Synthesis of hydrolysis-resistant pyridoxal 5′-phosphate analogs and their biochemical and X-ray crystallographic characterization with the pyridoxal phosphatase chronophin

Article abstract of DOI:10.1016/j.bmc.2015.02.049

A set of phosphonic acid derivatives (1-4) of pyridoxal 5′-phosphate (PLP) was synthesized and characterized biochemically using purified murine pyridoxal phosphatase (PDXP), also known as chronophin. The most promising compound 1 displayed primarily competitive PDXP inhibitory activity with an IC₅₀ value of 79 μM, which was in the range of the K_m of the physiological substrate PLP. We also report the X-ray crystal structure of PDXP bound to compound 3, which we solved to 2.75 ? resolution (PDB code 5AES). The co-crystal structure proves that compound 3 binds in the same orientation as PLP, and confirms the mode of inhibition to be competitive. Thus, we identify compound 1 as a PDXP phosphatase inhibitor. Our results suggest a strategy to design new, potent and selective PDXP inhibitors, which may be useful to increase the sensitivity of tumor cells to treatment with cytotoxic agents.

Full text of DOI:10.1016/j.bmc.2015.02.049

Accepted Manuscript
Synthesis of hydrolysis-resistant pyridoxal 5’-phosphate analogs and their bio-
chemical and X-ray crystallographic characterization with the pyridoxal phos-
phatase chronophin
Gunnar Knobloch, Nauras Jabari, Sven Stadlbauer, Hermann Schindelin, Maja
Köhn, Antje Gohla
PII:
S0968-0896(15)00153-4
DOI:
http://dx.doi.org/10.1016/j.bmc.2015.02.049
BMC 12112
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
16 January 2015
24 February 2015
25 February 2015
Please cite this article as: Knobloch, G., Jabari, N., Stadlbauer, S., Schindelin, H., Köhn, M., Gohla, A., Synthesis
of hydrolysis-resistant pyridoxal 5’-phosphate analogs and their biochemical and X-ray crystallographic
characterization with the pyridoxal phosphatase chronophin, Bioorganic & Medicinal Chemistry (2015), doi: http://
dx.doi.org/10.1016/j.bmc.2015.02.049
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis of hydrolysis-resistant pyridoxal 5’-phosphate analogs and their biochemical
and X-ray crystallographic characterization with the pyridoxal phosphatase chronophin
Gunnar Knobloch1,2,4, Nauras Jabari1,3,4, Sven Stadlbauer3, Hermann Schindelin2, Maja
Köhn3,*, Antje Gohla1,2,*
1Institute of Pharmacology and Toxicology, University of Würzburg, Versbacher Strasse 9,
D-97078 Würzburg, Germany
2Rudolf Virchow Center for Experimental Biomedicine, University of Würzburg, Josef-
Schneider-Strasse 2, D-97080 Würzburg, Germany
3EMBL Heidelberg, Genome Biology Unit, Meyerhofstrasse 1, D-69117 Heidelberg,
Germany
4Equal contributions
*Corresponding authors. Tel.: +49 931 3180099 (A.G.), +49 6221 3878544 (M.K.).
Email addresses: antje.gohla@virchow.uni-wuerzburg.de (A.G.), koehn@embl.de (M.K.)
1

ABSTRACT
A set of phosphonic acid derivatives (1-4) of pyridoxal 5’-phosphate (PLP) was synthesized
and characterized biochemically using purified murine pyridoxal phosphatase (PDXP), also
known as chronophin. The most promising compound 1 displayed primarily competitive
PDXP inhibitory activity with an IC50 value of 79 µM, which was in the range of the Km of
the physiological substrate PLP. We also report the X-ray crystal structure of PDXP bound to
compound 3, which we solved to 2.75 Å resolution (PDB code 5AES). The co-crystal
structure proves that compound 3 binds in the same orientation as PLP, and confirms the
mode of inhibition to be competitive. Thus, we identify compound 1 as a PDXP phosphatase
inhibitor. Our results suggest a strategy to design new, potent and selective PDXP inhibitors,
which may be useful to increase the sensitivity of tumor cells to treatment with cytotoxic
agents.
Keywords:
Cancer therapy
Chronophin
Haloacid dehalogenase-type phosphatase
Pyridoxal 5’-phosphate phosphatase
Non-hydrolyzable pyridoxal 5’-phosphate analogs
2

1. Introduction
Phosphatases of the haloacid dehalogenase (HAD)-type superfamily are emerging as enzymes
that are linked to cancer risk, cancer pathogenesis and sensitivity of cancer cells to treatment
with cytotoxic agents or γ-irradiation[1-6]. Despite their potential importance as therapeutic
target molecules, very few inhibitors of mammalian HAD phosphatases have been reported so
far[7-12]. Structurally, HAD phosphatases are typified by a modified Rossman fold that
positions the catalytic core residues. The first aspartate in the conserved DxDx(V/T) HAD
phosphatase signature motif serves as the nucleophile and phosphoryl group acceptor that
forms a phosphoaspartate intermediate during catalysis. This aspartate also coordinates the
catalytically essential Mg2+-ion[13]. HAD phosphatases are additionally equipped with
structurally highly diversified modules, the cap domains, which contribute to catalytic
efficiency and phosphatase specificity[13, 14].
Pyridoxal phosphatase (PDXP, also known as chronophin[15]) is a HAD phosphatase
dedicated to the metabolism of pyridoxal 5’-phosphate (PLP; Fig. 1) with a Km of 1.5 µM or
36 µM for the human or murine enzymes, respectively[14, 16]. PLP is the biologically active
form of vitamin B6 that functions as a cofactor in the catalysis of more than 140 different
enzymatic reactions, including trans- and deaminations, decarboxylations, racemizations,
aldol cleavage reactions, eliminations and substitutions at β- and γ-atoms of amino acid side
chains[17]. In mammals, enzymes using PLP as a prosthetic group are involved in the
synthesis of neurotransmitters, amino acid metabolism, glycogen breakdown, heme
biosynthesis and lipid metabolism[18].
A recent cohort study has demonstrated that increased metabolism and disposal of vitamin B6
is associated with a higher risk of developing malignant neoplasms, in particular lung
cancer[19]. Moreover, elevated circulating vitamin B6 levels are inversely correlated with the
risk of developing non-small cell lung cancer (NSCLC)[20]. Importantly, low expression
levels of the PLP-generating pyridoxal kinase (PDXK) can serve as a biomarker for poorer
disease outcome in NSCLC patients[21, 22]. The same study showed that a membrane-
permeable PLP precursor sensitizes lung cancer cell lines to lethal responses induced by
various types of physical stressors and chemotherapeutics, and that PDXK is required for
optimal cytotoxic effects[21]. Consistent with these findings, depletion of PDXP was
observed to sensitize cancer cells to death, and cisplatin-resistant cells exhibited a decreased
PDXK/PDXP ratio, either due to decreased PDXK or to increased PDXP protein levels[21].
3

