Synthesis of cyclobutane nucleoside analogues 3: Preparation of carbocyclic derivatives of oxetanocin

Article abstract of DOI:10.1080/15257770.2018.1500697

A synthesis of cyclobutene nucleoside analogs in which the nucleobase is tethered by a methylene group is described. The coupling of 6-chloropurine with 3-hydroxymethyl-cyclobutanone proceeds via its triflate to give both N-7 and N-9 regioisomers with relative yields corresponding to the calculated charge distribution of the 6-chloropurinyl anion. The stereoselective reduction of the N-alkylated ketones yielded quantitatively one stereoisomer in each case. The structural assignments were based on spectroscopic data and single crystal X-ray diffraction. Attempts to photoexcite the N-7 and N-9 ketones in order to promote ring-expansion did not ensue. Preliminary evidence suggests a photodecarbonylation to cyclopropanes took place.

Full text of DOI:10.1080/15257770.2018.1500697

Nucleosides, Nucleotides and Nucleic Acids
ISSN: 1525-7770 (Print) 1532-2335 (Online) Journal homepage: http://www.tandfonline.com/loi/lncn20
Synthesis of cyclobutane nucleoside analogues
3: Preparation of carbocyclic derivatives of
oxetanocin
Murtaza Hassan, Ayat Yaseen, Abdelaziz Ebead, Gerald Audette & Edward
Lee-Ruff
To cite this article: Murtaza Hassan, Ayat Yaseen, Abdelaziz Ebead, Gerald Audette &
Edward Lee-Ruff (2018): Synthesis of cyclobutane nucleoside analogues 3: Preparation of
carbocyclic derivatives of oxetanocin, Nucleosides, Nucleotides and Nucleic Acids, DOI:
10.1080/15257770.2018.1500697
To link to this article: https://doi.org/10.1080/15257770.2018.1500697
Published online: 06 Sep 2018.
Submit your article to this journal
Article views: 7
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=lncn20

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
https://doi.org/10.1080/15257770.2018.1500697
Synthesis of cyclobutane nucleoside analogues 3:
Preparation of carbocyclic derivatives of oxetanocin
Murtaza Hassana, Ayat Yaseena, Abdelaziz Ebeadb, Gerald Audettea, and
Edward Lee-Ruffa
b
aDepartment of Chemistry, York University, Toronto, Ontario, Canada; Chemistry, Arish
University, El Arish, Egypt
ABSTRACT
ARTICLE HISTORY
Received 5 March 2018
Accepted 10 July 2018
A synthesis of cyclobutene nucleoside analogs in which the
nucleobase is tethered by a methylene group is described.
The coupling of 6-chloropurine with 3-hydroxymethyl-cyclobu-
tanone proceeds via its triflate to give both N-7 and N-9
regioisomers with relative yields corresponding to the calcu-
lated charge distribution of the 6-chloropurinyl anion. The
stereoselective reduction of the N-alkylated ketones yielded
quantitatively one stereoisomer in each case. The structural
assignments were based on spectroscopic data and single
crystal X-ray diffraction. Attempts to photoexcite the N-7 and
N-9 ketones in order to promote ring-expansion did not
ensue. Preliminary evidence suggests a photodecarbonylation
to cyclopropanes took place.
KEYWORDS
Oxetanocin nucleoside
analogs; cyclobutanone;
cyclobutanol;
6-chloropurine; charge
delocalization of
6-chloropurine anion;
photochemistry
1. Introduction
Nucleosides represent an important class of antiviral and anticancer agents
interceding in various metabolic pathways of DNA/RNA transformation.[1]
With the emergence of drug resistant mutants, the quest for new nucleoside
analogs has become an active pursuit. In particular, the development of car-
bocyclic nucleoside analogs of natural nucleosides has shown promising
results. The replacement of a methylene group for oxygen in the ribose moi-
ety renders these compounds with greater hydrolytic stability, and greater
lipophilicity which favors absorption and penetration through the cell mem-
brane.[2] The cyclopentenyl derivatives Carbovir[3] and Abacavir[4] developed
in the 1990’s are some of the early examples of effective antiviral agents.
With the observation of potent antiviral activity of the natural occurring
OxetanocinA,[5] the development and commercialization of the synthetic car-
bocyclic analogs, Cyclobut-A and Cyclobut-G (Lobucavir) was realized.[6]
These derivatives have been shown to be more bioavailable as a result of
their enhanced stability towards hydrolysis.[6] We have been interested in
CONTACT Edward Lee-Ruff
leeruff@yorku.ca
Chemistry, York University, 4700 Keele Street, Toronto,
Ontario, M3J 1P3 Canada.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/lncn
ß 2018 Taylor & Francis Group, LLC

