to be the first member of a novel class of
APs, based on analysis of crystal structure and comparison with EcAP-
tionary distant from EcAP (~11% identity to PafA/PhoK) and NPP
(~18% identity to PafA/PhoK). While EcAP-like enzymes employ a
binuclear Zn2+-center accompanied by a Mg2+-ion as well as a serine
residue responsible for the formation of a covalent enzyme-phosphate
intermediate, in PafA and PhoK the role of the missing Mg2+ is thought
to be replaced by that of a protonated lysine, while a threonine appears
to replace the catalytic serine residue.[10b,d] In the case of PhoK, an
additional Ca2+-ion is suggested to aid in the dimerization of the pro-
tein, as inferred from analysis of the crystal structure.[10d] Further-
more, PhoK contains six cysteine residues, all of which are involved in
disulfide bond formation within the monomer and are responsible for a
rigid and compact structure. Remarkably, PhoK seems to recruit si-
multaneously several amino acid residues that are differently conserved
among other members of the AP superfamily, in particular Thr89,
Asn110, Lys171 and Arg173.
Beside biochemical significance, alkaline phosphatases have found
useful biotechnological applications, for instance in the synthesis of
farnesyl alcohol derivatives [11], in the analysis of sphingosine analo-
gues [12], in enzymatic chemiluminescent assays [13], and in a po-
tential bioremediation process involving precipitation of uranium in
form of phosphate salt [9]. Recently, alkaline phosphatases were used
in biocatalytic cascade reactions for the assembly of supramolecular
⁎ Corresponding authors.
E-mail addresses: melanie.hall@uni-graz.at (M. Hall), kurt.faber@uni-graz.at (K. Faber).
https://doi.org/10.1016/j.bbapap.2019.140291
Received 26 July 2019; Received in revised form 4 October 2019; Accepted 10 October 2019
Availableonline31October2019
1570-9639/©2019ElsevierB.V.Allrightsreserved.

M. Lukesch, et al.
B
B
A-ProteinsandProteomics1868(2020)140291
materials [14], for equilibrium displacement in disaccharide synthesis
[15], and in downstream processing of nucleotide-diphosphosugars
[16] and oligosaccharides [17]. Although alkaline phosphatases can
theoretically be used in the synthesis mode, similarly to the well-
exploited acid phosphatases [18], their phosphate-transfer ability has
barely been harnessed. One of the rare examples reported so far is the
use of calf intestine alkaline phosphatase employing inorganic oligo-
phosphates as phosphate donors (P-donors) and a range of short to
medium-chain alcohols as acceptors at high concentration (60 vol%)
[19]. The system was optimized for the large-scale production of gly-
cerol-1-phosphate in a continuous flow reactor using immobilized en-
zyme [7b]. More recently, immobilized bovine alkaline phosphatase
was used for the preparation of glycerol-1-phosphate, precursor of di-
hydroxy acetone phosphate, used in the synthesis of carbohydrates in a
cascade setup [20].
We recently became interested in exploiting acid phosphatases in
the synthesis mode for the biocatalytic selective phosphorylation of
alcohols using cheap inorganic oligophosphates as high-energy phos-
phate donors at pH 4–5 [21]. This required targeted efforts to reduce
unwanted product hydrolysis, which results in low yields of the phos-
phorylated compounds and is also connected with great variability in
optimum conditions to reach high product titer. Following similar
motivation, we aimed at studying the unexplored biocatalytic potential
of PhoK for synthetic applications by investigating both its phospho-
transferase and phosphate ester hydrolase activity (Scheme 1). A
number of structurally diverse molecules, precursors of valuable sugar-
phosphates, nucleotides and metabolites, such as mevalonate-phos-
phate, were therefore evaluated as suitable acceptors in transpho-
sphorylation reactions.
the enzyme was most likely replenished with metal and displayed im-
proved activity. Similar trends have been observed with carbox-
ypeptidase A, a zinc containing enzyme, in presence of phytic acid [24].
No activity was observed in metal-containing buffer in absence of en-
zyme (data not shown). As reported in the literature [10], the pair of
Zn2+ is essential for catalysis in the AP superfamily, not only activating
the nucleophilic residue but also stabilizing negatively charged oxygen
of the leaving phosphate during hydrolysis. Upon analysis of the crystal
structure, Ca2+ in PhoK has been proposed to play a structural role,
influencing dimerization, which likely explains the importance of suf-
ficient metal availability for activity. No other divalent metals were
tested. With more closely related PafA, the addition of Ca2+ was shown
to have no effect on activity [10b]. Interestingly, in the case of purple
acid phosphatase-type enzymes, which do not generate
a phos-
phoenzyme intermediate during catalysis, Ca2+ was also found essen-
tial for an active enzyme; with diesterase Rv0805 from Mycobacterium
tuberculosis, Ca2+ binding was proposed to regulate activity [25].
