Stereo- and regiospecificity of yeast phytases-chemical synthesis and enzymatic conversion of the substrate analogues neo- and L-chiro-inositol hexakisphosphate

Article abstract of DOI:10.1016/S0045-2068(02)00523-0

Phytases are enzymes that catalyze the hydrolysis of phosphate esters in myo-inositol hexakisphosphate (phytic acid). The precise routes of enzymatic dephosphorylation by phytases of the yeast strains Saccharomyces cerevisiae and Pichia rhodanensis have been investigated up to the myo-inositol trisphosphate level, including the absolute configuration of the intermediates. Stereoselective assignment of the myo-inositol pentakisphosphates (D-myo-inositol 1,2,4,5,6-pentakisphosphate and D-myo-inositol 1,2,3,4,5-pentakisphosphate) generated was accomplished by a new method based on enantiospecific enzymatic conversion and HPLC analysis. Via conduritol B or E derivatives the total syntheses of two epimers of myo-inositol hexakisphosphate, neo-inositol hexakisphosphate and L-chiro-inositol hexakisphosphate were performed to examine the specificity of the yeast phytases with these substrate analogues. A comparison of kinetic data and the degradation pathways determined gave the first hints about the molecular recognition of inositol hexakisphosphates by the enzymes. Exploitation of the high stereo- and regiospecificity observed in the dephosphorylation of neo- and L-chiro-inositol hexakisphosphate made it possible to establish enzyme-assisted steps for the synthesis of D-neo-inositol 1,2,5,6-tetrakisphosphate, L-chiro-inositol 1,2,3,5,6-pentakisphosphate and L-chiro-inositol 1,2,3,6-tetrakisphosphate.

Full text of DOI:10.1016/S0045-2068(02)00523-0

BIOORGANIC
CHEMISTRY
Bioorganic Chemistry 31 (2003) 44–67
www.elsevier.com/locate/bioorg
Stereo- and regiospecificity of yeast
phytases–chemical synthesis and
enzymatic conversion of the substrate analogues
neo- and L-chiro-inositol hexakisphosphate
Stephan Adelt,a Michael Podeschwa,b Guido Dallmann,a
b
Hans-Josef Altenbach, and Gunter Vogel
a,
*
€
a
Institut f€ur Biochemie, Fachbereich 9-Chemie, Bergische Universit€at Wuppertal,
Gaussstrasse 20, 42097Wuppertal, Germany
b
Institut f€ur Organische Chemie, Fachbereich 9-Chemie, Bergische Universit€at Wuppertal,
Gaussstrasse 20, 42097Wuppertal, Germany
Received 8 October 2002
Abstract
Phytases are enzymes that catalyze the hydrolysis of phosphate esters in myo-inositol hexa-
kisphosphate (phytic acid). The precise routes of enzymatic dephosphorylation by phytases of
the yeast strains Saccharomyces cerevisiae and Pichia rhodanensis have been investigated up to
the myo-inositol trisphosphate level, including the absolute configuration of the intermediates.
Stereoselective assignment of the myo-inositol pentakisphosphates (D-myo-inositol 1,2,4,5,6-
pentakisphosphate and -myo-inositol 1,2,3,4,5-pentakisphosphate) generated was accom-
D
plished by a new method based on enantiospecific enzymatic conversion and HPLC analysis.
Via conduritol B or E derivatives the total syntheses of two epimers of myo-inositol hexakis-
phosphate, neo-inositol hexakisphosphate and L-chiro-inositol hexakisphosphate were per-
formed to examine the specificity of the yeast phytases with these substrate analogues. A
comparison of kinetic data and the degradation pathways determined gave the first hints
about the molecular recognition of inositol hexakisphosphates by the enzymes. Exploitation
of the high stereo- and regiospecificity observed in the dephosphorylation of neo- and
L
-chiro-inositol hexakisphosphate made it possible to establish enzyme-assisted steps for the
* Corresponding author. Fax: +49-202-439-3399.
E-mail address: vogel@uni-wuppertal.de (G. Vogel).
0045-2068/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0045-2068(02)00523-0

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
45
synthesis of
phate and
Ó 2002 Elsevier Science (USA). All rights reserved.
D
-neo-inositol 1,2,5,6-tetrakisphosphate,
L-chiro-inositol 1,2,3,5,6-pentakisphos-
L
-chiro-inositol 1,2,3,6-tetrakisphosphate.
Keywords: Inositol stereoisomers; Inositol phosphates; neo-inositol phosphates;
phates; HPLC-analysis of inositol phosphates; Phytases; Yeast
L-chiro-inositol phos-
1. Introduction
In biological systems the hydrolysis of phosphate monoesters is a crucially
important process that is linked to energy metabolism, to metabolic regulation,
and to many signal-transduction pathways. Despite the apparent simplicity of
the task, the cleavage is accomplished by a diverse group of enzymes called phospha-
tases [1].
Among them a specific class of enzymes, the phytases (EC 3.1.3.8 and EC
3.1.3.26), has evolved which hydrolyze myo-inositol hexakisphosphate (phytic acid,
myo-InsP6) to less phosphorylated myo-inositol phosphates (in some cases to free
myo-inositol), releasing orthophosphate [2]. Most phytases belong to the high-
molecular-weight, acid-phosphatase subgroup. The reaction mechanism of these
enzymes involves a covalent base-stable, acid-labile phosphohistidine adduct, which
explains why the optimal pH for catalysis lies in the acidic range. Most phytases
sequenced so far share a highly conserved RHGXRXPmotif containing the active
center histidine [3].
Phytases are widespread in nature, occurring in microorganisms, plants, and some
animal tissues [2]. The role of this class of enzymes in these organisms seems to vary
and largely to depend on the function ascribed to the substrate myo-inositol hexakis-
phosphate. In plants, phytase is induced during germination to provide the growing
seedling simultaneously with orthophosphate and free myo-inositol, which is an im-
portant growth factor. The hydrolysis of phytate also releases bound cations, such as
Kþ, Mg2þ, Zn2þ, and Ca2þ. In microorganisms, phytase is most frequently induced
and sometimes secreted in response to phosphate starvation, and it releases phos-
phate from surrounding and/or internal myo-inositol hexakisphosphate. Besides
these two cases, in which myo-InsP6 serves merely as a storage molecule or nutrient,
there are hints that cellular myo-InsP6 and its breakdown products are involved in
signal-transduction pathways [4]. Two recent discoveries about myo-InsP6 focus on
potential functions in the nucleus: a requirement for myo-InsP6 in in vitro DNA re-
pair by non-homologous end-joining [5] and a relationship between myo-InsP6 levels
and nuclear mRNA export [6]. A phytase-like enzyme, multiple inositol polyphos-
phate phosphatase (MIPP), from mammalian tissue is capable of regulating the
cellular pools of myo-InsP6 and myo-Ins(1,3,4,5,6)P5 [7]. Furthermore, the MIPP-
generated downstream metabolites myo-Ins(1,3,4,5)P4 and myo-Ins(1,4,5)P3 are
themselves physiologically active in cellular Ca2þ homeostasis [8].
Despite all that is known about the biological importance of the ubiquitous myo-
inositol phosphates, there is little information available about the relevance of other

46
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
inositol stereoisomers and their derivatives. However, several publications have re-
ported the occurrence of low concentrations of scyllo-, chiro-, and neo-inositol in an-
imal tissue [9,10]. chiro-Inositol isolated from bovine-liver lipid was exclusively of the
L
-configuration [11]. The results of this study suggested that L-chiro-inositol was in-
corporated into the phosphatidylinositol pool. Mixtures of scyllo-, chiro-, and neo-
inositol are present in soil as their pentakisphosphate and hexakisphosphate esters
[12], but phosphate esters of these inositol stereoisomers have so far only rarely been
found in living nature. One exception seems to be the parasitic amoeba Entamoeba his-
tolytica, which possesses high levels of several neo-inositol polyphosphates [13]. The
question arises whether there exists a set of enzymes responsible for the metabolism
of the inositol isomers mentioned or their appearance is a consequence of a lack of
specificity in transformations catalyzed by the myo-inositol enzyme machinery.
In a first attempt to address this question, we have carried out the total synthesis
of two isomeric forms of myo-InsP6, neo-InsP6 (3), and -chiro-InsP6 (())-5) and in-
L
vestigated their phytase-catalyzed dephosphorylation with whole cells from the yeast
strain Saccharomyces cerevisiae and a secreted enzyme activity from Pichia rhodan-
ensis. A comparison with the conversion of myo-InsP6 revealed structural motifs im-
portant for molecular recognition. Taking advantage of the unexpected, well-defined
degradation routes observed for neo-InsP6 (3) and
L
-chiro-InsP6 (())-5), we estab-
-neo-Ins(1,2,5,6)P4 (10), -chiro-
L-chiro-Ins(1,2,3,6)P4 (12).
lished enzyme-assisted steps for the synthesis of
D
L
Ins(1,2,3,5,6)P5 (11) and
2. Materials and methods
2.1. General
The following myo-inositol phosphates needed as reference compounds for
HPLC-MDD (metal dye detection) were synthesized in our laboratories. For the
synthesis of myo-Ins(1,2,4,5)P4 and myo-Ins(1,2,4)P3 a chemoenzymatic approach
was used [14]. The isomers myo-Ins(1,2,5,6)P4, myo-Ins(1,2,4,6)P4, and myo–
Ins(1,2,3,4)P4 were prepared as decribed previously [15,16]. The meso-compound
myo-Ins(2,4,5,6)P4 was prepared by acid-catalyzed, non-specific dephosphorylation
of myo-InsP6 and purified by HPLC. The inositol pentakisphosphates of defined ste-
reochemistry (myo-Ins(1,2,4,5,6)P5, myo-Ins(2,3,4,5,6)P5, myo-Ins(1,2,3,4,5)P5, and
myo-Ins(1,2,3,5,6)P5) were synthesized (manuscript in preparation). Commercially
available compounds were from Sigma (myo-InsP6, myo-Ins(SO4Þ ), Alexis (myo-
6
Ins(1,2,3,5)P4), and Calbiochem (myo-Ins(2,4,5)P3). All other chemicals used were
of the highest purity available.
1
All NMR spectra were recorded on a Bruker ARX 400 spectrometer. Besides H
(400 MHz), 13C (101 MHz) and 31P(162 MHz) experiments, 2D COSY-( 1H–1H, 1H–
13C as well as 1H–31P), and DEPT spectra for the unequivocal correlation of the hy-
drogen, carbon and phosphorus atoms were recorded. The chemical shifts are given
in ppm, relative to the solvents as internal standard. The multiplicity is given by the
following symbols: s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), Wt

