close

Вход

Забыли?

вход по аккаунту

?

Stereocontrolled Synthesis of Oligo(nucleoside phosphorothioate)s.

код для вставкиСкачать
REVIEWS
Stereocontrolled Synthesis of Oligo(nuc1eoside phosphorothioate)s
Wojciech J. Stec,* Andrzej Wilk
Encouraging results obtained for modulation of gene expression by antisense
oligonucleotides and their analogues
have kindled hopes for a new generation
of therapeutics against viral infections,
cancer, and many other diseases. Among
such analogues, oligo(nuc1eoside phosphorothioate)s (Oligo-S) have generally
shown the highest efficacy in inhibiting
the biosynthesis of "unwanted" proteins.
The first clinical trials of antisense agents
are now in progress using Oligo-S against
genital warts and acute myeloid leukemia, and tests of Oligo-S against
AIDS should follow soon. Nevertheless.
their mechanism of action, internalization, cellular trafficking, subcellular localization, and interaction with cellular
proteins is still poorly understood. It is
assumed a priori that application involves rapid and efficient molecular
recognition of target RNA by Oligo-S;
however, the effects of the chirality of
Oligo-S have so far been unappreciated,
because Oligo-S has not yet been synthesized with stereocontrol. Indeed. the
diastereomeric composition of Oligo-S
has never been determined, primarily
because of the lack of appropriate analytical methods. Since each of the
diastereomers is a stereochemically
unique chemical entity, questions arise
as to which diastereomer is responsible
for an observed biological response, including positive (curative) or possibly
negative (toxic) side effects.
In this review we intend provide a perhaps somewhat speculative assessment of
the problems associated with the stereocontrolled synthesis of Oligo-S and to
discuss the state-of-the-art in this field
including strategies that may lead to
Oligo-S of predetermined chirality. This
article is not intended to discourage researchers from further studies of diasteromeric mixtures of Oligo-S as potential pharmaceuticals. Throughout the
history of medicinal chemistry numerous useful medicines were discovered,
developed, and employed without the
detailed knowledge of their structure.
Indeed, the composition of the vaccines
discovered by Pasteur is a subject of vigorous study still today.
1. Introduction
B=
Since the discovery that Oligo-S 1 can effectively protect cells
from the lethal action of the HIV-1 virus,['] a number of reports
have appeared concerning improvements in their synthesis."]
their purification and analysis,[31cellular uptake, pharmacokinetics and
mechanism of action,[51and potential therapeutic
In the first enzymatic['] and chemicalLs1
syntheses of Oligo-S it was recognized that stereodefined 3'-0deoxyri bosyl and 5'-O-deoxyribosyl moieties surround the
phosphorus atom of each phosphorothioate unit, and because
of the stereogenicity of phosphorus, new centers of chirality are
created in each molecule of 1.
However, relatively little attention has been paid to the structural and biochemical consequences of this diastereomerism.
The aim of the present review is to focus the attention of researchers, who are applying Oligo-S to antisense technology191
["I
Prof. DV.W. J. Stec, Dr. A. Wilk
Polish Academy of Sciences
Centre of Molecular and Macromolecular Studies
Departnienl of Bioorgdnic Chemistry
Sienkiewicaa 112. PL-90-363 Lodz (Poland)
Telefax' l n t . code +(42)815483
N
F
'N
)
(Ade)
I
0
NHZ
bH
X=O Y = S
x = s -Y = O
(all-Rp)-l
(all-Sp)-1
HNv
A
O
(Thy)
N
I
Scheme 1. Oligo(nuc1eoside phosphorothi0ate)s 1 and bases B.
and/or ribozyme technology,"'l on the problem of the polydiastereomerism of I, and to critically evaluate the potential
strategies for stereocontrolled synthesis of Oligo-S leading
to pure products with defined chirality at each phosphorus
atom.
W. .I.Stec and A. Wilk
REVIEWS
2. Nonstereocontrolled Synthesis of Oligo(nuc1eoside
phosphorothi0ate)s
Antisense Oligo-S 1 are relatively short oligonucleotides (ca.
15- to 30-mers) with a defined sequence of nucleobases complementary to R N A (or ssDNA) in which each internucleotide
phosphate is replaced by a phosphorothioate moiety. Thc sequence of nucleobases between the 5'-OH and 3'-OH end groups
in these molecules is responsible for recognition of a coniplementary target RNA and specific alignment governed by Watson -Crick hydrogen bond formation." ' I
Oligo-S have been prepared until now by primarily two
methodologies: a ) stepwise phosphoroamidite coupling["] fol-
"'"B'
I
O\
P-N
3
bR3
4
-I
lowed by sulfurizationl" I 3 l of the intermediate O-alkylphosphite 5, and b) tandem sulfurization of ohgo(nuc1eoside Hphosphonate)s 12". I 41 bound to the solid support (Schemc 3 ) .
Sulfurization of either O.O,O-trialkylphosphites or O,O-dialkyl-H-phosphonates is itself stereoretentive.[' Assuming that
both methodologies arc neither stereospecific nor stereoselective
and that each 0-alkyl phosphite internticleotide bond ( a s in 5) or
each H-phosphonatc bond (as in 6) is formed in a nonstercoselective manner. the resulting Oligo-S consists of a mixture of 2"
diastercomers. where 11 is the number of phosphorothioate internucleotide bonds. The content of any single diastereomer in this
mixture should therefore be T " .HOWKVKI-,
even from the pioneering works of Burgers and Eckstein1'61and Marlier and
Benkovic["] on the synthesis the stereoisomers 12. U,,A and
A,,A. respectively. it was known that the quantities of
diastereomers (R,) and (S,) were not equimolar. By applying
phosphorodichloriditc coupling (by means of 9) followed by
sulfurization of phosphite 10. both groups independently obtained diastereomers of diribonuclcoside 3'3-phosphorothioates 12 (Scheme 3) and assigned the configuration at the phosphorus atom. Moreover. recent studies h a w shown that condensation of 5'-O-DMT-nucleoside 3'-0-(0-2-cyanoethy1N'N-diisopropy1phosphoramidite)s with 5'-OH nucleosidc
bound to a solid support through the 3'-0 atom gives after
deprotection the diastereomeric mixtures of (Rp)- and (Sp)dinucleoside 3',5'-phosphorothioates in the ratios listed in
the Table 1. (This condensation must be performed according
to the generally accepted protocol for the synthesis of oligonucleotides, with the exception that the oxidation step with
iodine is replaced by sulfurization.[*,'2g1) It was demonstrated
earlier["' that the diastereomeric composition of the starting 5'-U-DMT-.W-benzoylcytidinc 3'-0-(O-methyl-N',iV'-diisopropylphosphoroamidites) does not influence the ratio of the
resulting diastereomers of dicytidiiie 3'.5'-phospliorothioatc.
The composition of the mixture of diastereomers obtained
with (R,) diastereosekctivity s and (s,)diastereoselectivity j'
(J. = 1 .Y) preserved at each of ti couplings can be calculated
from the polynomial Equation (a). The following example is
R'ow'
1
1
Scheme 2 . Synthesis of I with the phosphoroamidite ( a ) and H-phosphondte methods (b). a ) 1. Sulfurization. 2 . capping of the unreacted 5'-OH group. 3. remo!...il of
the S-trityl croup. 4. condensation with 2:tetmrole. 5 steps 1 4 are repeated ii
times. b) 1 Removal of the S'-trityl group. 2. condensation with 3:piv~rloyl
(adamanroyl) chloride, 3 . capping of the unreacted 5'-OH group. 4. stcps 1 . 3 are
repeated n times. 5 . final sulfurization. For both methods the product is then cleaved
from the solid support. deprotected. and purified. R 1 = 4.4-dimethox) tritql
(DMT). R 2 = CH,CH,CN. Me: R3 = long chain ainine controlled pore glass
(LCA-CPG); B'.B2 = AdeBr, GuaiBu, CytB7. Thq.
~
calculated for the pentamer (e.g. d[(C,,),C]) obtained with diastereoselectivity .Y = 0.6: (x + J , ) ~= s4 ~ S ' J , +
+
+
REVIEWS
Olipo( nucleoside phosph0rothioate)s
I/
R'OlOOdl
U
8
7
10
0
~ 03 ~
b)
I
4
,
-r
"'"W'
R'O
-
I
\
OR2
Hoo7coJs"
- o\p,sI
1
1/
/0
04
4'
'
U
0
~ 03 ~
0
0"
0"
OH
U
on
N h
dA
dG
dC
dT
dA
dG
51.7
53 4
53.8
dC
57.0
5X.6
32.6
51 .I)
38.6
55 2
59.1
60.0
62.1
59.6
50.8
f -
(Rp):(Sp)
is 52:48 in the case of dimer formation (see Table 1)
but increases roughly 1 YOper coupling. Thus, the second coupling occurs with 53 O h (R,) selectivity, the third with 54 YO,etc.
If this trend can be extrapolated, the 49th coupling should be
100 o/o (R,) diastereoselective. Perhaps the most convincing
demonstration of the effect of the variable diastereoselectivity of
the coupling process on the final composition of diastereomers
is given in Figure2. Here it is evident that even during the
/I
56.0
[a] Dinitcleo\ide j'.j'-phosphorothio;ites (N,,N' \\here N is at the 5'-end) \+ere
prepared xcordins to the standard coupling method (DNA synthesizer. Applied
Biosysteni\ modcl 380 B). rollowed hy sulfuri7ation with B 0.08 M solution of bis(diisopropo\4.phospliitiothioyl)disulfidc in ace1onitrtlc:pyridine (2:l) for 45 s (ca.
400 p L ) , The ~'-diiiiethoxytri(ylgroup was rcmoved in the final step in the synthesis.
Crude mixtiire\ were analyzed by H P L C (Supelcosil RP-18 column (5 p.
2.3 iiini x 250 min). acetonitrile gradient in 0.1 M TEAB fi-om 0%. 0 . 5 " h
('H,CN n i i i i r ' . no\+ rate 1 m L i n i n ~ ' ) . T h e p c r c e n t o f t h e ( / ~ , , ) i r o m e r w a s o b t a i n e d
froin pe:ik intcgrii~ion.The coupling efficiency was n o t lower than >9X%.
+
19 20 21
OH
l'ahle I . Proportionz c>f the (R,,) isomer in dinucleoside 3'.S'-phosphorothioiites
(N,,sN')i i i pcrcenl [a]
dl
~_-
17 18
30
t -
Scheme 3 l h e phosphite incthod used by Eckstcin (a) and Benkovic (b) for the
syiitlie~isof diribonuclcoside 3'.5'-phosphorotliio;ites [lh. 171. a) R ' = p-chlorophenow):icct) I: R' = methoxvtetrahydropyranyl, RZ.R' = O-methoxymcthylidene: R' = phenql: €3' = B' = Ura: B' = .V6.ben7oyl-Ade: B' = Ade. b) R 1 =
methoxqtt-11~1:R' = (1-nitrohenzyl: R'.RJ = beiizoyl. R' = methyl: B' = A"bcn/oyI-Adc: B' = ,V".N"-dibcnryl-Ade, B ' = B' = Ade.
55.3
20
Fig. 1 . Chromatogram froin a RP-HPLC separation of the four diactereomers of
d(C,,C,.,C) froin the sample prepared by the phosphoroamidite Inelhod. For cxperimcntal details see Table 1. a ) Entire chromatozram of crude iiiiiteriiil: h) cxpanded
section showing the baseline separated peaks for the diastereomers. Retention time
r iii min.
