Synthesis and anti-microbial activities of some pyridinium salts with alkoxymethyl hydrophobic group

Article abstract of DOI:10.1016/S0223-5234(01)01280-6

A novel class of functionalised cationic surfactant has been obtained. The work-up procedure of synthesis is very simple, the yield is high and the pyridinium salts with alkoxymethyl hydrophobic group are easily purified. All the salts examined showed anti-microbial activities. Some of them exhibited strong activity and wide anti-bacterial spectra similar to the activity of benzalkonium chloride. The relationship between chemical structure and anti-microbial activity was analysed by the QSAR method.

Full text of DOI:10.1016/S0223-5234(01)01280-6

Eur. J. Med. Chem. 36 (2001) 899–907
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01280-6/FLA
Laboratory note
Synthesis and anti-microbial activities of some pyridinium salts with
alkoxymethyl hydrophobic group
Juliusz Pernaka,*, Joanna Kalewskaa, Hanna Ksycin´skab, Jacek Cybulskib
aPoznan´ University of Technology, Skłodowskiej-Curie 2, 60-965 Poznan´, Poland
bPharmaceutical Research Institute, Rydgiera 8, 01-793 Warsaw, Poland
Received 5 June 2001; revised 25 September 2001; accepted 25 September 2001
Abstract – A novel class of functionalised cationic surfactant has been obtained. The work-up procedure of synthesis is very simple,
the yield is high and the pyridinium salts with alkoxymethyl hydrophobic group are easily purified. All the salts examined showed
anti-microbial activities. Some of them exhibited strong activity and wide anti-bacterial spectra similar to the activity of
benzalkonium chloride. The relationship between chemical structure and anti-microbial activity was analysed by the QSAR method.
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS
pyridium salts / alkoxymethylation / nicotinamide / anti-microbial activities / QSAR analysis
1. Introduction
is neatly complemented by their low toxicity for man
and other mammalian species. Such low mammalian
toxicity is an important factor in their broad use as
disinfecting or sanitising agents.
Pyridinium halides, invert soaps, possess surface-
active properties, detergency, anti-microbial proper-
ties and adsorption on negatively charged solids.
1-Alkyl-3-carbamoylpyridinium halides are a typical
example of functionalised surfactants, defined as com-
pounds containing reactive functions covalently
bound to the hydrophobic long chain and can aggre-
gate and exhibit chemical or biochemical reactivity at
the same time [1–3]. The anti-microbial activity of
1-alkylpyridinium salts dependent on the adsorptive
activities on the surface of bacterial cells and as a
consequence their destruction [4] and the acidic disso-
ciation constants (pKa) of the corresponding pyridines
[5]. In addition, it was proved that the factors which
control their anti-microbial activity, are molecular
hydrophobicity, adsorbability, surface activity, elec-
tron density of the ammonium nitrogen atom or
bacterioclastic activity [6]. The high toxicity of pyri-
dinium halides for microorganisms which is the basis
for their effective germicidal and deodouring activity
1-Alkyl- and 1-alkoxymethyl-3-carbamoylpyri-
dinium derivatives can be used as model compounds
of nicotinamide-adenine dinucleotide (NAD+). The
reaction of nicotinamide with chloromethyl methyl
ether is known [7] but no systematic study of 1-
alkoxymethyl-3-carbamoylpyridinium salts has so far
been carried out. Treatment of nicotinamide or isoni-
cotinamide with chloromethyl methyl sulphide in
CF3CO2H or MeSO3H yield the monomethylation of
amide, has been described as well [8]. In this study, we
will present the synthesis of novel biocides, which
possess carbamoyl or monosubstituted carbamoyl
group in a pyridinium ring and describe their anti-mi-
crobial activity. In our previous study of the synthesis
of some pyridinium chlorides with alkylthiomethyl or
alkoxymethyl hydrophobic group and their anti-mi-
crobial properties, we discovered they demonstrated
wide anti-microbial spectra against bacteria and fungi
[9–13]. Every day more and more information is
available about structure–activity relationships. The
purpose of the present investigation is to see if there is
* Correspondence and reprints
E-mail address: juliusz.pernak@put.poznan.pl (J. Pernak).

900
any
J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
Table I. Pyridinium salts (1–6) prepared.
correlation
between
1-alkoxymethylcar-
bamoylpyridinium chlorides and their anti-microbial
activity.