Furthermore, overexpression of PDXP transcripts has been observed in clinical NSCLC
specimens[23]. Although the mechanistic basis of each of these effects remains to be fully
understood, an increase in cellular PLP levels may sensitize cancer cells to cytotoxic insults
by targeting PLP-dependent enzymes that influence stress-relevant metabolic circuitries[17].
Together, these data suggest that PLP elevation in cancer cells may be beneficial as an
adjuvant pharmacological strategy in cancer therapy.
Attempts to increase PLP levels in cancer cells by inhibiting the PLP-catabolizing PDXP have
not yet been described. However, this approach may be valuable in cancers with PDXK
deficiency or PDXP overexpression, which is the case in a subset of NSCLC patients[21, 23].
As a first step towards this aim, we have synthesized hydrolysis-resistant PLP analogs (Fig. 1,
compounds 1-4) to target the active site of PDXP, and have characterized their inhibitory
potential using purified PDXP. Furthermore, we have solved the first X-ray crystal structure
of PDXP bound to one of these substrate analogs. Our results lay the groundwork for
structure-based drug design of active site-directed, potent and specific PDXP inhibitors.
2. Results and discussion
2.1. Chemistry
We designed the hydrolysis-resistant PLP analogs by replacing the 5'-phosphate group of PLP
with a phosphonate. The 4'-position of the PLP pyridine ring was either substituted with an
aldehyde (compound 1, Fig. 1), as in PLP, or with an alcohol (compound 3, Fig. 1), as in the
PLP precursor pyridoxine phosphate (PNP). Furthermore, a double bond was introduced at
the 5'-position of the pyridine ring in order to study the impact of the resulting structural
rigidity of the phosphonate residue on the inhibition potency of compounds 1 and 3
(compounds 2 and 4, Fig. 1).
The synthesis of the four compounds was based on published procedures[24, 25]. In the first
step, commercially available pyridoxine was dimethyl acetal-protected on the 3’- and 4’-
hydroxy groups (Scheme 1). Oxidation of the alcohol on the 5’-position furnished aldehyde 6,
which was then subjected to a Wittig-reaction with tetraethyl methylendiphosphonate to yield
phosphonate 7[24]. Deprotection of the hydroxyl groups led to phosphonate 8, and the four
reaction steps proceeded with good to excellent yields. The final compound 4 was obtained by
cleavage of the ethyl esters of the phosphonate group of compound 8 using trimethylsilyl
(TMS)-bromide[26]. Aldehyde 2 was synthesized by oxidation of the 4’-alcohol of compound
4

8 and subsequent TMS-Br mediated deprotection. To obtain final compounds 1 and 3, the
double bond of compound 8 was reduced to furnish 9, which was again deprotected using
TMS-Br to yield alcohol 3. Oxidation of compound 9 and subsequent TMS-Br mediated
deprotection provided aldehyde 1.
2.2. PDXP inhibition
The hydrolysis-resistant PLP analogs (compounds 1-4) were characterized biochemically in
dose-response and steady state kinetic assays using recombinantly expressed and highly
purified murine PDXP. Dose-response assays employing PDXP (100 nM) and its
physiological substrate PLP (20 µM) showed that the half-maximal inhibitory concentration
(IC50) of compounds 1 or 2 (containing an aldehyde in the 4’-position) towards PDXP was
reached at 79 1.2 µM or 246 1.1 µM, respectively. Thus, the presence of a C–C double
bond in the 5’-position in compound 2 increased the IC50 by approximately 3-fold compared
to compound 1. Compounds 3 and 4 (containing an alcohol at the 4’-position) had ~5−16-fold
higher IC50 values of around 1.3 mM (Fig. 2).
The kinetic constants derived from steady state kinetic measurements of PDXP-catalyzed
PLP dephosphorylation in the presence of compounds 1-4 demonstrated that compounds 1
and 2 increased the Km up to ~6-fold at the highest tested compound concentrations (2 mM).
In addition, the calculated vmax and the kcat values were reduced ~2-fold under these
conditions. As a result, the catalytic efficiency (kcat/Km) of PDXP was reduced to ~10% in the
presence of 2 mM compound 1 compared to buffer control conditions. Compounds 3 and 4
increased the Km up to ~3−5-fold, but did not markedly affect vmax and kcat. The catalytic
constants are summarized in Table 1. These data confirm that compounds 1-4 indeed act
primarily as competitive PDXP inhibitors, although the vmax and kcat changes observed in the
presence of compounds 1 and 2 may point to an additional inhibitory mechanism.
The results presented above indicate that the presence of an aldehyde group as compared to an
alcohol group at the 4’-position of the substrate analogs may result in a greater binding
affinity to PDXP. This is in line with a previous study reporting that an increase in
electronegativity at the 4’-position lowers the Km value of B6 vitamers when assayed with
human PDXP (Km pyridoxamine phosphate (PMP), 34 µM; Km pyridoxine phosphate (PNP),
5.19 µM; Km pyridoxal phosphate (PLP) 1.47 µM[16]). The crystal structures of human
PDXP in complex with PLP (PDB codes 2P69 and 2CFT) show that PLP is coordinated by an
5

extensive binding network in the active site, taking advantage of every electronegative atom
on the molecule. In addition, the PLP pyridine ring interacts with the imidazole ring of His182
in the PDXP substrate specificity loop by π-electron stacking (distance, 3.3−3.8 Å; Fig.
3)[14]. The structures do not reveal any strong binding interactions between the 4’-aldehyde
and its surrounding residues. Nevertheless, the 4’-aldehyde has the potential to build weak
hydrogen bonds with several adjacent residues through multiple water bridges, and it can
additionally form a weak hydrogen bond to Asn60 (mean distance, 3.4 Å; Fig. 3). Based on
this information, we hypothesize that the binding affinity and inhibitory efficacy of PLP
analogs might be further increased by substituents that bring the 4’-aldehyde moiety −or
another proton accepting functional group at this position− closer to Asn60. Additionally, the
presence of a more electronegative moiety in the 4’-position may induce a greater pulling
effect on the π-electrons of the pyridine ring, thus facilitating π-electron-stacking with
His182.
2.3. X-ray crystal structure of inhibitor 3 in complex with PDXP
For detailed structural information on the interactions of compounds 1-4 with murine PDXP,
we attempted to co-crystallize them with PDXP. Although several PDXP crystals were
obtained in the presence of all compounds, only a structure of PDXP in complex with
compound 3 contained electron density representing the entire substrate analog. The
PDXP·compound 3-complex crystallized in space group I23, and the structure was solved by
molecular replacement using the previously solved structure of murine PDXP/chronophin
(PDB code 4BX3). The structure was subsequently refined to 2.75 Å resolution with an Rcryst
of 19.0% and an Rfree of 24.7% (PDB code 5AES). The data collection and refinement
statistics are summarized in Table 2.
Figure 4 shows an overview of the PDXP·compound 3-complex and a detailed view of the
inhibitor bound to the active site of PDXP. As previously reported for murine PDXP, which
exists as a homo-dimer in solution, PDXP also crystallized as a homodimer in complex with
compound 3 [14]. The overall structure of murine PDXP remains unaltered by ligand binding,
as indicated by a root mean square (r.m.s.) deviation of 0.26 Å when the structure was aligned
with the previously solved structure of murine apo-PDXP (PDB code 4BX3). The
coordination of compound 3 within the catalytic pocket of murine PDXP is similar to the
coordination of PLP bound to human PDXP (PDB codes 2P69 and 2CFT). The residues that
are closest to the alcohol group in the 4'-position of compound 3 are Asp27 (mean distance,
6