The effect of previously reported inhibition of alkaline phosphatases
by phosphate [7b,9] was evaluated in the hydrolysis of p-NPP by PhoK
in the presence of increasing concentration of Pi. PhoK hydrolytic ac-
tivity decreased by 50% and 90%, respectively, in the presence of
100 mM and 300 mM phosphate, respectively (Fig. 2).
Inorganic oligophosphates are P-donors of choice in transpho-
sphorylation reactions catalyzed by acid phosphatases (for structures of
P-donors, see Table S1) [21]. However, oligophosphates are notorious
chelators and may cause inactivation of metal-dependent enzymes. The
effect of various oligophosphates and increasing concentration of zinc
and calcium ions was therefore studied in the p-NPP assay (Fig. 3).
PhoK displayed no activity in PPi and trimetaphosphate buffers in ab-
sence of added metal ions, however, zinc and calcium ion supple-
mentation resulted in enhanced specific activity, indicating partial re-
storation of the hydrolytic activity (up to 21 U mg−1). Phytic acid
seemed to be less detrimental as PhoK was reasonably active even
without added metal ions. The activity could be enhanced at low metal
ion concentrations (< 5 mM) but was reduced at elevated concentra-
tions (15 mM). In contrast, metal ions had no effect on the activity in
hexametaphosphate buffer, which remained comparatively low (max.
3.1 U mg−1).
PPi was selected for further investigations, based on previous re-
ports [21], and the more active protein obtained from expression using
IBA2 vector was chosen (PhoK-IBA2). The hydrolysis of 250 mM PPi
was first followed at various metal concentrations and pH values
(Figs. 4, 5). No conversion was observed without added metals, con-
firming the data obtained with p-NPP (Fig. 3). Addition of 5–10 mM
zinc and calcium could however restore the hydrolytic activity. Higher
concentrations (from 20 mM) strongly affected the activity of PhoK
enzyme. As expected for alkaline phosphatases, a pH study identified
pH 9 as optimum value for the hydrolysis of PPi (up to 90% conversion
after 20 h). Higher pH values resulted in lowest conversion levels
(< 10%). PPi could be replaced by PPPi, another cheap, high energy P-
donor, however, reduced reaction rate in the first 5 h was observed at
pH 9, in line with the observation made with acid phosphatase PhoN-Sf
[21]. After 20 h however, similar conversion level was achieved.
2. Results and discussion
2.1. Over-expression
The phoK gene was obtained through commercial synthesis (Uniprot
accession number: A1YYW7) and was initially cloned into pET30a
vector for expression with C-terminal His-tag. The vector was then
transformed into E. coli BL21(DE3) harboring chaperone 1 of the Takara
chaperone plasmid set, and the protein was over-expressed and purified
(see supporting information). The enzyme solution was stored at 4 °C
for several weeks after purification and dialysis to allow for proper
refolding of the protein. This step was necessary due to the use of de-
naturing conditions in the cell lysis step (0.5 M urea [10g], see sup-
porting information). The specific activity of PhoK-pET30a in the p-
nitrophenylphosphate (p-NPP) assay reached a final value of 39 U mg−1
at pH 9.0 in 200 mM Tris-HCl buffer, which is in the range of that of
non-specific acid phosphatases (NSAPs) expressed in E. coli [21].
During the course of the study, the expression of PhoK was further
optimized. PhoK was cloned into pASK-IBA2 vector allowing for peri-
plasmic expression, and purification via C-terminal Strep-Tag (see
supporting information). The purified protein PhoK-IBA2 showed a
significantly improved specific activity of 1690 U mg−1 (43-fold in-
crease compared to the activity of the previous preparation). Initial
characterization was performed with PhoK-pET30a before switching to
the more active PhoK-IBA2 (see below).
2.3. Transphosphorylation
2.2. Characterization
The transphosphorylation activity of PhoK was first probed with
500 mM D-glucose (1a) as substrate and 250 mM PPi in the presence of
5 mM ZnCl2 and 5 mM CaCl2 at various pHs (Fig. 6).