\
\o
\o +O<"
,2
4
10
4 . y ~+
' ~ = 0.1296 (4 x 0.0864) + (6 x 0,0576) + (4 x 0.0384)
+ 0.0256. Thus the mixture of sixteen diastereomers (24) contains 1?.96u/, (all-R,). 2.56% (all-s,), four possible "(all-R,)
but one" diastereomers, each formed in 8.64 YOyield. six fragments with equal numbers of (R,) and (S,) internucleotide
bonds, cach with a yield of 5.76%. and four possible "(all-S,)
but one" diastereomers, each formed in 3.84% yield. It is worth
mentioning that in the above example the ratio of (all-R,) to
(all-&) diastereomers is 5 : 1 .
Indeed, it has been demonstrated roughly that the ratios
of diastereomers of d(C,,C,,,C) are in agreement with calculated values: (R,.~,):(~S,,,R,):(R,,S,):(S,,S,)
= (0.6 x 0.6):(0.6x
0.4):(0.4 x O.h):(0.4: x 0.4) (Fig.
This assignment also indicates that the diastereoselectivity of subsequent steps of the
coupling process is, in the case of cytidine components, nearly
the same. That means that in oligo(cytidine phosphorothioate)s
the ratio o f phosphorothioates with (R,) configuration at a particular position i n the chain to those with (S,) configuration will
be close to 60:40. Further experimental data indicate that for
oligo(deoxyadenosine phosphorothi0ate)s the relative ratio
I
,
I
0
5
10
15
20
25
30
35
-__
21
22
23
21
Fig. 2. RP-HPLC triice of CI ude trimer d(A,,A,,A) prepared and :inalyrcd a s indicated in Figure 1. Complete baseline separation of ( R e , & ) and (S,.R,) diajtereom e n (peaks in the middle of b) was not possible. Retention time i in min.
synthesis ofthe trimer d(A,,A,,A) the ratio of the diastereomers
obtained (R,,R,): (RprSp):(S,.R,): (S,,S,) is 29. I :26. I : 22.7 :
22.1 .[I81
3. Biological Implications of the Chirality of Oligo-S
What are the consequences of a nonstatistical distribution of
individual diastereomers in such constructs as I ? One must accept the fact that if antisense oligonucleotides 1 are considered
as potential therapeutics, their contact with proteins is unavoidable after introduction to the body. Recent results indicate that
Oligo-S like d[(C,,),C] with n 2 1 4 can inhibit human DNA
polymerase and RNase activities.['"] Therefore. the most problematic consequence of varying the distribution of particular
diastereomers of I may be the batch-to-batch irreproducibility
of biological effects resulting from the different activity of individual diastereomers towards the variety of proteins in the inter71 I
W. J. Stec and A. Wilk
REVIEWS
and intracellular media. From the work initiated by studies of
Eck~tein.~"]Benkovic et a1..[201and FreyL2']it is known that
functional proteins such as polymerases and nucleases interact
with phosphorothioate analogues of biophosphates in a diastereoselective manner.[221However, the diastereoselectivity is
not predictable. since for each enzyme it results from different
spatial arrangements of functional groups in proteins (including
orientations of peptide bonds), which can interact with phosphate
oxygen atoms by means of hydrogen bonds and salt bridges.[231
Recently it has been hypothesized that the diastereoselectivity of
enzyme-catalyzed processes in the degradation of oligonucleotides bearing phosphorothioate functions results from directional two-point interactions between the phosphorothioate
unit and the protein (Scheme 4).[22f1This hypothesis is well
NH:
NH:
obtained with diastereomeric mixtures of 1. [ 4 - h 1 Although the
legal regulations[z61for a new potential drug insist on the precise
stereochemical description of the molecule when and if possible,
scholarly attitude alone requires chemists to work out the ideal
solution-to design a methodology for the synthesis of 1 in a
stereocontrolled manner which leads to a single desired diastereomer of l .
4. Stereocontrolled Chemical Synthesis
of Oligo(nuc1eotide phosphorothioate)s
One approach to the stereocontrolled synthesis of organic molecules is to use enzymes. For the synthesis of l polymerases,
transferases, and nucleases can be considered,[*'] since these
classes of enzymes can assist in the synthesis and degradation of
phosphorothioate moieties.
The first synthesis of 1 performed enzymatically (with polymerase) proved to be (R,) stereoselective (Scheme 5 ) . The use of
NTPaS + polymerasehnplate
coo-
doo-
Scheme 4. Representation of the hypothetical directional two-point interaction hetwecn thc phosphorothioate group and a protein. which is responsible for the
diastereoselective action of some nucleases. Left. arrangement leading to a "productive" interaction of a phosphorothioate function with a protein ( a s seen with the
natural oligonucleotide): right: arrangement leading to an "unproductwe" arrangement.
supported by the results of studies on the interactions between
oligonucleotides bearing the ' ' - O H ... GAATTC . . . 3 ' - 0 H sequence and the restriction enzyme EcoRT. Replacement of the
. . . A P T . . . internucleotide group with ...A,,T... having an (S,)
configuration improves the binding between oligonucleotide
and this particulai- enzyme.[22g1Moreover. the proposal that
cellular uptake of oligonucleotide involves a specific receptor or
transporting protein[4". 241 implies that the passage through the
cell membranes (either inward or outward) may be diastereoselective because of the chirality of both receptors and transporting
proteins. This diastereoselectivity could be cell- or species-dependent and could affect both pharmacokinetics and biodistribution.
Finally, the principle of antisense activity of 1 implies the
affinity of 1 towards target R N A (or DNA)."] It is tempting to
postulate. even on the basis of early, simplistic model studies,[251
that this affinity is also dependent on the configuration of the
phosphorus atom (vide infra). On the other hand, it is reasonable to consider, in light of the unpredictable influence of the
chirality of 1 on their interactions with proteins andjor RNA,
that the most reliable approach to obtaining active antisense
agents is to provide the living system with the pool of diastereomers of 1. Then the competitive. natural selection of effective
diastereomers in each step of the overall process would lead to
biological activity. This assumes that the other "incorrect" components of this diastereomeric mixture neither prevent activity
nor are harmful to the biological system. It should be emphasized that all biological results reported thus far, including those
on the inhibitory effects of Oligo-S on protein biosynthesis. were
71 2
T
(al'-Rp)
Scheme 5. Enzymatic synthesis of (all-R,)-poly(nuc1eoside phosphorothioate)s.
Template = DNA: R = H, B = Ade, Gua. Cyt. Thy or R = OH. B = Ade, Gua,
Cyt. Ura.
endonucleases for diastereoselective degradation of 1 having
undesired (R,) or (S,) configurations has been studied but is
limited in practice to short homopolymers like penta(cytidine
phosphorothiate)s, d[(C,,),C] .['** A diastereomeric mixture
of d[(C,,),C] was exposed to endonuclease Pl , which selectively
cleaves all phosphorothioate diesters with (S,) configuration.[22k1Besides intact (all-&)-d[ (C,,),C], the accompanying
diastereomerically pure 5'-phosphorothioylated units were converted by alkaline phosphatase to the corresponding 5'-OH
dimer, trimer, and tetramer having only (R,)-configurated internucleotide phosphorothioate groups (Fig. 3). Diastereoselective
degradation of Oligo-S with the "wrong" configuration at phosphorus is limited to phosphorothioates prepared diastereoselectively, in other words, Oligo-S with mainly (R,) or (S,) configuration. If this were not the case, diastereoselective enzymatic
digestion would be impractical for most purposes, because in
the statistically simplest case the desired product is only a small
fraction (2-")of the starting mixture. Moreover, since an endonuclease with diastereoselectivity opposite to that of nuclease
PI has not been found, recovery of oligonucleotide 1 of (all-S,)
configuration, even from a mixture of diastereomers of predominantly (S,) configuration, is not yet possible. Therefore, the
only rational approach to the synthesis of Oligo-S having a
predetermined sequence of nucleobases and predetermined chirality at each phosphorus atom is a stereocontrolled, and highly
stereoselective chemical method.
What are the potential strategies for the synthesis of Oligo-S
with defined stereochemistry at the phosphorus centers? MethodA17grw. (%ivii.
Inc. G I . €n,q/. 1994, 33, 709-722
REVIEWS
Oligo( nucleoside phosphoroth1oate)s
R'oScheme 6. R' = DMT, pixyl. etc.: R' = alkyl, aryl: X = S. 0
or R'X = Me, B' = protected Ade. Gna. Cyt. Thy: L =
leaving group.
dl
~~
0
10
5
15
20
25
30
213
should be removable by cleavage of the X-R2 bond. If R 2 X =
L, prochiral 13 offers the possibility for diastereoselective exchange (replacement) of (pro-R) or (pro-S) groups.
In early attempts at the diastereoselective synthesis of 1 [''I
substrates 13 (R' = DMT. R2X = MeO, L = iPr,N, B' =
CytBz) were obtained in diastereomerically pure form, but their
activation by tetrazole appeared to cause P-epimerization. Besides protonation, activation of 13 by tetrazole also involves the
substitution of the NiPr, group by tetrazole. Intermediate phosphorotetrazolidites 14 undergo fast P-epimerization in the presence of an excess of tetrazole, most probably by rapid exchange
of substituents (Scheme 7).
1 -
F I 3.
~ RP-HPLC traces of a) the parent tetramer d(C,,,C,,C,,Cj after digestion
with niicle;isc PI 'nlkaline pho4phatase; b) crude tetramer d(C,&,C,,C)
(mixture
of eight dia.;tereomers); cj crude trinier d(C,,C,,C) (four dinstereomers): d) crude
dimer d(C,.,C) (two diastereomers). Dimer. trimer. and tetlainer were prepared by
Ihc phosphoroaniidite.'sulfurization method and analyzed as described in Table 1
with thc cxception that the gradient was 0.66% m i n - l . Retention time i in min.
Peaks inasked with
ii
"'"
W'
bR3
4
t r i m ~ l ccorrespond to benzdmide.
ologies developed during 100 years of phosphorus chemistry["'
that may be employed for the stereocontrolled synthesis of diastereoinerically pure l are based on a) stereoselective or stereospecific nucleophilic substitution at tricoordinate phosphorus,
b) stereoselective synthesis or stereoselective sulfurization of intermediates containing internucleotide H-phosphonates, c) stereoselective or stereospecific nucleophilic substitution at tetracoordinate phosphorus.
In choosing a potential strategy. one must consider that the
diastereomeric purity (dp) of 1 is described b y Equation (b).
A
13 ( L = NiPr2,X = 0)
4.1. Nucleophilic Substitution
at Tricoordinate Phosphorus Centers
Stereoselective/stereospecific syntheses of 1 utilizing substrates
13 (Scheme 6). which contain a tricoordinate phosphorus center
(Pi"), have the following limitations: substrate 13 must be separable into its P enantiomers, the R' group must be an acid-labile
protecting group, X must be an oxygen or sulfur atom, and R2
14
-
R'ows'
1
-
If p, = p z = . . . = p,,.then dp = p". where n is the number of
internucleotide phosphorothioate units and p I , p 2 . . . . p , designate diaatereoselectivities of particular coupling steps. It is
obvious that the diastereoselectivity of any single coupling
must be greater than 0.95 for the effective application of this
strategy.
.
15
Scheme 7. Diastereomerically pure 13 undergoes P-epimerization in the presence of
tetrazole, which impedes the stereocontrolled synthesis of I by phosphoroamidite
methodology. a) I . S,, 2. deprotection. R 1 = DMT; Rz = Me; R3 = LCA-CPG:
B' = B' = CytBz.
Detailed mechanistic studies confirmed this early hypothesis.[301Accordingly, the resulting intermediate dinucleoside 3 ' 3 0-methylphosphite 15 consists of a mixture of diastereomers.
