3'-Azido-3'-deoxy-5'-O-isonicotinoylthymidine: a novel antiretroviral analog of zidovudine. II. Stability in aqueous media and experimental and theoretical ionization constants.

Article abstract of DOI:10.1081/NCN-120023273

Degradation of 3'-azido-3'-deoxy-5'-O-isonicotinoylthymidine (AZT-Iso), an antiretroviral derivative of zidovudine, was investigated in buffer pH 7.4, mu = 300 mOsm at 37, 50 and 60 degrees C, and in water (pH 6.6, 37 degrees C), giving zidovudine (AZT) and isonicotinic acid (INA) as products. The rate constants were determined by reversed-phase HPLC showing pseudo-first-order kinetics related to the residual amount of AZT-Iso. In this way, the studied compound was demonstrated to be 153 times more stable in water than in buffer solution at 37 degrees C. The analytical method was conveniently validated demonstrating to be a rapid and accurate stability-indicating technique. In addition, experimental and theoretical values of pKa were determined.

Full text of DOI:10.1081/NCN-120023273

This article was downloaded by: [University of Stellenbosch]
On: 09 October 2014, At: 16:57
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Nucleosides, Nucleotides and Nucleic Acids
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/lncn20
3′-Azido-3′-deoxy-5′-O-isonicotinoylthymidine: A
Novel Antiretroviral Analog of Zidovudine. II. Stability
in Aqueous Media and Experimental and Theoretical
Ionization Constants
Silvina A. Teijeiro a , Mónica A. Raviolo a , Marisa I. Motura a & Margarita C. Briñón a b
a Departamento de Farmacia, Facultad de Ciencias Químicas , Ciudad Universitaria,
Universidad Nacional de Córdoba , Córdoba, Argentina
b Departamento de Farmacia, Facultad de Ciencias Químicas , Ciudad Universitaria,
Universidad Nacional de Córdoba , Córdoba, 5000, Argentina
Published online: 07 Feb 2007.
To cite this article: Silvina A. Teijeiro , Mónica A. Raviolo , Marisa I. Motura & Margarita C. Briñón (2003) 3′-Azido-3′-
deoxy-5′-O-isonicotinoylthymidine: A Novel Antiretroviral Analog of Zidovudine. II. Stability in Aqueous Media and
Experimental and Theoretical Ionization Constants, Nucleosides, Nucleotides and Nucleic Acids, 22:9, 1789-1803, DOI:
10.1081/NCN-120023273
To link to this article: http://dx.doi.org/10.1081/NCN-120023273
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

NUCLEOSIDES, NUCLEOTIDES & NUCLEIC ACIDS
Vol. 22, No. 9, pp. 1789–1803, 2003
30-Azido-30-deoxy-50-O-isonicotinoylthymidine: A Novel
Antiretroviral Analog of Zidovudine. II. Stability in
Aqueous Media and Experimental and
Theoretical Ionization Constants
´
Silvina A. Teijeiro, Monica A. Raviolo, Marisa I. Motura,
*
´
and Margarita C. Brinon
˜
´
Departamento de Farmacia, Facultad de Ciencias Quımicas,
Ciudad Universitaria, Universidad Nacional de Cordoba,
´
´
Cordoba, Argentina
ABSTRACT
Degradation of 30-azido-30-deoxy-50-O-isonicotinoylthymidine (AZT-Iso), an
antiretroviral derivative of zidovudine, was investigated in buffer pH 7.4,
m ¼ 300 mOsm at 37, 50 and 60ꢀC, and in water (pH 6.6, 37ꢀC), giving zidovudine
(AZT) and isonicotinic acid (INA) as products. The rate constants were deter-
mined by reversed-phase HPLC showing pseudo-first-order kinetics related to
the residual amount of AZT-Iso. In this way, the studied compound was demon-
strated to be 153 times more stable in water than in buffer solution at 37ꢀC. The
analytical method was conveniently validated demonstrating to be a rapid and
accurate stability-indicating technique. In addition, experimental and theoretical
values of pKa were determined.
Key Words: Zidovudine analogs; Antiretroviral; Stability; pKa.
´
*Correspondence: Margarita C. Brin˜on, Departamento de Farmacia, Facultad de Ciencias
Quımicas, Ciudad Universitaria, Universidad Nacional de Cordoba, Cordoba 5000, Argen-
´
´
´
tina; Fax: 54-351-4334127; E-mail: macribri@dqo.fcq.unc.edu.ar.
1789
DOI: 10.1081/NCN-120023273
Copyright # 2003 by Marcel Dekker, Inc.
1525-7770 (Print); 1532-2335 (Online)
www.dekker.com

