Improvement of a critical intermediate step in the synthesis of a nitroxide-based spin-labeled deoxythymidine analog

Article abstract of DOI:10.3390/51201244

Methods to reduce the carboxylic acid moiety in 3-carboxy-2,2,5,5- tetramethyl-pyrrolin-1-oxyl to an alcohol as an intermediate toward the corresponding aldehyde have been explored and an improved method has been developed.

Full text of DOI:10.3390/51201244

Molecules 2000, 5, 1244-1250
molecules
ISSN 1420-3049
http://www.mdpi.org
Improvement of a Critical Intermediate Step in the Synthesis of
a Nitroxide-Based Spin-Labeled Deoxythymidine Analog
Jeannine H. Powell, Edward M. Johnson II and Peter M. Gannett*
Department of Basic Pharmaceutical Sciences, West Virginia University School of Pharmacy,
P. O. Box 9530, Morgantown, West Virginia, USA. Tel. (304)-293-1480, Fax (304) 293-2576.
*
Author to whom correspondence should be addressed; e-mail pgannett@hsc.wvu.edu
Received: 8 May 2000; in revised form 13 November 2000 / Accepted: 14 November 2000 /
Published: 18 December 2000
Abstract: Methods to reduce the carboxylic acid moiety in 3-carboxy-2,2,5,5-tetramethyl-
pyrrolin-1-oxyl to an alcohol as an intermediate toward the corresponding aldehyde have
been explored and an improved method has been developed.
Keywords: Nitroxide, spin label, reduction, carboxylic acid, alcohol, anhydride
Introduction
Spin-labeled nucleic acids have been proposed as probes for the study of DNA conformation and
dynamics [1,2]. A potentially very useful moiety is a thymidine analog coupled via an acetylene bridge
to a pyrrolin-N-oxyl molecule [2]. The synthesis of this spin-labeled nucleic acid begins with stable
nitroxides described by Rozantsev [3]; further elaboration of these compounds is described in several
sources [2,4]. In our hands, one of the most problematic steps in the overall synthesis is the reduction
of the carboxylic acid moiety in 1 to the alcohol 2, which is then oxidized to the aldehyde 3, without
perturbation of the nitroxide (Figure 1). Here we report several methods we have utilized for the
conversion of 1 into 2 and in one instance of 1 into 3.
©
2000 MDPI. All rights reserved.

Molecules 2000, 5
1245
Results and Discussion
In 1989, Spaltenstein et al [2] reported synthesis of a spin-labeled thymidine analog
phosphoramidite, from the aldehyde 3 (Figure 1). The synthesis of this aldehyde was described by
Hideg et al [4], who prepared it by oxidation of the alcohol, 2. Compound 2 was prepared by reduction
of 1 using sodium bis[2-methoxyethoxy]-aluminum hydride in 88% (Method 1). In our hands, we
found this reduction difficult to control. In some cases the reaction appeared not to occur while at other
times, the reaction occurred, but the reaction mixture became a gelatinous semi-solid from which it
was difficult to isolate 2. Therefore, we explored alternative reducing agents to affect the
transformation of 1 to 2 or 3.
We first explored the reduction of the mixed anhydride 4. Griffith et al [5] had previously reported
the preparation of the mixed anhydride 4 from 1 by treatment with ethyl chloroformate and
triethylamine in ether [6]. After isolation of 4 it was reduced to 2 with sodium borohydride in ethanol.
However, in our hands, this reduction was incomplete and was accompanied by partial hydrolysis of 4.
Separation of the reaction mixture on silica gel gave the desired alcohol 2, acid 1, and the mixed
anhydride 4. The yield of the alcohol was 18% (or 66% based on recovered 1 and 4, Method 2).
However, by using the conditions described by Ishizumi et al [7], reduction of 1 to 2 via 4 was
achieved in 55% yield (Method 3).

