Acid hydrolysis of myo-inositol 1,3,5-orthoesters, apart from orthoformates, exclusively affords the corresponding 2-O-acyl myo-inositol products via a 1,2-bridged five-membered ring dioxolanylium ion intermediate observed by NMR spectroscopy. These C-2-substituted inositol derivatives provide valuable precursors for rapid and highly efficient routes to 2-O-acyl inositol 1,3,4,5,6-pentakisphosphates and myo-inositol 1,3,4,5,6-pentakisphosphate with biologically interesting and anticancer properties. Deuterium incorporation into the α-methylene group of such alkyl ester products (2-O-C(O)CD2R), when the analogous alkyl orthoester is treated with deuterated acid, is established utilizing the novel orthoester myo-inositol 1,3,5-orthobutyrate as an example. Such deuterated ester products provide intermediates for deuterium-labeled synthetic analogues. Investigation into this selective formation of 2-O-ester products and the deuterium incorporation is presented with proposed mechanisms from NMR experiments.
Acid hydrolysis of myo-inositol 1,3,5-orthoesters, apart from orthoformates, exclusively affords the corresponding 2-O-acyl myo-inositol products via a 1,2-bridged five-membered ring dioxolanylium ion intermediate observed by NMR spectroscopy. These C-2-substituted inositol derivatives provide valuable precursors for rapid and highly efficient routes to 2-O-acyl inositol 1,3,4,5,6-pentakisphosphates and myo-inositol 1,3,4,5,6-pentakisphosphate with biologically interesting and anticancer properties. Deuterium incorporation into the α-methylene group of such alkyl ester products (2-O-C(O)CD2R), when the analogous alkyl orthoester is treated with deuterated acid, is established utilizing the novel orthoestermyo-inositol 1,3,5-orthobutyrate as an example. Such deuterated ester products provide intermediates for deuterium-labeled synthetic analogues. Investigation into this selective formation of 2-O-ester products and the deuterium incorporation is presented with proposed mechanisms from NMR experiments.
Inositol phosphatesare important intracellular
messengers that
play a vital role in many cellular functions including cell growth,
migration, differentiation, apoptosis, and endocytosis.[1,2] To facilitate biochemical investigations, it is vital that efficient
routes to these naturally occurring inositol phosphatesare established.[3−7] As the demand for key synthetic intermediates increases, so does
the need for differentially protected entities having free hydroxyl
groups at specific positions. myo-Inositol 1,3,5-orthoesters
have been extensively utilized as synthetic precursors for such inositol
derivatives,[8−21] since the O-1, O-3, and O-5 atoms can be protected in a single step
leaving O-2, O-4, and O-6 for further manipulation (Scheme 1). Also, varied patterns of hydroxyl group protection
can be obtained by the cleavage of orthoesters with reducing agents,[11−15] and in the case of the alkyl orthoesters, acid hydrolysis can introduce
an acyl ester at various positions on the inositol ring.[18,19]myo-Inositol orthoformate (1) has
been extensively investigated over the years.[11−17] Acid hydrolysis partially or completely removes the orthoformateester, since any intermediate formate ester produced is easily further
cleaved under acidicconditions. myo-Inositol orthoacetate
(2), which we[18] and others[19,20] have also explored, has the advantage that acid hydrolysis leaves
an acetate ester on one of the oxygen atoms of the inositol ring,
and this can be retained as a protecting group for further manipulation.
However, such acetate estersare prone to migration under acid conditions,
and therefore, we have recently investigated derivatives of myo-inositol orthobenzoate (3) as synthetic
precursors.[21,22] Acid hydrolysis of the orthobenzoate 3 gives a much more stable benzoate ester, which can be directed
to different positions on the ring, e.g., C-1 and C-3 by choosing
suitable conditions and protection of other OH groups[21] and in concert with chiral desymmetrization can lead to
optically active targets.
Scheme 1
In contrast, apart from orthoformate 1, acid hydrolysis
of C-2 unprotected myo-inositol orthoesters 2, 3, and 4 unexpectedly affords
the 2-O-acyl myo-inositol derivatives 5, 6, and 7, respectively (Scheme 2). In a preliminary communication,[22] we reported the surprising regioselective conversion of
orthobenzoate 3 to the 2-O-benzoyl derivative 6 and demonstrated that this proceeds via the intermediacy
of a 1,2-bridged 2′-phenyl-1′,3′-dioxolan-2′-ylium
ion. We now report full details and extension of this work including
mechanisticNMR and labeling experiments and show that the resulting
protected inositolcan be used to provide a highly efficient route
to two emerging anticancer agents.[23−25]
Scheme 2
Results and Discussion
Regioselective
Synthesis of myo-Inositol 2-O-Acetate
and Mechanistic Investigations
Orthoacetate 2 was prepared by treating myo-inositol with
triethyl orthoacetate and PTSA in DMF under reflux as previously reported.[18] Treatment of 2 with aqueous TFA
(10:1) gives 2-O-acetate 5 with complete
selectivity in quantitative yield. This reaction is extremely fast
and clean with no starting material observed by TLC even after 5 min
of commencing the reaction.Our finding that acid hydrolysis
of 2 leads to the selective formation of ester 5 was unexpected. Previous studies[21] on the hydrolysis of myo-inositol orthoesters have
shown that the ester group is directed to the C-1 or C-3 position,
although in these studies, the 2-OH group was protected. Interestingly,
it was previously reported that acid hydrolysis of (±)-4-deoxy-myo-inositol 1,3,5-orthopentanoate gave the 2-O-pentanoate, although the authors attributed this to acyl migration.[26] We reasoned, however, that acyl migration could
not explain the high regioselectivity in the formation of 5 from 2 and a series of NMR experiments was therefore
undertaken in an attempt to observe intermediates in this hydrolysis
reaction.A 1HNMR spectrum taken immediately after
treatment
of 2 with deuterated TFA and CDCl3 showed
only the starting material and some 2-O-acetate product 5, the latter presumably arising from adventitious water.
A small amount of CDCl3 was used due to difficulty in attaining
a lock signal in neat CF3CO2D. Although we were
unable to observe any intermediates at this initial stage, a small
change in the methyl peak of 5 was seen. Over time, the
disappearance of the methyl singlet at 1.52 ppm and the appearance
of new peaks next to it and their disappearance was observed by 1HNMR (Supporting Information,
Figure 1), leading us to believe that the methyl group was being deuterated
giving a mixture of −CH3, −CH2D, and −CHD2 peaks. The reaction was therefore
repeated, and further 2H and 13CNMR spectra
were taken at different times. DeuteriumNMR spectra showed an increase
in the deuterated methyl peaks (−CH2D, −CHD2, and −CD3) at 1.52 ppm over time, complementary
to the 1HNMR decrease in the methyl singlet (−CH3) proving that methyl group protons were exchanging (Supporting Information, Figure 2). This was further
proven by 13CNMR spectra taken on a different time scale.
Splitting patterns were seen progressively in the 13CNMR
spectrum with initially a singlet (−CH3) at δ
23 ppm, a mixture of the singlet and a triplet (indicating −CH2D), a mixture along with a quintet showing some −CHD2, and a septet showing −CD3.When
acidic hydrolysis of orthoacetate 2 was carried
out in other solvents such as ethanol, methanol, or acetonitrile,
or with use of mineral acids such as HCl, the observed selectivity
for 5 was reduced and longer reaction times were also
needed. Furthermore, under the selective reaction conditions using
10:1 TFA/water (50%) but in the presence of another solvent such as
methanol (50%), small amounts of 1/3-O-acetate were
also obtained.In order to further investigate this high regioselectivity
and
also to further understand the observed deuterium incorporation, both
the C-2hydroxyl-free and protected orthoacetates4,6-di-O-benzyl orthoacetate (12) and 2,4,6-tri-O-benzyl orthoacetate (13) were synthesized (Scheme 3).