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
47
(pseudotriplet for unresolved dd), and br (broad). The coupling constant J is given in
Hz. Melting points are uncorrected. IR spectra were obtained in KBr, and only note-
worthy absorptions (cmÀ1) are listed. Flash-column chromatography was performed
on Merck silica gel (40–63 lm). All organic extracts were dried over MgSO4, filtered
and concentrated with a rotary evaporator under reduced pressure. Only distilled
solvents were used. The building blocks (+)-1 and ())-1 and their precursors were
synthesized as described in a previously published method [16,17].
2.2. Cultivation of yeast strains
Saccharomyces cerevisae was obtained from a local grocery store (product of:
Deutsche Hefewerke, Nuremberg, Germany) and P. rhodanensis (CBS 5518) from
Centraalbureau voor Schimmelcultures (Utrecht, The Netherlands). Yeast strains
were grown in YPD medium consisting of 1% yeast extract (Oxoid, England),
2% peptone (Oxoid, England), and 2% glucose. The solution was autoclaved for
16 min at 121 °C before the addition of the glucose. The glucose stock solution
(20%) was autoclaved separately or filtered for sterility and added afterwards. S.
cerevisae was cultured at 30 °C, whereas P. rhodanensis was incubated at 25 °C,
both on a rotary shaker (150 rpm). The growth of the yeast cells was monitored
as optical density (OD) of the culture broth at 600 nm (S. cerevisiae OD
1 ꢀ 2:2 Â 108 cells=mL).
2.3. Precipitation of an extracellular phytase from P. rhodanensis
The cells were grown overnight and harvested at an OD600 of 20–25. The cells
were removed by centrifugation (4000g, 15 min, 4 °C), and the enzyme was precipi-
tated from the supernatant with 66% (v/v) ethanol precooled to )20 °C. The precip-
itation was carried out in an ice/NaCl bath under stirring, and ethanol was added
continuously. To preserve enzyme activity, freezing of the solution must be pre-
vented. When the addition of ethanol is complete, the mixture should be stirred fur-
ther for 15 min. The protein is sedimented by centrifugation (6000g, 15 min, 0 °C),
resuspended in 50 mM Tris/HCl, pH 7.7, aliquoted and frozen ()20 °C). Prepared
by this procedure, the enzyme maintains its activity for several months. A 50-mL cul-
ture of P. rhodanensis yielded 150–200 mU phytase activity with recoveries of more
than 80%.
2.4. P. rhodanensis phytase assay
The incubation mixtures contained 100 lL assay buffer (125 mM sodium acetate
(pH 4.6), 1.25 mM myo-InsP6) and 25 lL enzyme solution. The assays were incubated
at 56 °C, and at varying intervals 10-lL aliquots were withdrawn. The reaction was
stopped by the addition of 10 lL 2 N HCl to the aliquots, and the enzymatically
released inorganic phosphate was determined. In case of a high phosphate
background, the samples were analyzed with HPLC-MDD after neutralization.

48
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
Enzyme activities were calculated from the linear regions of kinetic data plots.
Protein was measured by an assay supplied by BIO-RAD (catalog no. 500-0112)
with bovine serum albumin as a standard.
2.5. Kinetic measurements with myo-InsP6, neo-InsP6 (3), and -chiro-InsP6 (())-5)
L
Measurements were performed by HPLC-MDD quantification of inositol hexa-
kisphosphate remaining after enzymatic conversion. S. cerevisiae cells were harvested
and washed twice with acetate buffer (3000g, 10 min., 4 °C). The cells (20–60 mg)
were resuspended in 700 lL of 75 mM sodium acetate (pH 4.6) containing 1 mM
of the corresponding substrate. In the case of P. rhodanensis, ethanol-precipitated
phytase (20 mU/mL, released Pi from 1 mM myo-InsP6, 56 °C) was used instead of
yeast cells. For inhibition studies the reaction mixtures also contained up to
100 lM inositol hexakissulfate. The mixture was incubated in a 24-well tissue-culture
plate at 45 °C in a water-bath shaker. To allow temperature equilibration, the plate
was preincubated for two minutes before the analysis was started. 50-lL aliquots
were taken, S. cerevisiae cells were removed by centrifugation (15,000g, 1 min,
4 °C) and the supernatants were frozen (liquid nitrogen). In the experiments with
the phytase of P. rhodanensis the samples were frozen immediately after removal.
2.6. Inorganic phosphate determination
A colorimetric assay was employed for the determination of nanomolar amounts
of inorganic phosphate [18]. Reducing the volume to a microplate scale made it pos-
sible to handle a great number of samples simultaneously. The method was applica-
ble to following the progress of the enzymatic phosphatase reactions and to
quantifying the mass of purified inositol phosphates. In the former case, the acidic
solution (10–50 lL) was mixed with 100 lL of malachite green reagent, and after
1 min 10 lL of 34% (w/v) sodium citrateÁ2H2O was added. For full color develop-
ment the mixture was incubated for 20 min at room temperature. Absorbance was
measured at 595 nm. The assay was linear in the range 0.3–3 nmol of phosphate.
Covalently bound organic phosphate had to be liberated by treatment with hot
sulfuric acid. In brief, a 10-lL aliquot of an inositol phosphate solution (approx.
1–2 mM) was pipetted into 50 lL of 10 N H2SO4, mixed and heated for 5 h at
170 °C (borosilicate glass tubes of hydrolysis class 1 are required; Macherey-Nagel).
The residue was diluted 100-fold with 1 N H2SO4, and 50-lL portions were assayed
for inorganic phosphate. Mass determinations for each substance were done in trip-
licate.
2.7. Purification and analysis of inositol phosphates (HPLC-MDD)
Inositol phosphates were purified and analyzed with the HPLC-MDD method de-
scribed previously [16,19]. The compounds were separated by anion-exchange chro-
matography on a MonoQ HR10/10 column (Pharmacia). To purify inositol
phosphates and identify enzymatic degradation products, a linear gradient of HCl

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
49
was applied (0 min 0.2 mM HCl; 70 min 0.5 M HCl; flow rate 1.5 mL/min). For the
kinetic evaluation of the enzymatic reactions with myo-InsP6, neo-InsP6 (3), and
L
-
chiro-InsP6 (())-5), a shorter linear acidic gradient was used (0 min 0.375 M HCl;
15 min 0.5 M HCl; flow rate 1.5 mL/min). In analytical runs photometric detection
at 546 nm was achieved with a metal-dye reagent (2 M Tris/HCl (pH 9.1), 200 lM
4-(2-pyridylazo)resorcinol (PAR), 30 lM YCl3, 10% (v/v) MeOH; flow rate
0.75 mL/min). In purification steps, where on-line detection was impossible, an anal-
ogous experiment was carried out on a microplate. In brief, a 0.2–2-lL portion of the
collected fraction was mixed with 100 lL of metal-dye reagent, and the absorbance
at 540 nm was measured.
To separate the products of degradation of myo-Ins(1,2,3,4,5)P5/myo-
Ins(1,2,3,5,6)P5 by the phytase of Dictyostelium discoideum, a modified, slightly alka-
line elution system was established. The gradient composed of 50 mM Tris/HCl (pH
8.5; solvent A) and 50 mM Tris/HCl, 0.4 M KCl (pH 8.5; solvent B) had the follow-
ing characteristics: 0 min, 30% B; 2 min, 40% B; 16 min, 42% B; 20 min, 50% B;
38 min, 60% B; 68 min, 75% B; 73 min, 100% B; 80 min, 100% B (flow rate 1.5 mL/
min). The metal-dye reagent consisted of 2 mM NH4OAc/AcOH (pH 5.0), 200 lM
PAR, 30 lM YCl3, and 10% (v/v) MeOH (flow rate 0.75 mL/min).
2.8. Stereochemical assignment of myo-inositol pentakisphosphates by means of a
phytase from D. discoideum
A Triton X-100 solubilized membrane-associated phytase from D. discoideum,
partially purified by sequential biochromatography (Q-Sepharose ff and Source
15Q, both from Amersham-Pharmacia) was used as the enzymatic tool [14].
A microplate assay was established that provided kinetic data. The incubation
mixture in one cavity contained 10 lL assay buffer (200 mM Mes/Naþ (pH 5.15),
0.2% (w/v) Triton X-100 hydr.), 10 lL water, 10 lL phytase (45 mU/mL, Pi-release
from 1 mM myo-InsP6) from D. discoideum and 10 lL of a 400-lM myo-inositol
phosphate solution. The samples were incubated at room temperature on a recipro-
cal shaker (80 rpm). The enzymatic reaction was terminated at varying intervals by
the addition of 10 lL of 2 N HCl, and the inorganic phosphate released was ana-
lyzed. For the stereochemical assignment of the myo-inositol pentakisphosphates
the assay was scaled up to a test volume of 120 lL. The reaction was stopped with
30 lL 2 N HCl when 24% Pi of the total organic phosphate had been released. After
neutralization, the samples were analyzed by HPLC-MDD with an acidic (myo-
Ins(1,2,4,5,6)P5 and myo-Ins(2,3,4,5,6)P5) or an alkaline (myo-Ins(1,2,3,4,5)P5 and
myo-Ins(1,2,3,5,6)P5) elution system.
2,3,4,5-Tetra-O-acetyl-neo-inositol,(+)-2. (1S,2R,3R,4S)-1,4-diacetoxy-2,3-dibromo-
cyclo-hex-5-ene (10 g, 28 mmol) (+)-1 was added to a hot solution of 10 g potas-
sium acetate or sodium acetate in 100 mL acetic acid and 5 mL water.The
mixture was stirred for 10 days at 130 °C. The solvent was removed under reduced
pressure, and the residue was dried under high vacuum. The residue was suspended
in anhydrous dichloromethane, and acetic anhydride (30 mL) and DMAP(200 mg)
were added. The mixture was stirred overnight. For work-up the mixture was