Molecules 2000, 5
1246
Next, we explored a different strategy. Instead of reducing 1 to 2 and then oxidizing 2 to 3, we
pursued an alternate route from 1 to 3. This route involved reducing the acid chloride 5 to the
corresponding aldehyde according to the method of Brown and Rao [8] by treating 5 with lithium tri-t-
o
butoxyaluminohydride in diglyme at –78 C (Method 4). This reaction worked well but removal of the
diglyme proved problematic due to the volatility of 3. We found that this reaction could be conducted
in THF with similar results and the product 3 was isolated in 60% yield.
Methods 2-4, however, did not give yields superior to those reported by Hideg and therefore, we
returned to the original route of 1 → 2 → 3. Rosen’s work suggested that the nitroxide moiety was
stable toward LAH reduction [9], in contrast with Rozantsev’s assertion that LAH reduces nitroxides to
hydroxylamines. We decided to explore this chemistry using LAH for the reduction of 1 → 2 and,
after several attempts, found that by limiting the quantity of 1 to 2 grams or less, we could achieve 97%
yields of 2 (Method 5). Occasionally, some over-reduction was observed yielding 6, in additon to the
desired product 2, especially if prolonged reaction times were used. However, 6 could be easily
separated from the desired product by simple acid extraction.
Finally, since LAH reduces esters more readily than acids, we prepared the ethyl ester 7 by
conversion of 1 to the acid chloride 5, followed by reaction with sodium ethoxide in ethanol to the
ester 7. Reduction of 7 with LAH did reduce the ester to an alcohol, but the nitroxide was also reduced
to an amine yielding 6 (Method 6).
Table 1. Comparison of Methods Used to Reduce Nitroxide 1.
{
Private}
Substrate
Conditions
Product (Yield)
Method
a
1
2
3
4
5
6
1
4
4
5
1
7
Sodium bis-(2-methoxyethoxy)-aluminum hydride/toluene/reflux
2 (88%)
b
NaBH /Ethanol
2 (66%)
4
NaBH /THF/water
2 (55%)
3 (60%)
4
Lithium tri-tert-butoxyalumino-hydride/diglyme
LAH/THF/reflux/24h
c
2 (97%)
LAH/THF/reflux/2h
6 (17%)
a
Published yield [4].
b
Yield is based on recovered acid 1 and mixed anhydride 4
c
Yields based on 2g or smaller scale.
Conclusions
We have presented a brief survey of various conditions which can be used for conversion of a
carboxylic acid functional group of 1 to the corresponding alcohol 2. The reduction of 1 to 2 is best
achieved with LAH in THF, and as long as the scale of the reaction is 2 grams or less, yields of 97%
can be realized.