Scheme 3
Orthoacetates in Deuterated TFA
Metal alkoxidechelation-controlled
regioselective protection of
C-4 and C-6hydroxyl groups was achieved in excellent yield when orthoacetate
triol 2 was treated with lithium hydride and benzyl bromide
in dry DMF. A higher yield was obtained when lithium hydride was used
in comparison to that using n-butyllithium.[27] Observation of 12 in deuteratedTFA and CDCl3 by NMR spectroscopy (1H, 2H, and 13C) again showed the disappearance of the
signal corresponding to the methyl group of the orthoacetate at 1.56
ppm (1HNMR) and an increase in the height of the deuterated
methyl peaks over time by 2HNMR spectra, once more confirming
that the protons of the methyl group were exchanging with solvent
deuterium. Although the major peaks seen were of the starting material
which was being deuterated, there was a very small amount of hydrolyzed
product, myo-inositol4,6-di-O-benzyl2-O-acetate, as well as a minute amount of some other
broad peaks, perhaps corresponding to an intermediate species. This
other species showed downfield broad signals at 5.91 and 5.72 ppm
corresponding to one proton each and a multiplet around 4.21–4.27
ppm corresponding to four protons, along with a broad singlet at 2.67
ppm corresponding to three protons, perhaps representing a methyl
group. However, since this species was observed only in trace amounts
by 1HNMR and not by 13CNMR, identification
of this potential intermediate was impossible.Tribenzyl orthoacetate 13, synthesized by conventional
benzylation of 2,[28] was treated
with deuterated TFA, and after 15 min, only starting material was
seen by NMR spectroscopy, and no intermediates or changes to the methyl
singlet were observed. Additionally, 2HNMR spectra taken
up to 5 days showed no peaks due to deuterium incorporation into the
methyl group of starting material 13.
Regioselective
Synthesis of myo-Inositol 2-O-Benzoate
(6)
Orthobenzoate 3 was synthesized
by transesterification of myo-inositol using DMSO
as the solvent.[21,22] Distilling
off the formed methanol (for example, by carrying out the reaction
in a rotary evaporator) shortened the reaction time. Initial attempts
to carry out the reaction in DMF at the usual temperatures for orthoformate 1 and orthoacetate 2 synthesis (100 °C)
gave very low yields, and higher temperatures (>140 °C) were
required.[21,22,29] When DMF was
used a large excess of both the acid catalyst (p-toluenesulfonic
acid) and trimethyl orthobenzoate was needed, perhaps due to thermal
decomposition. Use of DMSO as the reaction solvent also gives a much
cleaner reaction, making the purification easier.Acid hydrolysis
of 3 with aqueous TFA gave 2-O-benzoate 6, once more with complete selectivity and in quantitative
yield. Again, this reaction was extremely fast. However, when alcohols
such as ethanol or methanol were used as the solvent, along with longer
reaction times, selectivity was reduced and 1- or 3-O-benzoyl-myo-inositol (±14) was
also obtained as a minor product (Scheme 4).
Scheme 4
Acid Hydrolysis of 3 under Various Conditions
Investigation into the
Possibility of Acyl Migration
While we believed that acyl
migration could not be a result of the
high regioselectivity in the formation of 2-O-ester
products, further studies were undertaken to establish that 5 and 6 are not products of acyl migration of
their 1- or 3-O-acyl myo-inositol
regioisomers. We therefore tested the migratory abilities of the benzoyl
group in both 6 and (±)-14 under a
variety of reaction conditions.Both the (±)-14 and 6 regioisomers were tested individually for the
migratory ability of their benzoyl group under the same reaction conditions
used for the acid hydrolysis of orthoesters 2 and 3 (i.e., TFA/H2O 10:1). However, no products due
to migration were obtained even after extended reaction times, proving
that 6 is not a migratory product from (±)-14. Partial migration was observed only at elevated temperatures
and with much longer reaction times. For example, when (±)-14 was treated in 1 M HCl and ethanol at 80 °C for 10
h, a 6.5:1 mixture of (±)-14 and 6 was
obtained, while treatment of 6 under the same reaction
conditions gave a 4:1 mixture of 6 and (±)-14, respectively. Therefore, since more of the ester 6 is converted to (±)-14, the equatorial
benzoyl ester (±)-14 may be the thermodynamically
more stable product while the axial regioisomer 6 is
the kinetic product.
Role of a 1,2-bridged Dioxolanylium Ion Intermediate
in Orthoester
Opening
After establishing that 2-O-benzoate 6 is not a migratory product of acid hydrolysis of orthobenzoate 3, further NMR experiments were undertaken in an attempt to
observe the intermediates involved in the hydrolysis reaction. To
our surprise, the 1HNMR spectrum taken immediately after
the treatment of 3 with deuterated TFA showed a new species
along with product 6, the latter presumably arising from
adventitious water. (See Figure 3 in the Supporting
Information for a full spectrum. Expansions are shown in Figure 1.)
Figure 1
Proton NMR of the 1,2-bridged intermediate (±)-15 (labeled as I) and product 6 (labeled as P):
(A) expansion
of 3.90–6.45 ppm; (B) expansion of 7.40–8.50 ppm.
Proton NMR of the 1,2-bridged intermediate (±)-15 (labeled as I) and product 6 (labeled as P):
(A) expansion
of 3.90–6.45 ppm; (B) expansion of 7.40–8.50 ppm.No residual starting material
was observed, indicating complete
conversion to this novel intermediate, later identified as the 1,2-bridged
2′-phenyl-1′,3′-dioxolan-2′-ylium ion
(dioxolenium ion) (±)-15 (Scheme 5) possessing characteristic downfield signals at 6.12 and
6.29 ppm, corresponding to H-1 and H-2 of the inositol ring protons.
H-1 and H-2are strongly deshielded relative to the other inositol
ring protons, and the coupling constant between them is unusually
large (J1–2 9.0 Hz, while J1–2 1.9 Hz for orthobenzoate 3) owing to the incorporation of C-1 and C-2 into the five-membered
dioxolanylium ring. The signals corresponding to the phenyl ring protons
of the dioxolanylium ionare clearly distinguishable from those in
the starting material or product, being more deshielded [δH 8.40–8.37 (2H/Ar-ortho), 8.19–8.15
(1H/Ar-para) and 7.82–7.78 (2H/Ar-meta)] compared to those in 3 [δH 7.68–7.66 (2H) and 7.40–7.38 (3H)] and product 6 [δH 8.03–8.01 (2H/Ar-ortho), 7.74–7.70 (1H/Ar-para) and 7.54–7.50
(2H/Ar-meta)]. In addition, fully in line with the
values reported for similardioxolanylium ions and α,α-dialkoxybenzyl
cations,[30−32] the 13CNMR spectrum of (±)-15 also showed a signal attributable to C-2′ of the
dioxolanylium ion at 183.3 ppm (Supporting Information, Figure 4).
Scheme 5
Acid Hydrolysis of Orthobenzoate Derivatives Proceeds
Regioselectively
via a 1,2-Bridged Intermediate
It is likely that the dioxolanylium ion intermediate (±)-15 formed from myo-inositol orthobenzoate
(3) is more stable than the corresponding intermediate
for opening of myo-inositol orthoacetate (2). This would account for our finding that (±)-15 can readily be observed by 1H and 13CNMR.
Therefore, we went on to synthesize the 4,6-di-O-methyl
(16) and 4,6-di-O-benzyl (19) derivatives of 3, keeping the C-2hydroxyl group free
to allow formation of 1,2-bridged intermediate for further studies.Thus, protection of the C-2hydroxyl group of 3 as
its TBDMSether followed by C-4 and C-6 methylation afforded the compound 23 in good yield (Scheme 6). Finally,
removal of the TBDMS group gave the required 4,6-di-O-Me orthobenzoate 16.