Similar pH optimum as in the hydrolysis reaction was observed for
the phosphotransferase activity (pH 9). Product concentration reached
max. 16 mM at 91% consumption of PPi after 4 h, corresponding to
about 7% phosphate transfer efficiency; the major part of PPi was thus
preferentially hydrolyzed without phosphate transfer (Scheme 1, path
a) dominating). Beyond this time, significant product hydrolysis took
place. Lowering the pH resulted in reduced reaction rates and the
The hydrolytic activity of the enzyme was first tested in different
buffer systems. Metal ion supplementation (Zn2+ and Ca2+) was also
tested as a way to increase activity and overcome possible chelation
effects of Tris [22] and glycine [23] (Fig. 1). Surprisingly, the enzyme
activity was strongly affected by rising metal ion concentration in
carbonate, Tris and CHES (N-cyclohexyl-2-aminoethanesulfonic acid)
buffers but not in glycine buffer, in which the activity could be sig-
nificantly improved with up to 5 mM salt. Upon metal supplementation,
2

M. Lukesch, et al.
B
B
A-ProteinsandProteomics1868(2020)140291
Scheme 1. Competition between hydrolysis (red) and transphosphorylation (blue) catalyzed by alkaline phosphotase PhoK. Both pathways are proposed to proceed
through a phosphorylated threonine [10d]. R-OH: phosphate acceptors 1a-10a. Substrates investigated in this study are depicted 1a-10a and 1b/1c-13b; PPi:
pyrophosphate, Pi: monophosphate.
35
30
25
20
15
10
5
[CaCl2+ZnCl2]
0 mM
1 mM
2 mM
5 mM
10 mM
0
Carbonate
Tris
CHES
Glycine
Fig. 1. PhoK-pET30a specific activity based on p-NPP assay in various buffers (100 mM) at pH 9 at varying equimolar [ZnCl2 + CaCl2] concentration.
3

M. Lukesch, et al.
B
B
A-ProteinsandProteomics1868(2020)140291
(~29 mM) as well as reduced product hydrolysis rate, while 1.5 M 1a
resulted in substantially lower product concentration (Fig. S3). 31P
NMR analysis revealed that phosphorylation selectively took place at
the primary C6-OH moiety resulting in formation of D-glucose-6-phos-
phate (1c) [21c].
PhoN-Sf from Shigella flexneri, PiACP from Prevotella intermedia and
PhoN-Se from Salmonella typhimurium LT2 were used to compare the
transphosphorylation activity of PhoK with that of representative acid
phosphatases. These enzymes have been shown to exhibit a broad
substrate scope, high phosphotransfer activity and robustness in the
synthesis of phosphate esters at pH 4–5 [21a,b,26]. All three enzymes
were active on D-glucose (1a), PhoN-Sf and PiACP yielding
120–140 mM product (1c) in 23 h (~9-fold higher product titer com-
pared to that with PhoK), while PhoN-Se furnished moderate product
level (~45 mM) (Fig. S4).
Next, the substrate scope of PhoK was explored by testing additional
acceptors including 1,4-butanediol (2a), mono- and oligosaccharides
(3a-5a), nucleosides (6a-8a), rac-serine (9a) and rac-mevalonic acid
(10a) at pH 9 (Scheme 1 and Table 1). Unfortunately, only very low
phosphate transfer efficiencies (max. 5%) were observed with all tested
substrates at 70–80% PPi consumption after 2 h. In the case of inosine
(6a) and cytidine (9a), the formation of multiple products could be
monitored (31P NMR), indicating non-selective phosphorylation. De-
spite broad substrate acceptance, the moderate level of transpho-
sphorylation activity observed under the tested conditions does not
appear suited for preparative-scale synthetic applications using PhoK.
100
80
60
40
20
0
50
100
150
200
250
300
Pi (mM)
Fig. 2. Influence of Pi concentration on hydrolytic activity of PhoK-pET30a at
pH 9. 100% activity refers to enzyme activity in Tris-buffer (200 mM, pH 9) and
0 mM phosphate in the p-NPP assay.
25
20
15
PPi
TmP
10
5
HmP
PA
2.4. Hydrolysis
We then turned our attention to the hydrolytic activity of PhoK on a
set of phosphorylated compounds. Regioisomers of glycerol-phosphate
(2-phosphate: 11b and 1-phosphate: rac-12b) and D-glucose-phosphate
(1-phosphate: 1b and 6-phosphate: 1c) as well as phytic acid, ascorbic
acid-2-phosphate (13b), 4-hydroxybutyl phosphate (2b) and inosine-5′-
monophosphate (6b) were subjected to hydrolysis at pH 9 at
60–100 mM concentration (Table 2 and Fig. S5). Overall, the rate of
hydrolysis of phosphorylated secondary hydroxy group (1b, 11b, 13b)
was found higher than that of primary counterparts. PhoK displayed
however poor regiopreference in the hydrolysis of regioisomers of
glycerol 11b and 12b (Aspec = 629 vs. 520 U mg-1)
0
0
5
10
[Zn2++Ca2+] (mM)
15
Fig. 3. PhoK-pET30a hydrolytic activity based on p-NPP assay in buffers of P-
donor salts (100 mM) at pH 9 at varying equimolar [ZnCl2 + CaCl2] con-
centration (TmP: trimetaphosphate, HmP: hexametaphosphate, PA: phytic
acid).
consumption of PPi similarly slowed down. After 5 days, no remaining
product was detected at pH 8 and 9, while at pH 7, ~7 mM residual
product was measured. PPi was almost fully consumed in all cases. At
higher pH values (> 10), only traces of product could be observed
(< 5 mM) and a very slow rate of PPi consumption (data not shown).