1790
Teijeiro et al.
The first 20,30-dideoxynucleoside derivative approved for treatment of HIV-1
infections was 30-azido-30-deoxythymidine (AZT, zidovudine).[1,2] This drug needs
to be converted to its 50-O-triphosphate analog by cellular enzymes to exhibit inhi-
bitory activity against the reverse transcriptase (RT) of HIV-1, inducing immuno-
logic, virologic, and neurologic improvements in HIV-1 infected patients.[3] Some
of the major chemotherapy problems associated with AZT include bone marrow
toxicity resulting in anemia and leukopenia, low therapeutic index owing to inhibi-
tion of cellular polymerases, low localization in brain, and a short half-life in blood
which requires frequent AZT administration to maintain a therapeutic drug concen-
tration.[4,5] In spite of these undesirable effects, this nucleoside reverse transcriptase
inhibitor (NRTI) continues to play an important role in the therapy of AIDS extend-
ing the life expectancy of individuals with AIDS, particularly in combination with
other NRTI, non-nucleoside transcriptase reverse and protease inhibitors.[6–9] In
attempts to overcome the problems of rapid elimination of AZT and to increase
its therapeutic efficacy, numerous AZT prodrugs have been reported in the litera-
ture.[10,11]
In this way, we have been involved in a research program aimed at exploring
more convenient novel anti HIV-1 analogs of AZT. Thus, 30-azido-30-deoxy-50-O-
isonicotinoylthymidine (AZT-Iso, Fig. 1), a new prodrug of AZT,[12] has presented
interesting anti-HIV and cytotoxicity activity as well as lipophilicity and albumin
binding properties,[12–15] and therefore is an excellent candidate for preclinical
studies.
It is very important to point out that before formulating a new drug, it is essen-
tial to determine its chemical stability[16,17] as well as some fundamental physico-
chemical properties such as lipophilicity and pKa.[18] For this reason and taking
into account that AZT-Iso is an ester, in this paper we report stability studies of
this compound under the same conditions as those for which biological activity
is observed. In addition, the pKa values of AZT-Iso were determined since this
Figure 1. Structure and numeration of 30-azido-30-deoxy-50-O-isonicotinoylthymidine
(AZT-Iso).

Novel Antiretroviral Analog of Zidovudine. II
1791
parameter affects the bioavailability of a drug. Then, it could be possible to calculate
the percentage of a soluble drug molecule in the aqueous and the lipophilic media of
the organism.[19,20]
RESULTS AND DISCUSSION
Stability Studies
Identification of Degradation Products
The stress kinetics for the degradation of 30-azido-30-deoxy-50-O-isonicotinoyl-
thymidine (AZT-Iso) were followed at 70ꢀC in water. Identification of AZT-Iso
and their degradation products, zidovudine (AZT) and isonicotinic acid (INA),
was performed by TLC and by comparing their HPLC retention times with authen-
tic samples. A typical chromatogram is shown in Fig. 2. The peak designated as ‘‘3’’
is the parent compound (AZT-Iso, tr ¼ 4.25 min) and those designated ‘‘2’’ and ‘‘1’’
Figure 2. High-pressure liquid chromatograms of AZT-Iso degradation in water at 60ꢀC.
Key ¼ (1) isonicotinic acid (INA, tR ¼ 2.18 min); (2) zidovudine (AZT, tR ¼ 3.25 min);
(3) 30-azido-30-deoxy-50-O-isonicotinoylthymidine (AZT-Iso, tR ¼ 4.25 min).

1792
Teijeiro et al.
are the hydrolysis degradation products AZT (tr ¼ 3.25 min) and INA (tr ¼ 2.18 min),
respectively. Figure 2 shows an adequate separation among all compounds.
Calibration Graphs and Statistical Analysis
HPLC methods have been used extensively for the quantification of AZT.[21–26]
Calibration graphs and statistical analysis of the results are typically performed to
validate the analytical method.[27]
Calibration graphs were performed for each compound (AZT-Iso, AZT and
INA) showing in lock HPLC chromatogram, the absence of the other compounds.
Each AZT-Iso, AZT and INA HPLC peak areas were plotted vs. concentration to
verify the linearity of the calibration graph. The analyses were performed in tripli-
cate, showing that the relative standard deviation (RSD) for each experimental point
was less than 2%, demonstrating that there is negligible scatter of the experimental
points from the regression lines.
Tests of the significance of the regression line intercepts were performed to ver-
ify whether the intercepts ‘‘a’’ differed significantly from zero. Thus, the quantities
t ¼ a=Sa were determined (where Sa is the standard deviation of the intercept), and
then compared to the corresponding tabular data for t distribution.[28–31] The cal-
culated values for AZT-Iso (t ¼ 0.08), AZT (t ¼ 1.22) and INA (t ¼ 1.11) were smal-
ler than those tabulated at a 95% significance level with Student’s t (AZT-Iso,
t ¼ 2.31; AZT, t ¼ 2.36 and INA, t ¼ 2.57). This means that the intercepts of the
regression lines were not significantly different from zero. In addition, the confi-
dence intervals for the true intercepts and slopes values were calculated from the
expressions a t Sa and b t Sb, respectively,[27,32] where Sa and Student’s t are
the same as which the previous case and Sb is the standard deviation for the slopes.
These statistics lead us to infer that the analytical method, determining the concen-
trations of AZT, AZT-Iso and INA, is free of errors coming from other com-
pounds.
The accuracy, precision and sensitivity values of the proposed method[32–35] were
summarized in Table 1.
Based on the validation analyses, HPLC has been selected as a precise and accu-
rate analytical method to follow degradation kinetics of AZT-Iso.
Reaction Order and Rate Constants
Stability studies of AZT-Iso were carried out in a first step in buffer pH 7.4
plasma isotonic, m ¼ 300 mOsm at 37ꢀC, 50ꢀC and 60ꢀC. The linear relationship
(r > 0.99) between the natural logarithm (Ln) concentration of the intact molecule
drug vs time for all studied temperatures indicates pseudo-first order degradation
kinetics for AZT-Iso at constant pH, temperature and ionic strength (Fig. 3). The
same hydrolysis kinetics was observed when the study was performed at 37ꢀC in
water (pH 6.6).
The degradation rate constants of AZT-Iso, computed by the least-squares lin-
ear regression method together with those obtained by extrapolation from the Arrhe-
nius equation,[16] as well as t1=2 and t90 values, are listed in Table 2.

