Nucleoside H-phosphonates. XXI. Synthetic and 31P NMR studies on the preparation of dinucleoside H-phosphonoselenoates

Article abstract of DOI:10.1135/cccc20060820

Using ³¹P NMR spectroscopy as a tool, we investigated the condensation of H-phosphonoselenoate monoesters with a hydroxy component in the presence of various coupling agents. On this basis, synthetic protocols that eliminate or significantly su

Full text of DOI:10.1135/cccc20060820

820
Kullberg, Bollmark, Stawinski:
NUCLEOSIDE H-PHOSPHONATES. XXI. SYNTHETIC AND
31P NMR STUDIES ON THE PREPARATION OF DINUCLEOSIDE
H-PHOSPHONOSELENOATES
a
a
a1,b,
Martin KULLBERG , Martin BOLLMARK and Jacek STAWINSKI
*
a
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University,
S-106 91 Stockholm, Sweden; e-mail: js@organ.su.se
Institute of Bioorganic Chemistry, Polish Academy of Sciences, 61-704 Poznan, Poland
1
b
Received March 1, 2006
Accepted April 14, 2006
Dedicated to Professor Antonín Holý on the occasion of his 70th birthday.
Usin g 31P NMR spectroscopy as a tool, we in vestigated th e con den sation of H-ph osph on o-
selen oate m on oesters with a h ydroxy com pon en t in th e presen ce of various couplin g agen ts.
On th is basis, syn th etic protocols th at elim in ate or sign ifican tly suppress side reaction s
wh ich m igh t occur durin g th e con den sation in syn th esis of H-ph osph on oselen oate deriva-
tives, h ave been developed. Absolute con figuration of th e produced din ucleoside H-ph os-
ph on oselen oates h as been determ in ed by stereoch em ical correlation an alysis.
Keywords: H-Ph osph on oselen oates; H-Ph osph on ates; Couplin g agen ts; 31P NMR spectroscopy;
Absolute con figuration assign m en t; Din ucleosides; Oligon ucleotides.
As part of our research in H-ph osph on ate ch em istry1,2 directed toward de-
velopm en t of n ew syn th etic m eth ods for biologically im portan t ph osph ate
esters an d th eir an alogues, we h ave recen tly em barked on syn th etic studies
on H-ph osph on oselen oate esters. To provide an easy access to th is n ew class
of ph osph on ate an alogues, we h ave developed two efficien t m eth ods for
th e preparation of n ucleoside H-ph osph on oselen oate m on oesters3,4 an d
sh owed th at th ese com poun ds can be con verted to th e appropriate H-ph os-
ph on oselen oate diesters5 by con den sation with suitably protected n ucleo-
sides.
H-Ph osph on oselen oates are a relatively un explored class of ph osph orus
com poun ds6–8. Due to th e presen ce of selen ium boun d to a ph osph orus
atom at +3 oxidation state, th ese com poun ds can be con sidered as n ovel,
poten tially useful syn th etic in term ediates for th e preparation of n ew ph os-
ph ate an alogues or selen ium -con tain in g derivatives th at m igh t be rath er
difficult to obtain by classical procedures in volvin g oxidation of P(III) com -
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831
© 2006 Institute of Organic Chemistry and Biochemistry
doi:10.1135/cccc20060820