Stereoretentive sulfurization of 15 followed by deprotection of
amino and hydroxyl functions, and cleavage from the solid support gave dicytidine 3',5'-phosphorothioates with the ratio of
diastereomers (Rp):(Sp)
= 60:40. In similar studies on the diastereoselective synthesis of dinucleoside 3',5'-methanephosphonates, Lebedev et al.[3i1used diastereomeric substrates 13
(R'X = Me, L = iPr2N). The nucleoside component in this reaction was 5'-0-trifluoroacetyl-3'-O-acetylthymidine,and 4dimethylaminopyridine (DMAP) was used as the catalyst. After
mild oxidation and deprotection, the resulting dinucleoside
71 3
W J. Stec and A. Wilk
REVIEWS
3’,5’-methanephosphonate appeared to consist of a 1: 1 mixture
of diastereomers.
These discouraging attempts obviously d o not preclude
success with other P’l‘ compounds. Perhaps substrates 13
bearing bulky and/or chiral R 2 groups (but offering selective
R’-X bond cleavage) can be constructed. Compounds with a
different leaving group L may require a different type of
activation than that used for substrates with L = NR,. Alternatively, NR, groups that do not undergo ready exchange
should be employed. A potentially good leaving group is 4-nit r o p h e n ~ x y l , [but
~ ~ ]only under the condition that the resulting
4-nitrophenoxide ion is removed from the reaction medium immediately after release; otherwise, it may cause epimerization
of 13.
Selectivity may also be attained by the diastereoselective reactions of prochiral Pl” substrates (13, R2X = L) with the S-OH
function of the nucleoside component. Diastereoselectivity
within the coupling process is possible since both reagents ( 3 ’ 4 3
phosphitylated nucleoside and 5’-OH nucleoside) bear chiral
auxiliaries (2’-deoxyribosyl moieties), and one of the two competing routes (leading to the (S,) and (R,) diastereomers) can be
considered to be kinetically favored. In the course of the development of the phosphorodiamidite approach to oligonucleotide
synthesis[3315‘-0-DMT-thymidine 3’-0-phosphorodimorpholidite 16 was synthesized and coupled with 3’-0-acetylthymidine
4 in the presence of tetrazole as the catalyst (Scheme 8). Unfortunately, neither phosphorothioate 18 nor H-phosphonate 19
were enriched in one diastereomer to a meaningful extent.[”]
These results strengthen the hypothesis that any phosphoroamidite substrate like 13. which is activated by an acid nucleophilic
catalyst, is prone to P-epimerization.
“ ’ O W 1
“ ’ O W 1
4.2. Stereoselective Formation and Sulfurization
of Oligo(nuc1eoside H-phosphonate)s
The H-phosphonate approach to oligonucleotide synthesis
involves 5‘-0-DMT-nucleoside 3’-O-(H-phosphonate)s as Pprochiral substrates 3. Their activation with sterically demanding pivaloyl[35”1or adamantoyl[35‘1chloride may. in principle,
be stereoselective and lead to chiral oligo(nuc1eoside H-phosphonate)s bound to the solid support. These, after tandem sulfurization. cleavage from the support, and nucleobase deprotection, give Oligo-S.[3h1However, the ratio of diastereomers of
di(deoxyribonuc1eoside) 3’.5’-H-phosphonates 6 is known to be
close to 1 :
which dictates that a similar ratio of the corresponding diastereomeric dinucleoside 3’3-phosphorothioates results from subsequent ~ u l f u r i z a t i o n . [Very
~ ~ ~ recently
]
Stawinski
et al. reported that diribonucleoside 3’S-H-phosphonate C,,U
prepared by his ~nethod[~’Iconsists of a mixture of
diastereomers in the ratio 6: 1 .13*] Independently, it has been
demonstrated that diastereomers of dinucleoside 3‘,5’-II-phosphonates 6 can be separated by silica gel chromatography1”* 38]
and sulfurized with retention of configuration. This P-configurational stability of unsymmetrical 0.0-dialkyl H-phosphonates was, however. not surprising in light of classical studies on
the preparation of diastereomerically pure O-menthylphenylp h o ~ p h o n a t e s ’ ~and
~ ’ enantiomeric O-isopropylmethylphosphonates.[40] Moreover. sulfurization of P-chirdl 0,O-dialkyl
H-phosphonates is known to be a stereoretentive process.[’51
Therefore, stereocontrolled synthesis of ohgo(nuc1eoside Hphosphonate)s as convenient precursors of Oligo-S (or other
P-chiral analogues of oligonucleotides) seems to be a still-unexplored avenue that deserves more attention, especially in view of
results reported by Hata et al.[4’1and recently by Battistini
et ai.[421
Hata et al. found that degradation of dinucleoside 3’.5’-s(-ketophosphonates 20 by treatment with n-butylamine and 1 . M i azabicyclo[5.4,0]undec-7-ene(DBU) is stereoselective (Scheme 9).
In these studies intermediate H-phosphonates 21 were not isolated but treated with trimethylsilyl chloride and sulfurized with
4
OR3
16
17
I
“ ‘ O w B 1
nBuNH,/DBU
-
OD81
20
21
I
OR3
19
IS’
(Rpj-TpsT+ (Spl-TpsT
Scheme 8. The synthesis ol Oligo-S with the phosphoroamidite method could not
be conducted diastereoselecti\ely R ’ = DMT: R’ = Ac. B i . B 2 = Thy. (Kp)-!(Sp)18 and (&)-:(Sp)-TP,T are both formed in approximately 1 : 1 ratios.
714
Scheme 9. Thc DBU-catalyzed aminolysis of phosphonaks 20 wiis remarkably
diastereoselective, as shown by the diastereomeric ratio of the sulfurized product.
a) 1 . S,, 7. deprotection of OH Froups. b) 1 . deprotection of OH groups. 2.
Me,SiCI’Et,N + S,. B’ = Ade. Thy: B’ = Thy. Ar = phenyl. p-chlorophenyl.
A n , q w Ciiori. Irrr. Ed. Erifi/. 1YY4.
33,709 -722
REVIEWS
Oligo(nuc1eoside phosphorothioatefs
elemental sulfur. Depending on the kind of protective groups
and the sequence of deprotection/sulfurization steps, either pure
(R,)-dinucleoside 3’,5’-phosphorothioate o r a significantly enriched mixture of diastereomers [(S,):(R,) = 2.5:1] was obtained. As pointed out by the
in the case of d(T,,T),
when a benzoyl or tert-butyldimethylsilyl group was used as the
3’-terminal protecting group instead of 1,3-benzodithiol-2-yl
(DBT). the formation of the phosphorothioates was not
stereoselective. Because the starting a-ketophosphonates 20
were used as a mixture of diastereomers for the DBU-catalyzed
aminolysis, the conclusion drawn was that DBU influences the
configurational equilibration of the resulting H-phosphonates
21 (Scheme 10). Hata et al. postulatedL41b1
that DBU, generally
freedom than the cyclic 3‘,5‘-0,O- or 2l.3‘-0,O-1 .I .3,3-tetraisopropyl-I .3-disiloxanediyl protective groups, causes a reduction
in diastereoselectivity [(S,):(R,)= 70: 301.
These intriguing observations can be interpreted in terms of
diastereoselective activation of the H-phosphonate function of
protected 22 owing to the presence of sterically demanding protective groups at the nucleoside components. Presumably. for
steric reasons, acylation of anionic 22 occurs exclusively at the
( p r o - R ) oxygen atom, and anhydride 25 reacts stereospecifically
with 23 leading to (Rp)-2’,5’-A,,A 26; subsequent stereoretentive sulfurization gives (S,)-24 (Scheme 12). This explanation
,T
O\
/H
__
%
,,
(“ - ’J
pro-S
O\ /H
,.o
,.“,
//”
O\
H%
,
’7
II
0
P ~ - R
22
+-
25
considered to be a nonnucleophilic base,[43’ interacts with Hphosphonates to form pentacoordinated intermediates which
epimerize at the phosphorus center by pseudorotation to give
the thermodynamically more stable isomers. The thermodynamic preference is dictated by steric demands of bulky ligands
(augmented by sterically demanding protecting groups at the
5’-0 and 3‘-0 positions) .[441
A different explanation must be given for the stereoselectivity
in the synthesis of 2,5‘-(A,,A) reported by Battistini et al.L421
Condensation of N6-benzoyl-3’,5’-0,0-tetraisopropyldisiloxane
adenosine T-O-(H-phosphonate) (22) with 23 in the presence
of pivaloyl chloride provides exclusively (S,)-configurated
diadenosine 2’.5’-phosphorothioate 24 after sulfurization
(Scheme 11). Interestingly, introduction of a terr-butyldimethylsilyl protective group at 2’-0, which has more conformational
’7
24
26
(K,) - 2’,.5’ - A,,,A
Scheme 10. The effect of DBU in shifting the configurational equilibrium of 21.
P
O\
-s/s
( S , ) - 2’. 5’ A,$
~
Scheme 12. (I””-R)-O-AcyIation of 22 responsible for the diactereoselective synthesis of 2’.5’-A,A.
implies that nucleophilic substitution at the phosphorus center
in 25 occurs with an inversion of configuration; however, no
examples of stereochemical transformations of compounds of
type 25 have been reported thus far.
In light of the results of Hata et al. in the stereoselective synthesis of oligo(nuc1eoside phosphorothioate)s by DBU-catalyzed
stereoselective epimerization of the internucleotide H-phosphonate functions, one should consider introducing specially designed sterically and/or stereoelectronically demanding protective groups into the participating nucleosides (nucleotides). In
our own research a variety of potential catalysts for shifting the
( R,): (S,) equilibrium of 21 have been studied, including optically active and sterically demanding acids and bases. but so far the
process is not controllable.[34]
4.3. Diastereoselective Activation of Prochiral Substrates
with Tetracoordinate Phosphorus Centers
22
The P-prochiral substrates 27 (Scheme 13) are employed in
the phosphotriester approach to oligonucleotide synthesis.L4s1
When R2X was an alkoxy or aryloxy group, activation ofphosphate function led to the formation of P-chiral dinucleoside
3’.5’-O-aryl(alkyl)phosphates. Since the synthesis requires the
subsequent removal of the 0-aryl function, the diastereomeric
ratio of the intermediate triester was not of interest. However,
the prochirality of 27 suggests that the activation process could
23
0
is,)
Scheme 1 I . Stcreoselective synthesis of 2’.5‘-A,,A. B’
loyl.
Aitgrir
C7roii.
I n ! . E d Engl. 1994. 33. 709-722
=
-24
B’ = AdeBr; Piv = piva-
Scheme 13. R’ = DMT. pixyl. etc.; R z = alkyl. aryl: X
B‘ = protected Ade, Gua. Cyt. Thy.
=
0:
27
71 5
W J. Stec and A. Wilk
REVIEWS
be diastereoselective. Indeed. in one well-documented case,
Ohtsuka et al.[461proved that the condensation of N6,5'-O-bis(dimethoxytrity1)adenosine3'-0-[(0-2-chlorophenyl)phosphate]
with Nh.3'-O-bis(dimethoxytrityl)adenosine in the presence of
1-(2.4.6-triisopropylbenzenesulfonyl)-5-(pyridin-2-yl)tetrazole
as the activating agent gives, after deprotection, dideoxyadenosine 3',5'-0-(2-~hlorophenyl)phosphate
with (S,) configuration. This particular finding prompted Cosstick and
Williams[471to study the diastereoselective activation of 27
(B' = CytBz; Scheme 14). They demonstrated that reaction of
intermediates in the triester approach to oligonucleotide synthesis are pyrophosphates;[481their intermediacy and the possibility of equilibration at the phosphorus stereocenter may cancel
the effects of diastereoselective activation of 27.[491
4.4. Stereoselective Nucleophilic Substitution
at Tetracoordinate Phosphorus Centers
The most promising and straightforward approach to stereocontrolled synthesis of Oligo-S involves substrates 32 bearing
leaving group L (R2 = aryl, X = 0, Y = S or R2 = alkyl,
X = S, Y = 0; Scheme 15). which in the process of nucleophilic
28
27 (X= S, R2 = CH2CH2CW
O
\
/
'
Scheme 15. R ' = DMT.
