Synthesis and antimycobacterial evaluation of pyrazinamide derivatives with benzylamino substitution

Article abstract of DOI:10.1016/j.bmcl.2012.11.052

A series of 19 new compounds related to pyrazinamide were synthesized, characterized with analytical data and screened for in vitro whole cell antimycobacterial activity against Mycobacterium tuberculosis H37Rv, Mycobacterium kansasii and two types of Mycobacterium avium. The series consisted of 3-(benzylamino)-5-cyanopyrazine-2-carboxamides and 3-(benzylamino)pyrazine-2,5-dicarbonitriles with various substituents on the phenyl ring. RP-HPLC method was used to determine the lipophilicity of the prepared compounds. Nine compounds exerted similar or better activity against Mycobacterium tuberculosis compared to pyrazinamide (MIC = 6.25-12.5 μg/mL). 3-(Benzylamino)pyrazine-2,5-dicarbonitrile inhibited all of the tested mycobacterial strains with MIC within the range 12.5-25 μg/mL. Although not the most active, 4-NH₂ substituted compounds possessed the lowest in vitro cytotoxicity (hepatotoxicity), leading to selectivity index SI = 5.5 and SI >21.

Full text of DOI:10.1016/j.bmcl.2012.11.052

Bioorganic & Medicinal Chemistry Letters 23 (2013) 476–479
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and antimycobacterial evaluation of pyrazinamide derivatives
with benzylamino substitution
Jan Zitko a, , Pavla Paterová , Vladimír Kubícek , Jana Mandíková , František Trejtnar , Jirí Kuneš ,
⇑
b
a
a
a
a
ˇ
ˇ
a
ˇ
Martin Dolezal
a Faculty of Pharmacy in Hradec Králové, Charles University in Prague, Heyrovského 1203, Hradec Králové 500 05, Czech Republic
b Department of Clinical Microbiology, University Hospital, Sokolská 581, Hradec Králové 500 05, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 19 new compounds related to pyrazinamide were synthesized, characterized with analytical
data and screened for in vitro whole cell antimycobacterial activity against Mycobacterium tuberculosis
H37Rv, Mycobacterium kansasii and two types of Mycobacterium avium. The series consisted of 3-
(benzylamino)-5-cyanopyrazine-2-carboxamides and 3-(benzylamino)pyrazine-2,5-dicarbonitriles with
various substituents on the phenyl ring. RP-HPLC method was used to determine the lipophilicity of
the prepared compounds. Nine compounds exerted similar or better activity against Mycobacterium tuber-
Received 19 September 2012
Revised 9 November 2012
Accepted 13 November 2012
Available online 27 November 2012
Keywords:
Antimycobacterial activity
Benzylamines
Lipophilicity
Pyrazinamide derivatives
Cytotoxicity
culosis compared to pyrazinamide (MIC = 6.25–12.5
inhibited all of the tested mycobacterial strains with MIC within the range 12.5–25
l
g/mL). 3-(Benzylamino)pyrazine-2,5-dicarbonitrile
g/mL. Although not
l
the most active, 4-NH2 substituted compounds possessed the lowest in vitro cytotoxicity (hepatotoxicity),
leading to selectivity index SI = 5.5 and SI >21.
Ó 2012 Elsevier Ltd. All rights reserved.
In 2011 at least two studies proposing a specific subcellular tar-
get of pyrazinamide (PZA) and/or its metabolite pyrazinoic acid
(POA) were published in prestigious journals. Firstly, PZA and
POA were shown to bind to fatty acid synthase I (FAS I).1 This sup-
ports the previous studies which suggested that the inhibition of
this enzyme system was the mode of action of some PZA/POA
derivatives.2–5 Secondly, POA (but not PZA) was shown to bind to
ribosomal protein RpsA, leading to the blockage of trans-translation
(the process of removing ribosomes stalled upon mRNA).6 This
renewed interest in PZA can be attributed to its unique therapeutic
properties, for example the ability to kill metabolically low-active
mycobacteria which leads to shortening the course of therapy.