Nucleoside H-Phosphonates
821
poun ds with selen ium . Moreover, due to th e possibility to con trol stereo-
ch em istry at th e ph osph orus cen tre durin g th e oxidation step, P-ch iral
ph osph ate an alogues with defin ed stereoch em istry can be obtain ed.
Replacem en t of on e oxygen in H-ph osph on ate esters by selen ium h as im -
portan t stereoch em ical an d ch em ical con sequen ces. At th e m on oester level,
sim ilarly to H-ph osph on oth ioate m on oesters9,10, it in troduces ch irality at
th e ph osph orus an d th is m ay poten tially be exploited in stereospecific syn -
th esis of ph osph oroselen oates an d oth er ch iral ph osph ate an alogues. Th e
presen ce of selen ium , h owever, m akes two differen t n ucleoph ilic cen ters (O
an d Se) available for activation by a con den sin g agen t. Lack of ch em o-
selectivity in th is process m ay cause a partial elim in ation of selen ium in th e
course of th e reaction , wh ich can lead to form ation of side products. At th e
diester level, substitution of oxygen by selen ium at th e ph osph orus cen ter
causes th e P–H bon d in H-ph osph on oselen oates to becom e n oticeably m ore
reactive th an th at in th e oxygen con gen er. Th is, in prin ciple, can also be a
poten tial source of side reaction s.
Havin g in m in d possible application s of H-ph osph on oselen oates as pre-
cursor to various oligon ucleotide an alogues, we h ave un dertaken system atic
studies on th is class of com poun ds. In th is paper we address th e problem of
possible side reaction s wh ich m ay occur in preparation of H-ph osph on o-
selen oate diesters by con den sation of suitably protected n ucleoside 3′-H-
ph osph on oselen oate m on oesters with n ucleosides. Various couplin g agen ts
an d reaction con dition s for th e con den sation h ave been evaluated. Th ese
en abled us to develop optim al syn th etic protocols wh ich elim in ated or sig-
n ifican tly suppressed side reaction s durin g syn th esis of din ucleoside
H-ph osph on oselen oate derivatives. Absolute con figuration of th e produced
H-ph osph on oselen oate diesters h as been determ in ed by stereoch em ical cor-
relation an alysis.
RESULTS AND DISCUSSION
Din ucleoside H-ph osph on oselen oates 3 can be produced in a con den sation
reaction (Sch em e 1) usin g n ucleoside H-ph osph on oselen oates 1 as a n ucleo-
tidic m aterial. It was expected, h owever, th at th is rath er straigh tforward re-
action for H-ph osph on ates, m igh t be syn th etically m ore com plex in th e case
of H-ph osph on oselen oate derivatives. Sim ilarly to th e correspon din g th io
an alogues, H-ph osph on oselen oate diesters m ay be m ore pron e to subse-
quen t reaction s durin g con den sation , due to h igh er reactivity of th e P–H
bon d. In addition , th e startin g m on oesters m ay react with con den sin g
agen ts in two differen t ways (O vs Se activation ) producin g differen t prod-
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

822
Kullberg, Bollmark, Stawinski:
ucts. For th ese reason s, th e ch oice of a con den sin g agen t in H-ph osph on o-
selen oate diester syn th esis can be m ore critical th an in th e case of H-ph os-
ph on ate or H-ph osph on oth ioate derivatives.
Differen t classes of con den sin g agen ts, such as acyl ch lorides, ch loro-
ph osph ates, arylsulfon yl ch lorides, an d carbodiim ides, h ave been evaluated
from th e poin ts of view of th eir (i) efficien cy to prom ote form ation of H-
ph osph on oselen oate diesters, (ii) ch em oselectivity durin g th e activation of
H-ph osph on oselen oate m on oesters, an d (iii) reactivity toward th e P–H
bon d in H-ph osph on oselen oate diesters.
To h ave a un iform set of data, all con den sation s h ave been carried out in
pyridin e with variable am oun ts of a con den sin g agen t, an d progress of th e
reaction s was m on itored by 31P NMR spectroscopy.
SCHEME 1
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

Nucleoside H-Phosphonates
823
Pivaloyl Chloride as a Condensing Agent
Sin ce pivaloyl ch loride (Pv-Cl) is com m on ly used as th e couplin g reagen t in
th e autom ated syn th esis of DNA an d RNA fragm en ts via th e H-ph os-
ph on ate approach 11, it was our in itial ch oice for th e H-ph osph on oselen oate
diester form ation (Sch em e 1). Un fortun ately, in th e reaction of H-ph os-
ph on oselen oate 1 (1.1 equivalen ts) with n ucleoside 2 in th e presen ce of
3 equivalen ts of Pv-Cl in pyridin e, th e startin g m aterial 1 (δP 44.7 an d
45.2 ppm ) rapidly disappeared (<5 m in ) but th ere was n o sign als th at could
be assign ed to th e desired H-ph osph on oselen oate diester 3 (two reson an ces
1
expected at ca. δP 76 ppm with JPH ca. 650 Hz). In stead, th e m ajor products
form ed gave rise to a sign al at δP 140.0 (ca. 30%) an d two sin glets at δP 72.6
an d 72.3 (ca. 30%). Th e form er sign al was ten tatively assign ed to trin ucleo-
side ph osph ite 7 (Ch art 1), an d th e two latter sign als, on th e basis of th eir
ch em ical sh ifts an d th e absen ce of a P–H bon d, to diastereom eric P-acylated
derivatives 5. Am on g oth er side products observed in th is couplin g reaction
CHART 1
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