2
M. HASSAN ET AL.
small rings as starting points for developing the synthesis of compounds of
biological interest. The strain associated with these intermediates leads to
some unusual transformations not encountered in strain-free or less
strained ring systems. In particular, the four carbon ring derivatives of
cyclobutanes and cyclobutanones show interesting ground and excited state
chemistry which underscore their importance as synthetic intermediates.[7]
These strained intermediates, nevertheless, are readily prepared by simple
procedures.[8] We have shown that appropriately substituted derivatives of
cyclobutanone can act as glycosyl donors by way of photoisomerization to
a transient oxacarbene which on N-H insertion to nucleobases give deoxy-
ribonucleosides.[9] These same ketones can be reduced with hydride donors
to the corresponding cyclobutanols resembling Oxetanocin.[9i,9j] In a con-
tinuing effort to prepare cyclobutane nucleoside analogs we were interested
in attaching the nucleobase to the carbocyclic ring by a methylene pendant.
Similar cyclobutene analogs with a nucleobase attached by a methylene
pendant to the vicinal hydroxymethyl substituent did not exhibit significant
antiviral activity against HIV-1. It was thought that the proximity of the
nucleobase from the 5’-hydroxyl group may be influential in affecting bio-
logical activity. Other polyhydroxylated cyclobutanes were reported with an
adenine methylene pendant obtained from dihydroxylation of their cyclo-
butene precursors but did not exhibit antiviral activity.[10] The cyclobutanol
hydroxyl group in these derivatives were always disposed in a trans config-
uration due to the nature of the synthetic sequence.
In the current study we report the stereospecific synthesis of cyclobuta-
nols to which a nucleobase is attached by a methylene pendant to the C-3
position in a cis orientation to the alcohol function. Furthermore, the regio-
chemistry of both N-7 and N-9 isomers produced in the coupling reaction
were unambiguously assigned. The ketone precursors were investigated for
their photochemistry in an attempt to convert these to their isonucleoside
furanose analogs by ring-expansion and insertion of the corresponding
oxacarbene with water or alcohols.
2. Results and discussion
The starting point for the synthesis was the known 3-hydroxymethylcyclo-
butanone 1b which can be readily obtained from the corresponding 3-ben-
zoyloxycyclobutanone 1a.[9i] The nucleobase can be readily attached by a
variety of coupling protocols.
Debenzoylation using a variety of bases resulted in variable yields of 1b.
We proposed the possibility of decomposition of 1b in basic conditions.
This was supported by the observed decrease in yield to 47% when the
reaction time was extended to 3 hours. Debenzoylation of 1a using a

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
3
saturated solution of NaOMe in methanol[9j] yielded the deprotected alco-
hol 1b in minute quantities with the remaining being indiscriminate
decomposition products as evident from the formation of intractable mix-
tures. We decided to use K2CO3, a milder base, at room temperature which
afforded 1b in 75% yields. Standard protocols to replace the hydroxyl
group with a nucleobase such as the Mitsunobu reaction failed to give rea-
sonable quantities of the coupling product[11]. The optimum conditions for
this transformation was conversion of alcohol 1b to its triflate 1c and its
displacement with the anion of 6-chloropurine. Triflate 1c was not isolated
in pure form but directly used in the subsequent SN2 step. The 6-chloro-
purinyl potassium salt 3 was prepared by stirring 6-chloropurine with
potassium hydroxide in acetonitrile with 5% molar tris[2-(2-methoxyethox-
y)ethyl]amine (TDA-1). Subsequent addition of the crude solution of tri-
flate 1c, afforded a total of 88% N-alkylated products as the N-9 (4) and N-
7 (5) ketones in 51% and 37% yields, respectively. Previous attempts at N-
alkylation involved mesylation and tosylation of 1b followed by addition to
a solution of the 6-chloropurinyl salt. The yields were very low (<4%) even
under elevated temperatures of 60 ꢀC and extended reaction times.
The observation of N-7 and N-9 alkylations is known with purines due
to the negative charge delocalization for the purinyl and 6-chloropurinyl
anions. The computed charge of ꢁ0.59 for the 6-chloropurinyl anion 3 at
N-9 and ꢁ0.37 at N-7 is reflected in the product distribution of the two
regioisomers.
The regiochemical assignment of ketones 4 and 5 was based on 1 D and
1
2 D (HSQC, HMBC as well as H-15N HMBC) techniques. Unequivocal
assignments for ketone 5 was based on the X-ray single crystal structure
(Figure 1). Furthermore, the regiochemistry of ketone 4 was confirmed by
its stereospecific reduction to the corresponding alcohol 6 (vide infra) and
its corresponding single crystal structure (Figure 2).
Hydride reduction of ketones 4 and 5
Initial attempts to reduce ketones 4 and 5, with sodium borohydride and
lithium aluminum hydride, yielded two products in each instance as evi-
dent from the TLC exhibiting UV-active spots with very similar Rf values
which could not be separated by preparative chromatography. It was
assumed that both cis and trans alcohols were being formed for both
regioisomeric ketones. Reduction by sodium borohydride and lithium alu-
minum hydride produced two products, in cases where the ketones reacted
completely as was noted by the emergence of two methylene H-50 peaks in
1
the H NMR spectra of the mixture and two UV-active spots by TLC.

4
M. HASSAN ET AL.
Figure 1. X-ray single crystal structure of ketone 5.
LiAl(OtBu)3H
-78oC for 2 hrs
then r.t overnight
OH
O
Et2O
N9
N
N9
N
N
N
N
N
7
Cl
7
Cl
6
4
LiAl(OtBu)3H
O
OH
Cl
-78oC for 2 hrs
then r.t overnight
Cl
N 7
N 7
Et2O
N
N
N
N
N
N
9
9
5
7
Figure 2. X-ray single crystal structure of alcohol 6.
It stands to reason that the activation barrier for trans hydride addition
would be lower, forming the cis alcohol, the kinetic product. Thus, the use
of lower temperatures and bulkier but more reactive hydride donors should
increase stereoselectivity. Since milder reducing agents such as sodium cya-
noborohydride produced a very low product conversion we turned to