X.Y = S.0 ( X Y); B'
L = leaving group.
+
2Y
pixyl. etc.: R2 = alkyl, aryl;
= protected Ade, Gua, Cyt. T h y ;
ex'
L
'
32
substitution at the phosphorus atom can be replaced by the 5'O H function of the nucleoside or oligonucleotide in stereodefined
manner. Such substrates 32 (R2X = MeS, Y = 0, L =
4-N0,C,H40) were first described by Lesnikowski et aLL5'I The
separated diastereomers of 32 reacted with 3'-O-acetyl nucleosides in the presence of tBuMgCl to give dinucleoside 3'.5'(S-methy1)phosphorothioates 33 (Scheme 16). Cleavage of the
CH, -S bond (phosphorothioate deprotection) was achieved in a
later step with PhSH/NEt,. The coupling occurs with inversion of
configuration at phosphorus with a stereospecificity of > 95 YO
and furnishes 33 in about 70% yield.
31
I
30
0
~
3
Scheme 14. Diastereoselectibe activation of P-prochiral phosphorothioate 27 by
1-(2.4.6-triniethqIphenylsuIfonyl)-5-(pyridin-2-yl)tetrazole.
tBuMgCl
27 with 3'-O-acetylthymidine (4, R3 = Ac, B' = Thy) in the
presence of 28 leads to the formation of the corresponding dinucleoside 30, which after decyanoethylation at sulfur consists
of the diastereomeric mixture of d(C,,T) with a distinctive preponderance of the (S,) diastereomer. In the best case the ratio of
diastereomers (S,):(R,) was 79: 21. Such enrichment, although
not satisfactory in terms of a stereocontrolled synthesis of I, is
worth deeper consideration and perhaps more extensive study.
If the mechanism of activation of 27 involves stereoselective
formation of the mixed phosphorothioic sulfonic anhydride 29,
and this undergoes double nucleophilic substitutions (first by 28
with transient formation of phosphorotetrazolidate 31 and subsequent reaction with 4 to give phosphorothioate triesters 30),
attempts at introducing sterically demanding protective groups
on nucleobases and hydroxyl functions not participating in the
coupling process increase the probability of diastereoselective
formation of dinucleoside 3'3-phosphorothioates. This promising scenario must be taken with caution, because earlier studies
of Knorre, Zarytova. et al. indicated that the phosphorylating
71 6
"'ow1
'
0, ,SR2
4
32 (X = S , Y = 0,
L = C,H,NOZ)
Scheme 16. Stereoselectivenucleophilic substitution a t the phosphorus ccnter in 32.
This reaction IS used for the synthesis of short Ohgo-S. R ' = monomethoxytrityl
(MMT); R' = Me; R' = Ac; B' = BZ = T h y .
Although this methodology cannot be used in solid-phase synthesis, Lesnikowski et al. were able to demonstrate that preparation of stereochemically pure trimers d(T,,T,,T) and tetramers
d(T,,T,,T,,T)
is feasible if protective groups such as 4-chlorobenzyl are introduced at sulfur. Very recently, this strategy was applied to the synthesis of ( R p , R p ) -and (S,,Sp)-U,,Up,U.Lso'l In
spite of efforts to find other catalysts, tBuMgC1. which was introduced for oligonucleotide chemistry by Hayakawa et aI.l5" is the
catalyst of choice; other catalysts such as tBuOK, DMAP, and
DBU gave unsatisfactory yields or led to P-epimerizati~n.[~~]
REVIEWS
Oligo(nuc1eoside phosphorothioate)s
4.5. Oxathiaphospholane Method for the Synthesis
of Oligo-S
A novel strategy leading to stereodefined Oligo-S based on
nucleophilic substitution at tetracoordinate phosphorus centers
was elaborated following studies on the conversion of 0,Odialkylphosphorothioates into the corresponding [ '*O]-phosp h a t e ~ . [Okruszek
~~]
et al. observed that O-alkyl-O-arylphosphorothioates 34 undergo reaction with ['*O]styrene oxide/
[170]waterwith replacement of sulfur by ["O] and of the aryloxy group by [' 'OIOH, respectively. Both replacements occur
with full stereospecificity (Scheme 17).
35
34
+
DBU
-I
L
41
46
42
43
36
bR3
Scheme 18 Base-catalyzed alcoholysis of 1.3,2-oxathiaphospholanes 41. R 1 =
DMT; R 3 = Ac; B' = B2 = T h y .
RO
I
o/pO
l ;-
R=
Mmo,,42
40
-0
tive groups, dinucleoside 3',5'-phosphorothioate 43 was obtained
with > 95 % yield (Scheme 18). Separation of the diastereomers
of 41 (B' = AdeBz, GuaiBu, CytBz, Thy) was followed by studies of the stereoselectivity of the oxathiaphospholane ring opening, which was found to be higher than 98 YO.[^*]
Model studies of (RP,R,)-2-[N-(a-naphthyl)ethylamino-2thiol-I ,3,2-oxathiaphospholane 44[541have shown that the ring
opening with MeOH/DBU occurs by attack of the base-activated hydroxyl function of MeOH on phosphorus from the side
opposite to the endocyclic P - 0 bond (Scheme 19). Collapse of
Scheme 17. Proposed mechanism for the stereocontrolled synthesis of nucleoside
3'3' 60.1
'0, '80]phosphates 40.
Detailed mechanistic studies showed that the initial attack of
the sulfur atom of 34 on the carbon atom of styrene oxide results
in intermediate 35. Subsequent attack of the p-alkoxyl anion on
the phosphorus center and elimination of ArO- from 36 gives
intermediate 37. In the hydrolysis of oxathiaphospholane 37 by
[' 'O]H,O the five-membered ring opens exclusively by cleavage
of the P-S bond. Intermediate 38 is formed by attack of the
water molecule on 37 from the side opposite to the endocyclic
P - 0 bond. Pseudorotation then places the leaving group in an
apical position in 39. Cleavage of the P-S bonds is followed by
Fast elimination of the alkylene episulfide.
Following the stereocontrolled synthesis of monoalkyl[l6O,'70,1s0]phosphates40, the construction of oxathiaphospholanes such as 41 and their reaction with alcohols were of
interest.[**] 5'-O-DMT-nucleoside 3'-0-(2-thio-1,3,2-oxathiaphospho1ane)s 41 were prepared and treated with 3'-O-acetylthymidine. In this case, the attack of the hydroxyl group on the
phosphorus atom requires a basic catalyst. Although the reaction occurs in the presence of triethylamine (room,temperature,
24 h, 5 YOyield) and diisopropylamine (room temperature, 4 h,
38 YOyield), DBU was the catalyst of choice. With DBU as catalyst a single coupling required 5 min and, after removal of protecAn,q+t C / w i n Inr bd t n g l 1994, 33, 709 - 122
(Rp, Rc)-44
Scheme 19 DBU-catalyzed methanolysis of 44.m-C,,H,
(SP,Rc)-45
=
a-naphthyl.
the pentacoordinate intermediate requires, according to Westheimer's rules,[551 pseudorotation placing the endocyclic P-S
bond to be cleaved in the apical position. Cleavage of the P-S
bond is followed by fast elimination of ethylene episulfide. The
assignment of absolute configuration of substrates and products, namely (R,,R,)-44 and (S,,R,)-45 (resulting from Smethylation of the product of DBU-catalyzed methanolysis of
44), and the analysis of the stereochemical course of the conversion of 44 into 45 also allowed tentative assignment of the absolute configuration at phosphorus in 41. The quickly eluting
diastereomers of 41, which react with 5'-OH nucleosides to give
(S,)-d(N,,N'), have (S,) configuration at phosphorus, whereas
the slowly eluting diastereomers of 41 have (R,)configurati~n.['~]
Detailed knowledge of the stereochemistry of substrates and
products of the oxathiaphospholane method leading stereo717
W J. Stec and A. Wilk
REVIEWS
selectively to dinucleoside 3'S-phosphorothioates was still not
sufficient for the stereocontrolled synthesis of Oligo-S. Since the
succinyl linker typically used with solid supports is cleaved in
the prcsence of DBU.'"] a succinylsarcosinyl type of linker was
e~nployed,[~""
which survives in 0.1 M DBU in acetonitrile during the solid-phase synthesis of ol"onucleotides.'"''
All four
support-bound nucleosides 46 were prepared and used for the
DBU-catalyzed reaction with separated dinstereomers of 41
(Scheme 20). Penta(deoxycytidine-. octa(thymidine-, and dode-
'
p , l O V 1
0
\
%
,,
O
+
//s
R'owl
*[
$-''ow2
"
~
*
;&-oR3
d
0
R3
46
( K , 2 )- 41
a)
h'H,OH
47
-
(&!$)
-
5'-O-DMT-Oligo-S
48
Schcmc 20 Strreoconrrollcd oxathiapho\pholane mcihod Ibr the synthesis of I
a ) 1 ti : 2. 41 + DBU: 3. capping:4. step5 1-3 arc rcpeated I I time\. R ' = DMT,
R ' = LC'A-CPG. B' = B' = AdeB7. GusiBu. CqiBz. Thq
ca(deoxyadenosine phosphorothioate)s have been obtained in
diastereoinerically pure
'I Following stereocontrolled
synthesis of (aIl-R,)- and (all-Spj-d[(Aps)llA] it was shown that
the alignment of (all-.Sp)-d[(Aps)i
lA] to complementary dodecathymidylic acid is more effective (7;, = 34.4 C) than that of
(aII-Xl,)-d[(A~s)l
l A ] (T,, = 30.0 C).15y1
This experimental result
is in agreement with the analysis of model duplexes of €3 conformation, which indicated that duplexes formed by Oligo-S with
(X,) configuration and the sulfur atoms directed "inward" relative to the axis of the double helix are sterically less favorable
than those formed by (all-S,,)-Oligo-S in which all sulfur atoms
are directed "outward". The higher steric demands of phosphorothioates in comparison to phosphates were also cvident from
earlier results concerning the synthesis, molecular mechanics
calculations, and biological properties of adenosine-3',5'-cyclic
pllosphorodithioate.'~'*l Recent molecular mechanics and dynamics calculations on a simulated complete helical turn of
DNA duplexes formed by unmodified oligonucleotide as well as
hoinochiral (aII-R,)- or (all-S,)-Oligo-S'6'1 confirmed the results of intuitive analysis and experimental facts. Independently,
N M R spectroscopic["'] and T,[("] studies of oligonucleotides
containing phosphorodithioate functions also indicated a destabilizing effect of sulfur a t o m on double-stranded constructs.