824
Kullberg, Bollmark, Stawinski:
th ere were diastereom eric H-ph osph on ate diesters 4 (ca. 15%, δP 9.6 an d
1
8.0 ppm , JPH = 714 an d 7.17 Hz), din ucleoside acyl ph osph ite 8 (ca. 10%,
δP 133.8 an d 132.9 ppm ) an d probably P-acylated m ixed an h ydride 6
(ca. 10%, δP 53.6 an d 53.4 ppm ).
Th e observed product distribution in th is reaction was con sisten t with a
tran sien t form ation of H-ph osph on oselen oate 3, wh ich in th e presen ce of
pivaloyl ch loride un derwen t rapid P-acylation (to form 5) or activation on
th e selen ium cen tre to form ultim ately ph osph ite derivatives 7 an d 8. Sin ce
H-ph osph on oselen oate 1 was used in a sligh t excess, its un reacted part
could accoun t for th e presen ce of P-acylated m ixed an h ydride 6.
As to th e form ation of H-ph osph on ate diester 4, two reaction path ways
were possible: activation of th e selen ium cen tre in H-ph osph on oselen oate
m on oester 1 by pivaloyl ch loride an d elim in ation of selen ium upon cou-
plin g with a n ucleoside, or con version of pivaloyl ph osph ite 8 in to H-ph os-
ph on ate diester 4 by h ydrolysis with adven titious water or via deacylation .
Irrespective of th e reaction path ways in volved, it seem ed th at H-ph os-
ph on oselen oate diesters 3 were sign ifican tly m ore pron e to side reaction s
th an th e correspon din g H-ph osph on oth ioates12. For th e latter com poun ds,
un der an alogous reaction con dition s, th e side reaction s am oun ted to ca.
10% an d th e desired H-ph osph on oth ioate diesters were form ed as m ajor
products of th e reaction . Also, in th e case of H-ph osph on oth ioate diesters,
th ere were n o H-ph osph on ate diesters form ed, th e presen ce of wh ich m igh t
in dicate in com plete ch em oselectivity durin g activation .
A dram atic im provem en t was observed wh en th e above couplin g reaction
was carried out in th e presen ce of 1.5 equivalen ts of pivaloyl ch loride. In
1
th is case th e desired H-ph osph on oselen oate 3 (δP 74.8 an d 76.7 ppm , JPH
=
1
648 an d 644 Hz, JPSe = 877 an d 879 Hz) was form ed as a m ajor product
(ca. 80%) togeth er with sm all am oun ts of ph osph ite triester 7 (<5%),
H-ph osph on ate diester 4 (ca. 5%), an d com poun d 6 (ca. 5%). Sim ilarly to
H-ph osph on oth ioate diesters12, th e 31P NMR reson an ces due to 3 were also
accom pan ied by two very sm all sign als at th e h igh er an d lower field. Th ese
sign als (δP 70.3 an d 73.8 ppm ), wh ich in creased in th e course of tim e an d
am oun ted to ca. 5% of th e total n ucleotidic m aterial after 5 m in , were as-
sign ed to th e ligan d exch an ge products, sym m etrical din ucleoside H-ph os-
ph on oselen oates 9 an d 10.
Th e m ech an ism of th e form ation of 9 an d 10 is n ot kn own . We can on ly
speculate th at th ese com poun ds are form ed in a pyridin e-m ediated ligan d-
exch an ge process, wh ich leads to scram blin g of th e substituen ts at th e
ph osph orus cen tre of th e H-ph osph on oselen oate diesters 3 13. Con sisten t
with th is, th e ligan d exch an ge process was sign ifican tly suppressed upon
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