50
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
added slowly to a vigorously stirred aqueous solution of NaHCO3 (16 g, 0.19 mol,
100 mL water). The aqueous phase was extracted several times with dichlorome-
thane. The combined organic phase was washed with saturated aqueous
NaHCO3 and then with brine. After evaporation of the solvent the raw product
(10 g, oil) was purified by flash chromatogaphy (cyclohexane/EtOAc 3:2) to yield
(+)-2a (6.5 g, 70%) as a colorless oil which crystallized after one day. Rf 0.35 (cy-
20
20
clohexane/EtOAc 3:2); ½aŠ ¼ þ243° (c ¼ 1:27, CHCl3); [20] ½aŠ ¼ þ244° (c ¼ 1:6,
D
D
CHCl3). Racemic 2a was purified without flash chromatogaphy simply by recrys-
tallization from EtOH to yield pure 2a (6 g, 67%) as a colorless solid. The analyt-
ical data is presented elsewhere [21].
A solution of trifluoroperacetic acid was prepared by adding 1.4 mL H2O2 (85%,
42 mmol) slowly to an ice-cooled solution of 7 mL trifluoroacetic anhydride (7 mL,
50 mmol) in 20 mL dichloromethane. The solution was stirred for 1 h at RT. This so-
lution, freshly prepared, was added dropwise at 0 °C over 15 min to a suspension of
tetra-O-acetyl-conduritol-E (+)-2a (1.2 g, 3.8 mmol) and NaHCO3 (1.4 g, 16 mmol)
in 35 mL dichloromethane. The mixture was refluxed for 2 h and stirred for 12 h at
room temperature. The reaction mixture was added slowly to a cooled and stirred
mixture of 30 mL saturated aqueous NaHCO3, 8.4 g NaHCO3, 100 mL brine, and
200 mL dichloromethane. After addition of aqueous thiosulfate solution (50 mL,
20%) the mixture was extracted four times with dichloromethane or EtOAc. The
combined organic layer was washed with brine and then dried. Evaporation of the
solvent gave a yellow foam (1.2 g). The raw product was purified by fractionated sil-
ica-gel filtration (first cyclohexane/EtOAc 8:2 until educt and epoxide could not be
detected by TLC, then EtOAc). The EtOAc fraction was evaporated to yield (+)-2
(900 mg, 2.7 mmol, 71%) as a colorless foam. Rf 0.03 (cyclohexane/EtOAc 1:1);
20
1
½aŠ ¼ þ12:7° (c ¼ 1:1, CHCl3); H-NMR (CDCl3): d 1.99 (s, 6 H, 2 CH3), 2.13
D
(s, 6 H, 2 CH3), 3.27 (br, 2 H, OH), 4.01 (s, 2 H, H-1, H-6), 5.20 (s, 2 H, H-3, H-
4), 5.62 (s, 2 H, H-2, H-5); 13C-NMR (CDCl3): d 20.58 (2 CH3), 20.70 (2 CH3),
68.33 (2 CH), 68.57 (2 CH), 70.19 (2 CH), 170.17 (2 C@O), 170.66 (2 C@O); MS
(EI, 70 eV, m/z (%)): 271 (2), 228 (1), 210 (5), 168 (16), 157 (9), 126 (27), 115(16),
60 (18), 43 (100); calcd. for C14H20O10: C, 48.28; H, 5.79. Found: C, 48.18; H, 6.00.
neo-Inositol hexakisphosphate 3. 1.9 g (5.5 mmol) Tetra-O-acetyl-neo-inositol (+)-2
was suspended in anhydrous methanol (50 mL) under argon and cooled to 4 °C.
500 lL (2.25 mmol) of a 5.5 M sodium methanolate solution was added dropwise
over 10 min. The solution was allowed to warm to room temperature and stirred
for 12 h. The solvent was removed, and for complete turnover the residue was dis-
solved in 20 mL of 0.25 M aqueous NaOH and refluxed for 2 h. The hot solution
was neutralized by addition of ion exchanger (Hþ, Dowex 50-X) and filtered, and
the Dowex material was washed with boiling water. The filtrate was first concen-
trated in high vacuum and then lyophilized. Recrystallization from water (50 mL)
yielded free neo-inositol (1 g, 99%) as a colorless solid. The analytical data is pre-
sented elsewhere [22].
514 mg (2.26 mmol) 3-Diethylamino-2,4,3-benzodioxaphosphepane was added
to a suspension of 50 mg (0.27 mmol) neo-inositol thus obtained and 315 mg
(4.5 mmol) 1H-tetrazole in 20 mL anhydrous dichloromethane/acetonitrile (1:1);

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
51
the solution was stirred at 40 °C for three days. For work-up the solution was
cooled to )40 °C, and an anhydrous solution of 1.5 g (6 mmol) m-CPBA (70%)
in dichloromethane (15 mL) was added. The solution was allowed to warm to
room temperature, and the stirring was continued for another hour. The reaction
mixture was diluted with dichloromethane (150 mL) and washed consecutively with
aqueous sodium thiosulfate (20%, 2 Â 100 mL), saturated NaHCO3 (3 Â 100 mL)
and brine. After evaporation of the solvent the resulting colorless foam was puri-
fied by flash chromatography on silica gel (Rf 0.22 (CH2Cl2/MeOH 95:5)) to yield
hexa-O-(2-oxo-5,6-benzo-1,3,2-dioxaphosphep-2-yl)-neo-inositol (75 mg, 21%) as a
colorless foam. 1H-NMR (CDCl3): d 5.15 (m, 4 H, H-1, H-3, H-4, H-6), 5.58
(m, 2 H, H-2, H-5), 5.00–5.82 (m, 24 H, 6 (CH2Þ C6H4), 7.26–7.43 (m, 24 H,
2
Ph–H); 13C-NMR (CDCl3): d 69.19–69.63 (m, 6 (CH2Þ C6H4), 72.86 (m, 4 C, C-
2
1, C-3, C-4, C-6), 77.06 (2 C, C-2, C-5), 128.94–129.48 (Carom:), 134.54, 135.63,
135.77 (Cipso); 31P-NMR{1H} (CDCl3): d )1.48 (2 P, P-2, P-5), )2.39 (4 P, P-1,
P-3, P-4, P-6).
To a suspension of 50 mg (39 lmol) hexa-O-(2-oxo-5,6-benzo-1,3,2-dioxaphosp-
hep-2-yl)-neo-inositol in 30 mL ethanol/water (1:2) was added Pd/C (50 mg, Degussa
RW 10). The mixture was stirred at room temperature under H2 overnight. The cat-
alyst was filtered off, and the filtrate was concentrated in high vacuum and then ly-
ophylized to give 24 mg (99%) of 3 as a colorless, very hygroscopic foam.
Purification by HPLC assured purity >99% for the following reactions. 1H-
NMR (D2O, pH adjusted to 6 (ND4OD)): d 4.37 (d, J ¼ 4:6Hz, 4 H, H-1, H-3,
H-4, H-6), 4.89 (d, J ¼ 9:2, 2 H, H-2, H-5); 13C-NMR (D2O, pH adjusted to 6
(ND4OD)): d 74.24 (4 C, C-1, C-3, C-4, C-6), 76.57 (2 C, C-2, C-5). 31P-NMR{1H}
(D2O, pH adjusted to 6 (ND4OD)): d 1.89 (2 P, P-2, P-5þ), 2.07 (4 P, P-1, P-3, P-4,
P-6); HR-MS(MALDI-FTMS): m/z: 658.8541 ½M À HŠ calcd. for C6H17O24P6:
658.8527.
Alternative preparation: 70 mg (0.38 mmol) of neo-inositol was dissolved in
0.22 mL phosphoric acid at 60 °C. Polyphosphoric acid (4 mL) was added, and the
mixture was heated to 150 °C under vacuum for 12 h. The mixture was diluted with
water and purified by an ion exchanger (ClÀ-form, DEAE sepharose) and eluted
with a linear gradient of hydrochloric acid (0–1 N HCl). Purification by HPLC
yielded 80 mg (31%) of neo-InsP6 (3).
2,3,4,5-Tetra-O-acetyl-1-O-benzyl-
D
D
-neo-inositol ())-6. A solution of 350 mg
(1 mmol) tetra-O-acetyl- -neo-inositol (+)-2 in 8.5 mL of anhydrous dichlorome-
thane/cyclohexane (2:1) was cooled to 0 °C. 2,2,2-trichloro-acetamidic acid benzyl
ester (550 mg, 2.2 mmol) and trifluoromethanesulfonic acid (33 mg, 0.22 mmol) were
added slowly; the reaction mixture was allowed to warm to room temperature and
was stirred for another 5 h. The solvent was evaporated, and the residue was sus-
pended in 1.5 mL hexane/dichloromethane (2:1). The solid was filtered off, and the
filtrate was concentrated under reduced pressure. The crude product was purified
by flash chromatography to yield ())-6 (100 mg, 0.22 mmol, 22%, Rf 0.21 (cyclohex-
ane/EtOAc 1:1)) as a colorless solid. The dibenzyl product was also isolated, but was
20
D
still strongly contaminated with trichloroacetamide. ½aŠ ¼ À8:4° (c ¼ 0:54, CHCl3);
1H-NMR (CDCl3): d 2.01 (s, CH3), 2.03 (s, CH3), 2.13 (s, CH3), 2.14 (s, CH3), 3.81