Salts
R
Anion Y
Yield (%)
M.p. (°C)
1a
1b
1c
1d
1e
1f
C2H5
C3H7
C4H9
Cl−
Cl−
Cl−
97
96
93
94
92
98
96
98
95
97
97
90
90
95
96
98
98
98
97
99
98
98
99
99
141–143
140–142
145–148
146–148
157–160
151–153
160–162
165–168
165–168
162–165
165–168
165–168
168–170
170–173
171–173
173–175
157–160
105–108
76–78
C5H11 Cl−
C6H13 Cl−
C7H15 Cl−
C8H17 Cl−
C9H19 Cl−
C10H21 Cl−
C11H23 Cl−
C12H25 Cl−
2. Chemistry
1-Alkoxymethylcarbamoylpyridinium chlorides 1,
2, 6 were prepared in very good yields by the nucle-
ophilic attack of nicotinamide or isonicotinamide on
chloromethyl alkyl ethers according to the SN1 mech-
anism. In N-(hydroxymethyl)-3-pyridinecarboxamide,
the attack of chloromethyl alkyl ether directed at the
O and N positions and O- and N-alkoxymethylated
product, simultaneously, 5 was obtained (figure 1).
Pyridinium salts 3, 4 were prepared via metathesis
reactions from the corresponding pyridinium chlo-
rides 1, 2 in aqueous solution. All the prepared chlo-
rides are crystalline solids of the hydroscopic nature
with the exception of 6a–e, which are very hygro-
scopic. All the salts prepared are air-stable under
ambient conditions and may be handled under nor-
mal laboratory conditions. Pyridinium chlorides 1, 2,
5, 6 are soluble in DMF, DMSO and water. We
found that they are water-stable to the temperature
60 °C, above which they hydrolysed to alcohol, form-
aldehyde and hydrochloride of nicotinamide.
1g
1h
1i
1j
1k
2a
2b
2c
2d
2e
3a
3b
3c
3d
3e
3f
C5H9
Cl−
C6H11 Cl−
C7H13 Cl−
C8H15 Cl−
C12H23 Cl−
C12H25 Br−
C12H25 I−
C12H25 NO−3
C12H25 ClO−4
C8H17 BF−4
C10H21 BF−4
C11H23 BF−4
C12H25 BF−4
90–94
65–67
77–81
78–81
3g
3h
3i
80–83
C12H25 CH3COO− 96
149–152
162–165
167–169
105–108
168–170
167–170
109–111
112–116
111–115
110–114
112–115
114–117
116–119
119–120
115–118
119–121
Hygroscopic
Hygroscopic
Hygroscopic
Hygroscopic
Hygroscopic
163–165
162–165
161–163
162–164
166–169
4a
4b
4c
4d
4e
5a
5b
5c
5d
5e
5f
C10H21 CoCl24−
C10H21 CuCl24−
C10H21 ZnCl24−
C10H21 MgCl24−
C10H21 FeCl−4
97
96
97
98
96
70
72
73
72
70
65
78
78
79
78
89
90
91
90
90
91
90
92
92
93
Anhydrous salts were obtained by heating the sam-
ples at 40 °C under vacuum until no water signals are
observed in the infrared spectrum around 3500 cm−1.
Under such conditions, no decomposition has been
C3H7
C4H9
Cl−
Cl−
C5H11 Cl−
C6H13 Cl−
C7H
C8H
C9H
Cl−
Cl−
Cl−
1
15
17
19
observed as evidenced by H-NMR spectroscopy.
5g
5h
5i
C10H21 Cl−
C11H23 Cl−
C12H25 Cl−
5j
6a
6b
6c
6d
6e
6f
C3H7
C4H9
Cl−
Cl−
C5H11 Cl−
C6H13 Cl−
C7H15 Cl−
C8H17 Cl−
C9H19 Cl−
C10H21 Cl−
C11H23 Cl−
C12H25 Cl−
6g
6h
6i
6j
As the chain length increases, the van der Waals
interactions increase and produce an increase in the
melting point, which is shown in table I. Also follows
Figure 1. Synthesis of pyridinium salts 1–5.

J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
901
3. Anti-microbial activity
stabilised microphase separation of the hydrophilic
ionic headgroups and hydrophobic alkyl chains on
the solid state. For the mesomorphic salts 3 and 4, the
variations in melting points presented in table I are
greater compared with the increase in melting points
with chain length. It is remarkable that the melting
point of salts 3 is the anion influence. In the case of
pyridinium tetrafluoroborate and nitrate, the melting
points are lower. These observations follow the gen-
eral trends for prepared pyridinium salts that the
melting points decrease on changing the anion in the
order: [Cl−, CoCl42−, CuCl24−, MgCl42−, FeCl41−]>[Br−,
CH3COO−]>I− >ZnCl24− >ClO42− >[BF−4, NO−3].
All prepared pyridinium salts were tested for anti-
microbial activity against cocci, rods, fungi and
bacillus.