Unfortunately. the synthesis and isolation of satisfactorily
piire Oligo-S longer than 12-mers by using oxathiaphospholane
methodology is still difficult. Single-step coupling efficiency
sometimes substantially lower than 96 Yo may be responsible for
the presence of "failed" sequences (17 1.17 - 2.17 - 3 ...), espccially if the "capping" step is not I00 %, effcctive. Recently,
~
71 8
y
acetic anhydridelN-methylimidazole, which was originally reported['81 for the capping of unreacted 5'-OH groups by acylation, has been replaced by acetic anhydridei4-dimethylaminopyridine. Moreover, because the water content in solvents and
reagents is most likely a critical parameter (traces of water in the
presence of DBU can initiate ring-opening polymerization of
1,3,2-oxathiaphospholanes),special precautions and strictly anhydrous reagents have been used. Also, the time of the coupling
step has been shortened from 600 to 200 seconds. Although the
conditions and parameters for the oxathiaphospholane method
are still not completely optimized, Grajkowski et al. have accomplished. starting from diastereomerically pure substrates 41
(B' = CytBz). the synthesis of (alI-X,) and (aIl-S,) diastereomers of d[(C,,)27C] with 86% and 82Y0 chain integrity, respectively, as proved by polyacrylamide gel electrophoresis
(PAGE).["] Analysis of the crude product by 3 1 P N M R spcctroscopy did not show any evidence of phosphate contaminants,
which are unavoidable when Oligo-S is prepared by other methods."'] Grajkowski et al. also prepared. starting from diastereomeric mixtures of 41. a "mixed-sequence" 26-mer Oligo-S containing six dC, six dG, ten dT, and four dA. which by PAGE was
comparable to the 26-mer prepared by the phosphoramidite:
sulfurization method. Current efforts are being directed at further improving the yield of single steps and elucidating the structures of the products resulting from unexpected, obscure side
reactions.
What kind of side reactions may be envisioned'? An obvious
one starts with the formation of an internucleotide phosphorothioate triester function containing the charged CH,CH,S- group.
as in 42. This moiety. although expected to be eliminated. may
instead attack the nearest j'-O-pliosphorus-bound 5'-carbon
atom. such that the phosphorothioate-carbon skeleton is extended by one CHzCHzSunit. This side reaction may proceed
through a thermodynamically unfavorable seven-memberedringtransition state. and the enzymatic hydrolysis of the final
Oligo-S should give modified nucleosides(tides). which does not
appear to be the case. However, this possibility requires further
investigation. since these contaminants are characterized by the
same base composition and net negative charge as the expected
product. Also under investigation is the possibility of "decapping" the 5'-0-acetyl group with DBU. which could be
responsible for unwanted formation of (n - I ) , ( n - 2), ( n - 3)
Oligo-S.['"]
Other studies in progress focus on the preparation of substrates
like 41 without amide hydrogens, since activation by DBU and
creation of the possible branching sites for growing polymer
would be avoided. Efforts are also being made to construct the
C-substituted oxathiaphospholanes 41, which would still react
with 5'-OH group, but the ring opening should not be followed
by subsequent elimination of episulfide. S-Acylation of an intermediate like 42 would preserve the triester moiety until the end
of desired chain extention, but one would alsb expect an increase
in the undesired attack on the 5'-carbon atom as discussed
above.['"]
Although beyond of the scope of this review, it is worth mentioning that modification of the oxathiaphospholane method by
use of j'-O-DMT-nucleoside 3'-0-(2-thiono-l.3,2-dithiaphospho1ane)s has allowed the preparation of oligo(nuc1eoside phosphorodithioate)s S2.['']
Oligo(nuc1eoside phosphorothioate)s
4.6. Other Possible Strategies for the Stereocontrolled
Synthesis of Oligo(nuc1eoside phosphorothiate)s
Since we wish to direct the attention of phosphorus and biopolymer chemists to the problem of the stereocontrolled synthesis of Oligo-S and related DNA or R N A congeners, other, yet
unexplored routes to P-chiral oligonucleotide analogues should
be mentioned. A mixture of diastereomers of 5'-0-DMT-nucleoside 3'-0-[S-(2,4-dichlorobenzyl)phosphorodithioate]
49 was
treated with 2,4,6-triisopropylbenzenesulfonylchloride (TIPSCI) and AV-methylimidazoleto give 97 % activation of the oxygen atom of the phosphorodithioate moiety as opposed to activation of the sulfur atom (Scheme 21). Although reaction of 50
atom should lead to Oligo-S. However, if a pyrophosphorothioate intermediate[481 follows the formation of thiophosphoric sulfonic anhydrides, this approach to the stereocontrolled synthesis of Oligo-S may fail.
Another possibility involves acylation of appropriately protected nucleoside 3'-0-(0-alkyl)phosphorothioate53 with sterically demanding, substituted benzoyl chlorides and subsequent
reaction of the resulting mixed anhydride 55 with a 5'-OH function of a nucleoside or or oligonucleotide (Scheme 23). Al-
0, JJ
+
CI-c
R ' O b o q '
53
55
Schctne 23. Anhydridcs 55 are potential substrates for the stereocontrolled synthe\IS of I .
49
Scheme 2 1. Activation of' phosphorodithioate 49 by T1PS-CI~N-mcthylimi~~izole
occurs 111 U7"(, ill oxygen.
with 5'-OH nucleoside 4 gives a diastereomeric mixture of dinucleotide 51, these undergo S-deprotection and provide P-achiral
dinucleoside 3'.5'-phosphorodithioates 52.r66'
This chemoselective activation, unexpected in the light of earlier works on the activation of ambident phosphorothioate anions,'"'.
kindled some hopes that substrates 53 with O R Z
groups such as OMe. OEt, or OBn. after separation into diastereomeric species. could undergo selective activation of phosphorothioate anions at the oxygen center; subsequent reaction
at phosphorus with an OH-nucleoside may then lead to
diastereomerically pure (or highly enriched) dinucleoside 3',5'0-alkylphosphorothioates 54 (Scheme 22). Removal of the
phosphorothioate protecting groups by attack on the carbon
'''
R'owl
53
54
I
OR3
Schcmc 22. Phosphorothioates 53 are potential substrates [or the stereocontrolled
synthesis o l I . a ) 1 . T1PS-CI~N-methylimidazole:2. 4. R ' = DMT: R 2 = Me, Et.
Bz. allyl: I<' = Ac: 8' = B' = Thy.
though mixed phosphoric carboxylic anhydrides are known to
exhibit both acylating and phosphorylating properties.[681there
are few well-documented examples of their exclusive phosphorylating activity. Jackson et aI.LhR"l
have pointed out the important change in regioselectivity depending on the nature of the
nucleophile: aminolysis occurs by an attack at the carbonyl
carbon. whereas alcoholysis takes place by an attack at phosphorus. Moreover, Lambie['"] demonstrated that 0-pentachlorobenzoyl U,O-diethylphosphate reacts as a phosphorylating
reagent even with aniline. This distinctive phosphorylating ability
prompted Wozniak to synthesize P-diastereomerically pure forms
of nucleoside 3'-0-(methanethiophosphonic) carboxylic anhydrides as the potential substrates for the stereocontrolled synthesis of oligo(nuc1eoside methanethiophosphonate)s. So far,
reactions of aforementioned mixed anhydrides with S'-OH
nucleosides 3'-O-bound to the solid support lead to the desired
product in miserable yields that are unacceptable for repetitive
multistep synthesis. The search for an appropriate catalyst for
this reaction is in progress and, if successful, efforts will be undertaken to synthesize analogous substrates applicable to the synthesis of l.ro9I
5. Analysis of the Diastereomeric Purity of Oligo-S
In the introduction to this review we mentioned how difficult
it is to determine the diastereomeric purity of Oligo-S. Figure 4
presents our best results obtained to date for the RP-HPLC
separation of eight diastereomers of d[C,,),C] (Fig. 4a,b) and
an H P L chromatograni of d[(C,,),C] (Fig. 4c. d); both products were prepared by the nonstereoselective phosphoroamidite
method. Figure 4 b clearly indicates the limit of RP-HPLC for
the separation of small Oligo-S.r'81Our earlier results also indicated that in the case of Oligo-S with (R,,)
configuration at each
internucleotide phosphorothioate. digestion with endonuclease
P1 can be used to prove its structure. It can also be useful as ii
"purification" process because under appropriate conditions
every phosphorothioate of (S,) configuration is cleaved.[22h1
719
W. J. Stec and A. Wilk
REVIEWS
6. Conclusions
a)
__
0
I0
20
30
24
0
10
20
30
26
26
28
t
f-
7
28
30
30
32
-
Fig. 4. RP-HPLC traces of tetramer d(C,C,C,,C):
a) entire chromatogram:
bj expanded section showing the presence of dl1 eight diastereoniers. cj and d) show
For
the partial separation of 16 diastereomers of pentamer d(C,,C,C,,C,,C).
HPLC conditions see Table 1. Retention time f in min.
Moreover, exposure of d[(C,,),C] (n = 1-4, prepared as diastereonieric mixtures) to nuclease P1 followed by alkaline phosphatase gave the product, which was separated by RP-HPLC to
give Pure (R,)-d(C,sC) > (RP ,R,)-d(C,,C,,C) > (R,,R, J,)d(C,,C,,C,,C),
and (R,,R,,R,,R,)-d[ (CPs),C] . Similar treatment of the aforementioned diastereomeric mixtures of Oligo-S
by means of snake venom phosphodiesterase (SVPDE), which
is known to be (R,) selective,122i.j1
was limited to digestion of
3'-terminal (R,) internucleotide phosphorothioate because the
enzyme has no endonucleolytic activity.1701
Very distinctive differences in electrophoretic mobilities between (aIl-R,)- and (all-S,)-d[ (C,,),C] have been found under
nondenaturing conditions (Fig. 5 ) . This observation could pos-
10
11
12
13
14
15
t Fig. 5. High-performance capillary electrophoresis analyses were performed on an
Applied Biosystetns 270A-HT capillary electrophoresis instrument with MICROGEL,,, (Applied Biosystems Inc., Foster City, CA) gel-filled capillaries (50 pm ID,
effective length 27 cm; running buffer: 7 5 mM TRlS phosphate. pH 7.6, 10%
methanol). Detection was at 260 nm, voltage 15 kV. Two individual diastereomers
(all-&.)- and (all-S,)-d[(C,,),C] were mixed (b) and compared with the mixture of
both diastereomers added to the sample of d[C,,),C]
prepared by phosphoroamidite/sulfurization method (S12 diastereomers) (a).
sibly indicate different self-organization (e.g. by base stacking)
or interactions with a gel, which occur in P-homochiral Oligo-S,
and may be considered as an indirect example of the effect of the
chirality of Oligo-S on mobility. Important parameters such as
cell membrane permeability may be influenced as well.
720
Besides the biological consequences of the defined chirality of
Oligo-S, there is another unanswered question : how far are we
from solving the problem of the stereocontrolled synthesis of l ?
For oligopeptides it took nearly a century to solve the synthetic
problem, and new enzymatic methods are being developed for
enantiospecific synthesis.[711 As pointed out earlier, in every
multistep synthesis the diastereomeric purity dp of the n-mer is
roughlyp"-', such that for a 28-mer prepared withp = 0.98 the
dp = 0.9827 = 0.56. Therefore, is diastereoselective synthesis of
chiral Oligo-S possible? Eckstein et al.17]demonstrated that the
polymerase-based synthesis of 1 with x-dNTPS leads to (aIl-R,)poly(nuc1eoside phosphorothioate)s. This type of enzymatic
preparation of Oligo-S would be limited to the preparation of
(all-R,)-Oligo-S. In this review it has also been indicated that
chemically synthesized (all-R,)-Oligo-S can be stereochemically
purified by digestion of contaminating (S,)-phosphorothioates
with nuclease PI. The next few years may bring the discovery of
an effective (R,)-diastereoselective endonuclease, or an appropriate enzyme may be developed by biotechnology. Such an
endonuclease would not only be very useful for analytical purposes but would also be valuable for purification of (aIl-S,)Oligo-S contaminated with diastereomers possessing (R,)-phosphorothioates. However, even if such endonucleases become
available, their application in stereochemical degradation
would not allow practical production of (all-S,)-Oligo-S, since
the content of the fraction with the desired homochirality in the
diastereomeric mixture is approximately 2-". Efforts should also
be directed toward the search for conditions under which a
stereoselective endonuclease can work in a reversible manner.