Scheme 6
Synthesis of myo-Inositol 4,6-Di-O-methyl 1,3,5-Orthobenzoate 16
Treatment of 16 with deuterated TFA showed a similardioxolanylium ion intermediate by 1HNMR for a second time
with immediate consumption of starting material to give the 1,2-bridgeddioxolanylium ion 17 (Scheme 5) comprising downfield signals at 5.92 and 6.16 ppm, representing
strongly deshielded inositol ring protons H-1 and H-2 with a large J1–2 value of 8.9 Hz (Supporting Information, Figure 5). 13CNMR spectra
further confirmed the structure of the intermediate ion (±)-17 with a signal at 182.9 ppm corresponding to the carbocation
(Supporting Information, Figure 6). A trace
amount of 4,6-di-O-methyl 2-O-benzoate
product 18 was also observed along with the dioxolanylium
ion 17.Chelation-controlled regioselective protection
of C-4 and C-6hydroxyl
groups of 3 was achieved using benzyl bromide to afford
4,6-di-O-benzyl orthobenzoate 19 in
high yield. When 19 was treated with deuterated TFA,
along with hydrolyzed product di-O-benzyl 2-O-benzoate 21, a 1,2-bridgeddioxolanylium
ion intermediate 20 was also observed. Dioxolanylium
ion 20 exhibited spectroscopic data analogous to 15 and 17 with 1HNMR signals at 6.02
and 6.25 ppm and 13CNMR signal at 182.3 ppm. The aromatic
protons of the dioxolanylium ion were also clearly noticeable since
they were more deshielded in comparison to those in the starting material
or product. Vicinal proton proton coupling constants (3JH,H) are compatible with slightly twisted
boatconformation (see the SI for a computational
molecular dynamics study of dioxolanylium ion 20).Although it was not possible to observe the initial acid-catalyzed
opening of the orthobenzoatecage owing to the rapidity of the reaction,
it is most likely that the more stable 1,2-bridgeddioxolanylium ion
is produced by rearrangement of a 1,3-bridgeddioxanylium (dioxenium)
ion intermediate, when the 2-hydroxyl group in the starting material
is free for instant rearrangement. Once the 2-hydroxyl group is protected,
and there is therefore no possibility of rearrangement, only 1- or
3-O-subsituted product is obtained,[21] providing further evidence of the involvement of a 1,3-bridgeddioxanylium intermediate in the initial stages of the acid hydrolysis.
Therefore, 2,4,6-tri-O- methylated, 2,4,6-tri-O-benzylated, and 2,6-di-O-benzylated orthobenzoates
were synthesized to facilitate NMR experiments, with the anticipation
that the initial opening of the orthobenzoatecage might be observed
due to the absence of a 2-hydroxyl, thus preventing the formation
of any 1,2-bridged intermediates or even a potential transient 1,2,3-cage
structure.Conventional methylation of 3 afforded
the 2,4,6-tri-O-methyl orthobenzoate 24. However, when 24 was treated with deuterated TFA,
signals in the 1HNMR spectrum were broadened suggesting
a rapid equilibration. Inositol
ring signals [δH 5.47 (2H), 4.55–4.52 (2H),
4.37–4.36 (2H)] were more deshielded in comparison to those
in the starting material [δH 4.58–4.56 (3H),
4.28–4.26 (2H), 3.67 (1H)] in just CDCl3 (Supporting Information, Figure 7). The signals
corresponding to the phenyl ring protons [δH 8.09
(2H/Ar-ortho), 7.87 (1H/Ar-para)
and 7.63–7.59 (2H/Ar-meta)] were also more
deshielded, with broadened peaks for ortho and para protons from those in the starting material 24 [δH 7.65–7.63 (2H) and 7.34–7.31
(3H)]. Broadening of the signals in the 13CNMR spectrum
was also significant, along with the disappearance of some signals
corresponding to inositol ring carbons and aromatic ortho and para carbons as well as the orthoestercarbon
(O3CPh), again consistent with a dynamic
equilibrium (Supporting Information, Figure
8).2,4,6-Tri-O-benzyl 1,3,5-orthobenzoate 25 was also obtained from 3, and treatment of 25 with deuterated TFA again showed broad signals in the 1HNMR spectrum with deshielded inositol and aromatic protons,
especially
H-1 and H-3 at 5.26 ppm instead of 4.52 ppm and ortho and para protons of orthobenzoate at 8.15 ppm and
8.00 ppm respectively instead of 7.23–7.69 ppm in the starting
material 25, indicative of a rapid equilibration (Supporting Information, Figure 9). Significant
broadening and disappearance/reduction of some signals corresponding
to inositol ring carbons and the aromatic region of the orthobenzoate
were also clearly distinguishable in the 13CNMR spectrum
from those in 25, again demonstrating a dynamic equilibrium
(Supporting Information, Figure 10).Treatment of 26(21) with
deuterated TFA yet again showed broad signals in the 1HNMR spectrum, with deshielded inositol and aromatic protons along
with broadened or lost signals in the 13CNMR spectrum,
once more suggesting a rapid equilibration. This may be due to reversible
opening of the orthobenzoatecage[33] giving
a rapidly interconverting mixture of starting material (24, 25, and 26) and the respective 1,3-dioxan-2-ylium
ion (27, 28, and 29) (Scheme 7). Nonetheless, addition of water to this equilibrium
mixture did not give rise to the expected hydrolysis product, but
instead resulted in ring closure to reform the starting material.
This presumably results from destabilization of the dioxanylium intermediate
due to dilution of the acid concentration, and demonstrates that hydrolysis
is the rate limiting step of this reaction. Since the departing hydroxyl
group remains in close proximity to the cationiccenter after ring-opening,
the intramolecular ring closure will be much faster for a such 1,3-bridgeddioxanylium ion intermediate than addition of water.[34,35] Although in principal, the orthobenzoatecage could also be opened
to give dioxanylium ions bridged between O-5 and O-1/3 of the inositol
ring in the initial stage of acid hydrolysis, only 1/3-O-benzoate ester products were obtained from acid hydrolysis of all
2,6-di and 2,4,6-tri-O-benzylated and 2,4,6-tri-O-methylated orthobenzoates 26, 25, and 24, respectively. However, on one occasion acid
hydrolysis of 25 in TFA and DCM, gave a minor product 31 in 8% yield where a trifluoro acetyl group was substituted
at the C-5 position along with the 1- or 3-O-benzoyl
major product 30 (Scheme 8). This
unexpected acylation at the C-5 position and not on the C-1 or C-3
positions also implies that the orthobenzoatecage may only be opened
to give a bridge between O-1 and O-3, leaving the 5-hydroxyl group
free for esterification by solvent. It could be that O-5 may perhaps
be more easily protonated since O-1 and O-3are more hindered due
to the equatorial C-2 substituent.
Scheme 7
Reversible Opening of C-2 Hydroxyl-Protected Orthobenzoates
Scheme 8
Acid Hydrolysis of 2,4,6-Tri-O-benzyl Orthobenzoate 25 in TFA/DCM; 1:1
This may, in addition, be due
to the fact that symmetrical 1,3-bridged
2′-phenyl-1′,3′-dioxan-2′-ylium ions 27, 28, and 29 are thermodynamically
more stable, since conformationally all four hydroxyl groups of the
twisted boatcould attain a less-hindered equatorial orientation in
comparison to an alternative 1,5- or 3,5-bridged dioxanylium ion intermediate
that would possess two equatorial and two axial hydroxyl groups.Therefore, from the above observations we can postulate that the
mechanism of acid hydrolysis of orthobenzoate 3 takes
place via initial reversible ring-opening giving the 1,3-bridged 2′-phenyl-1′,3′-dioxan-2′-ylium
ion 32 (indirectly observed as broad signals in the 1H and 13CNMR spectra of compounds with the 2-OH
protected). This symmetrical, six-membered intermediate then rearranges
immediately to the more stable 1,2-bridged 2′-phenyl-1′,3′-dioxolan-2′-ylium
ion (±)-15, which is then followed by the rate-determining
attack by water, presumably from the less hindered face, to provide
the hemiorthoester (±)-33 (Scheme 9). Subsequent decomposition of (±)-33, under
stereoelectroniccontrol,[36−38] affords the product with an axial
benzoate ester and equatorial hydroxyl groups (6). However,
for substrates in which the 2-hydroxyl group is protected, rearrangement
to the presumed five membered dioxolanylium intermediate is not possible
and thus slow hydrolysis gives product (±)-35 with
the benzoate ester at O-1 or O-3, via a 1,3-bridgedhemiorthoester
intermediate 34.
