N,N-Dichlorotaurine: Chemical and Bactericidal Properties

Abstract

The biogenous antimicrobial agent N-chlorotaurine (NCT) converts by disproportionation to N,N-dichlorotaurine (NDCT) at a rate proportional to acidity. This occurs at appreciable amounts already in weakly acidic biological systems. To understand the consequences of NDCT formation, a thorough investigation of this undescribed compound was mandatory, which needed its synthesis. Differently from NCT, this was possible in the aqueous system using trichloroisocyanuric acid. While the free acid, Cl(2)HNCH(2)CH(2)SO(3)H, was not available in pure form, its sodium and potassium salts were analytically pure and showed melting points (decomposition) of 125-128 degrees C (potassium) and 162-164 degrees C (sodium). The sodium salt demonstrated unexpected long-term stability even at room temperature (8.4 % loss of activity within 4 months). The aqueous solutions of both salts exhibited a weak acid reaction, and they were less stable than NCT. With regard to chlorination of amines (transhalogenation), NDCT was, surprisingly, less efficacious than NCT, which manifested itself by a lack of reactivity at pH < 7, for which a mechanistic explanation is given. Compared on a molar scale, NDCT was more bactericidal than NCT against the gram-negative bacteria E. coli, P. aeruginosa and P. mirabilis, while there was no difference concerning the gram-positive ones, S. aureus and S. epidermidis. The increase of bactericidal activity at acidic pH was the same as observed with NCT and is attributed to a higher susceptibility of bacteria in this environment. Taken together, NDCT seems not to be suited to substitute NCT as a preparation fit for medical practice.

Full text

473
Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483 DOI 10.1002/ardp.200500146
N,N-Dichlorotaurine: Chemical and Bactericidal
Properties
Waldemar Gottardi, Magdalena Hagleitner, Markus Nagl
Department of Hygiene, Microbiology and Social Medicine, Division of Hygiene and Medical Microbio-
logy, Innsbruck Medical University, Innsbruck, Austria
The biogenous antimicrobial agent N-chlorotaurine (NCT) converts by disproportionation to N,N-
dichlorotaurine (NDCT) at a rate proportional to acidity. This occurs at appreciable amounts already
in weakly acidic biological systems. To understand the consequences of NDCT formation, a thorough
investigation of this undescribed compound was mandatory, which needed its synthesis. Differently
from NCT, this was possible in the aqueous system using trichloroisocyanuric acid. While the free
acid, Cl
2
HNCH
2
CH
2
SO
3
H, was not available in pure form, its sodium and potassium salts were ana-
lytically pure and showed melting points (decomposition) of 125⫺128°C (potassium) and 162⫺164 °C
(sodium). The sodium salt demonstrated unexpected long-term stability even at room temperature
(8.4 % loss of activity within 4 months). The aqueous solutions of both salts exhibited a weak acid
reaction, and they were less stable than NCT. With regard to chlorination of amines (transhalogen-
ation), NDCT was, surprisingly, less efficacious than NCT, which manifested itself by a lack of reactiv-
ity at pH < 7, for which a mechanistic explanation is given. Compared on a molar scale, NDCT was
more bactericidal than NCT against the gram-negative bacteria E. coli,P. aeruginosa and P. mirabilis,
while there was no difference concerning the gram-positive ones, S. aureus and S. epidermidis. The
increase of bactericidal activity at acidic pH was the same as observed with NCT and is attributed to
a higher susceptibility of bacteria in this environment. Taken together, NDCT seems not to be suited
to substitute NCT as a preparation fit for medical practice.
Keywords:N,N-Dichloro amines; Comproportionation; Transhalogenation; Bactericidal activity
Received: May 30, 2005; Accepted: July 11, 2005
Introduction
The highly intriguing molecule N-chlorotaurine (NCT)
plays a fundamental role in the human immune system, and
its synthesized sodium salt represents a promising antiseptic
in human medicine [1⫺3]. Until now, its main identified
functions are the bactericidal and the immune-controlling
activity [4⫺6].
Already the first investigators of NCT discovered its incli-
nation to disproportionate to form N,N-dichlorotaurine
(NDCT) and taurine [7]. Because N-chlorotaurine and N,N-
dichlorotaurine are strong acids, their salts are completely
dissociated. The abbreviations NCT and NDCT, therefore,
concern the anions ClHN-CH
2
-CH
2
-SO
3
⫺
and Cl
2
N-CH
2
-
CH
2
-SO
3
⫺
, respectively, which are responsible for the reac-
tions quoted in this paper. The free acid and the solid alkali
salts are specified by NDCT-H and NDCT-Na, NDCT-K,
respectively.
Correspondence: Waldemar Gottardi, Department of Hygiene,
Microbiology and Social Medicine, Division of Hygiene and Medi-
cal Microbiology, Innsbruck Medical University, Fritz-Pregl-Str. 3,
A-6020 Innsbruck, Austria. Phone: ⫹43 512 5073430, Fax ⫹43 512
5072870, e-mail: waldemar.gottardi@uibk.ac.at
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The rate of disproportionation (Scheme 1) increases with
acidity, while the equilibrium is shifted to the right side.
Thus, in 0.01 M H
2
SO
4
, the reaction is finished very fast
(< 1 min, t
0.5
艐2 s), yielding an equimolar mixture of
NDCT and taurine. At pH 5, the half-reaction time of the
1 % (0.1%) solution comes to t
0.5
艐1.1 min (艐11 min),
while it is t
0.5
艐12 min (艐120 min) at pH 7 [8]. However,
these values apply to well-buffered solutions where dispro-
portionation is impressively accelerated (see below). In fresh
und unbuffered 1% (0.1 %) solutions of NCT an equilib-
rium at pH 8.0⫺8.2 (7.0⫺7.1) is established, and dispro-
portionation is virtually absent [8].
From these facts, it can be derived that the presence of finite
concentrations of NDCT has to be considered in NCT solu-
tions at pH <7. In biological systems, which are generally
well buffered, a weakly acidic milieu is perfectly possible,
2 ClHNCH
2
CH
2
SO
3
⫺
⫹H
⫹
↔Cl
2
NCH
2
CH
2
SO
3
⫺
⫹H
3
N
⫹
CH
2
CH
2
SO
3
⫺
(1)
2 NCT ⫹H
⫹
↔NDCT ⫹taurine
Scheme 1. Disproportionation of NCT.