One can envision that matrix-directed synthesis of Oligo-S by
means of phosphorothioate transferases will become feasible if
studies are focused on, for example, the reversibility of the action of nucleases or phosphotransferases.
Alternatively, chemists must master the art of constructing
substrates that allow the synthesis of Oligo-S with a diastereoselectivity > 99 YOand a predetermined chirality at each internucleotide phosphorothioate function. However, progress seems
to be rather slow on this front. Ten years after the first stereocontrolled synthesis of both diastereomers of dithymidine 3',5'-phosp h ~ r o t h i o a t e " ~ the
'
demonstrated ability of stereocontrolled
synthesis of 1 has extended only to dodecamers.1281Undoubtedly,
economic interests may provide enough incentive for the solution of this very challenging problem. If Oligo-S or synthetic
ribozyme constructs containing nuclease-resistant components
I responsible for recognition of appropriate target sequence within mRNA or pre-mRNA, are further evaluated as useful oligonucleotide therapeutics, then the stereocontrolled synthesis of 1 will
be facilitated. Should that not be the case, there is-still intellectual curiosity, which has always been an important factor in
purely fundamental studies. Mission-oriented projects may not
always succeed if they are not enforced by intellectual challenge.
This review was completed during the one-year appointment of
one of the authors ( W J S . ) as Fogarty Scholar-in-Residence at
the National Institutes of Health ( N I H ) , Bethesda, M D ( U S A ) .
We thank reveral colleagues and collaborators who contributed to
the development of the first stereocontrolled methodv for the synAnjiew. Chrm Int. Ed. En,$. 1994, 33, 709 .722
Oligo(nuc1eoside phosph0rothioate)s
thesis of'Oligo-S; their names are given in the relevant r+rences.
Research projects were supported by the State Committee f o r
Scientific, Research (grant no. 0034/PZ/92/03). We kvould also
like to e.q~rcssour gratitude to Dr. Gerald Zon q f l y n x Therapeurics, l n c . , Foster City, C A ( U S A ) and to Dr. Jack Schmidt,
Fogart>,International Center, N I H , ,for critically reading several
"mutations" qf'this re vie^^, and to Ms. Malgorzata Stecjor typing
this manuscript. We thank an anonymous rejeree j o r suggestions
concerning another side reaction that may accompany the
o s c i thiciphospliolane ring- opening process .
Received: February 12, 1993 [A 935 IE]
German version- Angen. Chem. 1994. 106, 747
[I] a) M. Matsukura. K. Shinoruka, G. Zon, H. Mitsuya, M. Reitz, J. S. Cohen,
S. Broder. Proc. N u / / . Acad Sci. U S A 1987. 84. 7706; b) M. Matsukura. G.
Zon. K Shinozuka. M. Robert-Guroff, T. Shimada, C. A. Stein, H. Mitsuya,
F. Wong-Staal, J. S. Cohen. S. Broder. rbid. 1989, 86, 4244; c) S. Agrawal. J.
Goodchild. M. P. Civeira. A. H. Thornton, P. S. Sarin, P C. Zamecnik. ibrd.
1988. X i , 7079; d) S. Agrawal. T. Ikeuchi, D. Sun, P. S. Sarin, A. Konopka. J.
Maiiel. P. C Zamecnik. ibrd. 1989, 86. 7790.
[2] a ) G . Zon, W J. Stec in Oligonuclroridiv and Their Anukiguer: A Pructical
A p p r ( ~/ I~ (Ed.:
(
F. Eckstein), IRL. Oxford. 1991, p. X7: b) G. Zon, T. G. Geiser.
A n t i - C a n w r Drug Drs. 1991. 6.539; cj T. Geiser, Aids: Anli-HIV Agenfs, Thcrupre\ und kc me,^ (Ann. N . Y . A<.orl.Sci. N . Y. 1990, 616, 173).
[3] a j J. W, Efcavilch in GP/Elri~rrophorcwso f n i o i ~ l e i cAcid> (Eds.: D. Rickwood.
B. D. Hannes). IRL, Oxford, 1990, p. 125: b)V. Metelev. S. Agrawal, A n d .
B i o d i i m 1992. 200, 342; c ) B. J. Bergot. W. Egan, J Cliromutogr. 1992, 599, 35;
d ) G Zon in High-Performunee Liquid Chromarogruphv in Biutethnulogy (Ed.:
W. S. Hancock). Wiley. New York. 1990, p. 310; e) J. Tang, A. M. Roskey, S.
Agrawal. Awl. Biochem. 1993. 212. 134.
[4] a ) C A. Stein. K. Mori. S. L. Loke. K. Subasinghe, J. S. Cohen, L. M. Neckers.
Grn<,1988. 72, 333; b) S. L. Loke. C. A. Stein, X. H. Zhang. K. Mori, M.
Nakanishi. C. Subasinghe, J. S. Cohen, L. M Neckers, Pro?. N a f l . Acud. Sci.
L;SA 1989. 86, 3474; c ) P. L. Iversen, Anti-Cuncer Drug Des. 1991, 6 , 531 ;
d j P. L. Iwrsen. S Zhu. A. Meyer. G. Zon, A n r k n s e Res. D e i . 1992. 2. 211;
e) P. L. Iversen. D. Crouse, G. Zon, C . Perr). ibid. 1992,2. 223; f j G. Marti,
W. Egm. P Noguchi, G. Zon. M. Matsukura, S. Broder, ibid. 1992,2.27;g) S .
Agrawiil. J. Temsamani. J. Y. Tang. Proc. Nut/. Acud. Sci. U S A 1991, 88, 7595;
h) D. J. Chin. G. A. Green. G. Zon. F. C. Szoka. R. M. Straubinger. NCM.
Bid.
1990. 2, 1091; i ) W.-I. Gao. C. Storm, W. Egan, Y.-C. Cheng, M u / . Phurniucol.
1993,43.45: i) P. T. C. Ho, T. A. Bacon. E. Wickstrom. A. C. Sartorelli. J CeN
B i ~ l1989.
.
I O Y . 329, k) J. M. Campbell. T. A. Bacon, E. Wickstrom. J. Eiuchem.
Biophi.\ Mi,ihorl.s 1990,20. 259: I ) Q. Zhao, S. Matson, C. J. Herrera, E. Fisher,
H \iu. A. M. Krie'g. Anri.sense Res. Dev. 1993. 3. 53; m) C. A. Stein. A. M.
Clear!. L. Yakubov. S. Lederman. ihid. 1993, 3, 19.
[5] a ) C . Carenave. C. A. Stein. N. Loreau, N. T. Thuong, L. M. Neckers, C.
inghe. C. Helene, J. S. Cohen. J. J. Toulme, Nuclerc Acids Res. 1989. 17,
b) P. J. Furd0n.A. Dominski. R. Kole. ihrd. I989,17,9193,cj W.-Y Gao.
F. Han. C-. Storm. W. Egan. Y:C. Cheng. M u l . Pharmucol. 1991.41, 223; d) M.
Chiang. H. Chan. M. A. Zounes. S. M . Freier. W. F. Lima. C. F. Bennett, J
Biol. Ch(wi. 1991. 266. 18162:e) T. M. Woolf. D. A. Melton. C. B. G. Jennings.
P r m . ,VrrrI. Acud Sci. C S A 1992. 89, 7305; f) C. A. Stein. J. L. Tonkinson, L.
Yakubov. Phormut.u/. Tho.. 1991. 52. 3 6 5 ; g) M. K. Ghosh, K. Ghosh, J. S.
Cohcu. .4nti.\cn.\e Res. Dev. 1992, 2, 111.
[6] a) M. Z . Ratajuak. J. A. Kanl, S. M. Luger, N. Hijiya, J. Zhang, G. Zon. A. M.
Gewirtz. Proc. Nut/. Acud. Sci. C S A 1992, 89. 11823; b) M. Simons, E. R.
Edelman. J. DeKeyser. R. Langer. R. D. Rosenberg, Nature 1992. 352, 67:
c ) K. D. Kunible. P. L. Iversen, J. K. Vishwanatha, J Cell Sci. 1992. 101. 35;
d) M. Simons. R D. Rosenberg. Cfrc. Re.,. 1992, 70, 835: e) G. Citro, D.
Pel-rotti. C. Cucco, 1. D'Agnano, A. Sacchi, G. Zupi, B. Calabretta, Proc. Nut/.
A r u d . S r i C'SA 1992, 89, 7031 ; f) K. H. Schlingensiepen. W. Brysch in Gem
Regiilorwri. Biulo,g of Anlisrnse R N A and DNA (Eds.: R. P. Erickson, J. G.
Izant). Raven, New York. 1992, p. 317; g) Y. Shi. J. M. Glynn, L. J. Guilbert.
T. G Cotter. R . P. Bismonette, D. R. Green. Science 1992, 257. 212; h) P. H.
Watson. R. T Pon. R. P. Shiu. Cancer Rm. 1991. 51. 3996; i) L. Whitesell, D.
Geselowitr. C'. Chavany. B. Fahmy, S. Walbridge. I. R. Alger. L. M Neckers.
P r o < . .Xu//.A i d . Sci. USA 1993, 90, 4665; j ) J Lisziewicz, D. Sun, M. Klotman. S Agrawal. P. Zamecnik. R. Gallo. ihid. 1992. 89, 11 209: k j S. Agrawal,
J. Y. Tang. .An/i~seiiseR ~ s Drl'.
.
1992. 2. 261; 1) I. Katajima, 7.Shinohara, T.
Minor. L. Bibbs, J. Bilakovics. M. Nerenberg. J B i d . Chem. 1992,267.25 881:
in) M. Hikida. K.-i. Haruma, H. Ohmori, Immunoi. Lett. 1992, 34. 297; n) M.
Ebbcckc. C. Unterberg. A. Buchwald, S. Stohr. V. Wiegand. Basic Res. Car&
ol. 1992. K?. 5 8 5 ; o j B. J. Chiasson, M . L. Hooper. P R. Murphy. H. A.
Robertson. kin. J. Phurmucoi. 1992. 227.451 ; p j B. C. Paria, S. K . Dey. G. K.
Andrews. P r o < .iVatl. Acud Sti. USA 1992. 89. 10051; q) T. Lallier, M. Bronner-Frcw. Science 1993. 259. 692: r) A. Colige. B. P. Sokolov, P. Nugent. R.
Baserpa. D. J. Prockop, Brochemistry 1993. 32. 8; s) L. M. Cowsert. M. C. Fox,
A n ~ c i i -C
. ' h w / . 1/11 E d Engl. 1994. 33. 709-722
REVIEWS
G. Zon. C. K. Mirabelli. Anlimicrob. AgenrJ Chrmothrr. 1993. 37. 171. t ) ti.
Yokozaki, A. Budillon, G. Tomtora, S. Meissuer. S. L. Beaucage. K. Miki.
Y. S. Cho-Chung, Cuncer Res. 1993,53,868: u) 1. M. E. Leitler. S. Agrawal. P.
Palese. P. C. Zamecnik. Proc. Natl. Acad. Sci. liSA 1990. 87. 3430; v) J. W.
Jaroszewski, 0. Kaplan, J.-L. Syi, M. Sehested, P. L. Faustino. J. S. Cohen.
Cancrr Commun. 1990. 2, 287; w) E. Rapdport. K. Misiura. S. Agrawal. P.
Zamecnik, Proc. Null. Acud. Sci. USA 1992. 89, 8577; x) J c'. Reed. C. A.