Nucleoside H-Phosphonates
825
ch an gin g th e reaction m edium to aceton itrile–pyridin e (4:1, v/v), an d prac-
tically elim in ated upon replacem en t of pyridin e by 2,6-lutidin e. Th e latter
reaction was rapid an d rath er clean . Un fortun ately, alth ough H-ph os-
ph on ates 4 were form ed on ly in trace am oun ts (<2%), th ere was an in crease
in th e P-acylation of H-ph osph on oselen oate 3 (form ation of acylph os-
ph on ate 5, ca. 10%), m ost likely due to h igh er basicity of 2,6-lutidin e th an
th at of pyridin e.
By observin g tren ds in product distribution upon varyin g th e am oun t of
pivaloyl ch loride an d reaction con dition s, on e can con clude th at appar-
en tly th e activation step of H-ph osph on oselen oates 1 is n ot com pletely
ch em oselective (selen ium activation m igh t occurs to som e exten t) an d th at,
due to rath er h igh reactivity of th e P–H bon d, th is couplin g agen t m igh t
h ave a lim ited application in H-ph osph on oselen oate diesters syn th esis.
Chlorophosphates as Condensing Agents
Due to low reactivity of S-n ucleoph iles towards th e ph osph orus cen tres14,
we expected th at th e use of ch loroph osph ates to produce H-ph osph on o-
selen oate diesters sh ould alleviate problem s con n ected with subsequen t
reaction s of 3 with con den sin g agen ts. To th is en d we carried out con den -
sation s of H-ph osph on oselen oate 1 with a n ucleosidic com pon en t 2 in th e
presen ce of diph en yl ph osph oroch loridate15 (DPCP), dieth yl ph osph oro-
ch loridate16 (DECP), 2-ch loro-5,5-dim eth yl-2-oxo-1,3,2-dioxaph osph in an e16
(NEP-Cl), an d bis(2-oxo-oxazolidin -3-yl)ph osph in ic ch loride16,17 (OXP-Cl)
(Sch em e 1).
In deed, all th ese reagen ts, wh ich h ave previously been evaluated in th e
syn th esis of H-ph osph on ate16 an d H-ph osph on oth ioate diesters12, pro-
m oted also clean form ation of th e H-ph osph on oselen oates 3 wh en used in
a 3 equivalen ts excess over th e startin g m aterial 1. Possible side products,
wh ich m igh t h ave resulted from th e subsequen t reaction of 3 with th e con -
den sin g agen ts (form ation of ph osph ite triester 7 or h ypoph osph ates18),
were n ot observed. All th e in vestigated ch loroph osph ates sh owed also a
com plete ch em oselectivity in th e activation process of 1, as judged from
th e absen ce of H-ph osph on ate diesters of type 4 in th e reaction m ixtures.
Th e couplin g reaction s prom oted by all ch loroph osph ates in vestigated
were very fast in pyridin e (<5 m in ), but differed in kin etics wh en th e reac-
tion s were carried out in th e presen ce of lim ited am oun ts of pyridin e.
DPCP an d DECP as con den sin g agen ts gave in distin guish able results both
in term s of kin etics an d purity of th e produced H-ph osph on oselen oates 3.
Both of th em prom oted rapid (<5 m in ) con den sation s in n eat pyridin e an d
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

826
Kullberg, Bollmark, Stawinski:
in aceton itrile con tain in g on ly 5 equivalen ts of th e n ucleoph ilic catalyst.
In th e reaction s in pyridin e, traces of th e ligan d exch an ge products (com -
poun ds 9 an d 10) could be observed after 5 m in , but wh en th e am oun t of
pyridin e was reduced to few equivalen ts, th ese side products could n ot be
detected even after 30 m in (31P NMR spectroscopy). For th e DECP-
prom oted con den sation s in n eat pyridin e, form ation of un iden tified side
product (δP 101.7 ppm , ca. 5%) was usually observed.
A sterically h in dered ch loroph osph ate, NEP-Cl, in n eat pyridin e pro-
m oted rapid (<5 m in ) an d clean con den sation , but in aceton itrile con tain -
in g on ly 5 equivalen ts of pyridin e, th e con den sation tim e in creased to
60 m in . For both reaction s, th e ligan d exch an ge products 9 an d 10 were ob-
served (<5%), due to h igh con cen tration of pyridin e or th e exten ded reac-
tion tim e (for th e reaction s with lim ited am oun ts of pyridin e). Sim ilarly to
th e reaction s in n eat pyridin e in wh ich DECP was used, also in th is case, a
by-product reson atin g at δP 100.7 ppm (ca. 2%), was form ed.
Th e last ch loroph osph ate in vestigated, OXP-Cl sh owed like NEP-Cl, a
sen sitivity to th e am oun t of pyridin e presen t in th e reaction m ixture, but
to sm aller exten t. For th is con den sin g agen t, th e reaction tim e in aceton i-
trile in th e presen ce of 5 equivalen ts of pyridin e was ca. 30 m in , an d th is
apparen tly secured a clean couplin g with out form ation of n oticeable
am oun ts of th e ligan d exch an ge products (31P NMR spectroscopy).
TPS-Cl-Promoted Condensations
Th e reaction of H-ph osph on oselen oate 1 (1.1 equivalen ts) with n ucleoside
2 in pyridin e in th e presen ce of 3 equivalen ts of 2,4,6-triisopropyl-
ben zen esulfon yl ch loride (TPS-Cl) produced after 5 m in a com plex m ixture
of products, n on e of th em con tain in g a P–H bon d. In th e ligh t of our previ-
ous in vestigation s of TPS-Cl as a con den sin g agen t in th e syn th esis of
H-ph osph on ate diesters19, an d th e fact th at aren esulfon yl derivatives did
n ot sh ow ch em oselectivity in activation of th e correspon din g H-ph os-
ph on oth ioate derivatives12, studies with th is class of con den sin g agen ts
were n ot pursued furth er.
Carbodiimides as Coupling Agents
Two carbodiim ides, n am ely dicycloh exyl (DCC) an d 1-[3-(dim eth ylam in o)-
propyl]-3-eth yl (EDC) derivatives (Sch em e 1), were evaluated in reaction s of
H-ph osph on oselen oate m on oesters 1 (1.1 equivalen ts) with n ucleoside 2
prom oted by con den sin g agen ts. Th e reaction s were carried out in pyridin e
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