Although structural modifications of PZA have not yielded any clin-
ical agents thus far, the validation of a specific subcellular target of
PZA/POA could lead to modern strategies of structure-based drug
design which could enhance future analogue development.
As a part of our continuous research of new PZA derivatives with
antimycobacterial activity, we report a series of 19 compounds
derived from 3-chloro-5-cyanopyrazine-2-carboxamide (4) and
3-chloropyrazine-2,5-dicarbonitrile (6) by nucleophilic substitution
by various benzylamines. Similar compounds, derived from 4 and 6
by reaction with anilines (therefore missing the –CH2– bridge), were
prepared before and possessed moderate to good antimycobacterial
activity.7,8
Parent compounds 4 and 6 were prepared as described in liter-
ature7–10—see Scheme 1. Briefly, PZA was treated with peroxoace-
tic acid to yield the mixture of pyrazinecarboxamide-N-oxides.
This unseparated mixture was chlorinated with POCl3 to yield
three position isomers of chloropyrazine-2-carbonitrile (1–3). The
main product 6-chloropyrazine-2-carbonitrile (1) was separated
by flash chromatography and underwent homolytic amidation to
produce 3-chloro-5-cyanopyrazine-2-carboxamide (4), which was
dehydrated by POCl3 to 3-chloropyrazine-2,5-dicarbonitrile (6).
Alternatively, compound 6 can be prepared by dehydration of
6-chloro-5-cyanopyrazine-2-carboxamide (5) derived by amidation
of 3-chloropyrazine-2-carbonitrile (2) as described in our previous
paper.12
Final compounds 7a–7i and 8a–8j were prepared by nucleophilic
substitution by various benzylamines (Scheme 2). The substitution
of chlorine proceeded well at mild conditions (mostly at RT, reflux
in THF for benzylamines with electron withdrawing substituents –
CF3 or –NO2). Potassium carbonate was used as a base. The
prepared compounds (all of them yellow to pale yellow solid or
crystalline) were characterized with 1H NMR, 13C NMR, FT-IR spec-
troscopy, melting point and elementary analysis. The analytical
data were fully consistent with proposed structures and are in-
cluded in the Supplementary data.
Lipophilicity is a very important physicochemical property with
great impact on biological activity and in our previous studies we
have observed that antimycobacterial activity of trisubstituted
pyrazines was strongly dependent on their lipophilicity.8,11,12
⇑
Corresponding author. Tel.: +420 495067272; fax: +420 495067167.
E-mail address: jan.zitko@faf.cuni.cz (J. Zitko).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.11.052

J. Zitko et al. / Bioorg. Med. Chem. Lett. 23 (2013) 476–479
477
Scheme 1. Synthetic routes to parent compounds 4 and 6. Reagents: (a) (1) H2O2, AcOH; (2) POCl3; (b) (NH4)2S2O8, HCONH2; (c) POCl3.
culous mycobacteria (NTM; namely M. kansasii and two strains of
M. avium). On contrary, series 8 yielded six active compounds
and most of them exerted at least some activity against NTM as
well. Most significantly, compound 8a inhibited all of the tested
mycobacterial strains with MIC within the range 12.5–25
lg/mL
(53–106 M). The fact that pyrazinedicarbonitriles possessed bet-
l
ter activity against NTM in comparison to analogous cyanopyrazin-
ecarboxamides is in concordance with observations in our previous
series, for example compare Ref. 11,12. Compounds with electron-
donating substituents R = (4-CH3; 4-NH2) were active in both
series, whereas strongly electron-withdrawing substituents
R = (3-CF3; 3-NO2) produced active compounds only within series
7. Chlorine substituents R = (3-Cl; 4-Cl) were active in series 8,
but not in series 7. No direct correlation was found between anti-
mycobacterial activity (log (1/MIC)) and lipophilicity (logk). Also
Scheme 2. Synthesis of final compounds.
application of electronic Hammett constants
r of substituents R
on the aromatic ring did not produce a universal rule describing
the SAR applicable for both series.