4. Results and discussion
New pyridinium salts 1–6 listed in table I were
synthesised by the method shown in figures 1 and 2.
Quaternisation of pyridine with alkyl halides, the
Menschutkin reaction, has been the subject of numer-
ous mechanistic studies [14–17]. The reaction mecha-
nisms SN2 have been proposed based on kinetic
studies. Quaternisation of pyridine derivative (include
nicotinamide and isonicotinamide) by chloromethyl
alkyl ether or sulphide run according to the SN1
mechanism [18, 19], where the formation of cation by
ionisation is the slowest step. Chloromethyl alkyl
ether is an excellent reagent but very easily hy-
drolysed. In this situation, quaternisation should be
conducted under strictly anhydrous conditions.
The minimal inhibitory concentration (MIC) values
determined for prepared chlorides (1, 2, 5, 6) are
given in table II and for salts (3, 4) in table III. As
shown by the results in these tables, all studied com-
pounds are very active against cocci and fungi and
active against rods and baccillus. In general, quater-
nary ammonium halides such as 1-alkylpyridinium
halides exhibit strong bacteriostatic activities against
gram positive bacteria rather than those against gram
negative bacteria [20, 21]. The activity of studied
compounds depends on the length of the substituent.
The most favourable alkoxymethyl group is that
which contains 10–12 carbon atoms in 1-
alkoxymethyl(carbamoyl)pyridinium compounds (1–
4 and 6). Activities of the salts synthesised against
various microorganisms are significantly different.
They are the most effective against cocci and show
good activity against bacillus and fungi. Their activity
against rods is rather poor. Also activities depend on
the kind of anion in salts 3, 4. For anion without
metal, the mean values of MIC increase in the order
of anions
The new pyridinium salts were characterised by
1
their H- and 13C-NMR spectra and microanalysis
CHN. The results of elemental analysis were in agree-
ment with the theoretical values. Generally, in salts
1–4 protons of N+CH2O were observed as a singlet at
6.0 ppm, protons of the amid group as two singlets at
8.9 and 8.3 and protons of the aromatic ring as a
singlet at 9.7, a triplet at 8.4 with a coupling constant
J=6.7 Hz and two doublets at 9.3 and 9.2. In the O-
and N-alkoxymethylated product 5, protons of group
NHCH2OCH2OR were resonated as a triplet at 10.7
with a coupling constant J=6.2 Hz, as doublet at 4.9
and a singlet at 4.8, respectively. Examination of the
1H-NMR of chloride 6 shows modification in aro-
matic section, protons resonated at two doublets at
9.3 and 8.5. In the 13C-NMR spectra the chemical
shift of the carbonyl carbon appeared at 163 ppm, the
carbon of N+CH2O at 89 and the carbons of group
NHCH2OCH2OR were observed at 93 and 72 ppm,
respectively. Thus, these NMR data are consistent
with the proposed structures of compounds 1, 2, 5
and 6. The NMR spectra of salts 3 are identical as for
corresponding chlorides 1.
CH3COO− >[Br=NO3]− >Cl− >ClO4− >BF−4 >I−
and for anion with metal in the order
Figure 2. Synthesis of pyridinium chloride 6.
MgCl24− =CoCl24− >ZnCl42− >CuCl42− >FeCl41−

902
J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
Table II. The MIC values a of pyridinium chlorides (1, 2, 5, 6).
Strains
Chlorides
b
1i
1j
1k
2e
5e
5f
5g
5h
6h
6i
6j
BAC
Rods
P. aeruginosa
E. coli
P. 6ulgaris
K. pneumoniae
761
381
381
190
730
365
365
365
351
351
351
175
1410
1410
1410
1410
281
35
35
132
17
33
250
31
62
946
473
473
473
761
381
381
381
730
182
182
182
351
88
88
88
11
22
11
70
33
62
88
Cocci
S. aureus
M. luteus
E. faecalis
M. catarrhalis
95
95
95
48
23
11
11
44
44
44
11
176
705
705
176
9
18
18
4
4
17
17
2
4
31
31
4
237
118
59
186
186
186
48
182
23
23
88
5
11
11
6
11
22
11
0.2
59
23
Bacillus
B. subtilis
381
182
88
1410
18
17
16
59
381
182
88
11
Fungi
C. albicans
R. rubra
381
381
91
46
44
22
705
353
35
35
33
17
16
16
59
59
381
381
182
182
44
88
11
11
a In mM, the number of microorganisms in 1 mL range from 104 to 105.
b BAC, benzalkonium chloride.
Table III. The MIC values a of pyridinium salts (3, 4).