52
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
(dd, J ¼ 3:1Hz, J ¼ 10:2Hz, 1 H, H-1), 4.10 (dd, J ¼ 2:8Hz, J ¼ 9:9Hz, 1 H, H-6),
4.46, 4.74 (AB, J ¼ 10:7Hz, 2 H, Ph–CH2), 5.26 (Ws, 2 H, H-3, H-4), 5.69 (Wt,
J ¼ 2:3Hz, 1 H, H-5), 5.85 (Wt, J ¼ 2:3Hz, 1 H, H-2), 7.24–7.46 (m, 5 H, Ph–H);
13C-NMR (CDCl3): d 20.55 (2 CH3), 20.66 (CH3), 20.71 (CH3), 66.55 (C-2), 67.58
(C-6), 68.17 (C-3 or C-4), 68.27 (C-3 or C-4), 69.57 (C-5), 72.06 (Ph–CH2), 76.15
(C-1), 128.10, 128.52 (Carom:), 136.98 (Cipso), 169.93, 170.02, 170.06, 170.14 (C@O).
1-O-Benzyl- -neo-inositol ())-7. 55 mg (0.13 mmol) ())-6 was suspended in 10 mL
D
anhydrous methanol and cooled to 4 °C. A sodium methanolate solution (50 lL, 5.5
M, 0.28 mmol) was added; the solution was allowed to warm to room temperature
and was stirred for 3 h. The solution was neutralized by addition of ion exchanger
(Hþ, Dowex 50-X) and filtered, and the residue washed with methanol. The filtrate
was concentrated in high vacuum to yield ())-7 (35 mg, 0.13 mmol, 100%) as a col-
20
D
1
orless solid. ½aŠ ¼ À14:0° (c ¼ 0:35, DMSO); H-NMR (d6-DMSO): d 3.37 (dd,
J ¼ 3:1Hz, J ¼ 9:7Hz, 1 H, H-1), 3.40 (dd, J ¼ 2:8Hz, J ¼ 9:9Hz, 1 H, CH),
3.46 (dd, J ¼ 2:8Hz, J ¼ 9:9Hz, 1 H, CH), 3.66 (under HDO, 2 CH), 3.91 (Wt,
J ¼ 2:8Hz, 1 H, H-2), 4.51, 4.59 (AB, J ¼ 12:2Hz, 2 H, Ph–CH2), 7.24–7.46 (m, 5
H, Ph–H); 13C-NMR (d6-DMSO): d 68.98 (C-2), 69.40 (CH), 69.95 (CH), 70.08
(CH), 72.36 (CH), 70.98 (Ph–CH2), 78.49 (C-1), 127.42, 127.82, 128.31 (Carom:),
139.52 (Cipso).
D
-neo-Inositol 2,3,4,5,6-pentakisphosphate (+)-8. 3-Diethylamino-2,4,3,-benzodi-
oxaphosphepane (0.2 g, 0.9 mmol) was added to a solution of 1-O-benzyl-D-neo-
inositol ())-7 (30 mg, 0.11 mmol) and 1H-tetrazole (140 mg, 2 mmol) in anhydrous
dichloromethane (20 mL) under argon, and the mixture was stirred at room tem-
perature for 4 h. For work-up the solution was cooled to )40 °C, and an anhy-
drous solution of m-CPBA (1 g, 4 mmol, 70%) in dichloromethane (15 mL) was
added. The solution was allowed to warm to room temperature, and the stirring
was continued for another hour. The reaction mixture was diluted with dichlo-
romethane (100 mL) and washed consecutively with aqueous sodium thiosulfate
(20%, 2 Â 50mL), saturated NaHCO3 (3 Â 50mL) and brine. After evaporation
of the solvent the resulting colorless foam was converted without further purifica-
tion.
To this raw product in 40 mL ethanol/water (1:1) 50 mg Pd/C (Degussa RW 10)
was added. The mixture was stirred at room temperature under H2 for 12 h. The cat-
alyst was filtered off, and the filtrate was concentrated in high vacuum and then ly-
ophilized. Purification by HPLC yielded 30 mg (50%, purity >99%) of 8 as a
20
D
1
colorless, very hygroscopic foam. ½aŠ ¼ þ8:6° (c ¼ 0:51, H2O); H-NMR (D2O,
pH adjusted to 6 (ND4OD)): d 3.98 (dd, J ¼ 2:5Hz, J ¼ 10:2Hz, 1 H, H-1), 4.30
(dWt, J ¼ 2:8Hz, J ¼ 10:2Hz, 1 H, H-6), 4.36 (m, 2 H, H-3, H-4), 4.76 (d,
J ¼ 9:2Hz, 1 H, H-2), 4.89 (d, J ¼ 10:2Hz, 1 H, H-5); 13C-NMR (D2O, pH adjusted
to 6 (ND4OD)): d 70.69 (m, C-1), 74.50 (C-3 and C-4), 75.60 (d, J ¼ 5:1Hz, C-6),
76.56 (d, J ¼ 4:1Hz, C-5), 77.23 (d, J ¼ 5:1Hz, C-2); 31P-NMR{1H} (D2O, pH ad-
justed to 6 (ND4OD)): d 1.90 (2 P, P-5, P-3 or P-4), 2.08 (2 P, P-6, P-3 or P-4), 3.08
(P-2).
D
-neo-Inositol 1,2,5,6-tetrakisphosphate ())-10. S. cerevisiae was grown for 24 h
and harvested at a OD600 of about 15. The cells were washed twice with 75 mM

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
53
sodium acetate pH 4.6 (3000g, 10 min, 4 °C), and the cellular wet weight was deter-
mined (25 mL culture gives approximately 500 mg). The cells were resuspended in
75 mM sodium acetate (pH 4.6) containing 1 mM neo-InsP6 (3) to a final concentra-
tion of 30 mg/mL cells. The reaction was carried out at 45 °C in a water-bath shaker
and followed by HPLC-MDD. After 80–90 min the reaction was stopped by remov-
ing the cells (6000g, 10 min, 4 °C). The sediment was washed with doubly distilled
water, and the collected supernatants were combined and filtered for sterility. The
product was purified by HPLC. After freeze-drying, overall yields were in the range
of 75–85% for ())-10 (quantification based on inorganic phosphate determination).
20
D
½aŠ ¼ À12:8° (c ¼ 0:22, H2O); 1H-NMR (D2O, pH adjusted to 6 (ND4OD)): d 3.86
(s, 2 H, H-3, H-4), 4.37 (d, J ¼ 6:6Hz, 2 H, H-1, H-6), 4.74 (d, J ¼ 9:2Hz, 2 H, H-2,
H-5); 13C-NMR (D2O, pH adjusted to 6 (ND4OD)): d 71.89 (C-3 and C-4), 74.73 (C-
1 and C-6), 77.50 (C-2 and C-5); 31P-NMR{1H} (D2O, pH adjusted to 6 (ND4OD)):
d 1.78 (P-1 and P-6), 3.08 (P-2 and P-5).
L
-chiro-Inositol hexakisphosphate ())-5. Pd/C (40 mg) was added to a suspension
of 250 mg (0.7 mmol) 2,5-di-O-benzyl- -chiro-inositol ())-4 in 20 mL ethanol/water
L
(1:1). The mixture was stirred overnight at room temperature under H2. The catalyst
was filtered off, and the filtrate was concentrated in high vacuum and lyophilized to
give a voluminous foam (120 mg, 99%). Crystallization from EtOH yielded pure L-
20
chiro-inositol (125 mg, 99%) as a colorless solid. ½aŠ ¼ À70° (c ¼ 0:55, H2O) agrees
D
with [23]; further analytical data is presented elsewhere (22).
470 mg (2.0 mmol) 3-Diethylamino-2,4,3-benzodioxaphosphepane was added to a
suspension of 50 mg (0.27 mmol) L-chiro-inositol obtained as above and 226 mg
(3.2 mmol) 1H-tetrazole in 20 mL anhydrous dichloromethane; the solution was stir-
red at room temperature overnight. The product was worked up as described for 3.
Purification by flash chromatography on silica gel yielded hexa-O-(2-oxo-5,6-benzo-
1,3,2-dioxaphosphep-2-yl)-L-chiro-inositol (200 mg, 60%) as a colorless foam with a
purity >80%. 13C-NMR (CDCl3): d 69.03–69.74 (m, 6 ðCH2Þ C6H4), 73.51 (m, 2
2
CH), 73.81 (m, 2 CH), 76.28 (m, 2 CH), 128.4–129.4 (Carom:), 134.41, 134.51,
135.33, 135.47, 135.52, 136.88 (Cipso); 31P-NMR{1H} (CDCl3): d )1.30 (2 P),
)2.05 (2 P), )2.74 (2 P).
Hydrogenolysis of hexa-O-(2-oxo-5,6-benzo-1,3,2-dioxaphosphep-2-yl)-L-chiro-
inositol was carried out as described for 3. Purification by HPLC assured a purity
20
>99%. ½aŠ ¼ À24:4° (c ¼ 1:17, H2O); 1H-NMR (D2O, pH adjusted to 6
D
(ND4OD)): d 4.30 (AA0 , 2 H, H-3, H-4), 4.36 (MM0 , 2 H, H-2, H-5), 4.65 (under
HDO, XX0 , 2 H, H-1, H-6); 13C-NMR (D2O, pH adjusted to 6 (ND4OD)):d 72.9
(m, C-1 and C-6), 73.3 (m, C-2 and C-5), 76.5 (m, C-3 and C-4); 31P-NMR{1H}
(D2O, pH adjusted to 6 (ND4OD)): d 1.63 (P-1 and P-6), 1.91 (P-2 and P-5), 2.66
(P-3 and P-4).
L
-chiro-Inositol 1,2,3,5,6-pentakisphosphate (11),
L-chiro-Inositol 1,2,3,6-tetrakis-
phosphate (12). The dephosphorylation of -chiro-InsP6 (())-5) catalyzed by the phy-
L
tase from P. rhodanensis was carried out at 45 °C in 75 mM sodium acetate (pH 4.6)
with an initial substrate concentration of 1 mM and an enzyme activity of 8.5
mU/mL in a water-bath shaker. The reaction was monitored by HPLC-MDD and
stopped after 80–90 min, when the contents of
L-chiro-Ins(1,2,3,5,6)P5 (11) and