3.75 Å) and Asn60 (mean distance, 4.75 Å). These distances are too far for hydrogen
bonding, but could allow van der Waals interactions. Nevertheless, since the mean distance
between the 4'-aldehyde group of PLP and Asn60 in human PDXP (PDB code 2CFT) spans
only 3.5 Å and results in the formation of a weak hydrogen bond, replacing the alcohol group
in the 4'-position of compound 3 with an aldehyde function should likewise result in hydrogen
bonding interactions with Asn60. Asn60 is the terminal amino acid in the HAD consensus
motif II (56FVSNN60 in human and mouse PDXP); the strictly conserved Ser58 in this motif
helps to orient the substrate for the nucleophilic attack and to stabilize the phosphoaspartate
intermediate that is formed during catalysis[13] (Fig. 5).
The PDXP·compound 3 structure will facilitate the rational development of inhibitory
compounds with higher efficacy and specificity for PDXP. PLP analogs containing a 4’-
aldehyde group can be expected to affect the activities of all PLP-dependent enzymes. This is
because the use of PLP as a prosthetic group involves the formation of a covalent Schiff-base
with the 4’-aldehyde group of PLP and the ε-amino group of a lysine side chain in the active
site of the respective enzyme[17]. The introduction of electronegative functional groups other
than an aldehyde at the 4’-position of hydrolysis-resistant substrate analogs is expected to
increase the affinity of the compounds for PDXP[16], and at the same time to prevent binding
to PLP-dependent enzymes. It has been reported that PLP substituted with a phenylalanine at
the 4'-position [N-(5'-phospho-4'-pyridoxyl)phenylalanine] is a better PDXP substrate than
PNP[27]. Consistent with our structural data, this finding suggests that PDXP tolerates bulky
substituents at the 4’-position, which could be exploited to improve the specificity of second
generation PDXP inhibitors (Fig. 5). The same study also showed that the chemical
modification of arginines with phenylglyoxal led to PDXP inactivation, which was prevented
in the presence of PLP, thus suggesting that at least one arginine residue is involved in PLP
coordination[27]. Indeed, the X-ray crystal structures of both human and murine PDXP reveal
three arginines in the entrance of the active site (see Figs. 3 and 5; PDB codes 2P69, 2CFT
and 4BKM). This positively charged region (the ‘arginine mouth’) may be accessed by
replacing the PLP 4'-position with negatively charged substituents of sufficient length.
Other modifications that may increase the affinity of active site-directed PDXP inhibitors
described in the present study involve the insertion of a difluoromethylphosphonate group at
the 5'-position to form hydrogen bonds with Ser58 and Asn60 backbone atoms of the PDXP
catalytic domain. Finally, replacements of the 2'-methyl group with other hydrophobic
7

substituents may increase hydrophobic interactions with Tyr146 (Tyr150) of the capping
domain in murine (human) PDXP (Fig. 5).
3. Conclusions
The elevation of intracellular PLP levels by means of inhibiting the activity of the pyridoxal
phosphatase PDXP/chronophin is a novel concept to increase the sensitivity of cancer cells to
cytotoxic agents. We show here that PDXP inhibition is feasible by targeting its active site
using PLP-based antimetabolites. The synthesis of non-hydrolyzable PLP analogs, which
were prepared previously and assayed for their biological activity against tyrosine
decarboxylase apo-enzyme and aspartate aminotransferase[24], proceeded successfully
following published procedures[24-26]. Aldehyde 1 was the most potent inhibitor of PDXP-
catalyzed PLP hydrolysis, with an IC50 of 79 1.2 µM, which is in the range of the Km of the
physiological substrate PLP (Km = 42 µM PLP for murine PDXP in the current study). Thus,
the 4’-aldehyde group in the PLP analogs appears to be an important structural element for the
recognition by PDXP. The presence of compound 1 in kinetic measurements of PDXP-
mediated PLP hydrolysis led to an increase of the Km, but did not markedly change vmax
,
showing that compound 1 is primarily a competitive inhibitor of PDXP. While this compound
was not entirely visible in the crystal structure in complex with PDXP, the complex of PDXP
bound to another hydrolysis-resistant substrate analog (alcohol 3) was solved by X-ray
crystallography (2.75 Å, PDB code 5AES). The co-crystal structure proves that compound 3
binds in the same orientation as the natural substrate PLP does, and confirms the mode of
inhibition to be competitive. In summary, we identified compound 1 as the first inhibitor of
PDXP phosphatase activity. The PDXP·compound 3 structure can now be used to design new
inhibitors of PDXP that are less PLP-like, in order to potentially achieve selectivity over other
PLP-binding proteins.
4. Experimental protocols
4.1. Chemistry
Anhydrous solvents and all reagents were purchased from Sigma-Aldrich. All reactions
involving air- or moisture-sensitive compounds were performed under argon atmosphere
using dried glassware and syringes techniques to transfer solutions. Nuclear magnetic
resonance (1H NMR, 31P NMR, 13C NMR) spectra were recorded using a 400 MHz Bruker
Avance DPX. Chemical shifts (δ) are reported in parts per million (ppm) and the coupling
8

constants (J) are expressed in Hertz (Hz). Splitting patterns are designated as follows: s,
singlet; d, doublet; sept, septet; t, triplet; q, quadruplet; m, multiplet; br s, broad singlet; br m,
broad multiplet; dd, double of double. The assignment of exchangeable protons (NH) was
confirmed by the addition of D2O. Analytical thin-layer chromatography (TLC) was carried
out on Merck silica gel F-254 plates. HPLC analysis and purifications were carried out on a
Shimadzu High Performance Liquid Chromatograph/Mass Spectrometer LCMS-2010EV with
a UV/Vis photodiode array detector SPD-M20A Prominence. The analytical column was a
Macherey Nagel C18 EC 250/4.0 NUCLEODUR 100-5 C18 ec for gradients of 10–100%
CH3CN in water and a Macherey Nagel EC 250/4.6 NUCLEODUR C18 Pyramid for
gradients staring from 0%. For preparative separations a Macherey Nagel C18 VP 250/21
NUCLEODUR 100-C5 C18 ec column was used. For purity analysis and mass spectrometry
an Agilent Technologies UPLC 1290 Infinity system coupled with a 6120 Quadrupole mass
spectrometer was used.
3,4-O-isopropylidene pyridoxine 5
To a mixture of pyridoxine (3.5 g, 20.7 mmol) in CH2Cl2 (70 mL) dimethoxypropane (21.1
mL, 172 mmol) and 96% H2SO4 (1.25 mL, 23.5 mmol) were added under argon atmosphere.
The mixture was refluxed at 45 ˚C for 6 h, resulting in a deep brownish/red color. The
solution was then diluted with CH2Cl2, washed with sodium bicarbonate, and extracted with
CH2Cl2. The combined organic fractions were dried over sodium sulfate, and concentrated in
vacuo. The crude compound was re-dissolved in a mixture of Et2O/pentane (2:1) and refluxed
at 40˚C for 30 min. During this time, the desired product began crystallizing. The solution
was then tempered to room temperature and kept at –25˚C for 30 h. The precipitate was
collected via suction filtration and washed with cold (–25˚C) pentane to afford 5 as a fluffy,
white, crystalline solid (3.72 g, 86%): mp 106–107 °C; 1H NMR (400 MHz, CDCl3) δ 1.55 (s,
6H), 2.39 (s, 3H), 2.79 (s-broad, 1H), 4.56 (s, 2H), 4.94 (s, 2H), 7.86 (s, 1H). 13C NMR (100
MHz, CDCl3) δ 18.1, 24.7, 58.6, 60.0, 99.8, 126.1, 129.7, 138.4, 146.1, 147.5. HPLC-MS
(10–100% CH3CN in H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 5.8 min; m/z: 210.0
[M+H]+.
3,4-O-isopropylidene pyridoxal 6
MnO2 (27.6 g, 317 mmol) was added to a solution of 5 (3.70 g, 17.7 mmol) in CH2Cl2 (36
mL) and the mixture was stirred vigorously for 2 h at room temperature. Then the reaction
mixture was filtered through a short silica column to remove the MnO2 and the product was
9