Strains
Salts
b
3a
3b
3c
3d
3h
3i
4a
4b
4c
4d
4e
BAC
Rods
P. aeruginosa
E. coli
P. 6ulgaris
K. pneumoniae
312
156
156
156
1116
279
558
279
326
163
163
82
594
149
149
149
613
153
306
153
329
82
82
100
25
25
200
50
50
100
25
50
103
26
26
176
88
11
22
11
88
88
44
82
25
50
50
26
Cocci
S. aureus
M. luteus
E. faecalis
M. catarrhalis
78
19
78
19
279
140
140
70
82
41
82
20
149
74
149
37
306
306
306
77
82
41
82
10
6.3
6.3
13
1.6
6.2
25
13
25
3.2
44
22
44
11
6
11
22
11
6.2
6.2
3.1
6.5
6.5
0.8
3.1
Bacillus
B. subtilis
78
140
82
74
153
82
25
25
25
26
44
11
Fungi
C. albicans
R. rubra
78
78
70
140
82
82
74
149
153
153
82
82
6.3
13
13
25
25
25
0.5
13
44
44
11
11
a In mM, the number of microorganisms in 1 mL range from 104 to 105.
b BAC, benzalkonium chloride.
This result shows the effect on the size of anion of
the pyridinium salt and the anti-microbial activity.
The most favourable anions, as seen in table III, are
MgCl24−, CoCl24−, ZnCl42− and FeCl14−. Figure 3 shows
the dependency of the calculated average MIC values
for all the strains upon the alkoxymethyl chain length
for the O- and N-alkoxy-methylated product (5). The
anti-microbial activity is greatly affected by the
alkoxymethyl chain length in two positions. The sub-
stituent elongation in positions one and three have an

J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
903
It is supposed that a good estimation of chemical
compound lipophilicity may be obtained using high
performance liquid chromatography (HPLC) on hy-
drophobic stationary phases (e.g. C18 or C8) and
water and methanol mixture as an eluent. Capacity
factors for a compound, extrapolated to pure water,
according to the equation, give a log kw value which is
often referred to hydrophobic index.
log k=log kw −Sb
where k is the capacity factor [k=(t−t0)/t0 where t,
retention time of an investigation compound, to, dead
volume of a column], b, volume fraction of methanol
in a mobile phase, S, constant for a given solute–elu-
ent combination.
The linear extrapolation of retention data to 100%
water in chromatographic partition-like systems was
first suggested and theoretically verified by Soczewin´-
ski and Watmeister. With the very hydrophobic struc-
ture of the chlorides 5, it was necessary to replace the
C18 column to C8 stationary phase [22, 23].
In view of QSAR analysis, further modifications of
pyridinium chlorides should be made to increase their
lipophilicity. Of course, one can expect that the in-
crease of log(1/MIC) with log P or log kw will not be
unlimited. High log P or log kw values will cause a
gradual decrease of bioactivity as is generally ob-
served with chemotherapeutic agents. It is known that
the hydrophobic alkyl chain of anti-microbial quar-
ternary ammonium compounds has the important
function of adsorbing the surface of the bacterial cell,
and, on the other hand, interacts with environmental
hydrophobic materials [6].
Figure 3. Relation between alkoxy chain length and anti-mi-
crobial activity.
influence on the biological activity. The curves
demonstrated optimum anti-microbial efficiency, the
favourite alkoxy group is the one which contains
seven to nine carbon atoms. This activity should be
ascribed to the substituent elongation in position
three of the pyridine ring. The anti-microbial activi-
ties of prepared compounds 4a, 4d and 5f are similar
to the activity of commercially available benzalko-
nium chloride (BAC, Aldrich product, in which R
represents a mixture of alkyls from C8H17 to C18H37).
The results of QSAR analysis for anti-microbial
activities by prepared pyridinium chlorides (1, 2, 5
and 6) given in tables IV and V demonstrate the
importance of lipophilicity. Each of the 11 anti-micro-
bial activities increases with the lipophilicity of the
pyridinium chlorides.
Table IV. Slopes (a) and intercepts (b) of linear relationship between logarithm of reciprocal of MIC against individual microbial
strains and the theoretically calculated logarithm of octanol–water partition coefficient, log(1/MIC)=a+b·log P+c·log P2.