Molecules 2000, 5
1247
Experimental
General
The carboxylic acid 1 was prepared according to the method of Rozantsev [3] and purified by
recrystallization from tetrahydrofuran (THF). All reagents were purchased from Aldrich Chemical Co.
and solvents were purchased from Fisher Scientific. Purification of solvents was performed as follows:
toluene and benzene were washed with concentrated sulfuric acid and distilled; ether and THF were
distilled from lithium aluminum hydride; pyridine and triethylamine from calcium hydride; chloroform
was distilled from phosphorus pentoxide. NMR spectra were measured on a Varian Gemini 300
broadband spectrometer. Analysis of purity for compounds 1, 2, and 3 was made by HPLC analysis (C-
1
8 (5 ꢀP, 2.6 x 100 mm), mobile phase 90% sodium acetate (50 mM, pH 6), 10% MeOH, UV
detection (254 nM), flow rate 2.0 mL/min (retention times: 1, 2.68 min, 2, 3.60 min, 3, 3.51 min) and
by comparison of their melting points to literature values.
Method 1 – The acid 1 (1.84 g, 10 mmol) was suspended in dry toluene (10 mL), to which 11 mL of
a 70% solution of sodium bis[2-methoxyethoxy]-aluminum hydride was added dropwise. The solution
was heated at reflux for two hours and quenched by addition to 100 mL of a 5% solution of sodium
hydroxide. The organic phase was separated and dried over anhydrous sodium sulfate and the toluene
o
o
removed in vacuo to yield 1.6g (88%) 2 (mp 75-76 C; lit. 75-77 C[4])
Method 2 – The acid 1 (1.2 g, 6.5 mmol) and triethylamine (0.69 g, 6.8 mmol) were dissolved in
ether (50 mL) and cooled in an ice-salt bath. Ethyl chloroformate (0.71 g, 6.5 mmol) was added and the
reaction was stirred for 1.5 h. The reaction mixture was filtered to remove triethylamine hydrochloride,
the ether filtrate was washed with cold HCl (5%, 10 mL), water (10 mL), saturated NaHCO (10 mL),
3
and brine (10 mL). The ether layer was dried over Na SO , filtered and concentrated in vacuo to yield
2
4
4
, 1.04 g (63%). This product was suspended in ethanol (16.8 mL), to which was added NaBH (560
4
mg, 14.8 mmol) in two equal portions. The reaction was stirred for 30 minutes, the ethanol removed in
vacuo, the product dissolved in water (8 mL), and extracted with chloroform (3 X 25 mL). The organic
layer was washed with saturated NaHCO (10 mL) and water (10 mL), dried over Na SO , and
3
2
4
concentrated in vacuo. Separation on silica gel (1:1 ethyl acetate/hexanes) afforded three fractions
corresponding to compounds 1, 4 and 2 in a ratio of 1:3:1. After separation the yield of 2 (mp 75-
o
o
7
6 C; lit. 75-77 C[4]) was 0.22 g (18%, based on 1, 66% based on recovered 1 and 2).
Method 3 – The acid 1 (4.2 g, 23 mmol) and triethylamine (2.36 g, 23 mmol) were dissolved in THF
(40 mL). With cooling (ice-salt bath), ethylchloroformate (2.95 g, 27 mmol) in THF (7 mL) was slowly
added and the mixture stirred for 30 minutes. The triethylammonium chloride salts were filtered off
and washed with THF (10 mL). The filtrate and washings were added dropwise to NaBH (2.36 g, 62
4
mmol) in water (30 mL). This reaction was allowed to stir overnight. The mixture was acidified with

Molecules 2000, 5
1248
1
0% HCl to destroy excess NaBH and then the solution was saturated with solid NaCl, causing the
4