Overall, the transphosphorylation activity follows the same trend upon
variation of the pH value as observed in the hydrolysis of PPi. Enhanced
concentration of 1a (1.0 M) at pH 9 furnished double amount of product
regioisomers of glycerol 11b and 12b (Aspec = 629 vs. 520 U mg−1).
With glucose derivatives, 1b was converted at higher rate than 1c
(Aspec = 1417 vs. 477 U mg−1), indicating an almost 3-fold preference
for hydrolysis of 1-phosphate compared to 6-phosphate, which is
somewhat unexpected given the exquisite regioselective transpho-
sphorylation of glucose 1a toward 1c. Nearly full conversion was ob-
served with most substrates after 5 h incubation time except in the case
100
90
80
70
60
50
40
30
20
10
0
[CaCl2+ZnCl2]
0 mM
5 mM
10 mM
20 mM
50 mM
0
500
1000
1500
t (min)
Fig. 4. Hydrolysis of 250 mM PPi at pH 9 by PhoK-IBA2 (8.2 μg mL−1, ~0.14 μM) in the presence of various amounts of [ZnCl2 + CaCl2] (equimolar).
4

M. Lukesch, et al.
B
B
A-ProteinsandProteomics1868(2020)140291
100
90
80
70
60
50
40
30
20
10
0
pH 7
pH 8
pH 9
pH 9, PPPi
pH 10
pH 11
0
200
400
600
800
1000
1200
1400
t (min)
Fig. 5. Hydrolysis of 250 mM PPi or PPPi by PhoK-IBA2 (8.2 μg mL−1, ~0.14 μM) at various pHs in the presence of 5 mM [ZnCl2 + CaCl2] (equimolar).
pH 7
pH 8
pH 9
Table 2
18
16
14
12
10
8
100
90
80
70
60
50
40
30
20
10
0
Phosphohydrolase activity of PhoK-IBA2 on various
phosphate monoestersa.
Substrate
Aspec (U mg−1
)
1b
1c
1417
477
2b
443
11b
12b
13b
629
520
1114
6
4
a
Reaction performed at pH 9 in the presence of
5 mM ZnCl2 and 5 mM CaCl2. See supporting in-
formation for details.
2
0
0
200
400
600
t (min)
800
1000
1200
of 4-hydroxybutyl phosphate (2b, full conversion reached after over-
night incubation). Phytic acid was not accepted. The data clearly
highlights the potential of PhoK as an efficient phosphohydrolase, en-
abling rapid hydrolysis of primary and secondary phosphorylated hy-
droxy groups of diversely substituted substrates in the alkaline range,
with TTN up to 7.1 × 106 (0.14 μM biocatalyst used on 60–100 mM
substrate).
Fig. 6. Transphosphorylation activity of PhoK-IBA2 (8.2 μg mL−1, ~0.14 μM)
with 500 mM D-glucose (1a) and 250 mM PPi at pH 9 in the presence of 5 mM
ZnCl2 and 5 mM CaCl2 toward formation of D-glucose-6-phosphate (1c).
Table 1
Substrate-scope of PhoK-IBA2 in transphosphorylation mode using PPi as do-
nora.
ConsumptionPPi (%)d Efficiency (%)e
2.5. Mutational study
Substrate
[2–10a]
[2–10b]
(mM)b
(mM)c
The analysis of PhoK crystal structure indicates major differences
with the structures of other alkaline phosphatases, such as emblematic
EcAP from E. coli [10d]. Four residues located in the active site have
been identified as signature of a suggested new class of alkaline phos-
phatases – Thr89, Asn110, Lys171 and Arg173 – and the catalytic
function of each residue was predicted based on crystal structure
comparison. However, no mutational analysis was performed to clearly
identify the role of each residue with PhoK, which prompted us to
design single alanine mutations of PhoK at the indicated positions. In
addition, the alleged phosphorylating threonine residue was replaced
by other nucleophilic residues (Ser, Asp, Lys, His, Tyr) known to be
responsible for phosphate-transfer in other classes of phosphatases
[27].