Scheme 9
Proposed Mechanism for the Acid Hydrolysis
of Orthobenzoate 3
Acid Hydrolysis of myo-Inositol 1,3,5-Orthoformate
(1)
Orthoformate 1 was also explored
under the selective conditions used in opening orthoesters 2 and 3 onto the C-2 position. Orthoformate 1 was prepared using literature procedures.[39] It should be noted that in the 1HNMR spectrum of the
orthoformate, the signal that corresponds to the methylidyne proton
is a small doublet of 1.1 Hz due to a 5 bond long-range spin coupling
with C-2-H.An 1HNMR spectrum taken immediately
after treatment of 1 with deuterated TFA and CDCl3 showed only the starting material and no intermediate species
or hydrolyzed products. When the same NMR sample was analyzed after
being at room temperature for 15 h mainly the starting material was
seen, along with very small amounts of products. Also, no deuterated
starting material or products were seen even after 3 days. However,
when the reactions were carried out in deuterated TFA and CDCl3 with a drop of D2O, the spectrum taken 5 min after
the addition of D2O showed a 50:50 mixture of starting
material and products, though no other intermediates were seen. Among
the products, both the 2-O-formate and 1/3-O-formate products were present at a ratio of 2.8:1, the
2-O-formate product still being predominant. Analysis
of the same NMR sample after 30 min showed a starting material to
products ratio of 1: 9.3, with 2-O-formate product
to 1/3-O-formate product corresponding to a ratio
of 2.5:1.Acid hydrolysis of 1 was carried out
using 10:1 mixture
of TFA:water at room temperature for 15 min until all starting material
had been consumed and the resulting mixture was evaluated by NMR after
evaporation to establish product ratios. The 1HNMR spectrum
showed the presence of inositol (arising due to hydrolysis of the
formate ester group under the acidicconditions), 2-O-formate product and 1/3-O-formate product in a
ratio of 1:1.7:1.2.X-ray crystal structure of 7. Ellipsoids
are represented
at 30% probability.
Regioselective Synthesis
of myo-Inositol 2-O-Butanoate 9 and Mechanistic Investigations
Having studied orthoacetate 2, orthobenzoate 3, and orthoformate 1 opening under acid hydrolysis
and seeing their similarities and differences, we synthesized another
orthoester, this time with a longer alkyl chain to investigate further
deuterium incorporation and intermediate cation formation. The novel
orthoester, myo-inositol orthobutanoate (4), was prepared in high yield using analogous conditions for the
synthesis of orthoacetate 1 by treatment of myo-inositol with trimethyl orthobutyrate and a catalytic amount of
PTSA in DMF at 140 °C (this time due to shorter reaction time
than for orthobenzoate synthesis under the same reaction conditions).4 Was then treated with aqueous TFA to obtain myo-inositol 2-O-butanoate (7) in quantitative yield. Recrystallization of 7 in water
and methanol afforded long thin crystals for which an X-ray crystal
structure was obtained (Figures 2 and 3).
Figure 2
X-ray crystal structure of 7. Ellipsoids
are represented
at 30% probability.
Figure 3
Crystal packing diagram for compound 7 showing
the
extensive H-bonding network.
Crystal packing diagram for compound 7 showing
the
extensive H-bonding network.Therefore, from the above results, we can conclude that any
1,3,5-orthoester
apart from orthoformate 1 should selectively open to
give the axial 2-O-ester product under the above
acidic hydrolysis conditions as already established for orthoacetate 2 and orthobenzoate 3. This ties in well with
the rationalization offered by King and Allbutt[36] based on the differences in steric strain among the possible
transition states that fulfill the stereoelectronic requirements of
dialkoxycarbonium ion formation.Mechanistic investigations
of orthobutanoate 4 in
deuterated TFA again showed deuterium incorporation at the α-position.
However, unlike for orthobenzoate, only a minor species was seen by 1HNMR, perhaps attributed to a 1,2-bridged 2′-butyl-1′,3′-dioxolan-2′-ylium
ion intermediate akin to (±)-15 which could not
be fully identified due to the small amount present.Hence,
having observed α-methylene protons exchange for deuterium
in orthobutanoate 4, orthoacetate 2 and 12 using deuterated acid, we can postulate that the mechanism
of exchange proceeds via a ketene acetal 38 or (±)-40 or both (Scheme 10). Although, having observed only a small amount of deuterium exchanged
product and no such exchange in the starting material when tri-O-Bn orthoacetate 13 was treated with deuterated
acid, this suggests that when the C-2hydroxyl group is protected,
the equilibrium may lie quite far on the side of the 1,3-dioxan-2-ylium
ion 37 due to the rapid interconvertion between the starting
material 13 and the ion 37. Thus, when the
C-2hydroxyl group is free, the equilibrium being rapidly established
may lie greatly toward the side of the ketene acetal (±)-40 due to the immediate rearrangement of 1,3-dioxan-2-ylium
ion 37 to the1,2-bridged ion (±)-39.
Scheme 10
Proposed Mechanism for the Deuterium Exchange
Treatment of 4 in deuterated TFA
for 24 h followed
by the addition of water in the subsequent hydrolysis gave the 2-O-C(O)CD2CH2CH3myo-inositol as the major product (over 92%) with trace
amounts of 2-O-butanoate 7 and 2-O-C(O)CHDCH2CH3myo-inositol, perhaps arising due to adventitious water. Therefore,
any such alkyl orthoesters 36 could be treated in deuterated
acid under anhydrous conditions for the deuterium exchange to take
place before hydrolysis to yield 2-O-C(O)D2R ester products 44 with deuterated methylene group
at the α-position (Scheme 11). These
deuterium incorporated products could provide valuable intermediates
to many deuterium labeled synthetic analogues that could be used to
study biologically important metabolic pathways or chemical reactions.
Scheme 11
Proposed Mechanism for the Formation of α-Methyl Deuterated
Ester Products
We have also proven
that this deuterium incorporation only takes
place before the hydrolysis step and thus before the formation of
the hemiorthoester intermediate (±)-46 (Scheme 12). Therefore, once the products are formed deuterium
incorporation at the α-methylene position is not possible in
deuterated acid. To confirm this, 2-O-acetate and
butanoate products 5 and 7 were treated
in both deuterated TFA and CDCl3 and deuterated TFA and
D2O for 9 days and monitored by NMR spectroscopy for deuterium
exchange. No deuterium exchange was observed by 2HNMR
spectroscopy over 9 days; only decomposition and migration products
were seen by 1HNMR spectroscopy, further proving that
deuterium incorporation only takes place via the
ketene acetal (±)-40 and that no exchange takes
place once products are formed through a mechanism similar to that
occupy in migration of products involving a hemiorthoester intermediate
(±)-46.
Scheme 12
No Deuterium Incorporation after Formation
of Product
Synthesis of 2-O-Benzoyl myo-Inositol 1,3,4,5,6-Pentakisphosphate 9 and myo-Inositol 1,3,4,5,6-Pentakisphosphate 11
The efficient regioselective transformation observed
also facilitates exploitation in the synthesis of inositol polyphosphates
with 2-position substitutions that are of biological interest[23] as well as providing a proficient route to the
anticancer agent inositol pentakisphosphate 11.[24,25] Thus, 2-O-benzoyl myo-inositol 6 was phosphitylated using N,N-diethyl-1,5-dihydro-2,4,3-benzodioxaphosphepin-3-amine in the presence
of 5-phenyltetrazole. The commonly used 1H-tetrazole,
which has potential explosive properties at room temperature, has
recently become difficult to obtain due to shipping restrictions and
we found that 5-phenyltetrazole is equally effective as an activator
in such phosphitylation reactions. The intermediate pentakisphosphite
was subsequently oxidized in situ by m-CPBA to afford
the symmetrical pentakisphosphate 47 in 96% yield (Scheme 13). 6 could also be phosphorylated
with the phosphitylating agent bis(benzyloxy)diisopropylaminophosphine
in good yield (94%) using 5-phenyltetrazole followed by oxidation.