54
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
L
-chiro-Ins(1,2,3,6)P4 (12) were approximately equal. The products were isolated by
HPLC. After freeze-drying, overall yields were in the range of 30–40% each for 11
and 12.
20
D
1
11: ½aŠ ¼ À26° (c ¼ 0:5, H2O); H-NMR (D2O, pH adjusted to 6 (ND4OD)): d
3.86 (Wt, J ¼ 9:2Hz, 1 H, H-4), 4.25 (Wq, J ¼ 9:2Hz, 1 H, H-3), 4.33 (dWt,
J ¼ 9:7Hz, J ¼ 3:1Hz, 1 H, H-5), 4.39 (dWt, J ¼ 9:7Hz, J ¼ 2:5Hz, 1 H, H-2),
4.65 (under HDO, 2 H, H-1, H-6); 13C-NMR (D2O, pH adjusted to 6 (ND4OD)):
d 73.72 (dd, J ¼ 6:1Hz, J ¼ 2:0Hz, CH), 74.52 (d, J ¼ 4:1Hz, CH), 75.20 (m,
CH), 75.43 (d, J ¼ 4:1Hz, CH), 76.04 (Wt, J ¼ 5:1Hz, CH), 79.21 (Wt,
J ¼ 5:6Hz, CH); 31P-NMR{1H} (D2O, pH adjusted to 6 (ND4OD)): d 1.17 (P-1),
1.48 (P-5 and P-6), 2.09 (P-2), 2.48 (P-3).
20
D
12: ½aŠ ¼ À19:8° (c ¼ 0:63, H2O); 1H-NMR (D2O, pH adjusted to 6
(ND4OD)): d 3.82 (Wt, J ¼ 9:9Hz, 1 H, H-4), 3.91 (dd, J ¼ 9:9Hz, J ¼ 2:3Hz,
1 H, H-5), 4.23 (Wq, J ¼ 9:0Hz, 1 H, H-3), 4.39 (dWt, J ¼ 9:4Hz, J ¼ 2:5Hz, 1
H, H-2), 4.59 (dd, J ¼ 9:2Hz, J ¼ 3:6Hz, 1 H, H-6), 4.65 (under HDO, 1 H, H-
13
1);
C-NMR (D2O, pH adjusted to 6 (ND4OD)): d 70.29 (m, CH), 73.16 (s,
CH), 73.53 (m, CH), 73.57 (m, CH), 74.05 (d, J ¼ 4:1Hz, CH), 77.70 (m, CH);
31P-NMR{1H} (D2O, pH adjusted to 6 (ND4OD)): d 1.68 (P-1), 2.36 (P-2), 2.57
(P-6), 3.07 (P-3).
D
-chiro-Inositol 1,3,4,6-tetrakisphosphate (+)-9. 3-Diethylamino-2,4,3,-benzodiox-
aphosphepane (253 mg, 1 mmol) was added to a solution of 50 mg (0.13 mmol) dib-
enzyl- -chiro-inositol (+)-4 and 1H-tetrazole (120 mg, 1.7 mmol) in anhydrous
D
dichloromethane (10 mL); the mixture was stirred at room temperature for 12 h. Af-
ter work-up as described for 3 (650 mg m-CPBA (70%), 2.6 mmol in 15 mL CH2Cl2)
and purification by flash chromatography (EtOAc), pure (+)-2,5-di-O-benzyl-
1,3,4,6-tetra-O-(2-oxo-5,6-benzo-1,3,2-dioxaphosphep-2-yl)-D-chiro-inositol (45 mg,
20
D
30%) was isolated as a colorless solid. Rf 0.30 (EtOAc); ½aŠ ¼ þ16:1° (c ¼ 0:49,
1
CHCl3); H-NMR (CDCl3): d 3.99 (s, br, 2 H, H-2, H-5), 4.75 (d, J ¼ 11:2Hz, 2
H, Ph–CH2), 4.85 (d, J ¼ 11:2Hz, 2 H, Ph–CH2), 4.66-5.58 (m, 16 H,
(CH2Þ C6H4), 5.07 (m, 2 H, H-3, H-4), 5.30 (Wd, J ¼ 9:7Hz, 2 H, H-1, H-6), 7.13-
2
7.54 (m, 26 H, Ph–H); 13C-NMR (CDCl3): d 68.08–69.46 (m, 4 (CH2Þ C6H4),
2
71.89 (m, C-1 and C-6), 72.93 (Ph–CH2), 75.38 (m, C-2 and C-5), 77.65 (m, C-3
and C-4), 127.8–129.2 (Carom:), 134.76, 135.19, 135.31, 137.02 (Cipso); 31P-NMR{1H}
(CDCl3)þ: d 0.56 (br, 2 P), )1.71 (2 P). HR-MS (ESI-pos., TOF): m/z: 1089.2208
½M þ HŠ calcd. for C52H53O18P4: 1089.2182.
To a suspension of 35 mg (27 lmol) (+)-2,5-di-O-benzyl-1,3,4,6-tetra-O-(2-oxo-
5,6-benzo-1,3,2-dioxaphosphep-2-yl)-D-chiro-inositol, isolated as above, in 10 mL
ethanol/water (1:1) was added Pd/C (20 mg). The mixture was stirred overnight at
room temperature under H2. The catalyst was filtered off, and the filtrate was con-
centrated in high vacuum and lyophilized to give (+)-9 as a colorless, strongly hygro-
20
scopic foam (15 mg, 99%). ½aŠ ¼ þ13:6° (c ¼ 0:75, H2O); 1H-NMR (D2O, pH
D
adjusted to 6 (ND4OD)): d 4.02 (m, 2 H, H-2, H-5), 4.16 (m, 2 H, H-3, H-4), 4.52
(m, 2 H, H-1, H-6); 13C-NMR (D2O, pH adjusted to 6 (ND4OD)): d 72.10 (m, C-
2 and C-5), 75.76 (m, C-1 and C-6), 79.49 (m, C-3 and C-4); 31P-NMR{1H} (D2O,
pH adjusted to 6 (ND4OD)): d 2.17 (2 P), 3.03 (2 P).

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
55
3. Results and discussion
3.1. Stereo- and regiospecificity of myo-InsP6 dephosphorylation by the phytases from
S. cerevisiae and P. rhodanensis
Among the sources of phytase useful for the intended study, yeast strains seemed
to be especially promising. The cells are easy to cultivate with short generation times,
making fresh cellular material accessible at any time, even in large amounts; this is
important for their application as biocatalysts in enzyme-assisted synthesis. Whole
cell transformations of myo-InsP6 with phytase from bakerÕs yeast (S. cerevisiae) have
already been published [24,25]. Dephosphorylation of myo-InsP6 proceeds via a re-
markably stereo- and regiospecific route [26]. We chose this phytase as an example
of an enzyme with rather high specificity. Previous experiments with P. rhodanensis
were indicative of a phytase activity with a more complex dephosphorylation pat-
tern. This species secretes high levels of the enzyme into the culture medium. We es-
tablished a rapid method for precipitating the phytase from the culture supernatant
by the addition of ethanol. Recoveries of more than 80% of the total enzyme activity
can be achieved by this procedure. Being able to perform conversions with enriched
enzyme preparations instead of whole cells has several advantages, e.g., no risk of
contamination through cell lysis. Furthermore, work with living cells is often re-
stricted to narrow temperature and pH ranges.
The conditions for the myo-InsP6 conversion with S. cerevisiae were adopted from
the literature [25]. To correlate the data with those obtained for the crude phytase
from P. rhodanensis, we employed identical conditions in the two cases: a tempera-
ture of 45 °C, a pH of 4.6 (acetate buffer) and a starting substrate concentration of
1 mM myo-InsP6. In our experience S. cerevisiae remains viable for several hours un-
der these conditions. In a comprehensive study of yeast phytases, all of the enzymes
exhibited optimal activity in the pH range of 4–5 [27]. The reported KM values for the
partially purified phytases are 0.21 mM for the enzyme from S. cerevisiae [28] and
0.25 mM for the one from P. rhodanensis [27]. These data suggest that the concentra-
tion of 1 mM inositol hexakisphosphate used might lead to substrate saturation of
the enzymes. Commercially available myo-inositol hexakisphosphate was purified
by HPLC before use, to ensure that it contained no interfering lower phosphorylated
inositol phosphates. Aliquots of the salt-free, protonated form of myo-InsP6 were
subjected to sulfuric acid hydrolysis. As already described, the determination of re-
leased orthophosphate provides a reliable quantification of phosphorylated sub-
stances [16]. The chromatograms in Fig. 1 demonstrate how the phytase activities
from S. cerevisiae (trace 1a) and P. rhodanensis (trace 1b) catalyze the stepwise de-
phosphorylation of myo-InsP6. The achiral analytical method does not differentiate
between enantiomers, but for the phytase from S. cerevisiae the absolute configura-
tion of the accumulating myo-InsP3 isomer was determined unequivocally in earlier
work [24]. It is exclusively
with an interesting pharmacological profile [29]. As confirmed by cochromatography
with standards, the precursors are consequently -myo-Ins(1,2,4,5,6)P5, and -myo-
Ins(1,2,5,6)P4. This is in accordance with the dephosphorylation pathway previously
D-myo-Ins(1,2,6)P3 (alpha-trinositol, PP56), a compound
D
D

56
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
Fig. 1. HPLC-MDD analysis showing (a) myo-InsP6 dephosphorylation by the phytase of S. cerevisiae
and (b) the corresponding reaction catalyzed by the phytase of P. rhodanensis (I: Ins(1,2,4,5,6)P5/
Ins(2,3,4,5,6)P5; II: Ins(1,2,3,4,5)P5/Ins(1,2,3,5,6)P5; III: Ins(1,2,5,6)P4/Ins(2,3,4,5)P4; IV: Ins(1,2,4,5)P4/
Ins(2,3,5,6)P4; V: InsP3 isomers).
published by Greiner et al., who examined the purified phytase of S. cerevisiae [26].
Up to the inositol trisphosphate level it obviously makes no difference whether one
works with whole cells or the pure enzyme. For the phytase of P. rhodanensis the fol-
lowing intermediates of myo-InsP6 degradation were observed: Ins(1,2,4,5,6)P5/In-
s(2,3,4,5,6)P5, Ins(1,2,3,4,5)P5/Ins(1,2,3,5,6)P5, Ins(1,2,5,6)P4/Ins(2,3,4,5)P4, Ins(1,2,
4,5)P4/Ins(2,3,5,6)P4 and InsP3 isomers. The phytase activity of P. rhodanensis was
inhibited almost completely (P90%) by low concentrations of the pseudo-substrate
myo-inositol hexakissulfate (about 100 lM), as were the enzymatic conversions of
the substrate analogues (3) and (())-5) described below, an observation that has al-
ready been made for the myo-InsP6 conversion with phytases of Aspergillus ficuum
[30]. If there are different activities involved in the reactions investigated, they share
these characteristics. For an indication of whether the reactions found in the case of
P. rhodanensis are catalyzed by a single enzyme activity, the ethanol-precipitated en-
zyme preparation was subjected to size-exclusion chromatography. With recoveries
of more than 70%, the activity eluted as a single peak with an apparent molecular
weight of about 350 kDa (data not shown). HPLC-MDD revealed no differences be-
tween the intermediates of myo-InsP6 conversion observed before and after chroma-
tography, making it unlikely that different enzyme activities are present in the
preparation and contribute to this dephosphorylation pattern.
Knowing the exact route of enzymatic myo-InsP6 dephosphorylation, including the
absolute configuration of the intermediates, is valuable in studying the substrate spec-
ificity with the stereoisomers neo-InsP6 (3) and -chiro-InsP6 (())-5). Stereoselective
L
assignment of myo-inositol phosphates can be accomplished by a time-consuming
procedure starting with periodate oxidation (cleavage between two vicinal hydroxy
groups) and continuing with reduction and dephosphorylation. The polyol produced
is then incubated with L-polyol dehydrogenase (EC 1.1.1.14), and the products are