Strain
a
b
c
r
s
p
P. aeruginosa
E. coli
P. 6ulgaris
K. pneumoniae
S. aureus
−3.2384 (90.0426)
−3.1839 (90.0776)
−3.1810 (90.0703)
−3.1757 (90.0681)
−3.1197 (90.1063)
−3.1778 (90.1661)
−3.0618 (90.1091)
−2.9619 (90.1358)
−3.1741 (90.0848)
−3.1579 (90.0874)
−3.1874 (90.0887)
0.1490 (90.0272)
0.3021 (90.0495)
0.2922 (90.0448)
0.2968 (90.0434)
0.4353 (90.0677)
0.5885 (90.0964)
0.4627 (90.0695)
0.5977 (90.0865)
0.3575 (90.0540)
0.3898 (90.0557)
0.3405 (90.0565)
−0.0112 (90.0038)
−0.0267 (90.0069)
−0.0264 (90.0063)
−0.0271 (90.0061)
−0.0406 (90.0095)
−0.0609 (90.0119)
−0.0452 (90.0098)
−0.0589 (90.0122)
−0.0302 (90.0076)
−0.0341 (90.0078)
−0.0277 (90.0079)
0.733
0.740
0.758
0.771
0.747
0.791
0.751
0.762
0.773
0.785
0.750
0.207
0.378
0.342
0.331
0.517
0.509
0.531
0.661
0.413
0.425
0.432
510−5
510−5
510−5
510−5
510−5
510−5
510−5
510−5
510−5
510−5
510−5
M. luteus
E. faecalis
M. catarrhalis
C. albicans
R. rubra
B. subtilis
r, denotes the correlation coefficient; s, is the standard error of estimation; p, is the significance level of regression equations;
numbers in parenthesis are S.D. of regression coefficients.

904
J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
Table V. Slopes (a) and intercepts (b) of linear relationship between logarithm of reciprocal of MIC against individual microbial
strains and the logarithm of experimental parameter of lipophilicity, log(1/MIC)=a+b·log kw+c·log k2w.
Strain
a
b
c
r
s
p
P. aeruginosa
E. coli
P. 6ulgaris
K. pneumoniae
S. aureus
−3.3636 (90.0426)
−3.3857 (90.1777)
−3.3798 (90.1592)
−3.3713 (90.1568)
−3.3671 (90.2499)
−3.3872 (90.2115)
−3.3913 (90.2890)
−3.308 (90.3157)
−3.4148 (90.1987)
−3.3963 (90.2096)
−3.3973 (90.2034)
0.0286 (90.0061)
0.0469 (90.0122)
0.0450 (90.0110)
0.0449 (90.0108)
0.0663 (90.0172)
0.0599 (90.0185)
0.0737 (90.0171)
0.0888 (90.0217)
0.0571 (90.0137)
0.0591 (90.0144)
0.0556 (90.0140)
−0.0131 (90.0041)
−0.0381 (90.0056)
−0.0322 (90.0055)
−0.0298 (90.0059)
−0.0412 (90.0087)
−0.0594 (90.0108)
−0.0474 (90.0102)
−0.0602 (90.0118)
−0.0287 (90.0084)
−0.0372 (90.0065)
−0.0283 (90.0091)
0.675
0.601
0.627
0.632
0.603
0.612
0.645
0.625
0.634
0.626
0.614
0.249
0.497
0.445
0.439
0.699
0.632
0.697
0.883
0.556
0.587
0.569
8×10−5
7×10−4
3×10−4
3×10−4
7×10−4
2×10−4
2×10−4
4×10−4
3×10−4
4×10−4
5×10−5
M. luteus
E. faecalis
M. catarrhalis
C. albicans
R. rubra
B. subtilis
r, denotes the correlation coefficient; s, is the standard error of estimation; p, is the significance level of regression equations;
numbers in parenthesis are S.D. of regression coefficients.
Also salts 3, 4, 5 were tested against Mycobacterium
tuberculosis H37Rv (ATCC 27294). It was found in the
preliminary screening anti-tubercular activity, accord-
ing to an international program with the Tuberculosis
Antimicrobial Acquisition & Coordinating Facility
(TAACF) [24], that only 3i, 4d and 5g are active (for
3i, 4d, 5g, MIC>6.25 mg mL−1, inhibition 40%).
In summary, we found new functionalised cationic
surfactants, which have strong anti-microbial activi-
ties. The work-up procedure of synthesis is very sim-
ple, the yield is high and the pyridinium compounds
are easily purified. They are inexpensive and soluble
in water.
ence. Satisfactory elemental analyses were obtained, C,
0.31; H, 0.27 and N, 0.23. Melting points (m.p.)
were determined by using an electrothermal digital-melt-
ing-point apparatus model JA 9100.
5.1.1. General procedure for alkoxymethylation of
nicotinamide and isonicotinamide
5.1.1.1. Pathway A
To a solution of nicotinamide or isonicotinamide (16
mmol, 1.95 g) in dry DMF (30 mL) was added the
corresponding chloromethyl alkyl ether in equimolar
amount. The mixture was stirred at room temperature
(r.t.) for 1 h and dry ethyl acetate (30 mL) was added.
The solid substrate was filtered and then recrystallised
from EtOH.