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
5
OH
Cl
N 7
N
H
N
N
H
9
H
8
Figure 3. Crystal structure of over-reduced alcohol 8.
lithium tri-tert-butoxyaluminum hydride which has three bulky alkoxy
groups and is a single hydride donor. Thus we attempted the reduction of
ketones 4 and 5 at ꢁ78 ꢀC in Et2O which successfully gave one stereoiso-
mer quantitatively. We reasoned that a greater activation barrier for a cis-
hydride addition would be required due to steric congestion of the bulky
tert-butoxy groups and the 6-chloropurine substituent in the transition
state. Not surprisingly, trans-addition was observed yielding exclusively the
cis alcohols 6 and 7, respectively. The crystal structure of 6 showed
unequivocally the stereochemistry of the hydride addition (Figure 2).
The reduction of ketone 5 with lithium tri-tert-butoxyaluminum hydride
gave a single product which was assigned the structure of alcohol 7 on the
1
basis of its H- and 13C-NMR and mass spectra. The crystal structure of 7
could not be obtained since crystals of adequate size and uniformity could
not be grown even after considerable effort. However, in an attempt to
grow crystals of this alcohol, another experiment involving the reduction of
ketone 5 initially with LiAl(OtBu)3H, followed by addition of excess NaBH4
yielded the over-reduced alcohol 8 as confirmed by X-ray crystallography
(Figure 3). Similar over-reductions of purine derivatives have been
reported, where electron deficient purines are converted to dihydropurine
with NaBH4.[12] We speculate that the reason for over-reduction of the N-7
ketone 5 but not 4 may be associated with steric interaction of 6-chloro-
purine and the cyclobutanone ring, increasing the strain in the N-7 coupled
ketone.This compound was easier to crystalize than its precursor 7 and
confirmed the hydride addition to 5 occurring in a trans manner, similar
to its N-9 regioisomer 4, giving the cis-alcohol 7. In conclusion, all of the
evidence acquired support the molecular composition and stereochemistry
of 7.
On the formation of regioisomers 4 and 5
Purine as well as pyrimidine bases can generally alkylate at multiple sites.
Introduction of heterocycles via a simple SN2 reaction limits the regioselec-
tivity of the N-alkylation step by the electronic charge density and steric

6
M. HASSAN ET AL.
factors of substituents in proximity to nitrogen. For instance, the parent
purinyl anion has a very similar charge distribution for the N-7 and N-9
positions (ꢁ0.45:ꢁ0.56) and is not sterically hindered due to the lack of
substitution on the heterocycle. Thus, it has been shown to give roughly a
1:1 mixture of N-7 and N-9 alkylated products.[13] Substitution of purine
can alter the charge distribution as well as steric hindrance around N-7 and
N-9, thereby allowing for greater regioselectivity in alkylation.
In the case of purines, substitution at C-6 and N-3 promotes alkylation
at N-9 and N-7 respectively. Substitution at C-6 with chlorine, peri to N-7,
alters the charge distribution of the corresponding anion and localizes the
charge density on N-9 (ratio of charge localization for N-7 andN-9 is
ꢁ0.37:ꢁ0.59, respectively), and sterically hinders the N-7 position as can be
predicted by it’s A-value of 0.52 (free-energy difference between its axial
and equatorial conformers of substituted cyclohexanes). The presence of
chlorine at C-6 also increases the acidity at the N-9 center through induct-
ive effects as predicted by the Hammett meta r value of 0.37. As expected,
6-chloropurine gives the N-9 alkylated regioisomer for N-alkylation at C-2
in 3-benzoyloxycyclobutanone as the major product[9i] as was also the case
for n-alkylation of triflate 1c.
Several methods have been developed to promote greater regioselectivity
for N-alkylation using sterically hindered substituents at C-3 and C-6.
Bulky protecting groups such as 2, 3-dicyclohexylsuccinimidyl (Cy2SI) at C-
6 affords N-9 alkylated acyclic and cyclic products regioselectively with
various substituted purine derivatives.[14] Similarly, the formation of N-7
products is favored in the case of 6-chloro-3-deaza-3-methyl, a C-3 substi-
tuted purine.[15]
Photochemistry of ketones 4 and 5
Cyclobutanones can undergo a number of photochemical transformations.
Three principal pathways include ring-expansion, cycloelimination and
decarbonylation forming a furanose carbene intermediate by ring-expan-
sion, an alkene by cycloelimination, and a cyclopropane by decarbonyla-
tion, respectively. The objective for photolyzing 4 and 5 was to form ring
expansion products as isonucleoside analogs (Scheme 1).
O
O
O
hv
OMe
MeOH
Base
Base
Base
Scheme 1. Photochemical ring-expansion of cyclobutanones.