Novel Antiretroviral Analog of Zidovudine. II
Table 1. Statistical data of the analytical method.
1793
Statistical parameters
n
3
AZT
AZT-Iso
INA
Accuracy (%)
Precision
Repeatabilitya (RSD) Sol. A
Sol. B
100.5 0.4
88.57 8.58
97.04 2.12
5
5
2
2
2
2
1
18.11
1.36
0.25
0.78
–
0.76
17.54
0.45
4.66
–
29.32
8.18
–
10 mg=mL
0.1 mg=mL
50.2 mg=mL
–
0.40
1.91
36.15
1 mg=mL
–
9.44
–
1.81
Reproducibilityb (RSD) (5 days)
Sensitivity (10À4); mg=mL
LOD; mg=mL
LOQ; mg=mL
7.12 (n ¼ 9)
0.1 (n ¼ 9)
0.4 (n ¼ 9)
5.42 (n ¼ 10)
0.2 (n ¼ 10)
0.7 (n ¼ 10)
4.01 (n ¼ 7)
0.6 (n ¼ 7)
2.0 (n ¼ 7)
AZT-Iso ¼ 30-Azido-30-deoxy-50-O-isonicotinoylthymidine; AZT ¼ 30-Azido-30-deoxythymidine;
INA ¼ Isonicotinic acid; n ¼ number of dates; Sol. A ¼ Solution A is composed by a mixture of
high concentration of AZT-Iso, and low concentration of AZT and INA; Sol. B ¼ Solution B is
composed by a mixture of low concentration of AZT-Iso, and high concentration of AZT and
INA; RSD ¼ Relative Standard Deviation; LOD ¼ Detection limit; LOQ ¼ Quantification limit;
aPrecision intraday.
bPrecision interday.
This table shows that AZT-Iso has a t90 ¼ 1.14 h and t1=2 ¼ 7.55 h at 37ꢀC, under
conditions that model plasma, with an activation energy (Ea) calculated from Arrhe-
nius plots[16] of 21.76 kcal=mol. As it can be seen, this drug molecule is 153 times
more stable in water (t90 ¼ 7.23 days and t1=2 ¼ 48.04 days) than in buffer, and there-
fore AZT-Iso solutions of pH 7.4 should be stored at 0ꢀC (t90 ffi 6 days and t1=2 ffi 37
days) to minimize its water hydrolysis process.
Determinations of pKa Values
Experimental Values
The importance of using drug pKa during preformulation is well known. Poten-
tiometric analysis is a convenient method for dissociation constant determination,
with accuracy and reproducible results. In this way, the experimental pyrimidinic
base pKa values for AZT-Iso (pKa ¼ 8.89) and for the standards thymidine and
AZT compounds were pKa ¼ 9.43 and pKa ¼ 9.85, respectively. The measured
pKa value for thymidine was similar to that of literature (pKa ¼ 9.79),[36] showing
the usefulness of this methodology for this family of compounds.
Conversely, it was found that the potentiometric method does not have sufficient
sensitivity to detect the pKa values of either the pyridine ring nitrogen of AZT-Iso or
OH aliphatic sugar moiety of thymidine and AZT compounds. This fact is due to the
very weak base characteristic of those groups.

1794
Teijeiro et al.
Figure 3. Plots of the observed pseudo-first order kinetic degradation of AZT-Iso in buffer
solution pH 7.4 at different temperatures. Key (G) 37ꢀC; (ꢁ) 50ꢀC; (&) 60ꢀC.
It is important to point out that AZT-Iso at physiological pH (pH 7.4) is
96.87% non-ionized, and taking into account that AZT-Iso is 16 times more lipo-
philic than AZT,[13] this drug could display better absorption properties than
AZT.
Table 2. Hydrolysis rate constants, t1=2 values and t90 values of AZT-Iso in buffer (pH 7.4,
m ¼ 300 mOsm).
Temp. (ꢀC )
kobs  104 (minÀ1
)
r
t90 (h)
t1=2 (h)
0Ã
0.129 ( 0.004)
3.746 ( 0.026)
15.30 ( 0.09)
64.40 ( 0.13)
178.3 ( 0.1)
–
–
134.4
4.643
1.137
0.2701
0.09757
173.6
892.5
30.84
7.550
1.794
0.648
1153
25Ã
37
50
60
37a
À0.99236
À0.99722
À0.99914
À0.9876
0.1002 ( 0.0055)
kobs ¼ rate constants of the pseudo first-order kinetics of hydrolysis reaction; r ¼ linear correla-
tion coefficient; Ea ¼ 21.76 kcal=mol.
ÃExtrapolated value.
aStability in water.