separation of the THF and aqueous layers. The THF layer was removed, and the aqueous layer
extracted repeatedly with methylene chloride (15 mL aliquots) until no more yellow color was
extracted into the organic phase. The THF/methylene chloride extract was dried over MgSO and
4
o
o
concentrated to dryness in vacuo to give compound 2, yield 2.13g (55%) (mp 75-76 C; lit. 75-77 C[4]).
Method 4 – The acid 1 (1.84 g, 10 mmol) was suspended in dry benzene (25 mL) and cooled to 0-
o
5
C. Dry pyridine (1 mL) and then thionyl chloride (0.9 mL) were added dropwise. The reaction was
stirred at room temperature for 30 min and then concentrated in vacuo. The residue (acid chloride 5)
o
was dissolved in THF (10 mL), the flask flushed with nitrogen, cooled to –78 C, and maintained at that
temperature as lithium tri-tert-butoxyaluminohydride (2.1 eq, 1 eq to neutralize the pyridine
hydrochloride) was added dropwise. The reaction was stirred at room temp for one hour. The mixture
was quenched by the addition of NH Cl (10%, 10 mL), the layers separated and the aqueous layer
4
extracted with CH Cl (10 mL). The combined organic layers were dried, filtered and concentrated to
2
2
yield the crude aldehyde 3. The product 3 was purified by chromatography (silica gel, 3:1 hexane/ethyl
o
o
acetate). Compound 3 yield: 1.01g (60%) (mp 77-79 C; lit. 78-79 C [4])
Method 5 – To a solution of 1 (1.84 g, 10 mmol) in dry THF (10 mL) was added, dropwise, a
solution of LAH in THF (1.0 M, 10 mL), and the mixture stirred at room temp for 12 hours. The
reaction was quenched by the addition of water (0.4 mL), 15% sodium hydroxide (0.4 mL) and water
o
o
(
1.2 mL), filtered, and concentrated in vacuo to yield 1.65g of 2 (97%) (mp 75-76 C; lit. 75-77 C[4]).
Scale-up proved problematic as the same procedure when repeated using 2 g or 4 g of 1 resulted in
declining yields of 76% and 41%, respectively. In the latter case the product was a mixture of the
1
nitroxide 2 and a compound whose H-NMR was consistent with the corresponding amine 6.
Method 6 – The acid chloride 5 was prepared as described under Method 4 and then the solvents
removed in vacuo. A solution of sodium ethoxide prepared from 0.25 g of sodium metal and absolute
ethanol (8 mL) was added. The resulting mixture was heated at reflux for 2 hr, cooled to room temp
and the salts removed by filtration. The filtrate was concentrated to dryness, dissolved in CH Cl (50
2
2
mL), washed with 5% aq HCl, 5% aq NaOH, and 5 % NH Cl (10 mL each), dried (MgSO ), filtered
4
4
o
o
and concentrated to dryness to yield 7 (1.2 g, 70%) (mp 109-111 C; lit [3] 109.5-111 C). The crude
product was dissolved in dry THF (7 mL) and added to LAH (1M, 7 mL). The mixture was stirred at
rm temp for 2 h and then quenched by the addition of water (0.3 mL), 15% aq NaOH (0.3 mL) and
water (0.9 mL). The reaction mixture was then filtered and concentrated in vacuo. The residue was
dissolved in CH Cl (50 mL), washed with sodium hydroxide (5%, 2 x 10mL) hydrochloric acid (5%, 2
2
2
x 10mL), dried (MgSO ), filtered and concentrated in vacuo. The hydrochloric acid extract was made
4
basic by the addition of aqueous 10% sodium hydroxide, extracted with CH Cl (3 x 20 mL), dried
2
2
(
MgSO ), filtered and concentrated in vacuo. The latter extract contained the product 6 (Yield: 0.27g,
4
1
7%).