Regarding antimycobacterial activity against M. tbc H37Rv, the
most active compounds (7b, 7d, 8a, 8b, 8e) exerted similar or
slightly better activity compared to PZA standard, that is
Therefore much attention was paid to correct determination of the
lipophilicity of the prepared compounds. Lipophilicity parameters
ClogP (ChemOffice) and ACD/logP of the prepared compounds
were calculated using software routinely engaged in our labora-
tory.13 Calculated values indicated that the compounds from series
8 (dicarbonitrile derivatives) were less lipophilic than correspond-
ing compounds from series 7 (cyanocarboxamides) with equal sub-
stitution on the phenyl ring (see Table 1). This was in contrast both
with our expectations and to the Rf values observed during TLC.
Therefore the lipophilicity was measured experimentally by means
of chromatographic separation on RP-HPLC C18 column and ex-
pressed as logk value derived from retention time of individual
compounds.14
MIC = 6.25–12.5
higher molecular weights of prepared compounds in comparison
with PZA, compound 8b (MIC = 25 M) scored two times better
then PZA (MIC = 51–102 M) against M. tbc H37Rv. For MICs in
M) see Table 1—values in parentheses.
As mentioned before, the final compounds reported in this Let-
lg/mL. Taking into consideration significantly
l
l
(l
ter are methylene homologs of 3-(phenylamino)-5-cyanopyrazine-
2-carboxamides and 3-(phenylamino)pyrazine-2,5-dicarbonitriles
published before, whose antimycobacterial activity against M. tbc
Lipophilicity substituent constants
p for substituents R were
H37Rv was MIC = 12.5–25 lg/mL and MIC = 8–16 lg/mL, respec-
calculated both from measured (logk) and calculated (ClogP,
ACD/logP) values (Table 2). In both series 7 and 8 the measured
tively.7,8 Although the comparison is limited by minor differences
in the testing method and sometimes different substituents R on
the aromatic ring, we conclude that the insertion of –CH2– bridge
is not primarily connected with the loss of antitubercular activity,
for at least some of the prepared methylene homologs possess
activity comparable with phenylamino predecessors.
p
p
values for individual substituents R were very similar—compare
(logk)7 versus (logk)8 columns in Table 2. This indicates the
p
capability of our method to measure correct values. The plot
p
(logk) on p(ACD/logP)—see Figure 1—showed a good linear corre-
lation (R2 = 0.985, n = 17) with regression line running very close to
Drug-induced hepatotoxicity is a well-known side effect of
many of the first line antituberculosis agents (PZA, isoniazid, rifam-
picin)16,17 and unfortunately there is a synergy effect which in-
creases the toxicity when these drugs are administered in
combination. Human liver carcinoma cell line (Hep G2) has been
used as a model to study this toxicity synergism in vitro.18,19
According to WHO treatment guidelines,20 the basic regime for
the treatment of non-complicated tuberculosis lasts at least
6 months and combines all of the potentially hepatotoxic antituber-
culosis agents mentioned above. Due to this multidrug-treatment
approach it is very probable that a newly developed antitubercular
drug would be used in combination with at least some of these
hepatotoxic drugs; therefore the hepatotoxicity of new drugs
should be considered very cautiously.
the point [0;0] (which would be intersected if the correlation was
ideal). Similarly, the plot
p(logk) on p(ClogP) showed correlation
R2 = 0.966, n = 17 (not shown). This correlation indicates that both
ChemOffice ClogP and ACD/logP algorithms were accurate in cal-
culating
p increments based on the substituents of the phenyl ring.
According to experimentally determined logk, all dicarbonitriles
8a–8i were little more lipophilic compared to matching cyanocarb-
oxamides 7a–7i. The virtual exchange of 5-CONH2 group to 5-CN
conferred d (logk) = 0.034 0.016 (mean SD).
Prepared compounds were screened for antimycobacterial
activity against M. tuberculosis H37Rv, M. kansasii and two strains
of M. avium by microdilution panel method.15 For results see Ta-
ble 1. In series 7 there were four compounds active against M.
tuberculosis, but none of them was active against tested nontuber-

478
J. Zitko et al. / Bioorg. Med. Chem. Lett. 23 (2013) 476–479
Table 1
Summary of prepared compounds, their antimycobacterial activity expressed as MIC (
lg/mL) and comparison of calculated lipophilicity parameters with determined logk values
Compound Antimycobacterial activity ( g/mL) (
l
l
M)
Cytotoxicity
M) SI
n.d.