474
Gottardi et al. Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483
for example at inflammation sites [9, 10]. Therefore, also in
vivo considerable concentrations of NDCT can be present.
Beside disproportionation, also other features of NCT are
dependent on the pH, but they are not understood in all
details. While the bactericidal effect of NCT increases sig-
nificantly if the pH changes from 7 to 5 [5, 11], the reverse
behavior is observed with the capability to chlorinate bac-
terial surfaces, i.e. the formation of durable chlorine covers
[12].
To clarify these observations, a thorough investigation of
NDCT was necessary, which required its synthesis. Differ-
ently from NCT, which can be isolated only as an alkali
salt, namely in an alcoholic system, the free acid NDCT-H
and its alkali salts, NDCT-Na and NDCT-K, could be iso-
lated from an aqueous solution.
Results
Chemical properties
Thermal stability
The free acid and its salts exhibited unexpected thermal sta-
bility and withstood evaporation at 90°C. The melting
points of the salts were 125⫺128 °C (potassium) and
162⫺164 °C (sodium), both under decomposition.
Stability upon storage
With a loss of 8.4% within 133 days at room temperature
(RT), the sodium salt unfolded a surprising stability (Figure
1), whereas the potassium salt lost 66% of its oxidation ca-
pacity at RT and 21 % at 4 °C within the same period.
Stored at RT, both salts produced by the time an oxidizing
Figure 1. Stability of solid NDCT-Na and NDCT-K. At RT:
(䉫) NDCT-Na, (䊊) NDCT-K; at 2⫺4°C: (䊐) NDCT-Na;
(䉭) NDCT-K. Each value represents the mean ±SD of three
to five replicates of iodometric titration; p< 0.01 between
both temperatures for NDCT-K and between NDCT-K and
NDCT-Na.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
atmosphere (chlorine smell) in the container. This is not the
case with solid NCT-Na.
Behaviour in aqueous solution
A freshly prepared 1% aqueous solution of the salts exhib-
ited a pH 艐5.0, which soon decreased gradually to pH 4.7
after 10 min, pH 3.3 after 24 h, and pH 2.1 after 17 days at
RT in the dark. During this time, also the oxidation capacity
decreased, showing a distinct temperature dependency.
While at RT the whole oxidation capacity had vanished
after 175 days, upon storage in the refrigerator (2⫺4°C) the
decrease was only ~40 %. No difference between the solu-
tions of sodium and potassium salt was observed (Figure 2).
Spectra
The UV spectra of the free acid and both salts were identical
and agree with the one acquired by disproportionation of
NCT [8]. The IR spectrum was very similar to the one of
NCT [8], except for the range of 3000⫺4000 cm
⫺1
where
the NH-band at 3260 cm
⫺1
is absent and the presence of
crystal water bands at 3570 and 3471 cm
⫺1
can be observed
(Figure 3).
Influence of buffer concentration on the formation of
NDCT by disproportionation of NCT
Besides the already known impact of pH [8], also the buffer
concentration had an effect on the rate of dispro-
portionation, as shown with citrate at pH 6 and phosphate
at pH 7 (Figure 4). Using 0.5 M buffer, the equilibrium was
reached after ~ 90 min, while in 0.01 M buffer the degree of
disproportionation came to less than half within the same
time.
Figure 2. Stability of aqueous solutions of NDCT-Na and
NDCT-K (each 55 mmol). At RT: (䉫) NDCT-Na, (䊊) NDCT-K;
at 2⫺4°C: (䊐) NDCT-Na; (䉭) NDCT-K. Each value represents
the mean ±SD of three to five replicates of iodometric titra-
tion; p< 0.01 between both temperatures, p> 0.05 between
NDCT-K and NDCT-Na.