Nucleoside H-Phosphonates
827
or in aceton itrile–pyridin e (4:1, v/v) usin g pyridin e h ydroch loride
(3 equivalen ts) as acid catalyst. With 3 equivalen ts of th e con den sin g
agen ts, th e reaction s were com plete in less th an 5 m in affordin g as m ajor
product (98%) din ucleoside H-ph osph on ate 4, irrespective of th e kin d of
th e carbodiim ide or th e reaction con dition s used. Always, a sm all am oun t
of H-ph osph on oselen oate 3 (1–2%) was presen t in th e reaction m ixtures,
in dicatin g an in com plete ch em oselectivity durin g th e activation step. Oc-
casion ally, also variable am oun ts (0–1%) of th e correspon din g n ucleoside
H-ph osph on ate m on oester were form ed durin g th e couplin g reaction s,
m ost likely due to th e presen ce of adven titious water.
Sin ce EDC is com m ercially available in th e form of its h ydroch loride,
we evaluated th is con den sin g agen t with out extern al acid catalyst added.
Th e couplin g reaction of H-ph osph on oselen oate 1 an d n ucleoside 2 in
aceton itrile–pyridin e (4:1, v/v) un der such con dition s was sign ifican tly
slower (30 m in for th e disappearan ce of 1) but sh owed im proved ch em o-
selectivity. Un fortun ately, alth ough th e reaction produced H-ph osph on ate
4 (ca. 70%) with out detectable am oun ts of H-ph osph on oselen oate diester 3,
a sign ifican t portion of 1 (ca. 30%) was usually con verted to th e corre-
spon din g n ucleoside H-ph osph on ate m on oester.
To con clude th is part, on th e basis of th e above 31P NMR studies we
foun d th at am on g th e various con den sin g agen ts in vestigated on ly ch loro-
ph osph ates secured clean form ation of H-ph osph on oselen oate diesters 3 in
th e couplin g reaction between n ucleoside H-ph osph on oth ioate m on oesters
1 an d th e appropriate h ydroxy com pon en t 2 (Sch em e 1), with com plete
ch em oselectivity in th e activation step. To m in im ize th e ligan d exch an ge
process, th e con den sation tim e sh ould be as sh ort as possible an d th e reac-
tion s sh ould be preferably carried out in n eutral solven ts con tain in g lim -
ited am oun ts of pyridin e.
Takin g th is in to con sideration , din ucleoside H-ph osph on oselen oate 3
was syn th esised on a preparative scale in aceton itrile con tain in g pyridin e
(5 equivalen ts) usin g DPCP or DECP as con den sin g agen ts. After work-up
an d ch rom atograph y on a silica gel colum n , com poun d 3 (1:1 m ixture of
th e diastereom ers) was obtain ed in a yield h igh er th an 80% yield. Th e
P-diastereom ers of 3 sh owed sufficien tly large differen ces in m obility on
silica gel to separate th em by th e ch rom atograph y.
Absolute Configuration of Diastereomeric H-Phosphonoselenoates 3
As a fin al stage of th ese in vestigation s, assign m en t of th e absolute stereo-
ch em ical con figuration at th e ph osph orus atom in H-ph osph on oselen oate
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