After purification of the product, the phosphate protection was removed
by hydrogenolysis to afford pure 2-O-benzoyl pentakisphosphate 9. Finally, the cleavage of benzoate ester in concentrated
aqueous ammonia followed by simple removal of benzamide byproduct
after aqueous work up with dichloromethane provided the Ins(1,3,4,5,6)P511 as the ammonium salt in 86% isolated overall
yield from myo-inositol. Pentakisphosphate 11 could also be synthesized from 2-O-acetyl myo-inositol 5 or 2-O-butanoyl myo-inositol 7 in a similar method in high
yield. This sequence greatly benefits from involving only a single
chromatographic purification step thus, eliminating the need for the
usual ion-exchange chromatography and also allows easy access to multigram
scale quantities for in vivo studies. We anticipate that this route
can give efficient synthetic access to a range of 2-substituted inositolphosphate derivatives of potential biological interest and such work
is in progress.[23]
Scheme 13
Synthesis of 2-O-Benzoyl myo-Inositol
1,3,4,5,6-Pentakisphosphate 9 and myo-Inositol 1,3,4,5,6-Pentakisphosphate 11
Conclusions
In summary, we show
that acid hydrolysis of C-2 unprotected inositol-based
orthoesters, apart from the orthoformate, can lead to exclusive formation
of 2-O-acyl myo-inositol products.
This interesting regioselectivity is rationalized through the intermediacy
of a 1,2-bridged 2′-substituted-1′,3′-dioxolan-2′-ylium
ion that is observed by NMR spectroscopy and preceded by a 1,3-bridgeddioxanylium ion intermediate. Deuterium incorporation into the α-methylene
group at inositolC-2 of an alkyl ester (2-O-C(O)D2R) is possible when the corresponding alkyl orthoester is
treated in deuterated acid under anhydrous conditions, with deuterium
exchange taking place before hydrolysis. Furthermore, using these
observations, we describe a most efficient route for gram-scale synthesis
of Ins(1,3,4,5,6)P511 and 2-O-Benzoyl Ins(1,3,4,5,6)P59 via myo-inositol 1,3,5-orthobenzoate 3.
Experimental
Section
All reagents and solvents either were of commercial
quality or
were synthesized and purified in the laboratory using standard procedures.
Some solvents were redistilled and dried where necessary using standard
procedures or purchased in anhydrous form. Petroleum ether 40–60
°C is abbreviated as pet. ether. 1HNMR and 13CNMR chemical shifts were measured in ppm (δ) relative to
internal tetramethylsilane (TMS), and 31PNMR chemical
shifts are measured in ppm (δ) relative to phosphoric acid as
an external standard. Signals are expressed and abbreviated as s (singlet),
d (doublet), t (triplet), q (quartet), m (multiplet), br (broad),
and app (apparent). All 1HNMR and 13CNMR assignments
are based on gCOSY, gHMQC, gHMBC, and DEPT experiments. Coupling constants
(J) are given in hertz. TLC was performed on precoated
plates (aluminum sheets, silica gel 60 F254) with detection
by UV light or with phosphomolybdic acid in ethanol followed by heating.
Flash chromatography was performed on silica gel (particle size 40–63
μm).
2-O-Acetyl myo-Inositol (5)
A mixture of TFA (10 mL) and water (1 mL) was
added to 2 (1.5 g, 7.3 mmol) prepared as described,[18] and the solution was stirred for 5 min after
which time TLC (ethyl acetate) indicated the complete conversion of
starting material (R 0.5) to a product (R 0.0). The reaction mixture was then coevaporated with water followed
by DCM in vacuo to obtain 5 (1.63 g, quantitative) as
a white solid: 1HNMR (400 MHz, D2O) δ
2.12 (3H, s, CH3), 3.25 (1H, t, J4,5 = J5,6 = 9.2 Hz, C-5-H), 3.57
(2H, t, J1,6 = J3,4 = 10.2 Hz, C-4-H and C-6-H), 3.65 (2H, dd, J1,2 = J2,3 = 2.7 Hz, C-1-H
and C-3-H), 5.38 (1H, t, C-2-H); 13CNMR (100.6 MHz, D2O) δC 20.4 (q, CH3), 69.6 (d,
C-1 and C-3), 72.5 (d, C-4 and C-6), 74.1 (d, C-5), 74.3 (d, C-2),
173.8 (s, CO2CH3); HRMS (ESI-TOF) m/z [M + H]+ calcd for C8H15O7 223.0818, found 223.0812; HRMS
(ESI-TOF) m/z [M + Na]+ calcd for C8H14O7Na 245.0637, found
245.0632; 1HNMR (400 MHz, CF3CO2D and CDCl3) δ 2.25 (3H, s, CH3), 3.85
(1H, m, J = 7.4 Hz, C-5-H), 4.06–4.13 (4H,
m, C-1-H and C-3-H, C-4-H and C-6-H), 5.74 (1H, t, J1,2 = J2,3 = 2.4 Hz, C-2-H), 13CNMR (100.6 MHz, CF3CO2D and CDCl3) δC 20.2 (q, CH3), 70.4, 73.0
(2 × d, C-1 and C-3, C-4 and C-6), 73.8, 73.9 (2 × d, C-2
and C-5), 175.9 (s, CO2CH3).
Compound 13 was prepared as reported earlier:[28] mp
77.5–78.5 °C; 1HNMR (400 MHz, CF3CO2D and CDCl3) δ 1.54 (3H, s, CH3), 4.28 (1H, t, J = J = 1.6
Hz, C-2-H), 4.41 (2H, br s, Ins-H), 4.46–4.47 (3H, m, Ins-H),
4.64 (4H, AB, J = 11.7
Hz, C-4 and C-6CH2Ph), 4.76 (2H, s, C-2CH2Ph), 7.29–7.32 (4H, m, Ar-H),
7.40–7.46 (11H, m, Ar-H); 13CNMR (100.6 MHz, CF3CO2D and CDCl3) δC 23.2
(q, CH3), 67.6, 69.6, 72.1, 73.5 (2 × d, 2 ×
t, C-2, C-5, CH2Ph), 74.0, 74.4 (2 ×
d, C-1 and C-3, C-4 and C-6), 105.7 (s, O3CCH3), 129.3, 129.6 (2 × d, C-4 and C-6Ar-C and Ar-C), 129.7, 129.7, 129.8, 130.1 (4
× d, Ar-C), 135.6 (s, C-2Ar-C), 136.9 (s, C-4 and C-6Ar-C); HRMS (ESI-TOF) m/z [M + H]+ calcd for C29H31O6 475.2121,
found 475.2115.
myo-Inositol 1,3,5-Orthobenzoate
(3)
Trimethyl orthobenzoate (10 mL, 55 mmol)
was added to
a suspension of oven-dried myo-inositol (9.0 g, 50
mmol) and camphorsulfonic acid (232 mg, 1.00 mmol) in anhydrous DMSO
(30 mL). The resulting mixture was then heated at 60–80 °C
under a pressure of 260–280 mbar for 3 h on a rotary evaporator,
after which time TLC (ethyl acetate) indicated the complete consumption
of starting material (R 0.0) and the formation of a major product (R 0.4). The resulting clear solution was allowed
to cool, and the catalyst was neutralized by addition of triethylamine
(1.0 mL). The reaction mixture was concentrated under reduced pressure,
redissolved in hot ethyl acetate (500 mL), filtered through a pad
of silica gel, and washed further with hot ethyl acetate (2 ×
250 mL). The resulting filtrate was concentrated under reduced pressure
to about a volume of approximately 100 mL, and the solution was left
in the refrigerator overnight. Concentration of the mother liquor
and cooling gave a further crops of crystals to give a total yield
of 12.2g (92%) of colorless crystals of 3: mp 213-214
°C (ethyl acetate); 1HNMR (400 MHz, (CD3)2SO) δ 4.11 (1H, dt, J1,2 = J2,3 = 1.9 Hz and J = 6.3 Hz, C-2-H), 4.18–4.20
(2H, m, C-1-H and C-3-H), 4.23–4.25 (IH, m, C-5-H), 4.42–4.45
(2H, m, C-4-H and C-6-H), 5.38 (1H, d, C-2-OH), 5.57 (2H, d, J = 6.2 Hz, C-4-OH and C-6-OH),
7.34–7.41 (3H, m, Ar-H), 7.56–7.61 (2H, m, Ar-H); 13CNMR (100.6 MHz, (CD3)2SO) δC 58.2 (d, C-2), 67.7 (d, C-4 and C-6), 70.6 (d, C-5), 76.3