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
7
Previous studies[16] have shown that polar solvents, such as acetonitrile
and methanol, promote ring expansion, whereas non-polar solvents pro-
mote cycloelimination. Decarbonylation is promoted by the presence of
triplet photosensitizers such as acetone and electron-donating groups on
the a-position.[17] Thus, attempts at photolysis of N-9 and N-7 coupled
ketones 4 and 5, respectively, to promote ring expansion products were
conducted with methanol. Initial experiments included the use of pyrex
tubes. However, the N-9 coupled ketone 4 showed very little product con-
version even after 5 hours irradiation. Thus, further experiments were con-
ducted in quartz tubes since quartz is transparent down to 200 nm. The
carbonyl chromophore absorbs at 300 nm or below. Both ketones displayed
similar trends in their reactivity. Exposure for only 15 minutes gave a small
amount of product conversion as observed by the disappearance of the
cyclobutanone peaks (d3.0–3.28 ppm for 4 and d2.94–3.25 ppm for 5) and
the appearance of more shielded peaks (d1.95–2.75 ppm for the photoprod-
uct of 4 and d1.91–2.40 ppm for that of 5) in their NMR spectra. TLC indi-
cated only one major non-polar UV active spot for photolysis for both
regioisomers. Prolonged irradiation for 1 hours gave a crude mixture that
showed a product conversion of more than 60% as seen by the relative
integration ratios of the methylene protons for the nucleobase carbon
linker in the NMR spectra. Attempts at isolation of the product were
unsuccessful even after purification by column chromatography, followed
by preparative TLC. The products formed were neither cycloelimination
products, as seen by the absence of alkene protons, nor ring-expansion
products as noted by the absence of a singlet for the methoxy protons in
the proton NMR spectra. We speculate the possibility of the products being
cyclopropanes, 9 and 10 (Scheme 2), since methylene pendant peaks are
O
hꢀ
N9
N
MeOH
N9
N
N
N
N
7
Cl
N
7
9
Cl
4
O
hꢀ
Cl
N
Cl
N 7
N 7
MeOH
N
N
N
N
N
9
9
5
10
Scheme 2. Photochemistry of ketones 4 and 5.

8
M. HASSAN ET AL.
more deshielded while the remaining ring protons are more shielded in
1
their H-NMR spectra. Their relative non-polar nature also suggests that a
decarbonylation had occurred. Since formation of cyclopropanes usually
occurs by decarbonylation of cyclobutanones in their excited triplet state,
we speculate that perhaps the pendant 6-chloropurinyl moiety in ketones 4
and 5 may be acting as an intramolecular triplet sensitizer. Prolonged
irradiation led to decomposition of the ketones as seen by the broad NMR
signals appearing throughout the aliphatic region that did not have defini-
tive integration ratios related to the peaks at d4–5 ppm corresponding to
the pendant methylene protons, or to the aromatic proton peaks. The N-7
ketone 5 showed more rapid product conversion than 4, perhaps due to its
inherent strain leading to greater reactivity and decomposition.
In summary, we have shown that 6-chloropurine can be coupled to
3-hydroxymethylcyclobutanone via its triflate to give good yields of the
coupled product under basic conditions. Both N-7 and N-9 regioisomers
are produced with the latter formed as the major product with distribution
reflected by the computed charge localization of the anion. Their structures
were assigned unequivocally by NMR and single crystal X-ray diffraction.
The reduction of each regioisomeric ketones takes place stereospecifically
with the hindered hydride donor, LiAl(OtBu)3H, to give the cis alcohols in
excellent yields. Their structures including relative stereochemistry was
based on X-ray crystallography and NMR techniques. In the case of the
N-7 ketone, over-reduction of the nucleobase was observed. The presence
of chlorine at C-6 allows for their conversion to their more biologically
relevant adenine derivatives, as well as alkylated nucleoside analogs.
In an attempt to convert ketones 4 and 5 to their riboside derivatives by
a photochemical ring-expansion in methanol failed but instead led to what
appears to be decarbonylation products. Other non-photochemical proto-
cols for ring-expansion such as the Baeyer-Villiger oxidation and one car-
bon homologation with diazomethane will be explored.
Experimental
All solvents were dried and distilled prior to use. Reactions were carried
out in oven or flame-dried glassware under argon unless stated otherwise.
The melting points (mp) were measured on a Fisher-Johns melting point
apparatus and are uncorrected. FT-IR spectra were recorded on a Bruker
Alpha Platinum-ATR as a thin layer of liquid or a thin film of solid intro-
duced by dissolving the substrate in chloroform or dichloromethane and
applying as a thin layer, followed by evaporation.The mass spectrometry
analyses was performed at York University Centre for Research in Mass
Spectrometry on an Orbitrap Elite Hybrid Ion Trap-Orbitrap Mass