828
Kullberg, Bollmark, Stawinski:
diesters 3 was un dertaken . Th e diastereom eric H-ph osph on ate an d th e
H-ph osph on oth ioate diesters with kn own con figuration were used as refer-
en ces20–22. Th e absolute con figuration of H-ph osph on ate diesters was deter-
m in ed previously on th e basis of sulfurization (proceedin g with reten tion
of con figuration ), followed by stereospecific degradation of th e produced
ph osph oroth ioates with SVPDE an d P1 n uclease21,22. By em ployin g
desulfurization with m-ch loroperben zoic acid (m CPBA), th e con figuration
of H-ph osph on oth ioates was correlated with th at of th e correspon din g
H-ph osph on ates20.
For H-ph osph on oselen oate diesters, we perform ed a sim ilar stereo-
ch em ical correlation an alysis in order to determ in e th e absolute con figura-
tion of th e din ucleoside H-ph osph on oselen oates 3. Th us, treatm en t of th e
diastereom er of H-ph osph on oselen oate 3 dem on stratin g h igh er m obility
on silica gel (“fast” isom er) with m CPBA yielded th e correspon din g
H-ph osph on ate diester 11 with RP con figuration 20–22 (Sch em e 2). Sin ce
SCHEME 2
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

Nucleoside H-Phosphonates
829
deselen ization with m CPBA m ost likely occurs with reten tion of con figura-
tion aroun d th e ph osph orus atom , th is defin es RP con figuration for th e
“fast” isom er of 3.
To substan tiate th is assign m en t, th e “fast” H-ph osph on oselen oate 3 was
subsequen tly sulfurized usin g elem en tal sulfur. Th is afforded th e corre-
spon din g (RP)-ph osph oroth ioselen oate diester 13 20, wh ich was iden tical to
th e product obtain ed via selen ization of th e (SP)-H-ph osph on oth ioate 12
(Sch em e 2).
A sim ilar correlation an alysis perform ed on th e “slow” isom er of 3 con -
firm ed th ese assign m en ts. Th us, we can con clude th at th e “fast” H-ph os-
ph on oselen oate 3 h as th e RP con figuration , wh ile th e “slow” isom er, SP.
In con clusion , usin g 31P NMR spectroscopy we in vestigated th e con den sa-
tion of n ucleoside H-ph osph on oselen oate m on oester 1 with a n ucleosidic
com pon en t un der various experim en tal con dition s, an d on th is basis devel-
oped an efficien t protocol for th e preparation of din ucleoside H-ph osph on o-
selen oate diesters 3. Absolute con figuration of th ese P-ch iral com poun ds
was determ in ed by stereoch em ical correlation an alysis.
EXPERIMENTAL
Materials an d Meth ods
Pyridin e, 2,6-lutidin e, aceton itrile, an d trieth ylam in e (TEA) were refluxed with CaH2, th en
distilled an d stored over 4Å m olecular sieves or CaH2 (TEA). Pivaloyl ch loride (Pv-Cl), di-
eth yl ph osph oroch loridate (DECP), diph en yl ph osph oroch loridate (DPCP), 2,4,6-triisopropyl-
ben zen esulfon yl ch loride (TPS-Cl), bis(2-oxo-oxazolidin -3-yl)ph osph in ic ch loride (OXP-Cl),
dicycloh exylcarbodiim ide (DCC), an d 1-[3-(dim eth ylam in o)propyl]-3-eth ylcarbodiim ide
(EDC), were of com m ercial grade (Aldrich ).
Nucleoside H-ph osph on oselen oate m on oester 1 was prepared by ph osph in ylation of
5′-O-(4,4′-dim eth ytrityl)th ym idin e, followed by selen ization of th e produced n ucleoside
ph osph in ate5. 2-Ch loro-5,5-dim eth yl-2-oxo-1,3,2-dioxaph osph in an e (NEP-Cl) was obtain ed
usin g th e publish ed procedure23. TLC an alyses were carried out on Merck silica gel 60 F254
precoated plates usin g th e followin g eluen ts: toluen e–m eth an ol (9:1, v/v; system A); ch loro-
form –m eth an ol (9:1, v/v; system B). 1H, 13C an d 31P NMR specra were m easured in CDCl4;
ch em ical sh ifts (δ scale) are given in ppm , couplin g con stan ts (J) in Hz. Th e 31P NMR experi-
m en ts con cern in g th e form ation of H-ph osph on oselen oate diesters 3 from 1 an d 2 were car-
ried out in 10-m m tubes usin g 0.05 m m ol of ph osph orus-con tain in g com poun ds 1 in 2 m l
of a solven t. H3PO4 (2%) in D2O was used as extern al stan dard (coaxial in n er tube). Th e
am oun ts of th e h ydroxy com pon en t 2 an d a couplin g agen t an d th e solven t com position
are as in dicated in th e text. Th e values of th e ch em ical sh ifts for th e in term ediates produced
in situ in som e experim en ts varied (±1 ppm ) depen din g on th e reaction con dition s. A ten ta-
tive assign m en t of structure of in term ediate or by-products 4–10 observed durin g th e cou-
plin g reaction was don e on th e basis of kn own or expected 31P NMR ch em ical sh ifts, th e
observed splittin g pattern s or com parison with auth en tic sam ples.
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