Novel Antiretroviral Analog of Zidovudine. II
1795
Prediction Values
Since the pKa of N-pyridine and OH groups of AZT-Iso and pyrimidinic
nucleosides, respectively are too weak, to be experimentally determined, linear free
energy relationships (LFER) using the Hammett equation, were employed to predict
the pKa’s (Eq. (1)),
X
pKa ¼ pKaꢀ À r
s
ð1Þ
where pKaꢀ is the pKa of the unsubstituted acid or base, s is a constant for the par-
ticular substituent, and r is the constant for a particular equilibrium reaction.[37]
pKa for Imide Group. Although Hammett-type equations are available for
many families of compounds, no relations were found for pyrimidinic bases. For this
reason, as a first step to calculate AZT-Iso pKa values, 2-pyridone (Fig. 4a) was used
(Eq. (2)) as a model for pyrimidinic nucleosides.
X
pKa ¼ 11:65 À 4:28
s
ð2Þ
From Eq. (2), the pKa of 2-pyrimidone (Fig. 4b) could be calculated considering
the structural similarity of this compound with pyrimidinic nucleosides. In this way,
Eq. (3) was constructed, using s ¼ 0.71 for heterocyclic nitrogen according to
Hammett approximations, leading to a pKa ¼ 8.6 for 2-pyrimidone [pKa (exp) ¼
8.59].[37]
pKa2-pyrimidone ¼ 11:65 À 4:28 sN ¼ 8:6 ðcalc:Þ
ð3Þ
Assuming this pKa for 2-pyrimidone, the s value for the second carbonyl group
of pyrimidine base, s2-CO, was calculated applying to Eq. (2) uracyl compounds
(pKa ¼ 9.46),[36] using a value of s1-N ¼ 0.71 for the nitrogen at position 1, which
leads to Eq. (4). Then, a s2-CO ¼ À0.20 was obtained.
9:46 ¼ 11:65 À 4:28 ðs1-N þ s2-CO
Þ
ð4Þ
Figure 4. a) Chemical structure of 2-pyridone. b) Chemical structure of 2-pyrimidone.

1796
Teijeiro et al.
Table 3. Theoretical ionization constant calculations of the imide group of pyrimidinic
nucleosides.
pKa
Equation
pKa ¼ pKaꢀ À ra sb
P
Comp.
Calc. Exp. DpKa
Thymine
pKa ¼ pKa2-Pyridone À r(s1-N þ s2-CO þ s5-Me
)
9.77 9.90
10.05 9.79
9.77 9.30
0.13
0.26
0.47
c
Thymidine pKa ¼ pKa2-Pyridone À r(s1-N þ s2-CO þ s5-Me þ sSug
)
c
Uridine
pKa ¼ pKa2-Pyridone À r(s1-N þ s2-CO þ sSug
)
ar ¼ 4.28.
bs1-N ¼ 0.71; s2-CO ¼ À0.20; s5-Me ¼ À0.07; sSug ffi À0.07; pKa2-Pyridone ¼ 11.65.
cThe sSug ffi À0.07 is an approximated value, that adjust well to pKa prediction of thymidine
and uridine.
Table 3 shows the calculated and experimental pKa values for thymine, thymi-
dine and uridine, demonstrating that s2-CO ¼ À0.20 is a reliable value to predict the
pKa of pyrimidinic nucleosides, since no literature data were found for this moiety.
pKa for Aromatic Nitrogen. AZT-Iso has an additional basic pKa correspond-
ing to the isonicotinoyl aromatic nitrogen, which was calculated applying the
Hammett equation for pyridine compounds (Eq. 5) and employing sCO2Et ¼ 0.72
as a representative p-ester group. As can be seen, the calculated pKa of 0.932 corre-
sponds to an extremely weak base moiety.
X
pKaN-Ar¼ pKaꢀ
pKa ¼ 5:18 À 5:90sp-COOEt ¼ 5:18 À 5:90 Â 0:72 ¼ 0:932
r
s
PyrÀ
ð5Þ
pKa for Hydroxyl Group. Taking into consideration the aliphatic character of
deoxyribose ring, the Taft equation was applied to calculate the ionization constant
of 50-OH group of AZT and thymidine molecules. Then, the equation corresponding
to primary alcohols (RCH2OH) (Eq. (6)) was used.[37]
X
pKa ¼ 15:9 À 1:42
sÃ
ð6Þ
The validity of Eq. (6) was corroborated calculating the pKa of 50-OH group of
the thymidine (Eq. (7)) and AZT (Eq. (8)), taking into account that the effect of a
group as expressed in the sà constant is halved for each saturated carbon atom that