Lipo/hydrophobicity
No.
R
MW
253.26 >100
267.29 12.5 (47)
M. tbcH37Rva
M. kansasiib
M. aviumc
M. aviumd
IC50
(l
logk
ClogP
ACD/logP
7a
7b
7c
7d
7e
7f
7g
7h
7i
8a
8b
8c
8d
8e
8f
8g
8h
8i
H
4-CH3
4-OCH3 283.29 >100
4-NH2
3-Cl
>100
>100
>100
>100
>100
>100
>100
>100
>100
25 (106)
>100
25 (99)
>100
12.5 (46)
12.5 (46)
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
25 (106)
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
12.5 (53)
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
n.d.
54.6
n.d.
0.3260
0.5533
0.3074
À0.2005
0.5141
0.5421
0.7353
0.1331
0.5104
0.3350
0.5598
0.3259
1.350
1.849
1.269
0.123
2.063
2.063
2.656
1.093
2.233
1.165
1.664
1.084
2.98 0.62
3.44 0.62
2.90 0.63
1.70 0.62
3.57 0.63
3.57 0.63
4.04 0.64
2.71 0.63
3.55 0.65
2.26 0.67
2.72 0.67
2.18 0.68
0.98 0.68
2.86 0.68
2.86 0.68
3.33 0.69
1.99 0.68
2.84 0.70
2.72 0.67
1.2
n.d.
>21.3
n.d.
n.d.
n.d.
5.4
0.4
1.1
2.8
0.5
5.5
0.9
0.4
n.d.
n.d.
n.d.
n.d.
268.27 12.5 (47)
287.70 >100
287.70 >100
>1000
n.d.
4-Cl
n.d.
n.d.
3,4-Cl2 322.15 >100
3-NO2
3-CF3
H
298.26 25 (84)
321.26 25 (78)
235.24 12.5 (53)
249.27 6.25 (25)
457.0
27.8
60.8
70.8
95.9
548.9
40.0
35.5
n.d.
4-CH3
4-OCH3 265.27 50 (188)
4-NH2
3-Cl
250.26 25 (100)
269.69 12.5 (46)
269.69 25 (93)
À0.1621 À0.062
0.5456
0.5735
0.7830
0.2049
0.5615
0.5043
1.878
1.878
2.471
0.908
2.048
1.614
4-Cl
3,4-Cl2 304.13 >100
3-NO2
3-CF3
2-CH3
—
280.24 >100
303.24 >100
249.27 >100
123.11 6.25–12.5 (51–102) >100
137.14 1.56 (11)
>100
>100
>100
n.d.
n.d.
n.d.
8j
PZA
INH
79 Â 103e 775–1550e n.d.
À0.676 À0.37 0.27
—
12.5–25 (91–182) 6.25–12.5 (46–91) 3.13–6.25 (23–46) n.d.
n.d.
n.d.
À0.668 À0.89 0.24
Numbers in parentheses represent the MICs converted to molar concentrations (
l
M).
a
CNCTC My 331/88.
CNCTC My 235/80.
CNCTC My 80/72.
CNCTC My 152/73.
b
c
d
e
IC50 = 79.1 mM from literature18, in comparable Hep G2 cytotoxicity assay; SI calculated for M. tbc and MIC = 51–102
lM
Table 2
Lipophilicity constants
p
for substituents R derived from measured and calculated
parameters
R
p
(logk)7
p(logk)8
p
(ACD/logP)
p(ClogP)
H
0
0
0
0
4-CH3
4-OCH3
4-NH2
3-Cl
0.227
0.225
0.460
0.499
À0.019
À0.527
0.188
0.216
0.409
À0.009
À0.497
0.211
0.239
0.448
À0.130
0.226
—
À0.080
À1.280
0.595
0.595
1.065
À0.081
À1.227
0.713
0.713
1.306
4-Cl
3,4-Cl2
3-NO2
3-CF3
2-CH3
À0.193
À0.270
À0.257
0.184
0.575
0.883
0.169
0.460
0.449
Figure 2. Cytotoxicity of final compounds expressed as log(1/IC50) increases with
lipophilicity logk. The grey line represents possible coordinates of the point of 7d—
exact vertical coordinate not known.