(d, C-1 and C-3), 106.9 (s, O3C-Ar), 126.0
(d, Ar-C), 128.0 (d, Ar-C), 129.5 (d, Ar-C), 138.3 (s, Ar-C); HRMS
(ESI-TOF) m/z [M + H]+ calcd for C13H15O6 267.0869, found
267.0863. Anal. Calcd for C13H14O6 (266.25): C, 58.64; H, 5.30. Found: C, 58.80 H, 5.33.
2-O-Benzoyl myo-Inositol (6)
A mixture of TFA (10 mL) and water (1 mL) was
added to 3 (1.5 g, 5.6 mmol), and the solution was stirred
for 3 min after which time TLC (ethyl acetate) indicated the complete
conversion of starting material (R 0.4) to a product (R 0.0). The reaction mixture was then coevaporated with water
in vacuo to obtain 6 (1.6 g, quantitative) as a white
solid: mp 240–242 °C (ethanol/water); 1HNMR
(400 MHz, D2O) δ 3.26–3.30 (1H, m, C-5-H),
3.64–3.71 (4H, m, C-1-H, C-3-H, C-4-H and C-6-H), 5.57 (1H,
t, J1,2 = J2,3 = 2.7 Hz, C-2-H), 7.39–7.43 (2H, m, Ar-H), 7.54–7.58 (1H, m, Ar-H), 7.92 (2H, m, J = 8.2 Hz, J = 1.2 Hz, Ar-H); 13CNMR (100.6 MHz, D2O) δC 70.0, 73.0
(2 × d, C-1 and C-3, C-4 and C-6), 74.4 (d, C-5), 75.2 (d, C-2),
128.9 (d, Ar-C), 128.9 (s, Ar-C), 129.8 (d, Ar-C), 134.2 (d, Ar-C), 168.1 (s, CO2Ph); HRMS (ESI-TOF) m/z [M + H]+ calcd for C13H17O7 285.0974, found 285.0969. Anal. Calcd for C13H16O7 (284.26): C, 54.93; H, 5.67. Found: C,
54.90; H, 5.75.Data for 1,2-bridged 2′-phenyl-1′,3′-dioxolan-2′-ylium
ion intermediate (±)-15 observed by 1H and 13CNMR spectroscopy soon after the addition of
deuterated trifluoroacetic acid (0.5 mL) into a NMR sample tube containing
the dry myo-inositol 1,3,5-orthobenzoate 3 (50 mg, 0.19 mmol) and anhydrous deuteratedchloroform (0.2 mL).
A small amount of CDCl3 was used due to the difficulty
in attaining a lock signal in neat CF3CO2D: 1HNMR (400 MHz, CF3CO2D and CDCl3) δ 4.18 (1H, dd, J4,5 =
5.9 Hz, J5,6 = 10.2 Hz, C-5-H), 4.48 (1H,
app t, J3,4 = 5.1 Hz, C-4-H), 4.76 (1H,
dd, J2,3 = 3.5 Hz, C-3-H), 4.95 (1H, dd, J1,6 = 6.7 Hz, C-6-H), 6.12 (1H, dd, J1,2 = 9.0 Hz, C-1-H), 6.29 (1H, dd, C-2-H),
7.80 (2H, dd, J = 7.8 Hz, J = 8.6 Hz, Ar-H), 8.17 (1H, t, Ar-H), 8.38 (2H, dd, J = 1.2 Hz, Ar-H); 13CNMR (100.6 MHz, CF3CO2D and CDCl3) δC 69.6
(d, C-3), 71.5 (d, C-6), 73.9 (d, C-5), 74.9 (d, C-4), 87.1 (d, C-2),
90.2 (d, C-1), 115.9 (s, Ar-C), 131.2
(d, Ar-C), 134.8 (d, Ar-C), 143.5 (d, Ar-C), 183.3 (s, C-2′).2-O-Benzoyl myo-inositol (6) in CF3CO2D and CDCl3: 1HNMR (400 MHz) δ 4.08
(1H, t, J4,5 = J5,6 = 9.8 Hz, C-5-H), 4.33
(2H, dd, J1,2 = J2,3 = 2.7 Hz, J1,6 = J3,4 = 10.2 Hz, C-1-H and C-3-H), 4.46 (2H, t, C-4-H and
C-6-H), 6.12 (1H, t, C-2-H), 7.52 (2H, dd, J = 8.2 Hz, J = 7.8 Hz, Ar-H), 7.70–7.74 (1H, m, Ar-H), 8.02 (2H, dd, J = 1.2 Hz, Ar-H); 13CNMR (100.6 MHz) δC 71.0
(d, C-1 and C-3), 73.4 (d, C-4 and C-6), 74.3 (d, C-2 and C-5), 127.4
(s, Ar-C), 129.5 (d, Ar-C), 130.3 (d, Ar-C), 136.0 (d, Ar-C), 170.0 (s, CO2Ph).
Sodium
hydride (451 mg of a 60% dispersion in
oil, 11.3 mmol) was added portionwise to a solution of 3 (500 mg, 1.9 mmol) in dry THF (6 mL) at 0 °C. The resulting
mixture was stirred for 10 min, and methyl iodide (0.70 mL, 11.3 mmol)
was then added dropwise. Stirring was continued for a further 16 h,
after which time TLC (1:1, pet. ether/ethyl acetate) showed the complete
conversion of starting material (R 0.10) to a product (R 0.45). The reaction mixture was then diluted with ethyl acetate
(50 mL), washed with water (50 mL) and brine (50 mL), dried (MgSO4), and evaporated in vacuo. The resulting compound was purified
by column chromatography (pet. ether/ethyl acetate, 2:1) to afford 24 (562 mg, 97%) as a white solid: mp 121.5–122.8 °C
(ethyl acetate/hexane); 1HNMR (400 MHz, CDCl3) δH 3.48 (6H, s, C-4-OCH3 and C-6-OCH3), 3.53 (3H, s, C-2-OCH3), 3.67 (1H, br s, C-2-H),
4.26–4.28 (2H, m, C-4-H and C-6-H), 4.56–4.58 (3H, m,
C-1-H, C-3-H and C-5-H), 7.31–7.34 (3H, m, Ar-H), 7.63–7.65
(2H, m, Ar-H); 13CNMR (100.6 MHz, CDCl3) δC 56.8 (q, C-2-OCH3), 57.8 (q,
C-4-OCH3 and C-6-OCH3), 68.3, 68.3 (2 × d, C-2 and C-5), 70.6, 76.0 (2 ×
d, C-1 and C-3, C-4 and C-6), 107.8 (s, O3CPh), 125.3, 127.9 (2 × d, Ar-C and Ar-C), 129.3 (d, Ar-C), 136.9 (s, Ar-C); HRMS (ESI-TOF) m/z [M + H]+ calcd for C16H21O6 309.1338, found 309.1333. Anal.
Calcd for C16H20O6 (308.33): C, 62.33;
H, 6.54. Found C, 62.30; H, 6.43. 1HNMR (400 MHz, CF3CO2D and CDCl3) δH 3.62
(6H, s, C-4-OCH3 and C-6-OCH3), 3.63 (3H, s,
C-2-OCH3), 4.36 (2H, br s, Ins-H), 4.53 (1H, br d, Ins-H),
5.47 (2H, br s, Ins-H), 7.61 (2H, br t, Ar-H), 7.87 (1H, br s, Ar-H), 8.09
(2H, br s, Ar-H).