Novel Antiretroviral Analog of Zidovudine. II
1797
separates the OH center from the substituent.[37Ã] Thus, for thymidine, the sà values
of Eq. (7) are, soÃBu ¼ 1.68, sOÃ H ¼ 1.34=4 and s2-pyrimidinyl ¼ 1.67=16.
pKaThym ¼ 15:9 À 1:42ðsoÃBu þ sOÃ H þ s2Ã-Pyrimidinyl
Þ
pKaThym ¼ 15:9 À 1:42ð1:68 þ 0:335 þ 0:104Þ ¼ 12:89ðcalcÞ
pKaThym ¼ 12:85ðexpÞ
ð7Þ
DpKa ¼ 0:04
As it can be seen from Eq. (7), a DpKa of 0.04 was observed between experimen-
tal (pKa ¼ 12.85) and calculated (pKa ¼ 12.89) values. Applying this equation, the
50-OH pKa of AZT was calculated using Eq. (8), with a sNÃ 3 ¼ 2.62=4, obtaining a
pKa value of 12.44.
pKaAZT ¼ 15:9 À 1:42ðsoÃBu þ sNÃ 3 þ s2Ã-Pyrimidinyl
pKaAZT ¼ 15:9 À 1:42 ð1:68 þ 0:655 þ 0:104Þ ¼ 12:44ðcalcÞ
Þ
ð8Þ
Table 4 presents a summary with experimental and calculated pKa values for
AZT-Iso and the studied pyrimidinic nucleosides, showing a good relationship
between both methods.
The pKa for the NH imide group of AZT-Iso was calculated using the equation
for thymidine of Table 3 and the sigma Hammett constant s2-CO ¼ À0.20. This value
was pKa ¼ 10.05 (exp. pKa ¼ 8.89), showing a DpKa ¼ 1.16 since the ionization con-
stant was determined in a DMF=water medium, due to its sparing water solubility.
For this reason, the calculation of pKa value of AZT-Iso was very important.
Moreover, the similar pKa values for AZT, thymidine and AZT-Iso, corre-
sponding to the imide group (NH) of the pyrimidinic base, can be attributed to
the fact that structural differences are sufficiently separated from each other to have
only minor effects.
Table 4. pKa values of pyrimidinic nucleosides.
Calculated pKa
OH-sugar
Experimental pKa
Comp.
NH-base
N-Ar
Found
Lit.
DpKa
Thymine
Uridine
Thymidine
AZT
9.77
9.75
10.05
10.05
10.05
–
–
–
–
–
–
–
9.90a
–
–
9.30a
12.85
12.64
–
9.43a
9.85a
8.89a
9.79a, 12.89b
0.62c,À0.04d
–
0.932
–
–
0.20c
1.16c
AZT-Iso
aNH-base.
bOH-sugar.
cDpKa (NH-base) ¼ pKacalc. À pKaexp.
dDpKa (OH-sugar) ¼ pKacalc. À pKalit.

1798
Teijeiro et al.
CONCLUSIONS
HPLC has demonstrated to be simple, precise, specific and a good stability-indi-
cating technique due to the ability of the method to simultaneously and accurately
measure the analyte (AZT-Iso) response in the presence of its degradation products
(AZT and INA), which follows pseudo first-order degradation kinetics. Though in
plasma conditions, AZT-Iso hydrolyzes at 37ꢀC faster than in water, its chemical sta-
bility could give enough time to AZT-Iso to reach lipophilic tissues, before this pro-
drug converts into AZT by intracellular enzymes. It is important to point out that
the covalent bond between AZT and the isonicotinoyl moiety is labile enough to
easily regenerate the anti HIV parent compound.
In addition, because of the very weak acid character exhibited by the pyrimidinic
base of AZT-Iso, it will be a non-ionized form at physiological and gastric pHs
(96.87% and 100% respectively). On the other hand, as we can see from the calcu-
lated pKa value of the aromatic moiety (pKa 0.932, Eq. (5)) it has a base behavior
with 100% and 78.72% of non-ionized forms at physiological and gastric pHs,
respectively.
Previously, we pointed out that AZT-Iso (Log P ¼ 0.82) was 16.4 times more
lipophilic than AZT (Log P ¼ 0.05), which achieves a higher intracellular concentra-
tion by crossing biological membranes by passive diffusion.[12]
EXPERIMENTAL SECTION
Materials
30-Azido-30-deoxythymidine (AZT), a generous gift from Filaxis (Buenos Aires,
Argentina), was used without purification. Isonicotinoyl chloride hydrochlo-
ride and isonicotinic acid (INA) were purchased from ALDRICH and SIGMA
Companies, respectively. 30-Azido-30-deoxy-50-O-isonicotinoylthymidine (AZT-Iso)
was prepared as previously reported.[12] All chemicals and reagents were of analytical
grade and all solvents were purified by distillation prior to use. Standard buffer solu-
tions were purchased from Carlo Erba (Argentina). Water used for buffers and
for mobile phase in high-performance liquid chromatography (HPLC) was double
distilled and deionized through a Milli-Q water purification system (Millipore).
All the solvents used for HPLC were filtered through a 0.45 mm filter prior to use.
Silica gel type H, size 10–40 mm, without binder (Sigma), was used for flash chro-
matography and precoated 60 F 254 silica gel plates (Merck) for thin layer chroma-
tography (TLC).
Buffer Solution
A phosphate buffer solution (81 mL of Na2HPO4 0.2 M and 19 mL of NaH2PO4
0.2 M), pH 7.40 plasma isotonic (m ¼ 300 mOsm) was used in these investigations.
The solution was freshly prepared and the pH value was measured at the experimen-
tal temperature using a research pH meter with a SC-glass electrode.