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
9
Spectrometer (ThermoFisher, Waltham, MA) using an electrospray ioniza-
tion source and helium as the collision gas with 50/50 water/methanol solu-
tion. UV-vis spectra were determined using methanol as solvent on an
Ultraspec 4300 pro UV spectrometer. X-ray diffraction was carried out
using a Bruker APEX II CCD diffractometer at the University of Western
Ontario (UWO) using graphite-monochromated Mo Ka (k ¼ 0.71073 Ao)
for ketone 5 and CuKa(k ¼ 1.54178 Ao) for the alcohol 6.
The 1 D and 2D- NMR spectra were recorded on Bruker AV 300, Bruker
1
400 or Bruker 700 spectrometer. The H-NMR spectra are referenced rela-
tive to the solvent residual proton signals of CDCl3 (d 7.27), CD3OD (d
3.31) and DMSO-d6 (d 2.50). The chemical shifts are reported as delta (d)
values downfield from tetramethylsilane (TMS, d ¼ 0) in parts per million
(ppm). Data from 1 D –HNMR are reported in the following order: chem-
ical shifts (d in ppm), multiplicity (s ¼ singlet, d ¼ doublet, t ¼ triplet,
dd ¼ doublet of doublets, q ¼ quartet, m ¼ multiplet), coupling constant
(J in Hz), and lastly relative integration (number of protons).
Photolyses were carried in quartz tubes strapped around a water-cooled
quartz immersion well containing a 450-W Hanovia medium pressure mer-
cury arc lamp. This set-up was immersed in an ice-water bath. Samples
were purged with argon for 15 minutes, prior to irradiation.
Analytical thin layer chromatography (TLC) was done using commercial
Machery-Nagel 60 F 254 silica gel coated aluminum sheets and visualization
of spots was carried out with KMnO4 and/or under UV light (254 nm).
Glass plates coated with Silicycle silica gel 60 F 254 (40–63 lm particle size)
were used for preparative TLC, and Silicycle silica gel (60–200 lm particle
size) used for column chromatography.
Debenzoylation of ketone 1a
0.81 g (5.9 mmol) of K2CO3 was added to a solution of 1.00 g (4.9 mmol) of
1a[17] in 20 mL methanol and the mixture was stirred for 1 hat room tem-
perature. A solution of saturated aqueous solution of NaHCO3 (10 mL) was
added and the mixture was stirred for an additional 15 minutes. The solv-
ent was evaporated under vacuum, and the sample was purified by flash
column chromatography using ethyl acetate to provide 1b in 70% yield. Its
complete characterization was not carried out, since it was a known
compound.[9c,18]
Data for 1b
1H NMR (400 MHz, CDCl3)d 3.75 (d, J ¼ 6.2 Hz, 2 H, HOCH2),
3.14–3.06 (m, 2 H, H-2, H-4 or H-2’, H-4’), 2.88–2.82 (m, 2 H, H-2’, H-4’
or H-2, H-4), 2.69–2.56 (m, 1 H, H-4)
IR c ¼ 1775 cmꢁ1

10
M. HASSAN ET AL.
Triflation of alcohol 1b
A solution of 0.100 g (1.00 mmol) 1b in 10 mL of dry CH2Cl2 was cooled to
ꢁ78 ꢀC before adding 0.56 mL (3.20 mmol) of Hunig’s base, followed by
0.17 mL (1.0 mmol) of trifluoromethanesulfonic anhydride. The mixture
was stirred for 10 min, warmed to 0 ꢀC and stirred for 1 hour providing
quantitative conversion to 1c. The mixture was not purified and immedi-
ately used for the next step.
N-alkylation of 6-chloropurine
A mixture consisting of 0.866 g (5.61 mmol) of 6-chloropurine 2, 0.314 g
(5.61 mmol) of potassium hydroxide, 0.091 g (0.28 mmol) of tris[2-(2-
methoxyethoxy)ethyl]amine, and 2 g of magnesium sulfate was added to
100 mL of anhydrous acetonitrile. This mixture was stirred overnight, then
heated to 60 ꢀC for 5 hours and then cooled to room temperature. Triflate
1c, 1.303 g (5.61 mmol) was added directly as a crude solution from the
previous step, and the reaction mixture was stirred overnight. The insoluble
material was filtered off, and the solvent removed under vacuum. The resi-
due was purified using a short silica column and a mixture of 5% methanol
and 5% triethylamine in chloroform to yield a mixture of two UV active
compounds. The two UV active spots in the TLC were collected and sepa-
rated by flash column chromatography using ethyl acetate to provide
0.671 g (51%) of N-9 alkylated derivative 4 and 0.495 g (37%) of N-7 alky-
lated derivative 5.
Data for 4
1 H NMR (700 MHz, CDCl3)d 8.78 (s, 1 H, H-8), 8.17 (s, 1 H, H-2), 4.54
(d, J ¼ 7.6 Hz, 2 H, N-CH2), 3.28–3.22 (m, 2 H, H-2, H-4 or H-2’, H-4’ of
cyclobutane), 3.15–3.09 (m,1H, H-3 of cyclobutane), 3.04–3.00 (m, 2 H, H-
2’, H-4’ or H-2, H-4 of cyclobutane)
13C NMR (700 MHz, CDCl3)d 203.5 (C ¼ O), 152.2 (C-6), 151.8 (C-8),
151.5 (C-2), 144.5 (C-4 or C-5), 131.6 (C-5 or C-4), 51.2 (CH2-N), 48.9
(C-3 of cyclobutane), 24.5 (C-2 and C-4 of cyclobutane),
IR c ¼ 3446, 3075, 2926, 1779 cmꢁ1
HRMS Calculated for C10H10ON4Cl(Mþ)m/z 237.0539, 239.0510, found
237.0539, 239.0513
Melting point: 67.1–69.3 ꢀC
Data for 5
1 H NMR (400 MHz, CDCl3)
d 8.92 (s, 1 H, H-8), 8.32 (s, 1 H, H-2), 4.73 (d, J ¼ 7.6, 2 H, N-CH2),
3.32–3.25 (m, 2 H, H-2, H-4 or H-20, H-40 of cyclobutane), 3.21–3.08 (m,
1 H, H-3 of cyclobutane), 3.01–2.94 (m, 2 H, H-2’, H-4’ or H-2, H-4 of
cyclobutane)