Sodium hydride (3.15
g of a 60% dispersion in oil, 78.9 mmol) was added portionwise to
a solution of 3 (3.5 g, 13.1 mmol) in dry DMF (20 mL)
at 0 °C. The resulting suspension was stirred for 10 min, and
benzyl bromide (9.4 mL, 78.9 mmol) was added. Stirring was continued
for a further 2 h, after which time TLC (ethyl acetate) showed the
complete conversion of starting material (R 0.4) to a product (R 0.7), and the excess sodium hydride was destroyed
by the dropwise addition of methanol. The solvents were removed under
reduced pressure and the residue was dissolved in dichloromethane
(200 mL), washed with water (200 mL) and brine (200 mL), dried (MgSO4), and evaporated in vacuo. The resulting compound was purified
by column chromatography (hexane/ethyl acetate, 4:1) to afford 25 (6.7 g, 95%) as a white solid: mp 84–85 °C
(ethyl acetate); 1HNMR (400 MHz, CDCl3) δ
4.13 (1H, t, J = J = 1.6 Hz, C-2-H), 4.48
(2H, t, J = J = J = J = 3.9 Hz, C-4-H and C-6-H), 4.52 (2H, dd, J = J = 1.6 Hz, C-1-H and C-3-H), 4.53 (2H, d, J = 11.5 Hz, part of a AB
system of C-4 and C-6CH2Ph), 4.57–4.59
(1H, m, C-5-H), 4.66 (2H, d, J = 11.5 Hz, part of a AB system of C-4 and C-6CH2Ph), 4.70 (2H, s, C-2CH2Ph), 7.23–7.38 (16H, m, Ar-H), 7.41–7.43 (2H, m, Ar-H),
7.67–7.69 (2H, m, Ar-H); 13CNMR (100.6 MHz, (CDCl3) δC 66.1 (d, C-2), 69.0 (d, C-5), 71.2 (t,
C-2CH2Ph) 71.6 (t, C-4 and C-6CH2Ph), 71.9 (d, C-1 and C-3), 74.0 (d, C-4 and
C-6), 107.8 (s, O3CAr), 125.3 (d, O3CAr-C), 127.6, 127.7, 127.8,
127.9, 128.0, 128.2, 128.4 (7 × d, Ar-C), 129.4
(d, O3CAr-C), 137.1 (s,
C-2Ar-C), 137.6 (s, C-4 and C-6Ar-C), 138.0 (s, O3CAr-C); HRMS (ESI-TOF) m/z [M + H]+ calcd for C34H33O6 537.2277, found 537.2272.
1HNMR (270 MHz, CF3CO2D and CDCl3) δ 4.47–4.49 (3H,
m, C-1-H, C-3-H
and C-5-H), 4.51–4.54 (1H, m, C-2-H), 4.80 (2H, dd, J = 3.8 and 4.2 Hz, C-4-H and C-6-H), 5.64 (1H, d, J = 1.1 Hz, O3CH); 13CNMR (100.6
MHz, CF3CO2D and CDCl3; peaks referenced
to CDCl3) δC 61.4, 67.7 (2 × d, C-2
and C-5) 67.7, 73.6 (2 × d, C-1 and C-3, C-4 and C-6), 102.7
(d, O3CH); HRMS (ESI-TOF) m/z [M + H]+ calcd for C7H11O6 191.0556, found 191.0550; HRMS (ESI-TOF) m/z [M + Na]+ calcd for C7H10O6Na 213.0375, found 213.0370.A mixture of TFA (2 mL) and water (0.2 mL)
was added to 1 (100 mg, 0.53 mmol), and the solution
was stirred for 15 min after which time TLC (ethyl acetate) indicated
the complete conversion of starting material (R 0.5) to products (R 0.0). The reaction mixture was then coevaporated
with water in vacuo to obtain a white solid containing a mixture of
2-O-formyl myo-inositol, 1/3-O-formyl myo-inositol, and myo-inositol in a ratio of 1.7:1.2:1, respectively: 1HNMR
(400 MHz, D2O) δ 3.11 (1H, t, J4,5 = J5,6 = 9.4 Hz, inositolC-5-H),
3.17 (1H, t, J4,5 = J5,6 = 9.4 Hz, 2-O-formyl inositolC-5-H),
3.22 (1H, t, J4,5 = J5,6 = 9.4 Hz, 1/3-O-formyl inositolC-5-H),
3.36 (2H, dd, J1,2 = J2,3 = 2.7 Hz, J1,6 = J3,4 = 9.8 Hz, inositolC-1-H and C-3-H), 3.43–3.53
(6H, m, inositolC-4-H and C-6-H, 2-O-formyl inositolC-4-H and C-6-H, 1/3-O-formyl inositolC-1/3-H and
C-4/6-H), 3.59 (2H, dd, J1,2 = J2,3 = 2.8 Hz, J1,6 = J3,4 = 10.2 Hz, 2-O-formyl inositolC-1-H and C-3-H), 3.70 (1H, t, J = 9.4 Hz, 1/3-O-formyl inositolC-4/6-H), 3.89
(1H, t, inositolC-2-H), 4.03 (1H, t, J1,2 = J2,3 = 2.2 Hz, 1/3-O-formyl inositolC-2-H), 4.71 (1H, dd, J = 2.2 Hz
and J = 10.2 Hz, 1/3-O-formyl inositolC-1/3-H), 5.28 (1H, app s, 2-O-formyl inositolC-2-H); 13CNMR (100.6 MHz, D2O) δC 69.3
(d, 2-O-formyl inositolC-1 and C-3), 69.8 (d, 1/3-O-formyl inositolC-2), 70.0 (d, 1/3-O-formyl
inositolC-4/6), 70.5 (d, 1/3-O-formyl inositolC-3/4
or C-1/6), 71.0 (d, inositolC-1 and C-3), 72.0, 72.1 (2 × d,
inositolC-2 and 1/3-O-formyl inositol C-3/4 or C-1/6),
72.3 (d, inositolC-4 and C-6), 72.4 (d, 2-O-formyl
inositolC-4 and C-6), 73.7 (d, 1/3-O-formyl inositolC-1/3), 73.8 (d, inositolC-5), 74.0 (d, 2-O-formyl
inositolC-5), 74.2 (d, 1/3-O-formyl inositolC-5),
74.9 (d, 2-O-formyl inositolC-2), 162.9, 163.5 (2
× s, 2-O-formyl inositolCO2H and 1/3-O-formyl inositolCO2H); HRMS (ESI-TOF) m/z [M + H]+ calcd for 1/3/2-O-formyl myo-inositol, C7H13O7 209.0661,
found 209.0656; HRMS (ESI-TOF) m/z [M + Na]+ calcd for C7H12O7Na 231.0481, found 231.0475.
myo-Inositol
1,3,5-Orthobutanoate (4)
Trimethyl orthobutyrate
(1.0 mL, 6.11 mmol) was added
to a suspension of oven-dried myo-inositol (1.0 g,
5.55 mmol) and PTSA (316.7 mg, 1.67 mmol) in anhydrous DMF (10 mL),
and the resulting mixture was heated at 140 °C for 3 h. Extra
trimethyl orthobutyrate (0.5 mL, 3.05 mmol) was then added and heated
for a further 15 min at 140 °C, after which time TLC (ethyl acetate)
indicated the complete consumption of starting material (R 0.0) and the formation of a major product
(R 0.5). The resulting
solution was allowed to cool and the catalyst was neutralized by addition
of triethylamine (0.22 mL). The reaction mixture was then concentrated
under reduced pressure, and the resulting residue was purified by
column chromatography (ethyl acetate) to afford 4 (1.12
g, 87%) as a white crystalline solid: mp 139–140 °C (ethyl
acetate); 1HNMR (400 MHz, CD3OD) δ 0.88
(3H, t, J = 7.4 Hz, O3C(CH2)2CH3), 1.41–1.51 (2H,
m, O3CCH2CH2CH3), 1.58–1.62 (2H, m, O3CCH2CH2CH3), 4.08–4.11 (4H,
m, C-1-H, C-2-H, C-3-H and C-5-H), 4.35–4.37 (2H, m, C-4-H
and C-6-H); 13CNMR (100.6 MHz, CD3OD) δC 14.4 (q, O3C(CH2)2CH3), 17.0 (t, O3CCH2CH2CH3), 40.4 (t, O3CCH2CH2CH3), 60.3 (d, C-2/C-5),
69.0 (d, C-4 and C-6), 70.7 (d, C-2/C-5), 76.5 (d, C-1 and C-3), 110.2
(s, O3C(CH2)2CH3); HRMS (ESI-TOF) m/z [M
+ Na]+ calcd for C10H16O6Na 255.0845, found 255.0839; HRMS (ESI-TOF) m/z [M + H]+ calcd for C10H17O6 233.1025, found 233.1020. Anal. Calcd for C10H16O6 (232.23): C, 51.72; H, 6.94. Found: C,
51.70; H, 7.01.