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
57
characterized by HPLC [31]. Only myo-inositol bis-, tris- and tetrakisphosphates with
two vicinal hydroxy groups can be analyzed by this method, and because of the mul-
titude of steps the use of radioactively labeled compounds is recommended. In a new
method for the determination of the enantiomeric purity of myo-Ins(1,2,4,5,6)P5/myo-
Ins(2,3,4,5,6)P5 and myo-Ins(1,2,3,4,5)P5/myo-Ins(1,2,3,5,6) P5, we exploit a partially
purified phytase from D. discoideum as an enzymatic tool [14]. Only about 20 nmol of
the myo-inositol pentakisphosphate isomer is needed for comparison of its enzymatic
conversion with that of authentic, enantiomerically pure standards generated by ste-
reocontrolled chemical transformations starting from conduritol-B derivatives (man-
uscript in preparation). Theoretically, the dephosphorylation of each enantiomer
could lead to five myo-InsP4 isomers, but in no case did we find any indication of
hydrolysis of the axial phosphate group in position 2. Fig. 2A summarizes the results
obtained by HPLC-MDD for the optical antipodes myo-Ins(2,3,4,5,6)P5 and myo-
Ins(1,2,4,5,6)P5. As shown in chromatogram 2Ab, the first dephosphorylation of
myo-Ins(2,3,4,5,6)P5 takes place mainly in position 6 and leads to myo-Ins(2,3,4,5)P4.
The enantiomer myo-Ins(1,2,4,5,6)P5 is also preferentially attacked at position 6,
thereby producing clearly distinguishable dephosphorylation products (chromato-
gram 2Ac). It is important to note that the signals of the breakdown products of
myo-Ins(2,3,4,5,6)P5 and myo-Ins(1,2,4,5,6)P5 do not superimpose. Experiments with
artificial mixtures of the optical antipodes yielded a detection limit of approximately
2% for the less concentrated enantiomer (data not shown). When the purified inositol
pentakisphosphate with the retention time of Ins(1,2,4,5,6)P5/Ins(2,3,4,5,6)P5 from
the myo-InsP6 conversion with Pichia phytase was subjected to this procedure it was
unequivocally identified as
D
-myo-Ins(1,2,4,5,6)P5 (P98% ee, chromatogram 2Ad).
Fig. 2B shows a comparison of analytical runs for the enzymatically converted refer-
ence compounds myo-Ins(1,2,3,5,6)P5 (chromatogram 2Bb) and myo-Ins(1,2,3,4,5)P5
(chromatogram 2Bc). For better separation of the breakdown products in this case,
the samples were chromatographed with an alkaline elution system. There is a
clear-cut difference in the myo-inositol tetrakisphosphates observed. The main inter-
mediates formed are Ins(1,2,3,6)P4 from Ins(1,2,3,5,6)P5 and Ins(1,2,4,5)P4 from
Ins(1,2,3,4,5)P5. Analogous experiments performed with the second inositol pentakis-
phosphate isolated after myo-InsP6 conversion with Pichia phytase revealed that this
compound is -myo-Ins(1,2,3,4,5)P5 (P98% ee, chromatogram 2Bd).
D
On the basis of the absolute configuration determined for the inositol pentakis-
phosphates, it was possible with HPLC-MDD to unravel the degradation pathway
of myo-InsP6 up to the inositol trisphosphate level. Scheme 1 illustrates the complex
routes of myo-InsP6 dephosphorylation catalyzed by the phytase of P. rhodanensis.
The reaction sequence framed with a dotted line represents the comparatively simple
but quite specific counterpart catalyzed by the phytase of S. cerevisiae.
3.2. Chemical synthesis of the stereoisomers neo-InsP6 (3) and -chiro- InsP6 (())-5)
L
and the required reference compounds
Previous work led us expect that the enantiopure diacetoxy-dibromocyclohex-
5-ene (+)-1 and ())-1, which are readily accessible from p-benzoquinone in four steps

58
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
Fig. 2. Stereoselective assignment of myo-inositol pentakisphosphates by means of a phytase from D. dis-
coideum. HPLC-MDD chromatograms for the enantiomeric pairs Ins(1,2,4,5,6)P5/Ins(2,3,4,5,6)P5 (A;
acidic elution system) and Ins(1,2,3,4,5)P5/Ins(1,2,3,5,6)P5 (B; alkaline elution system). A: Standard mix-
ture of resolved isomers including the elution position of myo-Ins(2,4,5,6)P4 (a). Enzymatically converted
enantiopure Ins(2,3,4,5,6)P5 (b), Ins(1,2,4,5,6)P5 (c) and the corresponding isolated isomer from the
myo-InsP6/Pichia-phytase degradation (d). B: Standard mixture of resolved isomers including the elution
positions of myo-Ins(1,2,5,6)P4 and myo-Ins(1,2,3,5)P4 (a). Enzymatically converted enantiopure
Ins(1,2,3,5,6)P5 (b), Ins(1,2,3,4,5)P5 (c) and the corresponding isolated isomer from the myo-InsP6/Pi-
chia-phytase degradation (d).
Scheme 1. Main pathways of myo-InsP6 dephosphorylation catalyzed by the phytase of P. rhodanensis
(framed reaction sequence characteristic for S. cerevisiae).
on multigram scale [16,17] can also serve as starting material for conduritol systems
other than conduritol B. Our synthetic concept is based on the stereospecific conver-
sion of this building block into suitably protected enantiomerically pure conduritol

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
59
derivatives, which can subsequently be transformed to inositol isomers by simple
functionalization of the double bond [22].
Synthesis of neo-inositol. The introduction of two acetate groups in the C-2 and
C-3 position under inversion of the stereogenic centers might lead to a conduritol
E derivative. Heating the diacetate (+)-1 for ten days with sodium acetate in aqueous
acetic acid indeed provided conduritol E derivative (+)-2a (structure not shown in
Scheme 2) in 70% yield [22]. The epoxidation of (+)-2a with trifluoroperacetic
acid and in situ hydrolysis gave neo-inositol tetraacetate (+)-2. The opening of the
oxirane was observed to proceed regioselectively to only the neo-isomer. De-O-acety-
lation of (+)-2 by treatment with a catalytic amount of sodium methylate in metha-
nol did not run to completion because of the coprecipitation of partially deacylated
educt with neo-inositol. Full deprotection was carried out by refluxing the resulting
Scheme 2. Preparation of neo-inositol phosphates (3, (+)-8) and -chiro-inositol phosphates (())-5, (+)-9).
L
Reagents: (a) (i) NaOAc, AcOH (95%), 10 days, 130 °C then Ac2O; CH2Cl2, DMAP(cat.). (ii)
(CF3COÞ O, H2O2, CH2Cl2; NaHCO3; (b) (i) NaOMe, MeOH then 0.25 M NaOH. (ii) 3-Diethylamino-
2
2,4,3-benzodioxaphosphepane, 1H-tetrazole, CH2Cl2=CH3CN then m-CPBA. (iii) Pd/C, H2
EtOH=H2O. (c) (i) NaOBn, BnOH. (ii) ðCF3COÞ O, H2O2, CH2Cl2, Na2CO3. (iii) H2SO4, dioxane, water.
2
(d) (i) Pd/C, H2 ethanol/water. (ii) 3-Diethylamino-2,4,3-benzodioxaphosphepane, 1H-tetrazole, CH2Cl2
then m-CPBA. (iii) Pd/C, H2 ethanol/water. (e) 2,2,2-trichloro-acetimidic acid benzyl ester, CF3SO3H.
(f) NaOMe, MeOH. (g) (i) 3-Diethylamino-2,4,3-benzodioxaphosphepane, 1H-tetrazole, CH2Cl2 then
m-CPBA. (ii) Pd/C, H2 EtOH=H2O.

60
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
precipitate in aqueous NaOH. neo-Inositol was easily purified by recrystallization
from water [32].
The well known 2,5-di-O-benzyl- -chiro-inositol ())-4 was synthesized in three
L
steps from diacetate (+)-1 [22]. The spectral data of the solid obtained correspond
to those observed by Liu et al. [33]. Pd/C-catalyzed hydrogenation gave free
L
-chiro-inositol as a colorless solid.
The introduction of the six phosphate functionalities into free L-chiro- and neo-
inositol was achieved by treatment with 3-diethylamino-2,4,3-benzodioxaphosphe-
pane in dichloromethane/acetonitrile in the presence of 1H-tetrazole. After 12 h
for L-chiro-inositol and 36 h for neo-inositol the resulting phosphites were oxidized
with m-CPBA and purified by flash chromatography to give the protected hexakis-
phosphates in 50% yield at most. Because of the low solubility of free inositol, espe-
cially neo-inositol, the phosphorylation is quite difficult, and full conversion under
these conditions was not possible. Varying the solvents or applying longer reaction
times did not increase the yield. The subsequent Pd/C-catalyzed hydrogenolysis gave
the free inositol hexakisphosphates 3 and ())-5. Alternatively, neo-inositol was phos-
phorylated in polyphosphoric acid to give 3 in good yield. This method was, how-
ever, only of minor value because of the problem in work-up caused by the great
excess of polyphosphoric acid needed and the difficulties of removing this substance
completely. Purification of the resulting inositol hexakisphosphates by HPLC en-
sures purities suitable for biological experiments.
For further insight into the enzymatic dephosphorylation of substrate analogues
by yeast phytases it was essential to synthesize enantiomerically pure reference ma-
terials for the potential degradation products (details are discussed later). Our first
effort was the synthesis of neo-Ins(1,2,5,6)P4 (10). Earlier work led us to propose
that enantiopure tetra-O-acetyl-neo-inositol ())-2 would be an ideal starting mate-
rial for the synthesis of neo-Ins(1,2,5,6)P4 (10), if the two free hydroxy groups
could be functionalized with an orthogonal protecting group. At first sight, the
best protection seemed to be benzyl ether groups, because it should be possible
to remove these completely in one step at the end of the reaction sequence. We
tried to benzylate under mild conditions with 2,2,2-trichloro-acetimidic acid benzyl
ester, but even with an excess of reagent the monobenzylated neo-inositol tetraac-
etate (+)-6 was formed as the main product; it was isolated in pure form in 20%
yield. The dibenzyl-neo-inositol tetraacetate was only obtained impure and in very
low yield along with other products. This result forced us to try the synthesis of
another, even more readily accessible reference material, namely neo-
Ins(2,3,4,5,6)P5 (+)-8. This new target molecule was expected to help in the assign-
ment of the absolute configuration of the products of the degradation of neo-InsP6
(3) by yeast phytases.
For this purpose the enantiomer (+)-2 was treated with 2,2,2-trichloro-acetimidic
acid benzyl ester in dichloromethane/hexane to furnish the monobenzylated deriva-
tive ())-6. Alkaline methanolysis gave free 1-benzyl-neo-inositol ())-7. Phosphoryla-
tion with 3-diethylamino-2,3,4-benzodioxaphosphepane in the presence of
1H-tetrazole, with subsequent oxidation with m-CPBA and finally Pd/C-catalyzed
hydrogenation of ())-7, provided neo-Ins(2,3,4,5,6)P5 (+)-8.