Novel Antiretroviral Analog of Zidovudine. II
Apparatus
1799
Ultraviolet spectrophotometric (UV) studies were carried out with a Shimadzu
Model UV-160A spectrophotometer, using 1 cm quartz cuvettes. An Orion Model
5A 520 pH meter was used to measure the pH of the buffer solutions with a glass-
reference electrode. For kinetic measurements the constant-temperature bath was
regulated by a Haake D8 thermostat with a precision of 0.1ꢀC.
High-Performance Liquid Chromatography Analysis (HPLC)
The HPLC measurements were assayed on a Konik 500 G chromatograph, using
an UV detector at l 265 nm and a Waters column, Spherisorb 250 Â 4 mm, 5 mm
particle size, packed with a C18 (octadecyl silane) chemically bonded non-polar
stationary phase. The mobile phase was methanol:water (70:30) and the flow rate
was 0.85 mL=min. The samples were injected into the column with a Rheodyne
(Model 7125) injector of a 20 mL loop. The experiments were performed in triplicate
at room temperature.
Analytical Method Validation
The stock solutions were prepared each day from 10, 10 and 50 mg=mL of AZT,
AZT-Iso and INA respectively, using water as solvent. Corresponding aliquots
were diluted with water in order to obtain final concentrations in the range to
0.1À10 mg=mL for AZT and AZT-Iso, and 0.5–50 mg=mL for INA. The solutions
were analyzed by HPLC and the area of the peak vs. the concentration was plotted,
finding that the response to the concentration was linear over the range of studied
concentrations, in agreement with Beer0s law. The column was flushed at the end
of each day with water followed by methanol.
The accuracy of the proposed method was determined by analyzing three known
concentrations of synthetic mixtures of the assayed compounds, and it is expressed
as the recovery percentage standard deviation (snÀ1) between the obtained and
expected values. The concentrations of the standard mixtures were chosen to give
similar conditions to the stability studies. The precision was evaluated through their
RSD parameters. For intraday precision (or repeatability), the study was performed
in quintuplets from two different ‘‘real’’ solutions.[33] In this way, two solutions of
AZT-Iso (10 mg=mL) were stored under stress conditions at 68ꢀC, 10 min (sol. A)
and 5 h (sol. B). Thus, the sol. A has high AZT-Iso concentration and low AZT
and INA concentrations, and the sol. B has low concentration of AZT-Iso and high
concentrations of AZT and INA. To determine the interday precision (or repro-
ducibility), the same sample solution was analyzed during five consecutive days,
and injected three consecutive times each day. The sensitivity of the method was
analyzed from the calibration graph intercepts of AZT-Iso, AZT and INA. The limit
of detection (LOD) and the limit of quantification (LOQ) were calculated from
the confidence intervals of the intercepts plus three and ten values of Sx=y
,

1800
Teijeiro et al.
respectively.[32,34,35] Also, the precision of the method was analyzed by the measure-
ment of high and low concentration solutions of each compound, and each sample
was injected five consecutive times.
Kinetic Studies
Identification of Degradation Products
40 mg (1.07 Â 10À7 mol) of AZT-Iso were dissolved in 10 mL of water. The solu-
tion was stored at 70ꢀC, during 10 h. Then, the flask was removed from the bath,
immediately cooled in an ice bath, and assayed at room temperature in duplicate
by HPLC (methanol:water, 70:30 as the mobile phase at a flow-rate of 0.85 mL=min)
and by TLC (ethyl acetate-petroleum ether (80:20)) techniques. The reaction mixture
was assayed against standard solutions of AZT, AZT-Iso and INA.
Stability Assays
A stock solution of AZT-Iso (50 mg=mL) was prepared in water. Aliquots were
taken from the stock solution and diluted with buffer phosphate solutions of pH 7.4
(Na2HPO4–NaH2PO4, m ¼ 300 mOsm ) and water (pH 6.6) up to final concentration
of 5.37 Â 10À6 M. All sample solutions were filled into 7 mL flasks and stored at the
appropriate temperature (37ꢀC, 50ꢀC and 60ꢀC for buffer solutions and 37ꢀC for
water solutions) in a constant temperature bath, while the ‘‘zero time’’ sample was
maintained at À20ꢀC. The flasks were withdrawn at suitable time intervals, immedi-
ately cooled in an ice-bath and stored in a freezer until analyzed. All kinetic studies
were done in duplicate. Upon removal of the last sample, the stored solutions were
equilibrated to room temperature and analyzed by HPLC (methanol:water, 70:30
as the mobile phase at a flow-rate of 0.85 mL=min).
Potentiometric Determination of pKa
The pKa value of AZT-Iso was determined by a potentiometric method[38] using
a glass electrode and an automatic dossier (0.1 mL each 20 s). A solution of 15 mL
of sulfuric acid (0.0672 N) containing 0.3 mM (112 mg) of AZT-Iso in 5 mL of
water:DMF (10:7.65) was titrated with sodium hydroxide solution (0.0590 N).
As a first step, the thymidine pKa was determined as standard compound
(Lit. pKa ¼ 9.9)[36] in order to validate this methodology to pyrimidinic nucleoside
systems, which involve a weak acid.
ACKNOWLEDGMENTS
´
The authors gratefully acknowledge the Agencia Nacional de Promocion
Cientıfica (FONCYT), the Agencia Cordoba Ciencia and the Secretarıa de Ciencia y
´
´
´