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
11
13C NMR (400 MHz, CDCl3)d 202.7 (C ¼ O), 162.0 (C-6), 152.6 (C-8),
148.3 (C-2), 142.8 (C-4 or C-5), 122.2 (C-5 or C-4), 51.4 (CH2N), 50.7 (C-
3 of cyclobutane), 25.2 (C-2 and C-4 of cyclobutane)
IR c ¼ 3443, 3104, 3069, 2854, 1777 cmꢁ1
HRMS Calculated for C10H10ON4Cl(Mþ) m/z 237.0538, 239.0510, found
237.0538, 239.0512
Melting point: 115.4–117.7 ꢀC
Reduction of N-9 ketone 4
A solution containing 50 mg (0.21 mmol) of 4 in 10 mL of Et2O was cooled
to ꢁ78 ꢀC and 81 mg (0.32 mmol) of lithium tri-tert-butoxyaluminum
hydride was added. The mixture was allowed to warm to room temperature
after 2.5 hours and was left to stir overnight to provide quantitative conver-
sion to the cis-cyclobutanol 6.
Data for 6
1 H NMR (400 MHz, CDCl3)d 8.76 (s, 1 H, H-8), 8.10 (s, 1 H, H-2), 4.36
(d, J ¼ 7.0, 2 H, NCH2), 4.27–4.19 (m, 1 H, H-1 of cyclobutane), 2.53–2.47
(m, 2 H, H-2, H-4 or H-2’, H-4’ of cyclobutane), 2.44–2.34 (m, 1 H, H-3 of
cyclobutane), 1.81 (d, J ¼ 5.6, 1 H, OH), 1.81–1.74 (m, 2 H, H-2’, H-4’ or
H-2, H-4 of cyclobutane)
IRc ¼ 3358, 3113, 3075, 2973, 2929, 2853 cmꢁ1
HRMS Calculated for C10H20ON4Cl(Mþ) m/z 239.0694, 241.0665, found
239.0689, 241.00661
UVkmax ¼ 206 nm, e ¼ 10,676 Mꢁ1 cmꢁ1
Melting point: 173.5–175 ꢀC
Reduction of N-7 ketone 5
A of solution of 50 mg (0.21 mmol) of 5 in 10 mL of Et2O was cooled to
ꢁ78 ꢀC and 81 mg (0.32 mmol) of lithium tri-tert-butoxyaluminum hydride
was added. The mixture was allowed to warm to room temperature after
2.5 hours and was left to stir overnight to provide quantitative conversion
to the cis-cyclobutanol 7.
Data for 7
1 H NMR (400 MHz, CDCl3)d8.89 (s, 1 H, H-8), 8.21 (s, 1 H, H-2), 4.55
(d, J ¼ 7.0, 2 H, NCH2), 4.30–4.22 (m, 1 H, H-1 of cyclobutane), 2.54–2.48
(m, 2 H, H-2, H-4 or H-2’, H-4’ of cyclobutane) , 2.45–2.33 (m, 1 H, H-3),
2.05 (d, J ¼ 5, 1 H, OH), 1.79–1.72 (m, 2 H, H-2’, H-4’ or H-2, H-4 of
cyclobutane)
13C NMR (400 MHz, CDCl3)d161.8 (C-6), 152.4 (C-8), 148.6 (C-2), 142.9
(C-5 or C-4), 122.2 (C-4 or C-5), 62.9 (C-1 of cyclobutane), 52.3 (C-N),
36.9 (C-3 of cyclobutane), 26.7 (C-2 and C-4 of cyclobutane).

12
M. HASSAN ET AL.
IRc ¼3364, 3102, 3068, 2974, 2934, 2857 cmꢁ1
UVkmax ¼ 208 nm, e ¼ 19487 Mꢁ1 cmꢁ1
Melting point: 175.6–177.6 ꢀC
Photolysis of ketones 4 and 5
A solution of the N-7 ketone 5 (29 mg) in methanol (6 mL) was irradiated
for 30 min in a quartz tube. The solvent was removed under
reduced pressure.
1H NMR (300 MHz, CDCl3) d 8.90 (d, J ¼ 2.3 Hz, H-8), 8.26 (s, H-2),
5.20–5.03 (m), 4.58 (d, J ¼ 7.7 Hz), 4.44 (dd, J ¼ 13.3, 7.6 Hz), 4.23–4.09
(m), 3.98 (s), 3.49 (s), 3.43 (d, J ¼ 8.8 Hz), 3.43–3.24 (m), 3.16 (dd, J ¼ 6.9,
2.6 Hz), 3.14 (s), 2.70 (dt, J ¼ 15.1, 7.6 Hz), 2.43–2.29 (m), 2.08–1.88 (m),
1.26 (d, J ¼ 1.7 Hz, cyclopropane-H), 0.95–0.83 (m, cyclopropane-H), 0.08
(s, cyclopropane-H).
A solution of the ketone 4 (50 mg) in methanol (10 mL) was irradiated
for 30 minutes in a quartz tube. The solvent was removed under
reduced pressure.
1H NMR (300 MHz, CDCl3) d 8.78 (d, J ¼ 4.3 Hz, H-8), 8.37 (s, H-2),
8.23–8.07 (m, purine H), 6.17–5.98 (m), 5.40 (d, J ¼ 10.2 Hz), 5.29 (d,
J ¼ 17.2 Hz, NCH2), 5.08 (dd, J ¼ 10.4, 4.9 Hz), 4.98–4.85 (m), 4.55 (d,
J ¼ 7.1 Hz), 4.42 (dd, J ¼ 12.5, 8.4 Hz), 4.38–4.25 (m), 4.24–4.15 (m),
4.08–3.91 (m), 3.86–3.66 (m), 3.68 (s), 3.50 (s), 3.39 (s), 3.33 (s), 3.29 (d,
J ¼ 4.7 Hz), 3.31–3.06 (m), 3.15 (s), 3.10–2.96 (m), 2.80 (s), 2.77–2.64 (m),
2.41–2.24 (m), 2.14 – 1.92 (m), 1.91–1.69 (m, cyclopropane-H), 1.59–1.42
(m, cyclopropane-H), 1.41 (d, J ¼ 7.3 Hz, cyclopropane-H), 1.26 (s, cyclo-
propane-H).
Disclosure statement
No potential conflict of interest was reported by the authors.
References
[1] (a) De Clercq, E. Strategies in the Design of Antiviral Drugs. Nat. Rev. Drug
Discovery. 2002, 1, 13–25; (b) De Clercq, E. AIDS in the Third World: How to Stop
the HIV Infection? Verh. K. Acad. Geneeskd. Belg. 2007, 69, 81–104.
[2] Chen, Q.; Davidson, A. Synthesis, Conformational Study and Antiviral Activity of
L-Like Neplanocin Derivatives. Bioorg. Med. Chem. Lett. 2017, 27, 4436–4439.
[3] Coates, J. A. V.; Inggall, H. J.; Pearson, B. A.; Penn, C. R.; Storer, R.; Williamson,
C.; Cameron, J. M. Carbovir: The (-) Enantiomer Is a Potent and Selective Antiviral
Agent Against Human Immunodeficiency Virus in Vitro. Antiviral Res. 1991, 15,
161–168.