In vitro cytotoxicity (hepatotoxicity) evaluation was performed
for all compounds with antimycobacterial activity. The decrease of
viability of human liver Hep G2 cells was measured using a stan-
dard protocol21,22 based on colorimetric method measuring reduc-
tion of tetrazolium salt. According to in vitro cytotoxicity, the
tested compounds can be divided into three groups. Most of the
compounds fell into group with IC50 at the order of tens of
Compounds 7h and 8d form the group with moderate cytotoxicity
(IC50 at hundreds of M, approx. 500 M). The least cytotoxic
lM.
l
l
group comprised of compound 7d, for which IC50 could not be
determined within the used dilution range and lies above
1000 lM. Cytotoxicity of tested compounds was obviously
strongly dependent on their lipophilicity (logk). All of the com-
pounds with low (7d) and moderate (7h, 8d) cytotoxicity pos-
sessed the lowest lipophilicity (logk) at the same time—see
Figure 2. The increased cytotoxicity of more lipophilic compounds
Figure 1. Plot of measured
p
substituent constants p(logk) on calculated constants
p(ACD/logP) for aromatic substituents R.

J. Zitko et al. / Bioorg. Med. Chem. Lett. 23 (2013) 476–479
479
might be caused by the increase of their intracellular
References and notes
accumulation.
1. Sayahi, H.; Zimhony, O.; Jacobs, W. R., Jr.; Shekhtman, A.; Welch, J. T. Bioorg.
Med. Chem. Lett. 2011, 21, 4804.
2. Zimhony, O.; Cox, J. S.; Welch, J. T. Nat. Med. 2000, 6, 1043.
3. Boshoff, H. I.; Mizrahi, V.; Barry, C. E. J. Bacteriol. 2002, 184, 2167.
4. Zimhony, O.; Vilcheze, C.; Arai, M.; Welch, J. T.; Jacobs, W. R. Antimicrob. Agents
Chemother. 2007, 51, 752.
5. Ngo, S.; Zimhony, O.; Chung, W. J.; Sayahi, H.; Jacobs, W. R.; Welch, J. T.
Antimicrob. Agents Chemother. 2007, 51, 2430.
6. Shi, W.; Zhang, W.; Jiang, X.; Yuan, H.; Lee, J. S.; Barry, C. E.; Wang, H. H.; Zhang,
W. H.; Zhang, Y. Science 2011, 333, 1630.
7. Dolezal, M.; Hartl, J.; Lycka, A.; Buchta, V.; Odlerova, Z. Collect. Czech. Chem.
Commun. 1995, 60, 1236.
8. Palek, L.; Dvorak, J.; Svobodova, M.; Buchta, V.; Jampilek, J.; Dolezal, M. Arch.
Pharm. 2008, 341, 61.
9. Dlabal, K.; Palat, K.; Lycka, A.; Odlerova, Z. Collect. Czech. Chem. Commun. 1990,
55, 2493.
10. Jampilek, J.; Dolezal, M.; Kunes, J.; Satinsky, D.; Raich, I. Curr. Org. Chem. 2005,
9, 49.
Selectivity index (SI) defined as IC50/MIC in molar concentra-
tions was calculated for antimycobacterial activity against M.
tuberculosis. SI >10 is generally recognized as a safe value for a drug
candidate, granting the sufficient difference between efficient
(regarding the intended biological effect) and cytotoxic concentra-
tions. Among the prepared compounds, only compound 7d met
this criterion. PZA itself had been reported to possess relatively
low single dose Hep G2 cytotoxicity of IC50 = 79.1 mM18, which,
taking into account the antitubercular activity against M. tuberculosis
determined in our assay, would lead to estimated SI = 775–1550.