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
61
Compound (+)-4 was made from the optical antipode ())-1; it can also be made
from -chiro-inositol by a method reported by Liu et al. [33]. In analogy to the syn-
thesis of (+)-8, (+)-4 was phosphorylated to yield pure -chiro-Ins(1,3,4,6)P4 (+)-9,
D
D
another compound useful in unravelling the degradation pathways of interest here.
It should be mentioned that both enantiomers are accessible in all cases from the
corresponding enantiomerically pure conduritol derivatives, illustrating the versatil-
ity of this synthetic strategy.
3.3. Enzymatic conversion of neo-InsP6 (3) and -chiro-InsP6 (())-5)
L
The obvious difference in the molecular structures of myo-InsP6 and neo-InsP6 (3)
-chiro-InsP6 (())-5) lies in the orientation of a single substituent (confirmed by
NMR at pH 5–7, see Scheme 3). In contrast to myo-InsP6, the phosphate group in
5-position of the epimer neo-InsP6 (3) is axially oriented, and -chiro-InsP6 (())-5)
possesses an axial phosphate group in 3-position (in the scheme this corresponds
to the 6-position of -chiro-InsP6 (())-5)). Scheme 3 gives only an idealized picture
or
L
L
L
of the structures in aqueous solutions, the spatial arrangement of the phosphate
groups in each of these molecules being dependent inter alia on the protonation,
the hydration and the formation of intramolecular hydrogen bonds.
Scheme 3. Molecular structures and numbering of carbons for the inositol hexakisphosphate stereo-
isomers.

62
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
To get an idea whether the phytases from S. cerevisiae and P. rhodanensis tolerate
deviations from the myo-configuration and catalyze the dephosphorylation of the
epimers neo-InsP6 (3) and -chiro-InsP6 (())-5), too, we compared the initial rates
L
of the conversions characteristic for the first dephosphorylation. The HPLC-MDD
system calibrated for the three stereoisomers was used to quantify the substances.
With a starting concentration of 1 mM, kinetics were measured for no more than
30% inositol hexakisphosphate consumption. The curves were evaluated by linear re-
gression. Table 1 summarizes the results. For S. cerevisiae the phytase activity is gi-
ven in relation to the wet weight of the cells, while for P. rhodanensis the values
represent the specific activity of the crude enzyme. The results were somewhat unex-
pected: although -chiro-InsP6 (())-5) is not degraded by S. cerevisiae, the ‘‘natural’’
L
substrate myo-InsP6 is clearly dephosphorylated more slowly than neo-InsP6 (3). The
situation is different for the phytase of P. rhodanensis: the rates for neo-InsP6 (3) and
myo-InsP6 are nearly equal, whereas the conversion of -chiro-InsP6 (())-5) is about
L
twice as fast. As mentioned, all inositol hexakisphosphates investigated were purified
by HPLC prior to use. To exclude the possibility that myo-InsP6 isolated from corn
contains inhibitory impurities coeluting in the purification procedure, we repeated
the study with the compound isolated from rice. The data correlate within the toler-
ance with those shown in Table 1. Although not demonstrated before for any phy-
tase of eucaryotic origin, in view of the habitat of yeast and the fact that several
inositol hexakisphosphate isomers are constituents of soil [12] it is not surprising that
the enzymes are capable of dephosphorylating isomers other than myo-InsP6. One
would, however, expect the catalyzed reaction to be optimized for the compound re-
garded as the ‘‘natural’’ substrate.
In order to clarify whether the dephosphorylation of the substrate analogues is
specific or rather random, we isolated the breakdown products by HPLC and char-
acterized relevant intermediates by NMR spectroscopy. Fig. 3 shows the HPLC-
MDD analysis after conversion of the two synthetic inositol hexakisphosphate
isomers by S. cerevisiae. Whereas
L
-chiro-InsP6 (())-5) is not dephosphorylated
(chromatogram 3b), a neo-inositol tetrakisphosphate accumulated in the course of
the reaction with neo-InsP6 (3; chromatogram 3a). As a putative precursor, a single
neo-inositol pentakisphosphate was observed, indicating a specific enzymatic conver-
sion. Even after complete degradation of neo-InsP6 (3) the compound with the reten-
tion time of a neo-InsP4 isomer shows little tendency towards further
dephosphorylation (data not shown). The assay initially designed for the kinetic
Table 1
Enzyme activities towards different inositol hexakisphosphate stereoisomers
InsP6 stereoisomer
Phytase activity
Saccharomyces cerevisiae
Pichia rhodanensis
(mU/mg wet weight)
(mU/mg protein)
myo-InsP6
neo-InsP6 (3)
0:13 ꢁ 0:03
9:3 ꢁ 1
9:7 ꢁ 1:1
18:6 ꢁ 0:5
0:55 ꢁ 0:03
L
-chiro-InsP6 (())-5)
Not detectable

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
63
Fig. 3. HPLC-MDD chromatograms representative of (a) neo-InsP6 (3) dephosphorylation (I:
neo-Ins(1,2,4,5,6)P5/neo-Ins(2,3,4,5,6)P5 ((+)-8); II: neo-Ins(1,2,5,6)P4 (10)/neo-Ins(2,3,4,5)P4) and (b)
L
-chiro-InsP6 (())-5) dephosphorylation by the phytase of S. cerevisiae.
evaluation of the reaction was scaled up to get enough neo-InsP4 for a NMR study.
The spectra of the purified compound are representative of a C2-symmetric molecule
with both axial positions phosphorylated (characteristic chemical shift and coupling
pattern for the corresponding ring protons). The only structures in agreement with
all of the spectroscopic data are those of the enantiomeric pair neo-Ins(1,2,5,6)P4/
neo-Ins(2,3,4,5)P4 (Scheme 4). It is very likely that the phosphoester bond susceptible
to the initial attack is the one in position 3, as found for the highly stereospecific first
dephosphorylation of myo-InsP6 by the phytase of S. cerevisiae. In this case the prod-
uct should be enantiomerically pure neo-Ins(1,2,5,6)P4 (10), but the definitive assign-
ment of the absolute configuration required reference material of known
stereochemistry. We therefore synthesized the optical antipode (+)-8 (Scheme 2) of
Scheme 4. Conversion of neo-InsP6 (3) by the phytase of S. cerevisiae.

64
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
the supposed precursor neo-Ins(1,2,4,5,6)P5 and examined its enzymatic dephosphor-
ylation. The degradation of neo-Ins(2,3,4,5,6)P5 ((+)-8) occurs quite unspecifically
under concomitant appearance of three HPLC-resolvable neo-inositol tetrakisphos-
phate isomers, which are subsequently dephosphorylated further (data not shown).
This confirms our hypothesis that neo-Ins(1,2,4,5,6)P5 is the intermediate actually
observed in enzymatic neo-InsP6 (3) conversion. The delay after two dephosphoryla-
tions, leading to the accumulation of neo-Ins(1,2,5,6)P4 (10), may reflect a property
of most phytases: the cleavage of an axial phosphate group is poorly catalyzed [2,3].
In analogy to the dephosphorylation of myo-InsP6, neo-InsP6 (3) breakdown will
start in position 3, followed by position 4; then, in contrast to myo-InsP6, it nearly
stops, because the phosphate group in position 5 is axially oriented. In a search
for structural motifs which would be helpful in understanding substrate recognition
by the phytase, we divided myo-InsP6 into two characteristic parts, the all-cis 1,2,3-
motif with an equatorial–axial–equatorial arrangement of the phosphate groups and
the all-trans 4,5,6-motif bearing only equatorial phosphate groups. The C2-symmet-
ric neo-InsP6 (3) consists of two equivalent parts (1,2,3 and 4,5,6; see Scheme 3), each
resembling the all-cis 1,2,3-motif of myo-InsP6. -chiro-InsP6 (())-5; Scheme 3) lacks
L
such a constellation of phosphate groups. The phytase of S. cerevisiae initially
attacks the 3-position in the 1,2,3-motif of myo-InsP6, which may explain why
L
-chiro À InsP6 (())-5) is resistant to conversion.
Finally, the enzymatic conversion of neo-InsP6 (3) and -chiro-InsP6 (())-5) by the
L
phytase of P. rhodanensis was investigated (Fig. 4). Chromatogram 4a is representa-
tive of the breakdown of neo-InsP6 (3). The same intermediates are observed as in the
reaction with Saccharomyces phytase. This correlates with the results we obtained
for the routes of myo-InsP6 conversion by Pichia phytase (Scheme 1). A prominent
Fig. 4. HPLC-MDD chromatograms representative of (a) neo-InsP6 (3) dephosphorylation (I: neo-
Ins(1,2,4,5,6)P5/neo-Ins(2,3,4,5,6)P5 ((+)-8); II: neo-Ins(1,2,5,6)P4 (10)/neo-Ins(2,3,4,5)P4) and (b)
L
-chiro-InsP6 (())-5) dephosphorylation (i:
by the phytase of P. rhodanensis.
L-chiro-Ins(1,2,3,5,6)P5 (11); ii: L-chiro-Ins(1,2,3,6)P4 (12))