The variants were cloned, overexpressed and purified according to
the procedure applied to the wild-type protein (see supporting in-
formation), and characterized (Table 3). In the p-NPP assay, a drastic
loss of activity was observed upon exchanging any of the four targeted
amino acids. Thr89Ser, Lys171Ala and Arg173Ala only showed a re-
sidual specific activity (max. 1 U mg−1) three orders of magnitude
lower than that of the wild-type. Further Thr89 mutations (including
Ala) resulted in complete loss of activity (data not shown). Asn110Ala
variant could retain approx. 2% of the wild-type activity (38 U mg−1).
1,4-butanediol
(2a)
D-glucosamine
(3a)
raffinose (4a)
maltotriose (5a)
inosine (6a)
adenosine (7a)
cytidine (8a)
rac-serine (9a)
rac-mevalonic
acid (10a)
500
500
11
3
82
85
5
1
500
500
80
n.d.
n.d.
3f
80
85
73
72
67
74
51
n.a.
n.a.
2
n.a.
5
40
n.d.
300
300
280
9g
9
5
< 1
1h
a
250 mM PPi at pH 9 in the presence of 5 mM ZnCl2 and 5 mM CaCl2,
samples taken after 2 h.
b
Starting concentration of substrate.
c
Product concentration based on 31P NMR analysis.
d
Consumption of PPi.
Percentage of consumed PPi used for product formation (phosphate transfer
e
efficiency).
f
Total product amount, ratio of products 1:1.1.
Total product amount, ratio of products 4:1:5.5.
Sample taken after 1 h. n.d. dot detected; n.a. not applicable.
g
h
5

M. Lukesch, et al.
B
B
A-ProteinsandProteomics1868(2020)140291
Table 3
4. Experimental section
Comparison of hydrolytic activity of PhoK wild-type and variants.
4.1. Assay for phosphohydrolase activity (p-NPP assay)
Enzyme
A
specp-NPP (U mg−1
)
Aspec PPi (U mg−1
)
PPi hydrolysis (%)a
PhoK-WT
1359
1.0
38
< 1
1.1
818
3.1
14
89 (9 h)
78 (24 h)
79 (23h)
23 (19 h)
n.d.
Enzyme activities were assayed spectrophotometrically by mea-
suring the dephosphorylation of 4-nitrophenyl phosphate (p-NPP) via
release of p-nitrophenol (pNP). A protein solution (final concentration
1–200 μg mL−1, ~12.5 nM-2.5 μM) was added to an appropriate buffer
(carbonate, Tris-HCl, CHES - N-cyclohexyl-2-aminoethanesulfonic acid,
glycine, Pi, PPi, TmP, HmP or PA; 100 mM final concentration or as
indicated) supplemented with the appropriate amount of ZnCl2 and
CaCl2 at a given pH to a final volume of 480 μL. The solutions were pre-
incubated at 30 °C for 5 min followed by addition of 20 μL 250 mM p-
NPP in H2O (10 mM final concentration in buffer) and mixed at 30 °C
and 450 rpm. After 1 min incubation time, the reaction was quenched
with 500 μL of 1 M NaOH and the absorbance of p-NP was recorded at
405 nm (ε = 18,500 M−1 cm−1). The activity tests were always per-
formed in triplicate and values are mean values. Deviation from mean
values was below 5%. One unit of phosphatase activity (U) corresponds
to the amount of p-NP (micromoles) released per minute under assay
conditions. Specific activity (Aspec) represents the phosphatase activity
(U) of 1 mg protein.
PhoK-T89S
PhoK-N110A
PhoK-K171A
PhoK-R173A
4
< 0.1
a
Conversion of PPi over varying reaction time (indicated in h); n.d. not
detected.
The difference in catalytic activity between variants and wild-type was
also reflected in the hydrolysis of 100 mM PPi, with highest activity
observed with Asn110Ala (Aspec = 14 U mg−1). Despite lower specific
activity upon mutation, high conversion level could still be reached
after 23 h (79% compared to 87% after 5 h with the wild-type, see Fig.