Stein. C. Subasinghe, S. Haldar, C. M. Croce, S. Yum. J. S. Cohen, C'uncw Rc,.!.
1990, 50. 6565; y) J. Lisziewicz, D. Sun, V. Metelev. P. Zamecnik. R. C. Gallo.
S. Agrawal. Proc. Nurl. Acad. Sci. U S A 1993. 90. 3860; zj E. Bayever. M. R.
Boshop, G. Zon. J. Spinolo, A. Kessinger, Proi. Am. Crinrcr K P ~1993,34,593;
A. Gewirtz. ihid. 1993. 34. 595.
a) F. Eckstein, H. Gindl, Enr. J. Biochem. 1970. 13. 5 5 8 : b) F. Eckstein. W.
Armstrong. H. Sternbach, Prnc. Null. A m ( / . Sci. USA 1976. 73, 29x7: c ) T. A.
Kunkel, F Eckstein, A. S Mildvan. R. M. Koplitz. L. Loeb. ihid. 1981, 78.
6734: d) L. J. P. Latimer. K. Hampel, J. S. Lee. Nucleic . 4 m / i R r s 1989, 17.
1549.
a) W. J. Stec, G. Zon. W. Egan, B. Stec, 1 Am. Cheni. Soc. 1984, 106. 6077;
b) W. J. Stec. G. Zon. Trtruhedrun Lrrt. 1984. 25. 5275: cj ibrd. 1984. 25. 5279:
d) W. J. Stec. G. Zon, B. Uznanski. J Chromutugr. 1985. 326. 263
a) G. Zon. Pharm. Res. 1988. 5, 539; b) C. A. Stein, J. S. Cohen. Canter Rc.5.
1988. 48. 2659. c) J. S. Cohen in 0Lgonucieotide.s: Antiscnw Inliihitnrs <J/ Gcw
Eqmsrioii (Ed.: J. S. Cohen). Macmillan, London, 1989, p 12. d j E U h mann, A. Peyman, Chen7. Rev. 1990. 90. 544; e ) U . Englisch. D. H. Gauss,
Angrw. Cheni. 1991, 103,629; Angew. Chem. I n / . Ed. ElngI. 1991.30.6l3:f) M.
Ghosh. J. S. Cohen, f r o g . Nucleic. Acid. Re.s. .Mo/. Bid. 1992. 42. 79; gj S
Agrawal. Trend5 Eiotechno/. 1992. 10, 152.
a ) T. R. Cech, Curr. Opin. Strucf. B i d . 1992, 2, 605; b) S. M. Edgington. BIOfechnoiogy 1992, 10. 256; c) N . Sarver. An//sense Res. Dev. 1991. I , 373; d) J.
Goodchild, Nucleic Acid.<Res. 1992, 20. 4607; e) K. Taira. S. Nishikawa in
Gene Regulufion: Biology uf Antisense R N A und DNA (Eds.. R. P Erickson,
J. G. Izant), Raven, New York. 1992, p. 35.
a) W. Saenger. Principles of Nucleic Acid Structure, Springer, New York. 1984:
b) N . T. Thuong, C. Helene, Angrw. Chem. 1993. 105, 697; A n g m Cheni / n / .
Ed. Engl. 1993, 32. 666.
a) S . L. Beaucage. M. H. Caruthers, Te/ru/iedronLett. 1 9 8 1 . 2 . 1859; b) M. H .
Caruthers. Science 1985. 230. 281; cj S. L. Beaucage, R. P. lyer, E,/ruhc,[/run
1992, 48, 2233.
a) M. Vaman Rao. C. B. Reese, Z. Zhengyun. TefrahedronLetr. l992,33.4839;
b) R. P. lyer, W. Egan, J. 8. Ryan, S. L. Beaucage. J. Am. Ckem. So<. 1990.112.
1253, c) R. P. lyer, L. R. Phillips. E. Egdn. J B. Regan, S. L. Beaucage. J. 0 r . y . .
Cheni. 1990. 32, 4693: d) H. Vu, B. L. Hirschbein. Tetruhedroii L e i / . 1991, 32,
3005; ej H. C. P. F. Roelen. P. C. J. Kamer, H. van den Elst. 0.A . b a n dcr
Marel, J. H. van Boom, RecL Truv. Chmi. Puys-Bus 1991. 110, 325; f j P. C. 1.
Kamer, H. C. P. F. Roelen. H. van den Elst. G. A. van der Marel. J. H . van
Boom, Terrulicdron Lett. 1989, 30. 6757; g j W. J. Stec. B. Umanski, A . Wilk.
B. L. Hirschbein, K. L. Fearon, B. J. Bergot. ZJrrahedron Lcrr. 1993. 34.
5317.
a) B. C. Froehler, Tetrrihedron Lrrl. 1986, 27, 5575; b) B. C Froehler. M. D.
Matteucci. ibid. 1986, 27. 469.
a) L. Homer, Pure Appl. Chem. 1964, 225; b) W. C. McEuen in 7iipic.s i n
Pliosphorus Chetnistr!,, &)/. 2 (Eds.: M. Grayson, E. J. Griffth). Intersciencc.
New York. 1965. p. 25; cj W. Bentrude. L. H. Hargis. P. E. Rusek, J C/ii~n/,
Sot.. D 1969, 296; d) M. Mikolajczyk, J. Luczak. Tetruhedrun 1972. 28. 541 1 .
e) W. J. Stec, A. Okruszek, J. Michalski, Anjiew. Chrm. 1971. 83. 491; A n p w
Chc,m. I n / . Ed. Engl. 1971.10.494; f ) W. J. Stec. A. Okruszek. M. Mikobijczyk.
Z. Nufurforsc,h.B 1971, 26, 8 5 5 ; g) L. J. Siafraniec. L. L. Szafraniec. H. S.
Aaron. J Org. Chem. 1982, 47. 1936.
P. M. J. Burgers. F. Eckstein, E,/ruhedron L e f f .1978, 3x35.
J. F. Marlier. S. J. Benkovic, Tetruhedron L e r t . 1980, 21, 1121
A . Wilk, W. J. Stec. unpublished.
F. Eckstein, Annu. Rev. Biochem. 1985. 54. 376.
F. R. Bryant, S. J. Benkovic. Biochrmislrj~1979. 1X. 2825.
P. A. Frey in N e w Cnmprrhensiw Biochenrotry (Eds.: A. Neuberger. L L. L.
van Deenenj, Elsevier, Amsterdam, 1982. p. 204.
a) P. M. J. Burgers, F. Eckstein. Bioehemisrry 1979, 18,450; b) B. A. Connolly,
B. V. L. Potter. F. Eckstein, A. Pingoud, L. Grotphn, ibid. 1984.23.3443;c) G.
Slim. M. J. Gait, Nucleic Acids R ~ J1991,
.
19, 1183; d ) P. M . J. Burgers. F.
Eckstein, D. H. Hunneman. J B i d . Chem. 1979. 254, 7476; e) B. V. L. Potter.
1983.22, 1369; f ) M. Koziolkiewici.
B. A. Connolly. F. Eckstein. Bioch~~mrsfry
W. J. Stec. h i d . 1992, 31, 9460; gj D. R. Lesser, A. Gr;ijkowski, M. R.
Kurpiewski. M. Koziolkiewicz. W. J. Stec, L. Jen-Jacobson, J. Brol. Chrni. 1992.
267, 24510; hj A. Gupta. C. DeBrosse, S. J. Benkovic. ;bid 1982. 257. 7689;
i) P. M. J. Burgers, F. Eckstein. D. H. Hunneman, d i d 1979, 254. 7476;
j) B. V. L. Potter, B. A. Connolly, F. Eckstein. Biorhiwiutr! 1983, 22. 1369:
k) A. D. Grifiths. B. V. L. Potter, I. C. Eperon. Nurleic Acid\ Re.\. 1987. 15.
4145; I) A. P. Gupta, P. A Benkovic, S. J. Benkovic, ibid. 1984. 12. 5897.
Z. Otwinowski, R. W. Schervitz, R. G. Zhang. C. L. Lawson. A . Joachimiak.
R Q. Marmorstein, B. F. Luisi. P. B. Sigler. Nutirrr 1988, 335. 321.
L. A. Yakubov, E. A. Deeva. V. F. Zarytova. E. M. Ivanovd, A. S. Rytc, L. V.
Yurchenko. V. V. Vlassov. Proc. Null. Acud Sri. LISA 1989. Hh. 6454.
721
W. J. Stec and A. Wilk
REVIEWS
(151 G. Zon. M. F. Summers. K. A. Gallo. K.-L. Shao. M. Koziolkiewicr. B. ULnaiiski. W. J. Stec in 6iopho.sphore.r ond Their AriuIoguiv - Synrhmv. Sfrucfure.
Mc~iahr/li.\mi r r i r / A ( ~ / r i i(Eds.:
r~~
K. S . Bruzik. W. J. Stec). Elsevier. Amsterdam,
1987. p. 165.
1261 a ) FDA's Policy Statement for Development of Neu Stereoisomeric Drugr.
Food and Drug Administration: Rockville. MD. Miiy 1992: b) I . Amato.
S c i w < c 1992. 256. 964: c) J. Hodgron. B i o t c ~ h n o i o g1992,
~ ~ 10, 1093; d ) A.G.
Petsko. S ~ i i w c1992. 256. 1403: e ) S. C. Stinson, C ' h ~ ~ nEng.
r . NEW.\ 1992, 71.
46, f ) W. A. Nupent. T.V. RajanBabu. M. J Burk. S<ren(('1993.25Y.479: p) U.
Hacksell, S. Ahleniua, li.cnrl.\ Biorcchiiol. 1993. / I , 73.
[27] J. A Gerll. J. A . Coderrc. S. Mehdi, Arlr. En:i.mo/. 1983. 55. 279.
[28] a ) W.J. Stec. : l ; i r < ~ / m A(1d.s LS),rvp. .Sw. 1991. 2.1. 171, b) W.J. Slec. A. Grajko\vski. M. Koziolkiewic/. B Urnanski. ,Vid?r< A d \ Res. 1991. 19. 3 x 3 .
[29] D Valentine in A . \ w v w t r i Srnrhem lYK3-1Y84 1985. 4,263.
1301 a ) S Berner. K Muhlegger, H . Seliger. , V u d ~ i i A. c i r A Rcs. 1989. 17, 853:
h) 8. H. Dalil. J. Nielsen. 0. Dahl. hid. 1987. 15, 1729: c) Y. Watanabe. J.
C'/ivni. S m . Pcrkiri Trriiis I 1992. 2879.
[ i l l A.V. Lebedev. J. P. Rife. H. W. Selizsohn. G. R . Wensinger, E Wickstrom.
f i ~ i r u h d r o i rLcii. 1990. 31, 855
[3?] a ) T. Tanska. S. Tamattukuri. M. Ikehara. Te/nihrdrori Lctr. 1986. 27. 199, b) J.
Helinski. W. Dahkou,ski. J. M
[33] B. Uznanski. A. Wilk, W. J St
[34] B. Urnanski. A. Wilk, W. J. Stec, unpublished.
[is] a ) P. J. Garegg. T. Regberg. J. Stawinski. E. Stomberg. Chcni. S w . 1985, 25.
780; b) P. J. Garegg, I. Lindh. T. Regberg. J. Stawinski. R. Stroinberg, Terrohpi h w i Lerr. 1986. 27. 4051: c) A. Andrus, 1. W. Efcavitch. L. McBride. B. Gusti.
rhid. 1988, 29. 861
[36] a) 8. C. Froehlcr. P. G. Ng. M D. Matteucci. .Vii~/c,r( Acidr Rcs. 1986. 14,
160: b) C.A. Stein. C. Suhasinghc. K. Shinozuka. J. S. Cohen, h i d . 1988, 16.