5. Experimental protocols
5.1.1.2. Pathway B
5.1. Chemistry
Nicotinamide (20 mmol, 2.44 g) and an equimolar
amount of the appropriate chloromethyl alkyl ether was
mixed by stirring at r.t. for 1 h. The product was
purified by extraction with hexane and crystallised from
EtOH.
Chloromethyl alkyl ether was prepared by passing
HCl-gas through a mixture of formaldehyde and alco-
hol. The percentage of obtained ether in a crude product
was determined by an alkalimetric method, 1 g crude
product was added to 10 mL acetone at −40 °C. HCl-
substrate was quickly neutralised with 1% KOH in
EtOH and 3 mL water was added. The mixture was
stirred at 45 °C for 10 min. HCl as a product of
hydrolysis of chloromethyl alkyl ether was neutralised
with 2% KOH in EtOH. The crude product contained
96–92% chloromethyl alkyl ether.
Attention, during extraction and crystallisation the
temperature of the mixture must be lower than 60 °C.
5.1.1.3. Chloride (1g)
1H-NMR (DMSO-d6) l ppm=9.67 (s, 1H), 9.30 (d,
J=6.3 Hz, 1H), 9.15 (d, J=8.5 Hz, 1H), 8.92 (s, 1H),
8.37 (t, J=6.7 Hz, 1H), 8.27 (s, 1H), 6.03 (s, 2H), 3.63
(t, J=6.5 Hz, 2H), 1.52 (m, 2H), 1.23 (m, 10H), 0.84 (t,
J=6.5 Hz, 3H). 13C-NMR l ppm=162.6, 145.1, 144.8,
143.4, 133.6, 127.9, 88.8, 70.4, 31.1, 28.6, 25.2, 22.0,
13.9.
NMR spectra were recorded on a Varian model XL
1
300 spectrometer at 300 MHz for H and 75 MHz for
13C at 20 °C with tetramethylsilane as internal refer-

J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
905
5.1.1.4. Chloride (6i)
ATCC 9341, Enterococcus faecalis ATCC 29212,
Moraxella catarrhalis ATCC 25238, bacillus Bacillus
subtilis ATCC 6633 and yeast-like fungi; Candida albi-
cans ATCC 10231 and Rhodotorula rubra (Demml 1889,
Lodder 1934). Standard strains were supplied by Na-
tional Collection of Type Cultures (NCTC), London
and American Type Culture Collection (ATCC). The R.
rubra was obtained from the collection of the Depart-
ment of Pharmaceutical Bacteriology, K. Marcinkowski
University of Medical Sciences, Poznan´, Poland.
MIC was determined by the tube dilution method. A
series of pyridinium salts dilutions were prepared on
Mu¨ller–Hinton broth medium (bacteria) or Sabouraud
broth medium (fungi). The samples were incubated at
37 °C (bacteria) or at 26 °C (fungi) and after 24 h the
results were recorded. Growth of the microorganisms
was determined visually. The lowest concentration at
which there was no visible growth (turbidity) was taken
as the MIC.
1H-NMR (DMSO-d6) l ppm=9.27 (d, J=6.9 Hz,
2H), 8.54 (d, J=6.6 Hz, 2H), 6.06 (s, 2H), 3.71 (t,
J=6.4 Hz, 2H), 1.66 (m, 2H), 1.27 (m, 16 H), 0.89 (t,
J=6.7 Hz, 3H). 13C-NMR l ppm=165.6, 151.5, 145.0,
127.3, 90.7, 72.8, 33.1, 30.7, 30.6, 30.52, 30.5, 30.4, 26.9,
23.8, 14.6.
5.1.2. Preparation of N- and O-alkoxymethylated
product (5)
To a solution of N-(hydroxymethyl)-3-pyridinecar-
boxamide (20 mmol, 3 g) in dry DMF (30 ml) was
added the appropriate chloromethyl alkyl ether (50
mmol) with stirring and the mixture was stirred at r.t.
for 1 h. Then dry ethyl acetate (30 mL) was added and
stirred for a further 1 h. The solid product was filtered
off, washed with hexane and recrystallised from EtOH.
5.1.2.1. Chloride (5h)
1H-NMR (CDCl3) l ppm=10.67 (t, J=6.2 Hz, 1H),
10.43 (s, 1H), 9.35 (d, J=8.2 Hz, 1H), 9.32 (d, J=7.4
Hz, 1H), 8.21 (t, J=6.3 Hz, 1H), 6.22 (s, 2H), 4.98 (d,
J=6.3 Hz, 2H), 4.82 (s, 2H), 3.69 (t, J=6.6 Hz, 2H),
3.57 (t, J=6.7 Hz, 2H), 1.59 (m, 4H), 1.19 (m, 36H),
0.88 (t, J=6.2 Hz, 6H). 13C-NMR l ppm=161.9,
146.5, 143.7, 143.5, 134.2, 127.7, 93.5, 89.8, 72.3, 68.4,
67.2, 31.8, 31.79, 29.6, 29.53, 29.5, 29.4, 29.23, 29.18,
16.1, 25.7, 22.6, 14.0.