S6). Thr89Ser and Lys171Ala, which displayed even lower specific ac-
tivity (max. 4 U mg−1), also led to high conversion levels after 24 h
(78% with Thr89Ser). Arg173Ala was found inactive in the hydrolysis
of PPi. In comparison, mutation of conserved threonine to serine with
NPP led to increased activity on monoester and decreased activity on
diester. With PafA, mutation of Thr79 to serine resulted in decreased
catalytic activity, regardless of the ester type. Mutation of the three
other residues with PafA resulted in 50% of wild-type catalytic activity
with Asn100Ala on monoester, and a reduction in catalytic activity by 1
and 4 orders of magnitude with Arg164Ala and Lys162Ala, respectively
[10d]
4.2. Phosphotransferase and phosphohydrolase activity using inorganic
oligophosphates
A standard reaction mixture contained substrate and/or P-donor
(PPi or PPPi) in H2O supplemented with the appropriate amount of
ZnCl2 and CaCl2 at a concentration and pH indicated in the footnotes of
tables and captions of all figures in 1 mL final volume. The reaction was
initiated by adding a given amount of enzyme, obtained via periplasmic
expression of PhoK-IBA2: 8.2 μg mL−1 (~0.14 μM) wt, 100 μg mL−1
Thr89Ser, Arg173Ala and Asn110Ala, 40 μg mL−1 Lys171Ala. The
mixture was shaken in 1.5 mL screw-cap glass vial at 30 °C and 800 rpm
in an Eppendorf thermoshaker. Samples of 25 μL volume from the re-
actions with PPi, PPPi, substrates 1a, 1b-13b and 1c were taken at in-
tervals, diluted with 480 μL of 140 mM aq. H2SO4 and analyzed on
HPLC-RI. Reactions were performed in duplicate. Data points are mean
values of at least duplicate reactions. Deviation from mean values was
below 5%, and curves displaying time profile behavior (from duplicate
samples) were perfectly continuous (relative error of ≤5%). Conversion
values in the hydrolysis of PPPi were calculated by relating Pi con-
centration to theoretical maximum Pi concentration (i.e. 750 mM) due
to the insufficient separation of PPi and PPPi on the HPLC column.
In the substrate scope study in transphosphorylation mode with
substrates 2a-10a (Table 1), 31P NMR was used to analyze the reac-
tions. To a standard reaction was added 20 μL conc. HCl after 1 or 2 h,
then 600 μL sample was added to 100 μL D2O and 31P NMR spectrum
was recorded using inverse gated decoupling (ns 32, d1 = 30 s,
pw = 11 μs).
Residues Asn110 and Arg173, which are not conserved in EcAP-like
enzymes, can be found in nucleotide phosphodiesterases and phos-
phonate monoester hydrolases, and were proposed to play a role in
binding of the substrate phosphoryl group in PhoK [10d]. Lys171 is
thought to replace the third metal ion (Mg2+) of EcAP-like alkaline
phosphatases via the protonated ε-amino group. The replacement of
Asn110, Lys171 and Arg173 with Ala most likely affected the sophis-
ticated hydrogen-bonding network responsible for the stabilization of
the phosphate group during catalysis, as suggested with PafA [10d].
Finally, the drastic loss of activity upon mutation of Thr89 clearly
supports the suggested role as phosphorylating residue, previously de-
monstrated with NPP [28]. Only mutation to homologous serine, pre-
sent in EcAP, resulted in a functional, although less active, enzyme.
3. Conclusion
In this study, the catalytic potential of the first member of a recently
proposed new class of alkaline phosphatases relying on a rare catalytic
threonine, PhoK from Sphingomonas sp. BSAR-1, was evaluated for
synthetic applications. Parameters important for activity under process
conditions were identified: i) The deactivating effect on PhoK of oli-
gophosphates via chelation could be alleviated by supplementation of
essential divalent metal ions (Zn2+/Ca2+ in equimolar amounts); ii) pH
value of 9 was determined to be optimum for both phosphotransferase
and hydrolase activity; iii) High concentrations of pyrophosphate and
triphosphate (up to 250 mM) were well accepted in hydrolysis reac-
tions. Importantly, phosphotransferase activity was identified on a
variety of substrates (e.g., alcohols, nucleosides, sugars) using pyr-
ophosphate as donor, however, maximum product levels overall re-
mained below those typically obtained with acid phosphatases and calf
intestine alkaline phosphatase. A range of prim- and sec-phosphate
monoesters was finally successfully hydrolyzed by PhoK, thereby of-
fering a suitable alternative for use for instance in biocatalytic cascade
reactions in the alkaline range. Finally, results of the mutation study of
selected residues in the active site unambiguously assigned the key
catalytic role to threonine 89.
Details on cloning and characterization of enzymes and analytical
methods can be found in the supporting information.
Declaration of Competing Interest
There is no conflict of interest to declare.
Acknowledgements
Funding by the Austrian BMWFW, BMVIT, SFG, Standortagentur
Tirol, Government of Lower Austria and ZIT through the Austrian FFG-
COMET-Funding Program is gratefully acknowledged. Klaus Zangger
(University of Graz) is thanked for NMR measurements.