3209.
[37] a ) F. Seela. U.Kretschmer. ,V~K/NJ,S&!. V ~ i c / ~ ~ u /1991.
d ~ ~ . 10.
\ 711 : b) J. Chvm
SOC. C ' h w i i . C o n m u r i . 1990, 1154; c) J Org. Chcn?. 1991. 56. 3861.
[38] a ) J. Svabbinski. R. Stromherg. R. Zain, T m o h e d r o n Lrrr. 1992, 3 3 . 3182: h) H.
Almcr. 1. Stdwinski. R. Stroinberg. M . Tlielin. J. Org. Chcuri. 1992. 57. 6163.
[39] ii) W B. F;irnham. R. K . Murray. K Mislow. JT Am. Chrrii. Soc. 1970. 92.
5809: b) J Donohue, N . Mandel.
B. Farnhani. M. H Murray. K Mislow.
H. P. Benschop. ihrd. 1971. 93. 3792.
1401 a ) L. P. Reiff. H. S. Aaron. J A m . Cho?i. S I X 1970. 92. 5275: b) L. J.
Smfraniec. L. P. Reiff. H. S. Aaron. hid 1970. Y2, 6393
[41] a ) M. Fujii. K. Ozaki. A. Kume, M. Sekine. T. Hata. Z?.truhet/rorr Lrtr. 1986.
26. 935: b) M. Fuji. K . Ozaki. M. Sekine, T. Hata. l i w o h r d r u n 1987. 43. 3395.
[42] a ) C. Battistini. M. G. Brasca. S. Fustinoni, .Vuc/~~o.rid~~s
Nuc/?oIi&\ 1991, 10.
723: b) C.Battistini. M. G. Brasca, S. Fualiiioni. E. Larrari. 7&vuhcdrort 1992.
4 8 3209.
[43] H. Oediger, F. Moller. K. Eiter. S J , n t h d s 1972. 591
[44] The influence of DBU on configurational stability of P-chirnl organophosphates requires further study. Unpublished results from our laboratory (Z. J.
Lesnikowski. L Wozniak) indicate that diastereomers of dinucleoside 3 ' 3
metliaiiephosphonatcs are not cpinierized 111 the presence of DBU under conditions that lead to complete P-epimenzation of diastereomcrically pure
or nucleoside 3'-0nncleoside 3'-O-(O-4-nitrophenyln~ethanephosphonate)s
(Se-methyl-met1innephosphonate)s. Therefore. these last results can be explained in terms of DBU-catalyzed hydrolysis caused by traces of water: released p-nitrophenoxidc ion or methylselenenyl ion. respectively, may cause
epimerization of'pirent compounds in the presence of DBU by a ping-pong
type mechanism
1451 a ) C. B. Reese. Tc>rruhcrlr-on1978. 34. 3143: b) M.Ikehara. E. Ohtsuka. A. F.
M;irkh;im. 4rlv. C w h o / f ~ ~ iChon.
lr.
Biochem. 1978. 36. 135, c) R. Wu, C. Bahl,
S. A . Narang. Prog. Niuleic Acid R c . M o l . Brol. 1978. 21. 101: d ) K. Itakura.
A. D. Riggs. S w r i c r 1980. 2OY. 1401: e ) S. A. Narang. Terruhedron 1983, 39, 3:
f) B. S. Sproat. M. J. Gait in O/;gonudr,ofidc>,T?nlhrsr.\; A P r u c t i d Approrich
(Ed.: M .I. Gait). IRL. Oxford, 1984: g) J. E. Marugg. C. Van Den Berph, J.
Tromp, G. G. Van Der Marel, J. Van Zoest. J. H. Van Boom. iVucl6~icAcids Re,!.
1984. I2. 9095.
1461 E. Ohtsuka. M .Shiraishi. M . Ikehara, Tetrahedron L e r r . 1985. 41. 5271
[47] R. Cosqtick. D.M . Williams. . V n c l ~ i cAcrds Res. 1987. 15, 9921.
1481 a) D.G.Knorre, A. V. Lebedev. A. V. Zarytova. Nucleic. Acid\ Rm. 1976. 3,
1401 ; b) D. 0. Knorre. V. F. Zarytowa in Pho.sphoru.r C'h<mir\trJ~Direcred To-
722
wurdr B i o l o ~(Ed.:
~
W. J. Stec). Pergamon, Oxford. 1979. p 13; c) €. H.
Ivanova. L. M. Khalimskaya. V. P. Romanienko. V. F. Zarytova. Terruliedron
Lrir. 1982. 23. 5447. d ) V. E Zirytovd. D. G. Knorre. N u c h Ack1.r RK 1984,
f2,2091
[49] W. Niwiarowski. Z .I. Lesnikowski. A. Wilk. P.Guga. A. Okruszek. B. Uzniinski, W. 1. Stec. k t u Bi0dibn Pol 1987. 34. 217.
[50] a ) Z. J. Lesnikowski. A. Sibinska. Plrrrhetlron 1986, 42, 5025: b) 2. J
Lesnikowski. M. Jaworska. Tc~ruhedroii Lrrr. 1989. 30. 3821, c) Z. J.
Lesnikowski. >Vric/imic/<,s .Nw/eoridm 1992. i 1 . 1621.
[51] a ) Y. Hayakawa, M. Uchiyama. R. A. Noyori. Z,rruliedron Lcrr. 1984. 25,
4003: b) M. Uchiyama. Y Aso, R. Noyori. Y. Hayakawa. J. Org. Chem. 1993,
58. 373.
[S2] M Jaworska-Maslanka. Ph. D . 7 h w s . University of Lodz, 1992.
1531 a) P. Guga. W. J. Stec. Tetrahedron LPII. 1983, 24. 3899: h) A . Okruszek, P.
Guga, W. J Stec, J. Chcrn. So(. C'hem. Cun?rnun. 1985. 1225; c) A. Okruszek, P.
Guga, W. J. Stec in Biophosphu~r~
and Thew Anuhgrres - .S~~nrheti.v.
Sfruclurr.
Memholicm u n d A r t w i n (Eds.: K. S. Bruzik. W. J. Stec). Elsevler. Amsterdam.
1987. p. 247: d) A. Okruszek, P. Guga. W. J. Stec, J. Chern. Soc. Chern. Corn
m r n . 1987, 5957. e) Hcrarout Chern. 1991. 2. 561.
[54] €3. Urnanski, A. Grajkowski. 8 . Krryranowska. A. Kazmierkowska, W. J. Stec.
M. W. Wieczorek. J Blaszcryk. J. An7. Chrn?. Soc. 1992. 114. 10 197.
[55] F. H. Westheiiner in Reurrungrrnmrs it1 Ground & Erricd Starcs (Ed.: P. de
Mayo). Academic Press, New York, 1980. p. 229.
[Sh] a ) K:P. Stengele. W. Pfleidercr. N u d v i ~ Acid,r Sjrnp. Ser. 1989, 21. 101;
b) 7i,rruhedron Lt,rr. 1990, 31. 2549; c) T Broun. C. E. Pritchard. G. Turner.
S. A. S a l i h r q . J Chem. Soc. Chon. Cornrnim. 1989, 891 ; d) C. Lehman. Y-Z.
Xu. C. Christolonlor. Z. K. Tan. M. J. Gait. iVirclex Acids Res. 1989, f7. 2379.
[57] A. Suska. A. Grajkowski, A. Wilk. B. Uznanski, J. Blaszczyk. M . Wieczorek.
W. J. Stec. Pure App/. C h m 1993, 6.7, 707.
[SX] A. Grajkowski. A. Wilk. M. KoriolkiewicL, W. J. Stec. unpublished
[59] The T,, value For heteroduplex (nll-R,)-d[(A,,),,A],'T,, given in ref. [57] was
determined incorrectly. All measurements were repeated with diastereomeric
dodeca(adcnorine phosphorothioate)s of sire-homogeneity > 99% as checked
by capillary electrophoresis. The diastereomeric purity the of (all-R,)
diastereomcr M B S proven by its treatment with nuclease P1. Material recovered
by RP-HPLC was converted into sodium salt by precipitation. Measurements
were performed in HEPES buffer (pH 5.4. 100mM HEPES. 2OmM MgCI,.
200 mM NaCI) with direct temperature monitoring inside the UV cell. The
results are as follows: dA,zdT,, 37.4 C, d[(A,,),,A] (mixture of all
diastereomers. ca. 58% ( R p )configuration):dT,, 31.6 C; (all-R,)-d[(A,,),,A]:
dT,, 30.0'C, (all-S,)-dI(A,,),,A]:dT,, 34.4 C.
[6O] L. H.Parker-Botelho. L C. Webster. J D. Rothemel. J. Bardniak. W. J. Stec. J.
Biol. Chm7. 1988. 26.1. 5301
[61] J. W. Jaroszewski. J.-J. Syi, J. Maizel. J. S. Cohen, Ant-Cunccr Drug Des. 1992.
7. 253.
1621 M. E. Piotto. J. N. Granger, Y, Cho. N. Farshtschi. D G. Gorenstein. Teiruhrdron 1991, 47. 2449.
[63] a) M. H. Caruthers. G. Beaton, L. Cummins. Nucleosides Nucleotides 1991, /(I.
47, h) K. Bjergarde. B. H. Dahl, 0. Dahl, ibid. 1991, 10. 461; c) K. Bjerparde,
0 . Dahl. Nuc/cic A<rd\ R e . 1991. 19, 5843.
[64] B. Karwowski. A. Kobylanska. A. Sierzchala, A. Okruszek. W. J. Stec, unpubI ished .
[65] A. OkrusLek, A. Sierzchala, M. Sochacki. W. J. Stec, Tetruhedron Lett. 1992.33,
7585.
[66] E. K. Yan. Y-X. Ma. M. A. Caruthers. Trrtrahedron L r f i . 1990.31. 1953.
[67] W.Dabkowski. A. Lopusinski. J. Michalski. C. Radziejewski. Phosphorus S d
f u r Relo,. Elern. 1980. 8. 375.
1681 a) G.H. Smith. C. N Cauhlan. F. Ramirez, S. Glaser. P. Stern. J. Am. Chein.
Soc. 1974, 96, 2698. and references therein.; b) R. Ramage in Orgunophosphom.!Reupnt.s rn Orgonrc Sinthesis (Ed.: J. I. G. Cadogan), Academic Press.
1979. p. 51 1 ; c) A. G. Jackson, G. W. Kenner, G. A. Moore. R. Ramage. Terraheilron Lrir 1976. 3627: d) P. C. Crofts, Org. Phosphorus Comp. 1972-1976
1973. 6, 51; e) A. .I. Lambie, R,rrahedron L e t t . 1966, 3709: f) N.E. Jacobsen.
P. A . Bartlett. J. Am. Chew. S o ( . 1983. 105. 1613, g) ihrd. 1983. 105. 1639.
[69] L.Wozniak, W.J. Stec, unpublished.
[70] G. M. Richards, G. d u Vair. M . Laskowski. Sr., Biochernbrr~1965, 4,501.
1711 W.K ullmann, Eiizymritic Peptide S.vnrhesis. CRC. Boca Raton. FL. 1987.
[72] B. Urnanski, W.Niewiarowski, W. J. Stec. Tetrahedrorr Letr. 1982. 23, 4289.
Anyew. C'hmi.
In/.Ed. Engl. 1994. 33, 709 722
Документ
Категория
Без категории
Просмотров
9
Размер файла
1 487 Кб
Теги
synthesis, oligo, stereocontrolled, phosphorothioate, nucleoside
1/--страниц
Пожаловаться на содержимое документа