2-O-Butanoyl myo-Inositol
(7)
A mixture of TFA (5 mL) and water (0.5 mL)
was added to orthobutanoate 4 (300 mg, 1.29 mmol), and
the solution was stirred for 5 min after which time TLC (ethyl acetate)
indicated the complete conversion of starting material (R 0.6) to a product (R 0.0). The reaction mixture was then
coevaporated with water in vacuo to obtain 7 (323 mg,
quantitative) as a white solid: mp 143–161 °C (methanol
and water); 1HNMR (400 MHz, D2O) δ 0.72
(3H, t, J = 7.4 Hz, CO2(CH2)2CH3), 1.42 (2H, sextet, J = 7.4 Hz, CO2CH2CH2CH3), 2.24 (2H, t, J = 7.4
Hz, CO2CH2CH2CH3), 3.10 (1H, t, J4,5 = J5,6 = 9.4 Hz, C-5-H), 3.41 (2H, t, J1,6 = J3,4 = 9.8 Hz, C-4-H
and C-6-H), 3.49 (2H, dd, J1,2 = J2,3 = 2.7 Hz, C-1-H and C-3-H), 5.22 (1H, t,
C-2-H); 13CNMR (100.6 MHz, D2O) δC 12.8 (q, CO2(CH2)2CH3), 17.8 (t, CO2CH2CH2CH3), 35.7 (t, CO2CH2CH2CH3), 69.6 (d, C-1
and C-3), 72.5 (d, C-4 and C-6), 74.0 (d, C-2), 74.2 (d, C-5), 176.3
(s, CO2(CH2)2CH3); HRMS (ESI-TOF) m/z [M
+ H]+ calcd for C10H19O7 251.1131, found 251.1125; HRMS (ESI-TOF) m/z [M + Na]+ calcd for C10H18O7Na 273.0950, found 273.0945. Anal. Calcd for C10H18O7 (250.25): C, 48.00; H, 7.25. Found: C,
47.90; H, 7.31.
Compound 47 (800 mg, 0.67 mmol)
was dissolved in methanol (60 mL), and water (15 mL) and 10% palladium
hydroxide on activated charcoal (400 mg) were added. The resulting
suspension was stirred at room temperature overnight under an atmosphere
of hydrogen in a hydrogenator. The catalyst was filtered through a
PTFE syringe filter, and the filtrate was evaporated under reduced
pressure to give 9 (460 mg, quantitative) as a hygroscopic
white foam: 31PNMR (109.4 MHz, H-decoupled, D2O) δ −0.02 (2P, s), 0.79 (2P, s), 1.09 (1P, s, phosphate
at C-5); 1HNMR (270 MHz, D2O) δ 4.34
(1H, ap. quartet, dt, J4,5 = J5,6 = JH,P = 9.4 Hz, C-5-H),
4.44 (2H, ddd, J1,2 = J2,3 = 2.7 Hz, J1,6 = J3,4 = JH,P = 9.4
Hz, C-1-H and C-3-H), 4.57 (2H, ap. quartet, ddd, JH,P = 9.4 Hz, C-4-H and C-6-H), 5.90 (1H, t, C-2-H), 7.32–7.37
(2H, m, Ar-H), 7.47–7.53 (1H,
m, Ar-H), 7.86–7.88 (2H, m,
Ar-H); 13CNMR (100.6
MHz, D2O) δC 71.5 (d, C-2), 72.5 (m, C-1
and C-3), 76.6 (m, C-4 and C-6), 76.9 (m, C-5), 128.5 (s, Ar-C), 128.8 (d, Ar-C), 129.9 (d, Ar-C), 134.2
(d, Ar-C), 166.9 (s, CO2Ph); HRMS (ESI-TOF) m/z [M + H]+ calcd for C13H22O22P5 684.9291, found 684.9285.Catalytichydrogenolysis
of compound 48 (4.0g, 2.52 mmol) was also carried out
as above in methanol (110 mL) and water (10 mL) with 10% palladium
hydroxide on activated charcoal (400 mg) to afford 9 (1.75
g) in quantitative yield. The free acid 9 was then converted
to the triethyl ammonium salt by neutralization with 1 M triethylammonium
bicarbonate buffer followed by coevaporation with methanol to remove
excess buffer.
myo-Inositol 1,3,4,5,6-Pentakisphosphate
(11)
Compound 9 (600 mg, 0.88 mmol)
was
dissolved in concentrated aqueous ammonia solution (30 mL) and heated
at 60 °C overnight in a Pyrex pressure tube. After evaporation
of solution under vacuum, the residue was dissolved in water and the
benzamide byproduct was removed by washing with dichloromethane to
afford the pure ammonium salt of 11 (600 mg) as a hygroscopic
white solid: 31PNMR (109.4 MHz, H-decoupled, D2O) δ 1.93 (2P, s), 2.98 (1P, s, phosphate at C-5), 3.69 (2P,
s); 1HNMR (270 MHz, D2O) δ 3.87–4.04
(3H, m, J = 9.4 Hz, C-1-H, C-3-H and C-5-H), 4.26
(2H, app quartet, ddd, J = 9.4 Hz, C-4-H and C-6-H),
4.51 (1H, br s, C-2-H); HRMS (ESI-TOF) m/z [M – H]− calcd for C6H16O21P5 578.8872, found 578.8878.
Anal. Calcd for hexaammonium salt of 11 C6H35N6O21P5 (682.24):
C, 10.56; H, 5.17; N, 12.32. Found: C, 10.30; H, 5.38; N, 12.40.The ammonium salt of 11 was converted into the free
acid by quick filtration [since prolonged exposure causes migration]
through Dowex H+ resin (10-fold excess, previously washed
with miliQ water) and then to its hexasodium salt by titration to
pH 7.42 with 0.1 M sodium hydroxide followed by lyophilization: 31PNMR (161.9 MHz, H-decoupled, D2O) δ 1.05
(2P, s), 1.18 (2P, s), 1.46 (1P, s, phosphate at C-5); 1HNMR (400 MHz, D2O) δ 3.93–3.97, 4.22–4.26
(6H, 2 × m, C-1-H, C-2-H, C-3-H, C-4-H, C-5-H and C-6-H).
Authors: Stuart J Conway; James Gardiner; Simon J A Grove; Melloney K Johns; Ze-Yi Lim; Gavin F Painter; Diane E J E Robinson; Christine Schieber; Jan W Thuring; Leon S-M Wong; Meng-Xin Yin; Antony W Burgess; Bruno Catimel; Phillip T Hawkins; Nicholas T Ktistakis; Leonard R Stephens; Andrew B Holmes Journal: Org Biomol Chem Date: 2009-11-09 Impact factor: 3.876
Authors: Kana M Sureshan; Mysore S Shashidhar; Thoniyot Praveen; Rajesh G Gonnade; Mohan M Bhadbhade Journal: Carbohydr Res Date: 2002-11-29 Impact factor: 2.104
Authors: Enza Piccolo; Sara Vignati; Tania Maffucci; Pasquale F Innominato; Andrew M Riley; Barry V L Potter; Pier Paolo Pandolfi; Massimo Broggini; Stefano Iacobelli; Paolo Innocenti; Marco Falasca Journal: Oncogene Date: 2004-03-04 Impact factor: 9.867
Authors: M Falasca; D Chiozzotto; H Y Godage; M Mazzoletti; A M Riley; S Previdi; B V L Potter; M Broggini; T Maffucci Journal: Br J Cancer Date: 2010-01-05 Impact factor: 7.640
Authors: Hayley Whitfield; Andrew M Riley; Soulla Diogenous; Himali Y Godage; Barry V L Potter; Charles A Brearley Journal: Plant Soil Date: 2017-06-27 Impact factor: 4.192
Authors: Hayley Whitfield; Gaye White; Colleen Sprigg; Andrew M Riley; Barry V L Potter; Andrew M Hemmings; Charles A Brearley Journal: Biochem J Date: 2020-07-31 Impact factor: 3.857