NUCLEOSIDES, NUCLEOTIDES AND NUCLEIC ACIDS
13
[4] Daluge, S. M.; Good, S. S.; Faletto, M. B.; Miller, W. H.; Clair, M. H. S.; Boone,
L. R.; Tisdale, M.; Parry, N. R.; Reardon, J. E.; Dornsife, R. E.; et al. Unique
Intracellular Activation of the Potent anti-Human Immunodeficiency Virus Agent
1592U89. Antimicrob. Agents Chemother. 1997, 41, 1099–1107.
[5] Hoshino, H.; Shimizu, N.; Shimada, N.; Takita, T.; Takeuchi, T. Inhibition of
Infectivity of Human Immunodeficiency Virus by Oxetanocin. J. Antibiot. 1987, 40,
1077–1078.
[6] (a) Bisacchi, G. S.; Braitman, A.; Cianci, C. W.; Clark, J. M.; Field, A. K.; Hagen,
M. E.; Hockstein, D. R.; Malley, M. F.; Mitt, T.; Slusarchyk, W. A.; et al. Synthesis
and Antiviral Activity of Enantiomeric Forms of Cyclobutyl Nucleoside Analogs. J.
Med. Chem. 1991, 34, 1415–1421; (b) Bisacchi, G. S.; Singh, J.; Godfrey, J. D., Jr.;
Kissick, T. P.; Mitt, T.; Malley, M. F.; Marco, J. D.; Gougoutas, J. Z.; Mueller, R. H.;
Zahler, R. Regioselective Coupling of Tetraalkylammonium Salts of 6-Iodo-2-
Aminopurine to a Cyclobutyl Triflate: Efficient Preparation of Homochiral BMS-
180,194, a Potent Antiviral Carbocyclic Nucleoside. J. Org. Chem. 1995, 60,
2902–2905; (c) Singh, J.; Bisacchi, G. S.; Ahmad, S.; Godfrey, J. D., Jr.; Kissick, T. P.;
Mitt, T.; Kocy,O.; Vu, T.; Papaioannou, C.G.; Wong, M.K.; et al. A Practical
Asymmetric Synthesis of the Antiviral Agent Lobucavir, BMS-180194. Org. Process
Res. Dev. 1998, 2, 393–399; (d) Genovesi, E. V.; Lamb, L.; Medina, I.; Taylor, D.;
Seifer, M.; Innaimo, S.; Colonno, R. J.; Clark, J. M. Antiviral Efficacy of Lobucavir
(BMS-180194), a Cyclobutyl-Guanosine Nucleoside Analogue, in the Woodchuck
(Marmota monax) Model of Chronic Hepatitis B Virus (HBV) Infection. Anti Viral
Res. 2000, 48, 197–203; (e) Hanson, R. L.; Shi, Z.; Brzozowski, D. B.; Banerjee, A.;
Kissick, T. P.; Singh, J.; Pullockaran, A. J.; North, J. T.; Fan, J.; Howell, J.; Durand,
S.C.; et al. Regioselective Enzymatic Aminoacylation of Lobucavir to Give an
Intermediatefor Lobucavir Prodrug. Bioorg. Med. Chem. 2000, 8, 2681–2687; (f)
Darses, B.; Greene, A. E.; Coote, S. C.; Poisson, J. F. Expedient Approach to Chiral
Cyclobutanones: Asymmetric Synthesis of Cyclobut-G. Org. Lett. 2008, 10, 821–824.
[7] Lee-Ruff, E.; Mladenova, G. Enantiomerically Pure Cyclobutane Derivatives and
Their Use in Organic Synthesis. Chem. Rev. 2003, 103, 1449–1483.
[8] Lee-Ruff, E. Synthesis of Cyclobutanes. In Chemistry of Cyclobutanes, Rappoport, Z.,
Liebman, J. F. Eds.; Wiley: Chichester, UK, 2005; Vol. 1, pp 281–355.
[9] (a) Lee-Ruff, E.; Wells, D. Bicyclic Nucleoside Synthesis-A Photochemical Approach.
Nucleosides, Nucleotides Nucleic Acids. 2008, 27, 484–494; (b) Mladenova, G.; Lee-
Ruff, E. Cyclobutanone Photoadducts of HCN and Malononitrile: Useful
Intermediates for the Synthesis of C-Nucleosides. Tetrahedron Lett. 2007, 48,
2787–2789; (c) Ghazi, H.; Lee-Ruff, E. Synthesis of Fluorocyclobutanones and Their
Use in the Synthesis of Fluoronucleosides. J. Fl. Chem. 2005, 126, 1565–1569; (d)
Zhong, J. H.; Fishman, A.; Lee-Ruff, E. Photochemistry on Soluble Polymer
Supports: Synthesis of Nucleosides. Org. Lett. 2002, 4, 4415–4417; (e) Lee-Ruff, E.;
Wan, W.-Q.; Jiang, J.-L. A Novel Approach Toward the Synthesis of Chiral2,3-
Dideoxy Nucleosides and Their Carbocyclic Analogs. J. Org. Chem. 1994, 59,
2114–2118; (f) Lee-Ruff, E.; Xi, F.; Qie, J. H. Enantioselective Preparation of 2, 3
–Dideoxynucleosides and Their Analogs from Ring-Expansion of Cyclobutanones. 2.
Synthesis
of
2,
3
-Dideoxyribosides
and
(1S,3R)-1-Amino-3-
(Hydroxymethyl)Cyclopentane. J. Org. Chem. 1996, 61, 1547–1550; (g) Lee-Ruff, E.;
Ostrowski, M.; Ladha, A.; Stynes, D. V.; Vernik, I.; Jiang, J. L.; Wan, W. Q.; Ding,
S. F.; Joshi, S. Synthesis and HIV Inhibition Activity of 2, 3 –Dideoxy-3 -C-
Hydroxymethyl Nucleosides. J. Med. Chem. 1996, 39, 5276–5280; (h) Lee-Ruff, E.;