eluted with CH2Cl2. Subsequent evaporation yielded 6 as a pale yellow syrup (1.91 g, 51%).
Further purification by column chromatography (100% EtOAc) afforded a clear and viscous
liquid that formed solid white plates at room temperature or in a freezer, mp 57–59 °C; 1H
NMR (400 MHz, CDCl3) δ 1.56 (s, 6H), 2.51 (s, 3H), 5.18 (s, 2H), 8.47 (s, 1H), 10.04 (2,
1H). 13C NMR (100 MHz, CDCl3) δ 18.8, 24.7, 60.2, 100.5, 126.3, 144.8, 147.8, 153.5, 191.4.
HPLC-MS (10–100% CH3CN in H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 7.1 min;
m/z: 208.1 [M+H]+, 415.2 [M2+H]+.
Diethyl 2-(3,4-O-isopropylidene-2-methyl-5-pyridyl)ethenylphosphonate 7
A solution of CH2[PO(OEt)2]2 (4.73 mL, 19.0 mmol) in cyclohexane (15 mL) was added to a
suspension of NaH (399 mg, 60% in mineral oil, 9.98 mmol) in cyclohexane (15 mL) at room
temperature, resulting in vigorous production of gas. After the reaction mixture became clear,
a solution of 6 (1.88 g, 9.07 mmol) in 15 mL of cyclohexane, was added dropwise at room
temperature. After approximately 60% of the aldehyde had been added, a thick, gummy
precipitate formed at reaction scales over 1 g of 6. The resulting mixture was stirred for 1 h,
then diluted with a mixture of CHCl3/H2O (2:1), and extracted three times with CH2Cl2. The
combined organic fractions were concentrated to provide a slightly viscous amber liquid,
which was subjected to column chromatography (EtOAc/Cyclohexane (1:1)) to afford 7 as an
amorphous white solid (2.94 g, 95%). 1H NMR (400 MHz, CDCl3) δ 1.36 (t, 6H), 1.56 (s,
6H), 2.17 (s, 4H), 2.44 (s, 3H), 4.14 (m, 4H), 4.91 (s, 2H), 6.23 (t, 1H), 7.33 (dd, 1H), 8.23 (s,
1H). 31P NMR (162 MHz, CDCl3) δ 17.7 (s).
Diethyl 2-(3-hydroxy-4-hydroxymethyl-2-methyl-5-pyridyl)ethenylphosphonate 8
7 (3.61 g, 12.0 mmol) was refluxed in a 10% solution of HCO2H in H2O (20 mL) for 1 h,
resulting in a clear, yellow solution. Although starting material was left, it was crucial to stop
the reaction after 1 h in order to prevent formation of side-products. Upon completion, the
water solution was concentrated and the resulting residue purified via column
chromatography (CH2Cl2/MeOH (9:1)), which afforded 8 as pale yellow plates (2.3 g, 72%).
Mp 96.5–97.5 °C; 1H NMR (400 MHz, CDCl3) δ 1.34 (t, 6H), 2.52 (s, 3H), 4.09 (m, 4H), 5.10
(s, 2H), 6.18 (t, 1H), 7.46 (dd, 1H), 8.09 (s, 1H). 13C NMR (100 MHz, CDCl3) δ 14.6, 16.4
(3JCP = 5 Hz), 60.7, 63.0 (2JCP = 5 Hz), 100.0, 123.0, 124.7, 127.9, 130.8, 138.4, 143.5. 31P
NMR (162 MHz, CDCl3) δ 17.3 (s). UPLC-MS (5–100% CH3CN in H2O with 0.1% TFA;
flow rate 1.5 mL/min): tR = 1.13 min; m/z = 302.1 [M+H]+.
10

2-(3-hydroxy-4-hydroxymethyl-2-methyl-5-pyridyl)ethenylphosphonic acid 4
Under argon atmosphere bromotrimethylsilane (262 µL, 19.9 mmol) was added dropwise to a
solution of 8 (100 mg, 0.33 mmol) in CH2Cl2 (2 mL) and the resulting mixture was stirred for
6 h at room temperature. Then bromotrimethylsilane (87 µL, 0.66 mmol) was added again and
the mixture was stirred overnight at room temperature. The reaction was quenched by
addition of water and the mixture was extracted with water. The combined aqueous phases
were lyophilized and the product was obtained in quantitative yield. No further purification
was required [26]. Solubility in water was low when re-solubilizing the compound. Due to
this low solubility, 13C chemical shifts were obtained indirectly from HSQC experiments for
those C-atoms bearing protons. Mp >210 °C (dec); 1H NMR (400 MHz, D2O) δ 2.62 (s, 3H),
4.99 (s, 2H), 6.53 (t, 2H), 7.41 (dd, 1H), 8.29 (s, 1H). 13C NMR (100 MHz, D2O) δ 14.0, 56.3,
128.7, 129.4, 133.4. 31P NMR (162 MHz, CDCl3) δ 10.5 (s). UPLC-MS (5–100% CH3CN in
H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 0.19 min; m/z = 246.0 [M+H]+.
Diethyl 2-(3-hydroxy-4-hydroxymethyl-2-methyl-5-pyridyl)ethylphosphonic acid 9
To a solution of 8 (1.46 g, 4.85 mmol) in ethanol (40 mL), Pd on charcoal (163 mg, ca. 10%
Pd, 0.15 mmol) was added and the resulting suspension hydrogenated at 1 atm of H2 at room
temperature for 1 h. The black suspension was filtered and the solution was concentrated in
vacuo, followed by recrystallization in EtOAc/Et2O (1:1). 9 was obtained as a greyish-white
solid (0.78 g, 53%): mp 107.5–109 °C; 1H NMR (400 MHz, CDCl3) δ 1.29 (t, 6H), 1.88 (m,
2H), 2.43 (s, 3H), 2.79 (m, 2H), 4.03 (m, 4H), 4.86 (s, 2H), 7.77 (s, 1H). 13C NMR (100 MHz,
CDCl3) δ 14.5, 16.4, 16.5, 22.7, 25.3, 25.9 (d, 1JCP = 110 Hz), 59.6, 62.36, 62.42, 66.7, 130.6,
135.0, 135.1, 139.3, 141.6, 148.1, 154.3. 31P NMR (162 MHz, CDCl3) δ 30.2 (s). UPLC-MS
(5–100% CH3CN in H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 1.12 min; m/z = 304.1
[M+H]+.
Diethyl 2-(4-formyl-3-hydroxy-2-methyl-5-pyridyl)ethylphosphonate 10
MnO2 (1.21 g, 13.9 mmol) was added to a solution of 9 (0.30 g, 0.99 mmol) dissolved in
CHCl3 (24 mL) and the resulting suspension was stirred vigorously for 1.5 h at room
temperature. Then the suspension was filtered through a short silica column to remove the
MnO2 and the product was eluted with CHCl3. Subsequent evaporation yielded 10 as a highly
viscous deep golden-yellow syrup (280 mg, 94%).1H NMR (400 MHz, D2O) δ 1.34 (m, 6H),
2.07 (m, 1 H), 2.35 (m, 1H), 2.73 (s, 1H), 2.85 (s, 1H), 3.39 (m, 3 H), 4.13 (m, 4H), 8.30 (s,
1H), 10.58 (s, 1H). 13C NMR (100 MHz, D2O) δ 15.0, 16.4, 18.4, 21.7, 26.8, 62.4, 62.6,
11