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
65
Scheme 5. Conversion of -chiro-InsP6 (())-5) by the phytase of P. rhodanensis.
L
branch in the reaction scheme starting with the removal of the phosphate in 3-posi-
tion is identical to the specific sequential dephosphorylation catalyzed by the enzyme
of S. cerevisiae. As we discussed above, the first two steps in neo-InsP6 (3) dephos-
phorylation will take place at the corresponding positions. The lack of an all-trans
4,5,6-motif in the structure of neo-InsP6 (3) may be the reason for the defined de-
phosphorylation pattern, so that no pathway starting with the removal of the phos-
phate in 6-position (Scheme 1) could be detected. For
L
-chiro-InsP6 (())-5) the
situation is the opposite. According to IUPAC-IUB, the numbering of the ring car-
bons differs from that of neo-InsP6 (3) and myo-InsP6 [34]. To prevent confusion we
refer to the structures depicted in Scheme 3. The all-cis 1,2,3-motif of myo-InsP6 is
disturbed in the structure of
the 2,1,6-arrangement of phosphate groups. Thus an introductory step resulting in
the removal of the phosphate in the axially oriented 6-position of -chiro-InsP6
L
-chiro-InsP6 (())-5). In Scheme 3 this corresponds to
L
(())-5) is very unlikely. The 5,4,3-assembly of phosphates resembles the all-trans
4,5,6-motif of myo-InsP6, and for that reason we expected an attack at the 3-position
of
MDD analyses were indicative of a rather specific breakdown (Fig. 4b), and the
NMR spectra of the purified -chiro-inositol pentakisphosphate confirmed that
the initial product is indeed -chiro-Ins(1,2,4,5,6)P5 (Scheme 5). Because of the C2
symmetry of the educt -chiro-InsP6 (())-5; Scheme 5), it is impossible to distinguish
between a dephosphorylation in 3- or 4-position. These positions are homotopic, and
the resulting compounds -chiro-Ins(1,2,4,5,6)P5 and -chiro-Ins(1,2,3,5,6)P5 (11)
L
-chiro-InsP6 (())-5) as the predominant first dephosphorylation step. HPLC-
L
L
L
L
L
are identical. In accordance with IUPAC-IUB recommendations that the lowest
possible numerals be preferred in nomenclature [34], the correct abbreviation is
L
-chiro-Ins(1,2,3,5,6) P5 (11). As proven by NMR of the isolated L-chiro-inositol tet-
rakisphosphate, the enzymatic hydrolysis proceeds mainly, as in the case of myo-
Ins(1,2,3,4,5)P5 (Scheme 1), at the neighbouring position counterclockwise to the
free hydroxy group. In view of the detection limit of the NMR method, the com-
pound is P95% isomerically homogeneous L-chiro-Ins(1,2,3,6)P4 (12).
3.4. Potential applications
One focus of our research is the establishment of efficient new strategies in inositol
phosphate synthesis, which combine the advantages of stereocontrolled chemical

66
S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
transformations and of regio- and stereospecific enzyme-assisted steps [14–16]. In the
present paper we demonstrate the usefulness of phytases from S. cerevisiae and P.
rhodanensis for the synthesis of structural analogues to naturally occurring, biolog-
ically active myo-inositol phosphates.
Although not based upon myo-inositol, neo-Ins(1,2,5,6)P4 (10) can be considered
a relative of myo-Ins(1,2,6)P3, a compound showing anti-inflammatory and analgesic
effects [29,35]. The potential of neo-Ins(1,2,5,6)P4 (10) to act biologically as a myo-
Ins(1,2,6)P3 surrogate has to be investigated.
of myo-Ins(1,2,3,6)P4, a compound which may function physiologically as iron car-
rier and calcium absorption enhancer [36,37]. The -chiro isomer may further our
L-chiro-Ins(1,2,3,6)P4 (12) is an epimer
L
understanding of the structure–activity relationship.
Acknowledgments
We thank Dr. W.V. Turner for critical reading of the manuscript. We are grateful
to the Bayer AG Company for supporting this project in many aspects, especially for
€
preparing the HR-MS spectra. We also thank the Max-Planck Institut fur Kohlen-
forschung, in particular Dr. W. Schrader and M. Blumenthal, for further HR-MS
spectra. This work was supported by the Deutsche Forschungsgemeinschaft (Grant
VO 348/3-1) and the Fonds der Chemischen Industrie.
References
[1] J.B. Vincent, M.W. Crowder, B.A. Averill, Trends Biochem. Sci. 17 (1992) 105–110.
ꢀꢁ
ꢁ
[2] J. Dvorakova, Folia Microbiol. 43 (1998) 323–338.
[3] J. Kerovuo, J. Rouvinen, F. Hatzack, Biochem. J. 352 (2000) 623–628.
[4] S.B. Shears, Cell. Signal. 13 (2001) 151–158.
[5] L.A. Hanakahi, M. Bartlet-Jones, C. Chappell, D. Pappin, S.C. West, Cell 102 (2000) 721–729.
[6] Y. Feng, S.R. Wente, P.W. Majerus, Proc. Natl. Acad. Sci. USA 98 (2001) 875–879.
[7] A. Craxton, J.J. Caffrey, W. Burkhart, S.T. Safrany, S.B. Shears, Biochem. J. 328 (1997) 75–81.
[8] P. Van Dijken, J.R. De Haas, A. Craxton, C. Erneux, S.B. Shears, P.J.M. Van Haastert, J. Biol.
Chem. 270 (1995) 29724–29731.
[9] W.R. Sherman, M.A. Stewart, M.M. Kurien, S.L. Goodwin, Biochim. Biophys. Acta 158 (1968)
197–205.
[10] W.R. Sherman, S.L. Goodwin, K.D. Gunnell, Biochemistry 10 (1971) 3491–3499.
[11] K.S. Bruzik, A.A. Hakeem, M.-D. Tsai, Biochemistry 33 (1994) 8367–8374.
ꢁ
[12] B.L. Turner, M.J. Paphazy, P.M. Haygarth, I.D. McKelvie, Philos. Trans. R. Soc. Lond. B 357
(2002) 449–469.
[13] J.-B. Martin, T. Laussmann, T. Bakker-Grunwald, G. Vogel, G. Klein, J. Biol. Chem. 275 (2000)
10134–10140.
[14] S. Adelt, O. Plettenburg, G. Dallmann, F.P. Ritter, S.B. Shears, H.-J. Altenbach, G. Vogel, Bioorg.
Med. Chem. Lett. 11 (2001) 2705–2708.
[15] O. Plettenburg, S. Adelt, G. Vogel, H.-J. Altenbach, Tetrahedron: Asymmetry 11 (2000) 1057–1061.
[16] S. Adelt, O. Plettenburg, R. Stricker, G. Reiser, H.-J. Altenbach, G. Vogel, J. Med. Chem. 42 (1999)
1262–1273.
[17] O. Block, G. Klein, H.-J. Altenbach, J. Org. Chem. 65 (2000) 716–721.

S. Adelt et al. / Bioorganic Chemistry 31 (2003) 44–67
67
[18] P.A. Lanzetta, L.J. Alvarez, P.S. Reinach, O.A. Candia, Anal. Biochem. 100 (1979) 95–97.
[19] G.W. Mayr, in: R.F. Irvine (Ed.), Methods in Inositide Research, Raven Press, New York, 1990, pp.
83–108.
[20] C. Sanfilippo, A. Patti, M. Piattelli, G. Nicolosi, Tetrahedron: Asymmetry 8 (1997) 1569–1573.
[21] T. Tschamber, F. Backenstrass, H. Fritz, J. Streith, Helv. Chim. Acta 75 (1992) 1052–1060.
[22] Podeschwa, M., Plettenburg, O., vom Brocke, J., Block, O., Adelt, S., Altenbach, H.-J., Eur. J. Org.
Chem., submitted.
[23] H. Takahashi, H. Kittaka, S. Ikegami, J. Org. Chem. 66 (2001) 2705–2716.
ꢁ
[24] Siren, M., U.S. Patent No. 4,851,560.
[25] C. Blum, S. Karlsson, G. Schlewer, B. Spiess, N. Rehnberg, Tetrahedron Lett. 36 (1995) 7239–7242.
[26] R. Greiner, M. Larsson Alminger, N.-G. Carlsson, J. Agric. Food Chem. 49 (2001) 2228–2233.
[27] Y. Nakamura, H. Fukuhara, K. Sano, Biosci. Biotechnol. Biochem. 64 (2000) 841–844.
[28] N.R. Nayini, P. Markakis, Lebensm. Wiss. u. Technol 17 (1984) 24–26.
ꢁ
[29] H. Siren, in: A. Reitz (Ed.), Inositolphosphates and Derivatives, Synthesis, Biochemistry and
Therapeutic Potential, ACS Symposium Series, vol. 463, ACS, Washington, DC, 1991, pp. 103–110.
[30] A.H.J. Ullah, K. Sethumadhavan, Biochem. Biophys. Res. Commun. 251 (1998) 260–263.
[31] L.R. Stephens, in: R.F. Irvine (Ed.), Methods in Inositide Research, Raven Press, New York, 1990,
pp. 9–30.
[32] S.J. Angyal, D.C. Craig, Carbohydr. Res. 263 (1994) 149–154.
[33] C. Liu, R. Davis, S. Nahorski, S. Ballereau, B. Spiess, B.V.L. Potter, J. Med. Chem. 42 (1999)
1991–1998.
[34] IUPAC and IUPAC-IUB Recommendations, Eur. J. Biochem. 57 (1975) 1–7.
[35] A.B. Malmberg, B. Fallgren, T. Hedner, Eur. J. Pharmacol. 294 (1995) 609–615.
[36] I.D. Spiers, C.J. Barker, S.-K. Chung, Y.-T. Chang, S. Freeman, J.-M. Gardiner, P.H. Hirst, P.A.
Lambert, R.H. Michell, D.A. Poyner, C.H. Schwalbe, A.W. Smith, K.R.H. Solomons, Carbohydr.
Res. 282 (1996) 81–99.
[37] X. Shen, C.M. Weaver, A. Kempa-Steczo, B.R. Martin, B.Q. Phillippy, R.P. Heaney, J. Nutr.
Biochem. 9 (1998) 298–301.