Novel Antiretroviral Analog of Zidovudine. II
1801
´
´
Tecnica de la Universidad Nacional de Cordoba (SECYT-UNC) of Argentina for
financial support. The authors also wish to express their sincere thanks to Dra. L.
Alassia (FILAXIS Laboratories, Buenos Aires, Argentina) for supplying zidovudine.
REFERENCES
1. Mitsuya, H.; Weinhold, K.J.; Furman, P.A.; St Clair, M.H.; Lehrman, S.N.;
Gallo, R.C.; Bolognesi, D.; Barry, D.W.; Broder, S. 30-Azido-30-deoxythy-
midine (BW A509U): An antiviral agent that inhibits the infectivity and
cytopathic effect of human T-lymphotropic virus type III=lymphadenopathy-
associated virus in vitro. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 7096–7100.
2. Fischl, M.A.; Richman, D.D.; Grieco, M.H.; Gottlieb, M.S.; Volberding, P.A.;
Laskin, O.L.; Leedom, J.M.; Groopman, J.E.; Mildvan, D.; Schooley, R.T.;
Jackson, G.G.; Durack, D.T.; King, D. The efficacy of azidothymidine
(AZT) in the treatment of patients with AIDS and AIDS-related complex.
A double-blind, placebo-controlled trial. N. Engl. J. Med. 1987, 317, 185–191.
3. Yarchoan, R.; Mitsuya, H.; Myers, C.E.; Broder, S. Clinical pharmacology of
30-azido-20,30-dideoxythymidine (zidovudine) and related dideoxynucleosides.
N. Engl. J. Med. 1989, 321, 726–738.
4. Richman, D.D.; Fischl, M.A, Grieco, M.H.; Gottlieb, M.S.; Volberding, P.A.;
Laskin, O.L.; Leedom, J.M.; Groopman, J.E.; Mildvan, D.; Hirsch, M.S.;
Jackson, G.G.; Durack, D.T.; Nusinoff-Lehrman, S.N. The toxicity of azido-
thymidine (AZT) in the treatment of patients with AIDS and AIDS-related
complex. A double-blind, placebo-controlled trial. N. Engl. J. Med. 1987,
317, 192–197.
5. Fletcher, C.V.; Collier, A.C. Principles and management of human immuno-
deficiency virus infection. In Pharmacotherapy. A Pathophysiologic Approach,
3rd Ed.; Dipiro, J.T., Talbert, R.L., Yee, G.C., Matzke, G.R., Wells, B.G.,
Posey, L.M., Eds.; Appleton & Lange, 1997; 2361–2363.
6. De Clercq, E. Toward improved anti-HIV chemotherapy: Therapeutic strate-
gies for intervention with HIV infections. J. Med. Chem. 1995, 38, 2491–2517.
7. De Clercq, E. Reverse transcriptase inhibitors as anti-HIV drugs. In Antiviral
Against AIDS; Unger, R.E., Kreuter, J., Ro¨bsamen-Waigmann, H., Eds.; Mar-
cel Dekker: New York, 2000; 107–150.
8. Darbyshire, J. Therapeutic interventions in HIV infection – a critical view.
Trop. Med. Int. Health 2000, 5, A26–A31.
9. Kukhanova, M.; Krayevsky, A.; Prusoff, W.; Cheng, Y.C. Design of anti-HIV
compounds: From nucleoside to nucleoside 50-triphosphate analogs. problems
and perspectives. Curr. Pharm. Des. 2000, 6, 585–598.
10. Tan, X.; Chu, C.K.; Boudinot, F.D. Development and optimization of anti-
HIV nucleoside analogs and prodrugs: A review of their cellular pharmacology,
structure-activity relationships and pharmacokinetics. Adv. Drug Deliv. Rev.
1999, 39, 117–151.
11. Parang, K.; Wiebe, L.I.; Knaus, E.E. Novel approaches for designing 50-O-
ester prodrugs of 30-azido-20,30-dideoxythymidine (AZT). Curr. Med. Chem.
2000, 7, 995–1039.

1802
Teijeiro et al.
0
´
´
12. Motura, M.I.; Moroni, G.N.; Teijeiro, S.A.; Salomon, H.; Brin˜on, M.C. 3 -
Azido-30-deoxy-50-O-isonicotinoylthymidine, a new prodrug of zidovudine.
synthesis, solid state characterization and anti-HIV-1 activity. Nucleos.
Nucleot. Nucl. Acids. 2002, 21, 217–230.
´
13. Teijeiro, S.A.; Moroni, G.N.; Motura, M.I.; Brin˜on, M.C. Lipophilic character
of pyrimidinic nucleoside derivatives: Correlation between shake flask, chroma-
tographic (RP-TLC and RP-HPLC) and theoretical methods. J. Liq. Chrom.
Rel. Technol. 2000, 23, 855–872.
´
14. Quevedo, M.A.; Moroni, G.N.; Brin˜on, M.C. Human serum albumin binding
of novel antiretroviral nucleoside derivatives of AZT. Biochem. Biophys. Res.
Comm. 2001, 288, 954–960.
´
´
15. Turk, G.; Moroni, G.N.; Pampuro, S.; Brin˜on, M.C.; Salomon, H. Antiretro-
viral activity and cytotoxicity of novel zidovudine (AZT) derivatives and the rela-
tion with their chemical structure. Int. J. Antimicrob. Agents 2002, 20, 282–288.
16. Connors, K.A.; Amidon, G.L.; Stella, V.J. Chemical Stability of Pharmaceuti-
cals, 2nd Ed.; Wiley Sons: New York, 1968.
17. Cartensen, J.T.; Rhodes, C.T. Drug stability. In Principles and Practices, 3rd
Ed.; Marcel Dekker: New York, 2000.
18. Wells, J.I. Pharmaceutical preformulation. In The Physicochemical Properties
of Drug Substances; Ellis Horwood: New York, 1993; 13–20.
19. Jambhekar, S.S. Biopharmaceutical properties of drug substances. In Principles
of Medicinal Chemistry, 4th Ed.; Foye, W.O., Lemke, T.L., Williams, D.A.,
Eds.; Williams and Wilkins: Baltimore, Philadelphia, 1995; 12–24.
20. Gringauz, A. Introduction to medicinal chemistry. In How Drugs Act and Why;
Willey-VCH: New York, 1997; 1–6.
21. Good, S.S.; Reynolds, D.J.; Miranda, P. de. Simultaneous quantification of
zidovudine and its glucuronide in serum by high-performance liquid chromato-
graphy. J. Chromatogr. 1988, 431, 123–133.
22. Mathes, L.E.; Muschik, G.; Demby, L.; Polas, P.; Mellini, D.W.; Issaq, H.J.;
Sams, R. High-performance liquid chromatographic determination of 20,30-
dideoxycytidine and 30-azido-30-deoxythymidine in plasma using a column-
switching technique. J. Chromatogr. 1988, 432, 346–351.
23. Unadkat, J.D.; Crosby, S.S.; Wang, J.P.; Hertel, C.C. Simple and rapid high-
performance liquid chromatographic assay for zidovudine (azidothymidine) in
plasma and urine. J. Chromatogr. 1988, 430, 420–423.
24. Irth, H.; Tocklu, R.; Welten, K.; Jong, G.J. de; Brinkman, U.A.; Frei, R.W.
Trace-level determination of 30-azido-30-deoxythymidine in human plasma by
preconcentration on a silver (I)-thiol stationary phase with on-line reversed-
phase high-performance liquid chromatography. J. Chromatogr. 1989, 491,
321–330.
25. Lacroix, C.; Hoang T.P.; Wojciechowski, F.; Duwoos, H.; Nouveau, J.;
DiquetB. Simultaneous quantification of zidovudine and its metabolites in
serum and urine by high-performance liquid chromatography using a col-
umn-switching technique. J. Chromatogr. 1990, 525, 240–245.
26. Stretcher, B.N.; Pesce, A.J.; Geisler, B.A.; Vine, W.H. A coupled HPLC=
radioimmunoassay for analysis of zidovudine metabolites in mononuclear cells.
J. Liq. Chromatogr. 1991, 14, 2261–2272.