113.0, 122.4, 132.7, 137.8, 147.5, 154.7, 194.5. 31P NMR (162 MHz, D2O) δ = 32.4. HPLC-
MS (10–50% CH3CN in H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 7.6 min; m/z:
302.0 [M+H]+.
2-(3-hydroxy-4-hydroxymethyl-2-methyl-5-pyridyl)ethylphosphonic acid 3
Under argon atmosphere bromotrimethylsilane (522 µL, 3.96 mmol) was added dropwise to a
solution of 9 (200 mg, 0.66 mmol) in CH2Cl2 (4 mL) and the resulting mixture was stirred for
6 h at room temperature. Then bromotrimethylsilane (174 µL, 1.32 mmol) was added again
and the mixture was stirred overnight at room temperature. The reaction was quenched by
addition of H2O and the product extracted into the aqueous layer, followed by lyophilization.
Although after bromotrimethylsilane-treatment no further purification was required due to the
side products being either volatile or remaining in the organic layer [26], we purified the
resulting residue with preparative HPLC (0–30% CH3CN in H2O, flow rate 5 ml/min) to
ensure its purity, which led to the loss of a large amount of the compound due to solubility
issues. Fractions containing the product were combined, the solvent evaporated and the
aqueous solution lyophilized, affording 3 as white solid powder (11.8 mg, 7.2%). 1H NMR
(400 MHz, D2O) δ 1.84 (m, 2H), 2.56 (s, 3H), 2.94 (m, 2H), 4.97 (s, 2H), 8.02 (s, 1H). 13C
NMR (100 MHz, D2O) δ 14.1, 23.7, 27.6, 28.9, 56.8, 131.1, 138.2, 140.9, 152.8. 31P NMR
(162 MHz, D2O) δ 22.6 (s). HPLC-MS (NUCLEODUR C18 Pyramid column, 0–30%
CH3CN in H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 2.2 min; m/z: 248.1 [M+H]+,
495.2 [M2+H]+.
2-(4-formyl-3-hydroxy-2-methyl-5-pyridyl)ethylphosphonic acid 1
Under argon atmosphere bromotrimethylsilane (612 µL, 4.64 mmol) was added dropwise to a
solution of 10 (233 mg, 0.77 mmol) in CH2Cl2 (4.5 mL) and the mixture was stirred for 6 h at
room temperature. Then bromotrimethylsilane (78 µL, 0.59 mmol) was added again and the
mixture was stirred overnight at room temperature. The reaction was quenched by addition of
H2O and the product extracted into the aqueous layer followed by lyophilization, and the
product was obtained in quantitative yield. No further purification was required [26].
Solubility issues were encountered when trying to re-dissolve the compound, and it was not
possible to obtain sufficient NMR spectra. HPLC-MS (NUCLEODUR C18 Pyramid column,
0–50% CH3CN in H2O with 0.1% TFA; flow rate 1.5 ml/min): tR = 2.1 min; m/z: 246.0
[M+H]+.
12

Diethyl 2-(4-formyl-3-hydroxy-2-methyl-5-pyridyl)ethenylphosphonic acid 11
MnO2 (1.24 g, 14.3 mmol) was added to a solution of 8 (310 mg, 1.03 mmol) in CHCl3 (25
mL) and the resulting mixture was stirred vigorously for 1.5 h at room temperature, filtered
through a short silica column to remove the MnO2, and the product was eluted with CHCl3.
Subsequent evaporation yielded 11 as a light greenish-yellow syrup (0.27 g, 88%). 1H NMR
(400 MHz, CDCl3) δ 1.36 (t, 6H), 2.55 (s, 3H), 4.16 (m, 4H), 6.31 (t, 1H), 7.95 (dd, 1H), 8.24
(s, 1H), 10.42 (s, 1H). 13C NMR (100 MHz, CDCl3) δ 16.4, 19.0, 62.3, 118.8, 122.1, 124.0,
129.0, 137.8, 139.8, 153.1, 195.4. 31P NMR (162 MHz, CDCl3) δ 16.0 (s).
2-(4-formyl-3-hydroxy-2-methyl-5-pyridyl)ethenylphosphonic acid 2
Under argon atmosphere bromotrimethylsilane (524 µL, 3.97 mmol) was added dropwise to a
solution of 11 (198 mg, 0.66 mmol) in CH2Cl2 (4 mL) and the resulting mixture was stirred
for 6 h at room temperature. Then bromotrimethylsilane (175 µL, 1.33 mmol) was added
again and the mixture was stirred overnight at room temperature. The reaction was quenched
by addition of H2O and the product was extracted into the aqueous layer followed by
lyophilization, and the product was obtained in quantitative yield. No further purification was
required [26]. 13C chemical shifts were obtained indirectly from HSQC experiments for those
C-atoms bearing protons. However, the signal for the aldehyde group was not visible. Mp
>150 °C (dec); 1H NMR (400 MHz, MeOD4) δ 2.57 (s, 3H), 6.53 (dd, 1H, 3JHH = 16 Hz, 2JHP
= 16 Hz), 7.40 (dd, 1H, 3JHH = 20 Hz, 3JHP = 16 Hz), 8.22 (s, 1H), 10.52 (s, 1H). 13C NMR
(100 MHz, MeOD4) δ 138.7, 134.7, 126.0, 16.1. 31P NMR (162 MHz, MeOD4) δ 9.6 (s).
UPLC-MS (5–100% CH3CN in H2O with 0.1% TFA; flow rate 1.5 mL/min): tR = 0.19 min;
m/z = 262.1 [M+H2O+H]+.
4.2. Protein expression and purification
Murine PDXP/chronophin cDNA was reverse-transcribed from adult mouse brain tissue.
Total RNA was isolated using TRIzol (Invitrogen) according to the manufacturer’s
instructions, and cDNA was obtained with the High Fidelity RNA PCR Kit (Takara) and
oligo dT primers. The PCR product was subcloned into the bacterial expression vector
pETM11 (EMBL) to create N-terminally His6-tagged chronophin for in vitro studies. His6-
tagged PDXP was transformed into BL21(DE3) cells (Stratagene), and expressed for 18 h at
20°C after induction with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). Cells were
harvested at 8,000 × g for 10 min and lysed in 100 mM triethanolamine (TEA), 500 mM
NaCl, 20 mM imidazole, 5 mM MgCl2; pH 7.4, in the presence of protease inhibitors (EDTA-
13

free protease inhibitor tablets, Roche) and 150 U/ml DNase I (Applichem) using a cell
disruptor (Microfluidizer Processor M-110 P, Microfluidics). Cell debris was removed by
centrifugation for 30 min at 30,000 × g. For purification, cleared supernatants were loaded on
a HisTrap HP column operated on an ÄKTA liquid chromatography system (GE Healthcare)
in binding buffer (50 mM TEA, 500 mM NaCl, 20 mM imidazole, 5 mM MgCl2; pH 7.4), and
His6-tagged proteins were eluted using a linear gradient up to 50% elution buffer (50 mM
TEA, 250 mM NaCl, 500 mM imidazole, 5 mM MgCl2; pH 7.4). Fractions containing His6-
tagged PDXP were pooled, and the His6-tag was cleaved with tobacco etch virus (TEV)-
protease for 2 days at 4°C. Subsequently, cleaved protein was separated from uncleaved
protein and from the His6-tagged TEV-protease on a HisTrap HP column. Untagged PDXP
was further purified on a HiLoad 16/60 Superdex 200 pg size exclusion chromatography
column (GE Healthcare) in buffer A [50 mM TEA, 250 mM NaCl, 5 mM MgCl2, 5% (v/v)
glycerol; pH 7.4].
4.3. In vitro PDXP phosphatase inhibitor assays
Pyridoxal 5’-phosphate (PLP) dephosphorylation assays were conducted in the absence and
presence of substrate analogs in 96- or 384-well microtiter plates (for kinetic or dose-response
assays, respectively). One hundred nM of PDXP per well were pre-incubated for 10 min at
22°C in buffer A supplemented with 0.001% (v/v) Triton X-100, followed by 4 min of
incubation with the respective compound (dissolved in 50 mM TEA, 250 mM NaCl, 5 mM
MgCl2; pH 7.4). The reactions were started by the addition of PLP. The final PLP
concentration was set to 20 µM in dose-response assays, or ranged between 0–500 µM in
kinetic assays. The final volume of each reaction was 50 µL for kinetic assays, and 25 µL for
dose-response assays. Kinetic measurements were stopped after 2 min, and dose-response
assays were stopped after 5 min by the addition of 50 or 100 µL Biomol Green, respectively
(Enzo Life Sciences). Color was allowed to develop for 15–20 min before the absorbance of
the resulting phosphomolybdate complex was read at 620 nm on an Envision 2104 multilabel
microplate reader (Perkin Elmer). Free phosphate release was quantified using phosphate
standard curves, and vmax and Km values were calculated using GraphPad Prism (GraphPad
Software Inc.), version 6. The lines were fitted by nonlinear regression using the least squares
fitting method. For dose-response assays, loginhibitor versus response was calculated for a Hill
slope of -1.
4.4. Crystallization and data collection
14