830
Kullberg, Bollmark, Stawinski:
5′-O-(4,4′-Dim eth oxytrityl)th ym idin -3′-yl 3′-O-(4,4′-Dim eth oxytrityl)th ym idin -5′-yl
H-Ph osph on oselen oate (3)
5′-O-(4,4′-Dim eth oxytrityl)th ym idin 3′-H-ph osph on oselen oate 1 (trieth ylam m on ium salt,
0.22 m m ol) an d 3′-O-(4,4′-dim eth oxytrityl)th ym idin e 2 (0.2 m m ol) were dried by repeated
evaporation of th e added pyridin e. Th e oily residue was dissolved in aceton itrile (4 m l) con -
tain in g pyridin e (1 m m ol), an d th en diph en yl ch loroph osph ate (0.5 m m ol) was added. After
5 m in , th e reaction was quen ch ed with saturated sodium ch loride solution (1 m l) an d th e
m ixture was partition ed between toluen e (2 × 50 m l) an d brin e (20 m l). Th e organ ic ph ase
was dried, evaporated an d th e residue was purified on a silica gel colum n usin g eth yl ace-
tate–toluen e (1:1, v/v) con tain in g 0.01% trieth ylam in e as eluen t. Th e product was isolated
as a wh ite solid in 80% yield (purity > 98%, ca. 1:1 m ixture of diastereoisom ers; 1H NMR
spectroscopy). FAB HRMS [M]+, foun d: 1221.3145; C62H63N4NaO14PSe requires 1221.3141.
Com poun d 3 could be separated in to P-diastereom ers. Th ese are referred to as “faster” an d
“slower” m ovin g diastereom er, accordin g to th eir ch rom atograph ic m obilities on silica gel.
Faster m ovin g diastereoisom er (RP)-3, RF 0.22 (system A): 1H NMR: 9.61 (s, 1 H, NH); 9.51
1
(s, 1 H, NH); 8.18 (d, JPH = 649, 1 H); 7.36 (m , 20 H); 6.85 (m , 8 H); 6.38 (dd, 3J = 8.3 an d
6.0, 1 H); 6.29 (dd, 3J = 8.2 an d 6.1, 1 H); 5.53 (m , 1 H); 4.28 (m , 1 H); 4.17 (m , 1 H); 4.1
(m , 1 H); 3.79 (s, 6 H); 3.78 (s, 6 H); 3.62 (m , 2 H); 3.44 (m , 2 H); 2.36 (m , 2 H); 2.02 (m , 1 H);
3
1
1.88 (s, 3 H); 1.77 (m , 1 H); 1.48 (s, 3 H). 31P NMR: 71.36 (1JPH = 649, JPH = 9.95, JPSe
=
878). 13C NMR: 164.17, 164.16, 159.09, 159.05, 150.79, 150.65, 145.10, 144.39, 136.22,
136.19, 135.95 135.45, 135.28, 130.52, 130.45, 130.29, 128.57, 128.51, 128.35, 127.50,
113.74, 113.73, 113.65, 112.00, 111.57, 87.74, 87.54, 86.34, 84.94 (J = 6.44), 84.66, 84.22
(J = 7.73), 78.41, 74.32, 66.83, 66.77, 63.41, 55.52, 39.43, 39.12, 12.78, 12.04.
Slower m ovin g diastereoisom er (SP)-3, RF 0.18 (system A): 1H NMR: 9.57 (s, 1 H, NH); 9.37
1
(s, 1 H, NH); 8.25 (d, JPH = 645, 1 H); 7.38 (m , 20 H); 6.84 (m , 8 H); 6.38 (m , 1 H); 6.18
(dd, 3J = 8.6 an d 5.9, 1 H); 5.53 (m , 1 H); 4.37 (m , 1 H); 4.21 (m , 1 H); 4.09 (m , 1 H); 3.98
(m , 1 H); 3.78 (s, 12 H); 3.66 (m , 1 H); 3.45 (m , 1 H); 3.35 (m , 1 H); 2.53 (m , 1 H); 2.38 (m ,
3
1 H); 1.89 (m , 1 H); 1.83 (s, 6 H); 1.70 (m , 1 H). 31P NMR: 73.24 (1JPH = 645, JPH = 10.6,
1JPSe = 879). 13C NMR: 162.22, 159.06, 158.96, 158.94, 150.77, 145.37, 137.39, 136.57,
136.51, 136.17, 135.42, 130.51, 130.47, 130.30, 128.57, 128.53, 128.46, 128.37, 128.30,
128.23, 127.50, 127.33, 113.68, 113.58, 111.17, 87.46, 86.85, 85.17, 74.54, 62.76, 55.52,
55.49, 38.95, 12.67.
Financial support from the Swedish Research Council is gratefully acknowledged.
REFERENCES AND NOTES
1. Stawinski J. in: Handbook of Organophosphorus Chemistry (R. Engel, Ed.), p. 377. Marcel
Dekker, New York 1992.
2. Stawinski J., Kraszewski A.: Acc. Chem. Res. 2002, 35, 952.
3. Kullberg M., Stawinski J.: Nucleosides Nucleotides Nucleic Acids 2005, 24, 1627.
4. Kullberg M., Stawinski J.: Synthesis 2005, 1668.
5. Bollmark M., Kullberg M., Stawinski J.: Tetrahedron Lett. 2002, 43, 515.
6. Michalski J., Krawiecki C.: Rocz. Chem. 1957, 31, 715.
7. Krawiecki C., Michalski J., Jones A. Y., Katritzky A. R.: Rocz. Chem. 1969, 43, 869.
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831