6

M. Lukesch, et al.
B
B
A-ProteinsandProteomics1868(2020)140291
Appendix A. Supplementary data
e) F. Sunden, A. Peck, J. Salzman, S. Ressl, D. Herschlag, eLife (2015)
e06181–e06211;
f) P.J. O'Brien, J.K. Lassila, T.D. Fenn, J.G. Zalatan, D. Herschlag, Biochemistry 47
(2008) 7663–7672;
g) K.S. Nilgiriwala, S.C. Bihani, A. Das, V. Prashar, M. Kumar, J.-L. Ferrer,
S.K. Apte, M.V. Hosur, Acta Crystallogr. Sect. F: Struct. Biol. Cryst. Comm. 65
(2009) 917–919.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bbapap.2019.140291.
References
[11] T. Koyama, H. Inoue, S.I. Ohnuma, K. Ogura, Tetrahedron Lett. 31 (1990)
4189–4190.
[12] J.-K. Min, H.-S. Yoo, E.-Y. Lee, W.-J. Lee, Y.-M. Lee, Anal. Biochem. 303 (2002)
167–175.
[13] B. Edwards, A. Sparks, J.C. Voyta, R. Strong, O. Murphy, I. Bronstein, J. Organomet.
Chem. 55 (1990) 6225–6229.
[1] M.Y. Galperin, A. Bairoch, E.V. Koonin, Protein Sci. 7 (1998) 1829–1835.
[2] S. Jonas, F. Hollfelder, Pure Appl. Chem. 81 (2009) 731–742.
[3] J.K. Lassila, J.G. Zalatan, D. Herschlag, Annu. Rev. Biochem. 80 (2011) 669–702.
[4] a) A. Pabis, S.C.L. Kamerlin, Curr. Opin. Struct. Biol. 37 (2016) 14–21;
b) V. López-Canut, M. Roca, J. Bertrán, V. Moliner, I. Tuñón, J. Am. Chem. Soc.
133 (2011) 12050–12062;
[14] J.K. Sahoo, C.G. Pappas, I.R. Sasselli, Y.M. Abul-Haija, R.V. Ulijn, Angew. Chem.
Int. Ed. 56 (2017) 6828–6832.
[15] K. Schmölzer, A. Gutmann, M. Diricks, T. Desmet, B. Nidetzky, Biotechnol. Adv. 34
(2016) 88–111.
c) B. van Loo, S. Jonas, A.C. Babtie, A. Benjdia, O. Berteau, M. Hyvonen,
F. Hollfelder, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 2740–2745;
d) A. Kim, M.M. Benning, S. OkLee, J. Quinn, B.M. Martin, H.M. Holden,
D. Dunaway-Mariano, Biochemistry 50 (2011) 3481–3494;
e) S. Jonas, B. van Loo, M. Hyvönen, F. Hollfelder, J. Mol. Biol. 384 (2008)
120–136;
f) R. Gijsbers, H. Ceulemans, W. Stalmans, M. Bollen, J. Biol. Chem. 276 (2001)
1361–1368;
g) M.Y. Galperin, M.J. Jedrzejas, Proteins Struct. Funct. Genet. 45 (2001) 318–324;
h) L.F. Olguin, S.E. Askew, A.M.C. O'Donoghue, F. Hollfelder, J. Am. Chem. Soc.
130 (2008) 16547–16555;
[16] M. Lemmerer, K. Schmölzer, A. Gutmann, B. Nidetzky, Adv. Synth. Catal. 358
(2016) 3113–3122.
[17] H. Yu, Y. Li, J. Zeng, V. Thon, D.M. Nguyen, T. Ly, H.Y. Kuang, A. Ngo, X. Chen, J.
Organomet. Chem. 81 (2016) 10809–10824.
[18] a) R. Wever, T. van Herk, K. Drauz, H. Gröger, O. May (Eds.), Enzyme Catalysis in
Organic Synthesis, Wiley-VCH, 2012, pp. 1001–1033;
b) R. Wever, L. Babich, A.F. Hartog, K. Faber, W.-D. Fessner, N. Turner (Eds.),
Science of Synthesis: Biocatalysis in Organic Synthesis, vol. 1, Georg Thieme Verlag,
2015, pp. 223–254.
[19] A. Pradines, A. Klaebe, J. Perie, F. Paul, P. Monsan, Tetrahedron 44 (1988)
6373–6386.
i) J.G. Zalatan, T.D. Fenn, A.T. Brunger, D. Herschlag, Biochemistry 45 (2006)
9788–9803;
[20] L. Babich, J.L.V.M. Peralta, A.F. Hartog, R. Wever, Int. J. Chem. 5 (2013) 87–98.