Obviously, the benzylamine derivatives presented in this paper
exerted significantly higher levels of in vitro Hep G2 cytotoxicity.
However, the SI indexes of compounds 7d, 7h and 8d indicate that
the antimycobacterial activity of these derivatives is rather specific
than nonselective based on general cytotoxicity.
11. Zitko, J.; Jampilek, J.; Dobrovolny, L.; Svobodova, M.; Kunes, J.; Dolezal, M.
Bioorg. Med. Chem. Lett. 2012, 22, 1598.
To sum up, nine of the prepared compounds exerted similar or bet-
ter activity against M. tuberculosis compared to PZA. Compound 8b
possessed the best antimycobacterial activity against M. tuberculosis;
8a inhibited all of the tested strains and had the broadest activity
spectrum; 7d combined good antimycobacterial activity against M.
tuberculosis with relatively low cytotoxicity (hepatotoxicity). In our
opinion these compounds are suitable as a subject to further
structural modifications.
12. Zitko, J.; Dolezal, M.; Svobodova, M.; Vejsova, M.; Kunes, J.; Kucera, R.; Jilek, P.
Bioorg. Med. Chem. 2011, 19, 1471.
13. CS ChemOffice Ultra ver. 12.0 (CambridgeSoft, Cambridge, MA, USA) and ACD/
logP v.12.01 (Advanced Chemistry Development Inc., Toronto, Canada).
14. Logk is the logarithm of capacity factor k calculated according to the formula
k = (tR À tD); tR is the retention time of individual compound, tD is the dead
time of the system determined as the retention time of KI methanol solution.
See Supplementary data for experimental details.
15. Microdilution panel method. M. tuberculosis H37Rv CNCTC My 331/88, M.
kansasii Hauduroy CNCTC My 235/80, M. avium ssp. avium Chester CNCTC My
80/72 and M. avium CNCTC My 152/73 were obtained from Czech National
Collection of Type Cultures (CNCTC), National Institute of Public Health,
Acknowledgments
Prague, Czech Republic. MIC [lg/mL] was determined visually as the lowest
concentration of tested compound that inhibited the growth of mycobacteria.
See Supplementary data for experimental details.
16. Yew, W. W.; Leung, C. C. Respirology 2006, 11, 699.
17. Tostmann, A.; Boeree, M. J.; Aarnoutse, R. E.; de Lange, W. C. M.; van der Ven,
A.; Dekhuijzen, R. J. Gastroenterol. Hepatol. 2008, 23, 192.
18. Tostmann, A.; Boeree, M. J.; Peters, W. H. M.; Roelofs, H. M. J.; Aarnoutse, R. E.;
van der Ven, A.; Dekhuijzen, P. N. R. Int. J. Antimicrob. Agents 2008, 31, 577.
19. Singh, M.; Sasi, P.; Rai, G.; Gupta, V. H.; Amarapurkar, D.; Wangikar, P. P. Med.
Chem. Res. 2011, 20, 1611.
This publication is a result of the project implementation: ‘Sup-
port of establishment, development, and mobility of quality re-
search teams at the Charles University’, project number CZ.1.07/
2.3.00/30.0022, supported by The Education for Competitiveness
Operational Programme (ECOP) and co-financed by the European
Social Fund and the state budget of the Czech Republic. This study
was also supported by the Ministry of Education, Youth and Sports
of the Czech Republic (SVV-2012-265-001) and Ministry of Health
of the Czech Republic—IGA NT 13346 (2012).
20. Treatment of tuberculosis: guidelines—4th ed. WHO document no. WHO/HTM/
TB/2009.420, 2010. ISBN 978 92 4 154783 3.
21. Promega Corporation. CellTiter 96Ò AQueous one solution cell proliferation
assay. Owen, T.C. U.S. Patent, 5185,450, 1993.
22. Kratky, M.; Vinsova, J.; Volkova, M.; Buchta, V.; Trejtnar, F.; Stolarikova, J. Eur. J.
Med. Chem. 2012, 50, 433.
Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.
11.052.