The procedure of the preliminary screening against
the virulent strain M. tuberculosis ATCC 27294 was
reported earlier [25].
5.3. Chromatographic results
5.3.1. Apparatus
The chromatographic system consisted of a Varian
9000 liquid chromatograph equipped with a UV–vis
variable wavelength detector and a gradient pump
(Varian Associates, Palo Alto, CA, USA) was used with
Valco valve injector (10 mL loop). The HPLC system
was operated under the STAR WORKSTATION 4.0
program.
5.1.3. General procedure for the preparation of salts (3,
4)
A warm solution of the appropriate inorganic salt (10
mmol) in water (10 mL) was added to a rapidly stirred
solution of 1-alkoxymethyl-3-carbamoylpyridinium
chloride (10 mmol) in water (10 mL). After 3 h, the
crystal product was filtered and was recrystallised from
5.3.2. Columns
Supelcosil LC18 column from Supelco (Bellefonte,
PA, USA), 5 cm×4.6 mm ID, particle diameter 5 m, was
used for chlorides 1 and 6; Supelcosil LC8DB column
from Supelco, 5 cm×4.6 mm ID, particle diameter 5 m,
was used for chlorides 5. The solvents used were of
HPLC grade from J.T. Baker (Deventer, Holland).
EtOH. Salts with the larger-size anions as CoCl42−
,
CuCl24− and FeCl14− are a colourless solid. The chemical
purity of the prepared pyridinium chlorides with long
alkoxymethyl chain (7–13 carbon atoms) was deter-
mined by a direct two-phase back titration (EN ISO
2871-2).
5.3.3. Chromatographic conditions
The chromatographic experiments were carried out
isocraticly at ambient temperature using a flow rate of 2
mL min−1 and UV detection at 220 nm. Methanol–0.05
M potassium phosphate buffer pH 7.4 different mixtures
were used as mobile phases to determine extrapolated
capacity factors for the chlorides 1 and 6. Methanol–
water different mixtures were used as mobile phases to
5.2. Anti-microbial activity
Microorganisms used: eleven standard strains repre-
sentative of rods; Pseudomonas aeruginosa NCTC 6749,
Escherichia coli ATCC 25922, Proteus vulgaris NCTC
4635, Klebsiella pneumoniae ATCC 33495, cocci;
Staphylococcus aureus NCTC 4163, Micrococcus luteus

906
J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
determine extrapolated capacity factors for the chlorides
5.
Clavemont, CA, USA). Experimental parameter of
lipophilicity (log kw) were determined chromatographi-
cally [26, 27]. Respective data are collected in table VI.
There is evident intercorrelation between the two
lipophilicity parameters considered. To derive QSAR
equations a multiple regression procedure was applied
within the Statgraphic Plus software (Manugistic,
Rockville MI, USA). The dependent variables were the
log(1/MIC) data determined in vitro against 11 micro-
bial strains. As the regressors the log P and its square
and log kw and its square were used. Quadratic QSAR
equations derived (tables IV and V) were highly signifi-
cant statistically as evidenced by significance levels (P<
0.001). As a matter of fact correlation coefficients (r)
were not that high but unquestionably excluded chance
correlations.
5.3.4. Methods
Dead volumes of the columns were determined with
0.001 M aqueous potassium nitrate as a non-retained
compound.
Capacity factors (k) obtained from at least three
(usually four) different chromatographic conditions (dif-
ferent mobile phases) were used to calculate lipophilicity
indexes for investigated compounds.
5.4. Methods of QSAR
Theoretical lipophilicity parameters (log P) for the
pyridinium chlorides were calculated by the Hansch–
Leo program (ClogP for Windows, BioByte Corp.,
Acknowledgements
Table VI. Theoretical (log P) and experimental (log kw)
parameter of lipophilicity for the pyridinium chlorides (1, 5,
6).
This investigation received financial support from the
Polish Committee of Scientific Research Grant KBN
No. 3 T09B 010 15. Anti-mycobacterial data were pro-
vided by the Tuberculosis Antimicrobial Acquisition &
Coordinating Facility (TAACS) through a research and
development contract with the US National Institute of
Allergy and Infections Diseases.