PDXP was concentrated to 8−10 mg/mL (as determined by absorption at 280 nm using a
calculated molar extinction coefficient of 18,450 M-1 · cm-1) in 10 mM TEA, 100 mM NaCl, 1
mM MgCl2 using 10 kDa MWCO centrifugal filter devices (Amicon Ultra-15, Millipore). The
crystals were grown at 20°C using the hanging-drop vapor-diffusion method, by mixing equal
volumes of protein solution with reservoir solution. The protein was crystallized in 0.1 M
imidazole, 0.2 M NaCl and 1 M sodium tartrate, supplemented with 5 mM compound 3. The
cubic crystals appeared within hours after setting up the crystallization experiment. All
crystals were cryoprotected for flash-cooling in liquid nitrogen by soaking in mother liquor
containing 30% (v/v) glycerol. PDXP diffraction data were collected on an R-axis HTC
imaging plate detector mounted on a Micromax HF-007 rotating anode X-ray generator. Data
were processed using iMosflm[28] and scaled with Scala from the CCP4 program suite[29].
The structure was solved by molecular replacement with the program Phaser [30] using the
previously solved murine PDXP/chronophin structure (PDB entry 4BX3) as a search model.
Ligand restraints for the refinement of compound 3 were generated using the electronic
Ligand Builder and Optimization Workbench (eLBOW[31]) and the Restraints Editor
Especially Ligands (REEL) of the Phenix program suite[32, 33]. The PDXP structure in
complex with compound 3 was refined at 2.75 Å resolution with Phenix[34], incorporating
torsion angle non-crystallographic symmetry (ncs) restraints. Data collection and refinement
statistics are summarized in Table 2. The figures were generated with PyMOL (The PyMol
Molecular Graphics System, version 1.5.0.4, Schrödinger, LLC). The X-ray crystal structure
of murine PDXP/chronophin bound to compound 3 has been deposited in the Protein Data
Bank under accession code 5AES.
15

FIGURE LEGENDS
Figure 1. PLP and the non-hydrolyzable analogs used in this study.
Scheme 1. Synthesis of compounds 1- 4. Reagents and conditions: a) dimethoxypropane,
H2SO4, CH2Cl2 reflux, 6h, argon; b) MnO2, CH2Cl2, r.t., 2h; c) CH2[PO(OEt)2]2, NaH,
cyclohexane, r.t., 2h; d) 10% HCO2H in H2O, reflux, 1h; e) 10% Pd/C, H2, ethanol, r.t., 1h; f)
MnO2, CHCl3, r.t., 1.5h; g) TMS-Br in CH2Cl2, r.t., 6h, argon, then again TMS-Br overnight
(see the experimental protocols for the low yield of compound 3). r.t. = room temperature.
Figure 2. Dose-response curves of PDXP-mediated PLP dephosphorylation in the presence of
compounds 1-4 (A) or compound 1 (B). The results represent mean values S.E. of three
independently performed experiments conducted with three independently purified batches of
protein. Controls were conducted in the absence of compounds. Control values of each
experiment in A were normalized to the mean of n=3 measurements. Errors bars not seen are
hidden within the symbols.
Figure 3. Coordination of PLP in the active site of human PDXP (PDB code 2CFT). Water
molecules are represented as red spheres.
Figure 4. Structure of murine PDXP in complex with compound 3 (PDB code 5AES). (A)
One protomer of the homodimeric PDXP is shown as a cartoon representation (rainbow
color), the other protomer is depicted as a surface representation (gray). Compound 3 is
shown in green and the cofactor Mg2+ as a magenta sphere. A detailed structure of the bound
inhibitor and adjacent residues of the active site is shown in the boxed area. An Fo-Fc omit
electron density map of the inhibitor is contoured at an r.m.s. deviation of 3 in red and is
superimposed with the refined model. (B) Structural comparison of murine PDXP in complex
with compound 3 (rainbow color) and human PDXP in complex with PLP (gray, PDB code
2CFT). The boxed area shows the overall structure and positioning of the ligands (green,
compound 3; gray, PLP).
Figure 5. Structure of compound 3 with adjacent residues of murine PDXP that might be
targeted to increase the specificity and affinity of the PDXP inhibitors described in this study.
16

Sites for possible modifications are indicated by color-coded dashed lines. For details, see 2.3.
Water molecules are omitted for clarity.
17

TABLES
Table 1. Steady-state kinetic constants of PDXP-catalyzed PLP hydrolysis in the presence of
compounds 1-4.
Inhibitor concentration [mM]
control
39 1.7
3,687 44
2.0 0.02
5.0 × 104
38 1.7
0.1
72 2.7
3,588 45
1.9 0.02
2.7 × 104
78 3.6
0.5
170 9.6
3,278 79
1.7 0.04
1.0 × 104
1.0
197 14.4
2,620 86 2,131 161
2.0
246 38.7
Km [µM]
vmax [nmol/min/mg]
Cmp
1
kcat [s-1]
1.4 0.05
1.1 0.09
kcat/Km [s-1·M-1]
Km [µM]
0.7 × 104
0.5 × 104
170 13.3
248 36.6
224 70.0
Cmp
vmax [nmol/min/mg]
3,839 49
2.0 0.03
5.4 × 104
49 3.0
3,577 55
1.9 0.03
2.4 × 104
59 8.1
2,986 99 2,301 163 1,644 238
kcat [s-1]
1.6 0.05
0.9 × 104
83 4.6
1.2 0.09
0.5 × 104
98 11.0
0.9 0.13
2
kcat/Km [s-1·M-1]
Km [µM]
0.4 × 104
136 10.9
vmax [nmol/min/mg]
3,843 70 3,739 159
3,680 71 3,440 140 3,148 101
Cmp
3
kcat [s-1]
2.0 0.04
4.2 × 104
40 2.7
2.0 0.08
3.3 × 104
44 5.6
2.0 0.04
2.4 × 104
77 10.7
1.8 0.07
1.7 0.05
kcat/Km [s-1·M-1]
Km [µM]
1.9 × 104
1.2 × 104
112 18.5
181 28.0
vmax [nmol/min/mg]
3,535 66 3,304 121 3,226 151 3,148 195 3,074 206
Cmp
4
kcat [s-1]
1.9 0.04
1.8 0.06
1.7 0.08
1.7 0.10
1.6 0.11
kcat/Km [s-1·M-1]
4.6 × 104
4.0 × 104
2.2 × 104
1.5 × 104
0.9 × 104
The data are mean values S.E. of three independent experiments performed with three
independently purified batches of proteins. Cmp, compound. Km, Michaelis-Menten constant;
vmax, maximum enzyme velocity; kcat, turnover number; kcat/Km, specificity constant. The kcat
values were calculated from the maximum enzyme velocities using a molecular mass of
31.828 kDa for PDXP. Control measurements were conducted in buffer without inhibitor.
18