[21] a) G. Tasnádi, M. Lukesch, M. Zechner, W. Jud, M. Hall, K. Ditrich, K. Baldenius,
A.F. Hartog, R. Wever, K. Faber, Eur. J. Org. Chem. (2016) 45–50;
b) G. Tasnádi, M. Hall, K. Baldenius, K. Ditrich, K. Faber, J. Biotechnol. 233 (2016)
219–227;
j) J.K. Lassila, D. Herschlag, Biochemistry 47 (2008) 12853–12859.
[5] a) O. Khersonsky, D.S. Tawfik, Annu. Rev. Biochem. 79 (2010) 471–505;
b) M.E. Glasner, J.A. Gerlt, P.C. Babbitt, Curr. Opin. Chem. Biol. 10 (2006)
492–497;
c) F. Baier, J.N. Copp, N. Tokuriki, Biochemistry 55 (2016) 6375–6388.
[6] a) C. Pandya, J.D. Farelli, D. Dunaway-Mariano, K.N. Allen, J. Biol. Chem. 289
(2014) 30229–30236;
c) G. Tasnádi, M. Zechner, M. Hall, K. Baldenius, K. Ditrich, K. Faber, Biotechnol.
Bioeng. 114 (2017) 2187–2195.
[22] C. Jacob, W. Maret, B.L. Vallee, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 3489–3494.
[23] C.B. Murphy, A.E. Martell, J. Biol. Chem. 226 (1957) 37–50.
[24] C.J. Martin, W.J. Evans, J. Inorg. Biochem. 35 (1989) 267–288.
[25] M.M. Pedroso, J.A. Larrabee, F. Ely, S.E. Gwee, N. Mitic, D.L. Ollis, L.R. Gahan,
G. Schenk, Chem. Eur. J. 22 (2016) 999–1009.
b) J.A. Gerlt, P.C. Babbitt, Annu. Rev. Biochem. 70 (2001) 209–246;
c) H. Renata, Z.J. Wang, F.H. Arnold, Angew. Chem. Int. Ed. 54 (2015) 3351–3367.
[7] a) N. Strater, W.N. Lipscomb, T. Klabunde, B. Krebs, Angew. Chem. Int. Ed. 35
(1996) 2024–2055;
b) P.J. O'Brien, D. Herschlag, Biochemistry 41 (2002) 3207–3225.
[8] F. Berlutti, C. Passariello, L. Selan, M.C. Thaller, G.M. Rossolini, Microbiology 147
(2001) 2831–2839.
[26] a) T. van Herk, A. Hartog, A.M. van der Burg, R. Wever, Adv. Synth. Catal. 347
(2005) 1155–1162;
b) L. Babich, A.F. Hartog, L.J.C. van Hemert, F.P.J.T. Rutjes, R. Wever,
ChemSusChem 5 (2012) 2348–2353.
[9] K.S. Nilgiriwala, A. Alahari, A.S. Rao, S.K. Apte, Appl. Environ. Microbiol. 74
(2008) 5516–5523.
[27] a) Y. Zhao, L. Wu, S.J. Noh, K.L. Guan, Z.Y. Zhang, J. Biol. Chem. 273 (1998)
[10] a) H. Wiersma-Koch, F. Sunden, D. Herschlag, Biochemistry 52 (2013) 9167–9176;
b) F. Sunden, I. Alsadhan, A.Y. Lyubimov, S. Ressl, H. Wiersma-Koch, J. Borland,
C.L. Brown, T.A. Johnson, Z. Singh, D. Herschlag, J. Am. Chem. Soc. 138 (2016)
14273–14287;
5484–5492;
b) D. Barford, Trends Biochem. Sci. 21 (1996) 407–412;
c) D.J. Rigden, Biochem. J. 409 (2008) 333–348;
d) W.R. Wang, H.S. Cho, R. Kim, J. Jancarik, H. Yokota, H.H. Nguyen,
I.V. Grigoriev, D.E. Wemmer, S.H. Kim, J. Mol. Biol. 319 (2002) 421–431;
e) P.V. Attwood, Biophys, Acta: Proteins Proteomics 1834 (2013) 470–478.
[28] H. Wiersma-Koch, F. Sunden, D. Herschlag, Biochemistry 52 (2013) 9167–9176.
c) F. Sunden, I. AlSadhan, A. Lyubimov, T. Doukov, J. Swan, D. Herschlag, J. Biol.
Chem. 292 (2017) 20960–20974;
d) S.C. Bihani, A. Das, K.S. Nilgiriwala, V. Prashar, M. Pirocchi, S.K. Apte, J.-
L. Ferrer, M.V. Hosur, PLoS One 6 (2011) e22767;
7