Chlorides
log kw
log P
1a
1b
1c
1d
1e
1f
–
−1.803
−1.274
−0.745
−0.216
0.313
0.842
1.371
1.900
2.429
2.958
3.487
−3.386
0.672
1.30
2.788
3.846
4.904
5.962
7.020
8.078
0.71
1.26
1.58
2.08
2.78
3.44
4.0
3.70
3.96
4.32
–
1g
1h
1i
References
[1] Fisicaro E., Pelizzetti E., Barbieri M., Thermochim. Acta 168
1j
(1990) 143–159.
1k
5a
5b
5c
5d
5e
5f
[2] Fisicaro E., Pelizzetti E., Lanfredi E., Savarino P., Progr. Coll.
Polym. Sci. 84 (1991) 474–482.
–
[3] Fisicaro E., Barbieri M., Pelizzetti E., Savarino P., J. Chem.
Soc. Faraday Trans. 87 (1991) 2983–2987.
[4] Kourai H., Takechi H., Horie T., Uchiwa N., Takeichi K.,
Shibasaki I., J. Antibact. Antifung. Agents 13 (1985) 3–10.
2.90
3.46
4.07
4.32
4.55
4.37
4.94
5.50
1.02
1.36
1.81
2.48
2.89
3.58
4.20
3.74
4.13
4.54
[5] Kourai H., Takechi H., Horie T., Takeichi K., Shibasaki I., J.
5g
5h
5i
Antibact. Antifung. Agents 13 (1985) 245–253.
[6] Maeda T., Manabe Y., Yamamoto M., Yoshida M., Okazaki
K., Nagamune H., Kourai H., Chem. Pharm. Bull. 47 (1999)
1020–1023.
5j
9.136
6a
6b
6c
6d
6e
6f
−1.274
−0.745
−0.216
0.313
0.842
1.371
1.900
2.429
2.958
3.487
[7] Gu¨ndel W.H., Z. Naturforsch. B 28 (1973) 471–474.
[8] Bernardi L., de Castiglione R., Scarponi U., J. Chem. Soc.
Chem. Commun. 9 (1975) 320–321.
[9] Pernak J., Szymanowski J., Pujanek M., Kucharski S., J. Am.
Oil Chem. Soc. 56 (1979) 830–833.
[10] Pernak J., Michalak L., Krysin´ski J., Pharmazie 49 (1994)
532–534.
6g
6h
6i
[11] We˛glewski J., Pernak J., Krysin´ski J., J. Pharm. Sci. 80 (1991)
91–95.
6j
[12] Pernak J., Indian J. Heterocy. Chem. 7 (1997) 97–100.

J. Pernak et al. / European Journal of Medicinal Chemistry 36 (2001) 899–907
907
[13] Pernak J., Rogoz
;
a J., Mirska I., Eur. J. Med. Chem. 36 (2001)
[21] Kourai H., Kume M., Takeichi K., Ohtsuru Y., Ohizumi M.,
313–320.
Shibasaki I., J. Antibact. Antifung. Agents 14 (1986) 435–442.
[14] Deady L.W., Finlayson W.L., Korytsky O.L., Aust. J. Chem.
[22] Kaliszan R., Quant. Struct. Act. Relat. 9 (1990) 83–87.
32 (1979) 1735–1742.
[23] Altomare C., Cellamare S., Carotti A., Ferappi M., Quant.
[15] Schung J.C., Viers J.W., Seeman J.I., J. Org. Chem. 48 (1983)
Struct. Act. Relat. 12 (1993) 261–268.
4892–4899.
[24] Tuberculosis Antimicrobial Acquisition & Coordinating Facil-
ity, website (www.taacf.org).
[16] Kondo Y., Ogasa M., Kusabayashi S., J. Chem. Soc. Perkin
Trans. II (1984) 2093–2097.
[25] Sanna P., Carta A., Rahbar Nikookar M.E., Eur. J. Med.
[17] Kondo Y., Uematsu R., Nakamura Y., Kusabayashi S., J.
Chem. 35 (2000) 535–543.
Chem. Soc. Perkin Trans. II (1988) 1219–1224.
[26] Kaliszan R., TRAC-Trend Anal. Chem. 18 (1999) 400–
[18] Pernak J., Walerowicz W., Pol. J. Chem. 55 (1981) 1109–1114.
[19] Pernak J., Pol. J. Chem. 59 (1985) 439–446.
410.
[27] Kaliszan R., in: Valko K. (Ed.), Separation Methods in Drug
Synthesis and Purification, Elsevier, Amsterdam, 2000, pp.
503–534.
[20] Kourai H., Takechi H., Kume M., Takeichi K., Shibasaki I., J.
Antibact. Antifung. Agents 14 (1986) 55–63.