Table 2. Data collection and refinement statistics of the PDXP·compound 3 structure.
Data collection
Wavelength (Å)
Space group
Refinement
1.5418
I23
Wilson B-factor (Ų)
Average B-factor (Ų)
macromolecules
solvent
67.0
50.8
50.9
Unit cell parameters
a = b = c (Å)
α = β = γ (°)
166.81
90
35.5
e
Rcryst
0.1895
(0.2628)
0.2466
(0.2902)
Resolution range (Å) a
Rfree
e
32.71−2.75
(2.90−2.75)
0.098 (0.952)
0.039 (0.380)
13.0 (1.9)
b
Rsym
Number of non H-atoms
macromolecules
ligands
4486
4403
40
c
Rp.i.m.
<I / σI> d
Completeness (%)
Multiplicity
Total reflections
Unique reflections
100 (100)
7.3 (7.2)
147756
20219 (2925)
water
43
Rms deviations in
bond lengths (Å)
bond angles (°)
planar groups (Å)
dihedral angles (°)
Coordinate error (Å) f
Ramachandran statistics g
favored (%)
0.006
0.932
0.004
14.18
0.34
98.44
0.87
0.69
8.02
allowed (%)
outliers (%)
MolProbity clashscore h
a Numbers in parentheses refer to the respective highest resolution data shell in the data set.
b Rsym = ΣhklΣi|Ii− < I > |/ΣhklΣi Ii, where Ii is the ith measurement, and < I > is the weighted
mean of all measurements of I.
c Rp.i.m. = Σhkl (1/(n-1))1/2 Σi|Ii - < I >|/ΣhklΣiIi , where n is the multiplicity of the observed
reflection.
d <I / σI>: Indicates the average of the intensity divided by its standard deviation.
e Rcryst = Σ|Fo-Fc|/Σ|Fo| where Fo and Fc are the observed and calculated structure factor
amplitudes. Rfree, same as Rcryst for 5% of the data randomly omitted from the refinement.
f Estimated coordinate error based on Rfree
.
g Ramachandran statistics indicate the fraction of residues in the favored, allowed and
disallowed regions of the Ramachandran diagram, as defined by MolProbity[35].
h number of serious clashes per 1000 atoms (Reference: see g).
19

ACKNOWLEDGEMENTS
This work was supported by the Deutsche Forschungsgemeinschaft (SFB688 to A.G., and
FZ82 to A.G. and H.S.). M.K. thanks the Deutsche Forschungsgemeinschaft for support
within the Emmy Noether program. G.K. was supported by a career development fellowship
from the Graduate School of Life Sciences at the University of Würzburg, and N.J. was
jointly supported by the SFB688 and EMBL.
20

REFERENCES
[1] R. Possemato, K.M. Marks, Y.D. Shaul, M.E. Pacold, D. Kim, K. Birsoy, S. Sethumadhavan, H.K.
Woo, H.G. Jang, A.K. Jha, W.W. Chen, F.G. Barrett, N. Stransky, Z.Y. Tsun, G.S. Cowley, J. Barretina,
N.Y. Kalaany, P.P. Hsu, K. Ottina, A.M. Chan, B. Yuan, L.A. Garraway, D.E. Root, M. Mino-Kenudson,
E.F. Brachtel, E.M. Driggers, D.M. Sabatini, Functional genomics reveal that the serine synthesis
pathway is essential in breast cancer, Nature, 476 (2011) 346-350.
[2] A. Seifried, J. Schultz, A. Gohla, Human HAD phosphatases: structure, mechanism, and roles in
health and disease, The FEBS journal, 280 (2013) 549-571.
[3] I.M. Wilson, E.A. Vucic, K.S. Enfield, K.L. Thu, Y.A. Zhang, R. Chari, W.W. Lockwood, N. Radulovich,
D.T. Starczynowski, J.P. Banath, M. Zhang, A. Pusic, M. Fuller, K.M. Lonergan, D. Rowbotham, J. Yee,
J.C. English, T.P. Buys, S.A. Selamat, I.A. Laird-Offringa, P. Liu, M. Anderson, M. You, M.S. Tsao, C.J.
Brown, K.L. Bennewith, C.E. MacAulay, A. Karsan, A.F. Gazdar, S. Lam, W.L. Lam, EYA4 is inactivated
biallelically at a high frequency in sporadic lung cancer and is associated with familial lung cancer risk,
Oncogene, 33 (2014) 4464-4473.
[4] K. Wu, Z. Li, S. Cai, L. Tian, K. Chen, J. Wang, J. Hu, Y. Sun, X. Li, A. Ertel, R.G. Pestell, EYA1
phosphatase function is essential to drive breast cancer cell proliferation through cyclin D1, Cancer
research, 73 (2013) 4488-4499.
[5] E. Tadjuidje, R.S. Hegde, The Eyes Absent proteins in development and disease, Cellular and
molecular life sciences : CMLS, 70 (2013) 1897-1913.
[6] M. Weinfeld, R.S. Mani, I. Abdou, R.D. Aceytuno, J.N. Glover, Tidying up loose ends: the role of
polynucleotide kinase/phosphatase in DNA strand break repair, Trends in biochemical sciences, 36
(2011) 262-271.
[7] G.K. Freschauf, F. Karimi-Busheri, A. Ulaczyk-Lesanko, T.R. Mereniuk, A. Ahrens, J.M. Koshy, A.
Rasouli-Nia, P. Pasarj, C.F. Holmes, F. Rininsland, D.G. Hall, M. Weinfeld, Identification of a small
molecule inhibitor of the human DNA repair enzyme polynucleotide kinase/phosphatase, Cancer
research, 69 (2009) 7739-7746.
[8] A.B. Krueger, S.J. Dehdashti, N. Southall, J.J. Marugan, M. Ferrer, X. Li, H.L. Ford, W. Zheng, R.
Zhao, Identification of a selective small-molecule inhibitor series targeting the eyes absent 2 (Eya2)
phosphatase activity, Journal of biomolecular screening, 18 (2013) 85-96.
[9] J.E. Hawkinson, M. Acosta-Burruel, P.L. Wood, The metabotropic glutamate receptor antagonist L-
2-amino-3-phosphonopropionic acid inhibits phosphoserine phosphatase, European journal of
pharmacology, 307 (1996) 219-225.
[10] E. Tadjuidje, T.S. Wang, R.N. Pandey, S. Sumanas, R.A. Lang, R.S. Hegde, The EYA tyrosine
phosphatase activity is pro-angiogenic and is inhibited by benzbromarone, PloS one, 7 (2012)
e34806.
[11] H. Park, S.K. Jung, K.R. Yu, J.H. Kim, Y.S. Kim, J.H. Ko, B.C. Park, S.J. Kim, Structure-based virtual
screening approach to the discovery of novel inhibitors of eyes absent 2 phosphatase with various
metal chelating moieties, Chemical biology & drug design, 78 (2011) 642-650.
[12] M. Zhang, E.J. Cho, G. Burstein, D. Siegel, Y. Zhang, Selective inactivation of a human neuronal
silencing phosphatase by a small molecule inhibitor, ACS chemical biology, 6 (2011) 511-519.
[13] K.N. Allen, D. Dunaway-Mariano, Markers of fitness in a successful enzyme superfamily, Current
opinion in structural biology, 19 (2009) 658-665.
[14] C. Kestler, G. Knobloch, I. Tessmer, E. Jeanclos, H. Schindelin, A. Gohla, Chronophin dimerization
is required for proper positioning of its substrate specificity loop, The Journal of biological chemistry,
289 (2014) 3094-3103.
[15] A. Gohla, J. Birkenfeld, G.M. Bokoch, Chronophin, a novel HAD-type serine protein phosphatase,
regulates cofilin-dependent actin dynamics, Nature cell biology, 7 (2005) 21-29.
[16] M.L. Fonda, Purification and characterization of vitamin B6-phosphate phosphatase from human
erythrocytes, The Journal of biological chemistry, 267 (1992) 15978-15983.
21