14
M. HASSAN ET AL.
Margau, R. Photochemical Synthesis of Novel Dideoxynucleosides. Nucleosides,
Nucleotides Nucleic Acids. 2001, 20, 185–196; (i) Ebead, A.; Fournier, R.; Lee-Ruff,
E. Synthesis of Cyclobutane Nucleosides. Nucleosides, Nucleotides Nucleic Acids.
2011, 3, 391–404; (j) Afifi, H.; Ebead, A.; Pignatelli, J.; Lee-Ruff, E. Synthesis of
Cyclobutane Nucleosides 2-Preparation of Thymine and Uracil Analogues.
Nucleosides, Nucleotides Nucleic Acids. 2015, 34:786–798.
[10] Marsac, Y.; Nourry, A.; Legoupy, S.; Pipelier, M.; Dubreuil, D.; Aubertin, A.-M.;
Bourgougnon, N.; Benhida, R.; Huet, F. Synthesis of Mono- and Polyhydroxylated
Cyclobutane Analogs. Tetrahedron. 2005, 61, 7607–7612.
[11] Hubert, C.; Alexandre, C.; Aubertin, A.-M.; Huet, F. A Short Route to New
Nucleoside Analogues. Tetrahedron. 2002, 58, 3775–3778.
[12] Aarhus, T. I.; Fritze, U. F.; Hennum, M.; Gundersen, L. L. Sodium Borohydride
Mediated Reduction of N-Boc Protected Purines and Applications in the Synthesis
of 7-Alkyladenines and Tetrahydro[1,4] Diazepino-[1,2,3-gh]Purines. Tetrahedron
Lett. 2014, 55, 5748–5750.
[13] Ebead, A. Synthesis of Cyclobutane Nucleosides and Related Analogues. Ph.D
Dissertation, York University, Toronto, ON, 2011.
[14] Pal, A.; Salandria, K. J.; Arico, J. W.; Schlegel, M. K.; McLaughlin, L. W. 2,3-
Dicyclohexylsuccinimide as a Directing/Protecting Group for the Regioselective
Glycosylation or Alkylation of Purines. Chem. Commun. (Camb). 2013, 49,
2936–2938.
[15] Irani, R. J.; SantaLucia, J. The Synthesis of Anti-Fixed 3-Methyl-3-Deaza-2’-
Deoxyadenosine and Other 3H-Imidazo[4,5-c] Pyridine Analogs. Nucleosides,
Nucleotides Nucleic Acids. 2002, 21, 737–751.
[16] Hayes, I. E. E.; Kazarians-Moghaddam, H.; Lee-Ruff, E. Photocycloeliminations of
Bicyclic Cyclobutanones. I. Solvent Effects on Distribution of Products. Struct.
Chem. 1991, 2, 175–183.
[17] Ramnauth, J.; Lee-Ruff, E. Photochemical Preparation of Cyclopropanes from
Cyclobutanones. Can. J. Chem. 2001, 79, 114–120.
[18] Lopp, M.; Paju, A.; Kanger, T.; Pehk, T. Asymmetric Baeyer-Villiger Oxidation of
Cyclobutanones. Tetrahedron Lett. 1996, 37, 7583–7586.