Novel Antiretroviral Analog of Zidovudine. II
1803
27. Bolton, S. Pharmaceutical statistics. In Practical and Clinical Applications,
3rd Ed.; Marcel Dekker: New York, 1997; 216–264.
28. Morelli, B. ‘‘Zero-crossing’’ derivative spectrophotometric determination of
mixtures of cephapirin sodium and cefuroxime sodium in pure form and in
injections. Analyst 1988, 113, 1077–1082.
29. Morelli, B. Second-derivative spectrophotometric assay of mixtures of di-
cloxacillin sodium and ampicillin sodium in pharmaceuticals. J. Pharm. Sci.
1988, 77, 1042–1046.
´
30. Dabbene, V.G.; Brin˜on, M.C.; Bertorello, M.M. de. Stability determination of
3-bromo-2-hydroxy-N-(3,4-dimethyl-5-isoxazolyl)-1,4-naphthoquinon-4-imine
in ethanol by first-derivative spectrophotometry. J. Pharm. Sci. 1994, 83, 1617–
1621.
´
31. Dabbene, V.G.; Brin˜on, M.C.; Bertorello, M.M. de Determination of 3-chloro-
N-chloro-N-(3,4-dimethyl-5-isoxazolyl)-4-amine-1,2-naphthoquinone in pre-
sence of its degradation products using second derivative spectrophotometry.
Anal. Chim. Acta 1996, 318, 221–228.
32. Miller, J.N. Basic statistical methods for analytical chemistry. Part. 2. Calibra-
tion and regression methods. A Review. Analyst 1991, 116, 3–14.
33. Edwardson, P.A.; Bhaskar, G.; Fairbrother, J.E. Method validation in phar-
maceutical analysis. J. Pharm. Biomed. Anal. 1990, 8, 929–933.
34. ACS Subcommittee on Environmental Analytical Chemistry, Guidelines for
data acquisition and data quality evaluation in environmental chemistry. Anal.
Chem., 1980, 52, 2242–2249.
35. Mehta, A.C. The validation criteria for analytical methods used in pharmacy
practice research. J. Clin. Pharm. Ther. 1989, 14, 465–473.
36. a) Christensen, J.J., Rytting, J.H., Izatt, R.M., Thermodynamics of proton dis-
sociation in dilute aqueous solution. VIII. pK, DHꢀ, and DSꢀ values for proton
ionization from several pyrimidines and their nucleosides at 25ꢀ. J. Phys.
Chem., 1967, 71, 2700–2705; b) Christensen, J.J., Rytting, J.H., Izatt, R.M.,
Thermodynamics of proton dissociation in dilute aqueous solution. Part XIV.
pK, DHꢀ and DSꢀ values for proton dissociation from several pyrimidines and
their nucleosides at 10 and 40ꢀC. J. Chem. Soc. B. Phys. Org. 1970, 1643–1646.
37. Perrin, D.D. Prediction of pKa values. In Physical Chemical Properties of
Drugs; Yalkowsky, S.H., Sinkula, A.A., Valvani, S.C., Eds.; Marcel Dekker:
New York, 1980; Vol. 10, 1–48.
38. Benet, L.Z.; Goyan, J.E. Potentiometric determination of dissociation con-
stants. J. Pharm. Sci. 1967, 56, 665–680.
Received November 12, 2002
Accepted April 9, 2003