[17] R. Percudani, A. Peracchi, A genomic overview of pyridoxal-phosphate-dependent enzymes,
EMBO reports, 4 (2003) 850-854.
[18] A.C. Eliot, J.F. Kirsch, Pyridoxal phosphate enzymes: mechanistic, structural, and evolutionary
considerations, Annual review of biochemistry, 73 (2004) 383-415.
[19] H. Zuo, P.M. Ueland, S.J. Eussen, G.S. Tell, S.E. Vollset, O. Nygard, O. Midttun, K. Meyer, A. Ulvik,
Markers of vitamin B6 status and metabolism as predictors of incident cancer: The Hordaland Health
Study, International journal of cancer. Journal international du cancer, (2014).
[20] M. Johansson, C. Relton, P.M. Ueland, S.E. Vollset, O. Midttun, O. Nygard, N. Slimani, P. Boffetta,
M. Jenab, F. Clavel-Chapelon, M.C. Boutron-Ruault, G. Fagherazzi, R. Kaaks, S. Rohrmann, H. Boeing,
C. Weikert, H.B. Bueno-de-Mesquita, M.M. Ros, C.H. van Gils, P.H. Peeters, A. Agudo, A. Barricarte, C.
Navarro, L. Rodriguez, M.J. Sanchez, N. Larranaga, K.T. Khaw, N. Wareham, N.E. Allen, F. Crowe, V.
Gallo, T. Norat, V. Krogh, G. Masala, S. Panico, C. Sacerdote, R. Tumino, A. Trichopoulou, P. Lagiou, D.
Trichopoulos, T. Rasmuson, G. Hallmans, E. Riboli, P. Vineis, P. Brennan, Serum B vitamin levels and
risk of lung cancer, Jama, 303 (2010) 2377-2385.
[21] L. Galluzzi, I. Vitale, L. Senovilla, K.A. Olaussen, G. Pinna, T. Eisenberg, A. Goubar, I. Martins, J.
Michels, G. Kratassiouk, D. Carmona-Gutierrez, M. Scoazec, E. Vacchelli, F. Schlemmer, O. Kepp, S.
Shen, M. Tailler, M. Niso-Santano, E. Morselli, A. Criollo, S. Adjemian, M. Jemaa, K. Chaba, C.
Pailleret, M. Michaud, F. Pietrocola, N. Tajeddine, T. de La Motte Rouge, N. Araujo, N. Morozova, T.
Robert, H. Ripoche, F. Commo, B. Besse, P. Validire, P. Fouret, A. Robin, N. Dorvault, P. Girard, S.
Gouy, P. Pautier, N. Jagemann, A.C. Nickel, S. Marsili, C. Paccard, N. Servant, P. Hupe, C. Behrens, P.
Behnam-Motlagh, K. Kohno, I. Cremer, D. Damotte, M. Alifano, O. Midttun, P.M. Ueland, V. Lazar, P.
Dessen, H. Zischka, E. Chatelut, M. Castedo, F. Madeo, E. Barillot, J. Thomale, Wistuba, II, C. Sautes-
Fridman, L. Zitvogel, J.C. Soria, A. Harel-Bellan, G. Kroemer, Prognostic impact of vitamin B6
metabolism in lung cancer, Cell reports, 2 (2012) 257-269.
[22] L. Galluzzi, E. Vacchelli, J. Michels, P. Garcia, O. Kepp, L. Senovilla, I. Vitale, G. Kroemer, Effects of
vitamin B6 metabolism on oncogenesis, tumor progression and therapeutic responses, Oncogene, 32
(2013) 4995-5004.
[23] J. Hou, J. Aerts, B. den Hamer, W. van Ijcken, M. den Bakker, P. Riegman, C. van der Leest, P. van
der Spek, J.A. Foekens, H.C. Hoogsteden, F. Grosveld, S. Philipsen, Gene expression-based
classification of non-small cell lung carcinomas and survival prediction, PloS one, 5 (2010) e10312.
[24] T.L. Hullar, Pyridoxal phosphate. I. Phosphonic acid analogs of pyridoxal phosphate. Synthesis via
Wittig reactions and enzymic evaluation, Journal of medicinal chemistry, 12 (1969) 58-63.
[25] R.F. Struck, Y.F. Shealy, J.A. Montgomery, Vitamin B6 analogs. 4. 4-Desoxyisopyridoxal and the
phosphonic acid analog of 4-desoxypyridoxine phosphate, Journal of medicinal chemistry, 14 (1971)
568-571.
[26] R.J. Kubiak, K.S. Bruzik, Comprehensive and uniform synthesis of all naturally occurring
phosphorylated phosphatidylinositols, The Journal of organic chemistry, 68 (2003) 960-968.
[27] G.J. Gao, M.L. Fonda, Kinetic analysis and chemical modification of vitamin B6 phosphatase from
human erythrocytes, The Journal of biological chemistry, 269 (1994) 7163-7168.
[28] T.G. Battye, L. Kontogiannis, O. Johnson, H.R. Powell, A.G. Leslie, iMOSFLM: a new graphical
interface for diffraction-image processing with MOSFLM, Acta crystallographica. Section D, Biological
crystallography, 67 (2011) 271-281.
[29] P. Evans, Scaling and assessment of data quality, Acta crystallographica. Section D, Biological
crystallography, 62 (2006) 72-82.
[30] A.J. McCoy, R.W. Grosse-Kunstleve, P.D. Adams, M.D. Winn, L.C. Storoni, R.J. Read, Phaser
crystallographic software, Journal of applied crystallography, 40 (2007) 658-674.
[31] N.W. Moriarty, R.W. Grosse-Kunstleve, P.D. Adams, electronic Ligand Builder and Optimization
Workbench (eLBOW): a tool for ligand coordinate and restraint generation, Acta crystallographica.
Section D, Biological crystallography, 65 (2009) 1074-1080.
[32] P.D. Adams, P.V. Afonine, G. Bunkoczi, V.B. Chen, I.W. Davis, N. Echols, J.J. Headd, L.W. Hung,
G.J. Kapral, R.W. Grosse-Kunstleve, A.J. McCoy, N.W. Moriarty, R. Oeffner, R.J. Read, D.C. Richardson,
J.S. Richardson, T.C. Terwilliger, P.H. Zwart, PHENIX: a comprehensive Python-based system for
22

macromolecular structure solution, Acta crystallographica. Section D, Biological crystallography, 66
(2010) 213-221.
[33] P.D. Adams, P.V. Afonine, G. Bunkoczi, V.B. Chen, N. Echols, J.J. Headd, L.W. Hung, S. Jain, G.J.
Kapral, R.W. Grosse Kunstleve, A.J. McCoy, N.W. Moriarty, R.D. Oeffner, R.J. Read, D.C. Richardson,
J.S. Richardson, T.C. Terwilliger, P.H. Zwart, The Phenix software for automated determination of
macromolecular structures, Methods, 55 (2011) 94-106.
[34] P.V. Afonine, R.W. Grosse-Kunstleve, P.D. Adams, V.Y. Lunin, A. Urzhumtsev, On macromolecular
refinement at subatomic resolution with interatomic scatterers, Acta crystallographica. Section D,
Biological crystallography, 63 (2007) 1194-1197.
[35] V.B. Chen, W.B. Arendall, 3rd, J.J. Headd, D.A. Keedy, R.M. Immormino, G.J. Kapral, L.W. Murray,
J.S. Richardson, D.C. Richardson, MolProbity: all-atom structure validation for macromolecular
crystallography, Acta crystallographica. Section D, Biological crystallography, 66 (2010) 12-21.
23

24

25

26

27

28

29