Nucleoside H-Phosphonates
831
8. Stec W. J., Uznanski B., Michalski J.: Phosphorus 1973, 2, 237.
9. Stawinski J., Thelin M., Westman E., Zain R.: J. Org. Chem. 1990, 55, 3503.
10. Zain R., Strömberg R., Stawinski J.: J. Org. Chem. 1995, 60, 8241.
11. Stawinski J., Strömberg R.: Trends Org. Chem. 1993, 4, 31.
12. Zain R., Stawinski J.: J. Org. Chem. 1996, 61, 6617.
13. Studies on ligand exchange process in H-phosphonothioate and H-phosphonoselenoate
diesters are subjects of separate investigations in this laboratory.
14. Dostrovsky I., Halmann M.: J. Chem. Soc. 1953, 502.
15. Hall R. H., Todd A., Webb R. F.: J. Chem. Soc. 1957, 3291.
16. Strömberg R., Stawinski J.: Nucleic Acids Symp. Ser. 1987, 18, 185.
17. Cabre-Castellvi J., Palomo-Coll A., Palomo-Coll A. L.: Synthesis 1981, 616.
18. Stec W., Zwierzak A., Michalski J.: Bull. Pol. Acad. Sci., Chem. 1969, 17, 587.
19. Regberg T., Stawinski J., Strömberg R.: Nucleosides Nucleotides 1988, 7, 23.
20. Stawinski J., Thelin M.: Tetrahedron Lett. 1992, 33, 3189.
21. Stawinski J., Strömberg R., Zain R.: Tetrahedron Lett. 1992, 33, 3185.
22. Seela F., Kretschmer U.: J. Chem. Soc., Chem. Commun. 1990, 1154.
23. McConnell R. L., Coover H. W.: J. Org. Chem. 1959, 24, 630.
Collect. Czech. Chem. Commun. 2006, Vol. 71, No. 6, pp. 820–831