Molecules 2000, 5
1249
NMR data - [Note: all nitroxides were reduced to hydroxylamines for NMR analysis with 1.1
equivalents Na S O in a 1:1 mixture of D O:d -acetone.]
2
2
4
2
6
1
Carboxylic acid 1: H-NMR (1:1 d -acetone, D 2ꢀꢁ/ ppm 1.57 (s, 6 H, 2 -CH ), 1.67 (s, 6 H, 2 -
6
2
3
1
3
CH ), 6.52 (s, 1H, =CH-); C-NMR (1:1 d -acetone, D 2ꢀꢁ/ ꢂꢃꢄꢅꢃꢁꢆꢂꢁ&ꢇꢁꢂꢁꢈ&+ ), 27.35 (2 C, 2 -CH ),
3
6
2
3
3
6
8.36 (1 C, quaternary), 70.87 (1C, quaternary), 137.0 (=C-CO H), 148.8 (=CH-), 167.4 (=C-CO H).
2 2
1
Alcohol 2: H-NMR (1:1 d -acetone, D 2ꢀꢁ/ ppm 1.45 (s, 6 H, 2 -CH ), 1.47 (s, 6 H, 2 -CH ), 4.31
6
2
3
3
1
3
(
s, 1 H, -CH -), 4.32 (s, 1 H, -CH -), 5.81 (s, 1H, =CH-); C-NMR (1:1 d -acetone, D 2ꢀꢁ/ ꢂꢉꢄꢊꢋꢁꢆꢂꢁ&ꢇ
2
2
6
2
2
1
-CH ), 24.80 (2 C, 2 -CH ), 58.21 (1 H, -CH -OH), 70.13 (1 C, quaternary), 72.20 (1C, quaternary),
3 3 2
28.5 (=C-CH OH), 144.5 (=CH-).
2
1
Aldehyde 3: H-NMR (1:1 d -acetone, D 2ꢀꢁ/ SSPꢅꢄꢋꢋꢁꢆVꢇꢁꢅꢂꢁ+ꢇꢁꢋꢁꢈ&+ ), 7.11 (s, 1H, =CH-), 9.69
6
2
3
1
3
(
-CHO); C-NMR (1:1 d -acetone, D Oꢀꢁꢁ/ ppm 24.37 (2 C, 2 -CH ), 24.52 (2 C, 2 -CH ), 68.47 (1 C,
6
2 3 3
quaternary), 68.58 (1C, quaternary), 146.31 (=C-CHO), 156.66 (=CH-), 191.02 (-CHO).
1
Mixed anhydride 4: H-NMR (1:1 d -acetone, D 2ꢀꢁ/ ppm 1.473 (s, 6 H, 2 -CH ), 1.51 (t, 3H, -
6
2
3
CH CH , J = 7.1 Hz), 1.55 (s, 6 H, 2 -CH ), 4.42 (q, 2 H, -OCH CH , J = 7.1 Hz), 6.92 (1H, =CH-).
2
3
3
2
3
1
Ester 7: H-NMR (1:1 d -acetone, D 2ꢀꢁ / ppm 1.47 (s, 6H, 2-CH ), 1.50 (t, 3H, J=7.1 Hz,
6
2
3
1
3
CH CH O), 1.54 (s, 6H, 2-CH ), 4.41 (q, 2H, J=7.1 Hz, CH CH O), 6.91 (s, 1H, CH); C-NMR (1:1
3
2
3
3
2
d -acetone, D 2ꢀꢁ / ppm 13.68 (CH CH O), 24.32 (2 C, 2 -CH ), 24.41 (2 C, 2 -CH ), 60.98
6
2
3
2
3
3
(
CH CH O), 68.05 (1 C, quaternary), 69.72 (1 C, quaternary), 136.7 (=C-CO ), 146.6 (=CH-), 164.8 (-
3
2
2
COOEt).
References
1
2
. Strobel, O. K.; Kryak, D. D.; Bobst, E. V.; Bobst, A. M. Preparation and Characterization of Spin-
Labeled Oligonucleotides for DNA Hybridization. Biocon. Chem. 1991, 289-95.
. Spaltenstein, A.; Robinson, B. H.; Hopkins, P. B. Sequence- and Structure-Dependent DNA Base
Dynamics: Synthesis, Structure and Dynamics of Site and Sequence Specifically Spin- Labeled
DNA. Biochemistry 1989, 28, 9484-9495.
3
4
. Rozantsev, E. G. Synthesis of Some Stable Radicals and the Most Important Intermediates. In:
Free Nitroxyl Radicals ; Plenum Press:New York, 1970; pp 203-246.
. Hideg, K.; Hankovszky, H. O.; Lex, L.; Kulcsar, G. Nitroxyls VI. Synthesis and Reactions of 3-
Hydroxymethyl-2,2,5,5-tetramethyl-2,5-dihydropyrrole-1-oxyl and 3-Formyl Derivatives. Synthesis
1
980, 911-914.
5
6
. Griffith, O. H.; Keana, J. F. W.; Noall, D. L.; Ivey, J. L. Nitroxide Mixed Carboxylic-carbonic Acid
Anhydrides. A New Class of Versatile Spin Labels. Biochim. Biophys. Acta 1963, 14, 8583-8585.
. Keana, J. F. W. Newer Aspects of the Synthesis and Chemistry of Nitroxide Spin Labels. Chem.
Rev. 1978, 78, 37-4.

Molecules 2000, 5
1250
7
. Ishizumi, K.; Koga, K.; Yamada, S.-I. Chemistry of Sodium Borohydride and Diborane. IV.
Reduction of Carboxylic Acids to Alcohols with Sodium Borohydride through Mixed Carbonic-
Carboxylic Acid Anhydrides.Chem. Pharmaceut, Bull. (Japan) 1968, 16, 492-497.
. Brown, H. C.; Rao, B. C. S. A New Aldehyde Synthesis - the Reduction of Acid Chlorides by
Lithium Tri-t-butoxyaluminohydride. J. Amer. Chem. Soc. 1958, 80, 5377-5380.
. Rauckman, E. J.; Rosen, G. M. Synthesis of Spin Labeled Probes: Esterification and Reduction.
Syn. Commun. 1976, 6, 325-329.
8
9
Sample Availability: Samples are available from the authors (or from MDPI)
2000 by MDPI (http://www.mdpi.org)
©