475
Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483 Properties of N,N-Dichlorotaurine
Figure 3. IR spectrum of NDCT-Na.
Figure 4. Rate of disproportionation of NCT (0.055 M) in ci-
trate buffer (pH 6.0) and phosphate buffer (pH 7.0). Citrate:
(䉭)0.01M,(䊐) 0.5 M. Phosphate: (*) 0.01 M, (䊊) 0.5 M. Each
value represents a single UV measurement.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
Absence of hydrolysis
The weakly acidic milieu (initially pH 艐5) that both alkali
salts generate in aqueous solution points out that NDCT is
not prone to undergo hydrolysis (Scheme 2). This is in con-
trast to NCT which produces a weakly alkaline solution
[8]. The equilibrium of Scheme 2 lies therefore far on the
left side.
Equilibration of NDCT with taurine (compro-
portionation)
The position of the comproportionation equilibrium
(Scheme 3) at the final pH (which is more acidic than the
initial one) expressed by the measured NCT concentration
fitted well to values calculated with the reported equilibrium
constants of the disproportionation of NCT [8, 13] (see
Table 1).
Cl
2
NCH
2
CH
2
SO
3
⫺
⫹H
2
O↔HCl
2
N
⫹
CH
2
CH
2
SO
3
⫺
⫹OH
⫺
(2)
NDCT ⫹H
2
O↔NDCTH
⫹
⫹OH
⫺
Scheme 2. Hydrolysis of NDCT.
Cl
2
NCH
2
CH
2
SO
3
⫺
⫹H
3
N
⫹
CH
2
CH
2
SO
3
⫺
↔2 ClHNCH
2
CH
2
SO
3
⫺
⫹H
⫹
(3)
NDCT ⫹taurine ↔2 NCT ⫹H
⫹
Scheme 3. Equilibration of NDCT with taurine (comproportio-
nation).

476
Gottardi et al. Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483
Table 1. Reaction of NDCT (0.0275 M) with taurine (0.055 M)
in0.2MNa
2
HPO
4
pH %NCT
†
Time
‡
%NCT/h
Initial Final Exptl. Calc.
‡‡
Calc.
§§
h
7.6
§
6.9 50.7 53.7 70.4 1.5 34
8.1
#
7.6 71.5 71.0 83.4 2.0 36
9.2
††
8.5 92.7 87.6 93.6 3.5 27
†
% NCT refers to the maximally possible 0.055 M.
‡
For attaining the equilibrium.
§
Adjusted with H
2
SO
4.
#
pH without adjustment.
††
Adjusted with NaOH.
‡‡
K
dispr
⫽(4.5 ±0.8) ⫻10
6
[8].
§§
K
dispr
⫽(1.07 ±0.15) ⫻10
6
[25].
Contrary to disproportionation in a weakly acidic milieu
[8], the relative rate of comproportionation, expressed
as %NCT/h, remained approximately constant within the
pH range 6.9⫺8.5. As observed with disproportionation
(Figure 4), also the rate of comproportionation increased
with buffer concentration (Figure 5).
Comparison of the reaction of NDCT with different
N⫺H compounds
Under the same conditions, i.e. c(R-NH
2
)/c(NDCT) ⫽2
and 0.2 M phosphate buffer adjusted to pH 8.0, the initial
rate of equilibration (Scheme 4) measured by the decrease
of the concentration of NDCT was taurine >> α-alanine >
glycine > β-alanine >> NH
4
⫹
(Figure 6). The shape of the
graph of α-alanine points out that in this case an equilib-
rium is not approached, because of decomposition. From
experiments with Chloramine T [14], it can be derived that
the stability of the N-chloro derivatives of the investigated
Figure 5. Rate of the reaction of NDCT (0.0275 M) with
taurine (0.055 M) in 0.02 and 0.2 M phosphate buffer at
pH 8: (䉭)0.02M;(䉫) 0.2 M. Each value represents a single
UV measurement.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
Cl
2
NCH
2
CH
2
SO
3
⫺
⫹H
3
N
⫹
-R ↔ClHNCH
2
CH
2
SO
3
⫺
⫹ClHN-R ⫹H
⫹
(4)
R⫽H, CH
2
CO
2
⫺
, CH(CH
3
)CO
2
⫺
,CH
2
CH
2
CO
2
⫺
,CH
2
CH
2
SO
3
⫺
Scheme 4. Reaction of NDCT with different N⫺H com-
pounds.
amino-carbonic acids decreases in the order b-alanine > gly-
cine > α-alanine. During the monitored reaction time (60
min) the pH dropped to 7.65⫺7.50, in case of NH
4
Cl only
to 7.90. This pH effect can be attributed to the transhal-
ogenation reaction formally laid down in Scheme 4.
The same experiment conducted at pH 6.0 (0.2 M phos-
phate) showed not only a strong rate reduction but signifies
also a shift of the equilibria of Eq. 4 (Scheme 4) towards
the left side. Here, again, a clear difference between NH
4
Cl
and taurine is evident (Figure 6).
Figure 6. Rate of the reaction of NDCT (0.0275 M) with
NH
4+
,β-alanine, glycine, α-alanine and taurine. (䉭)NH
4+
,
(䊊)β-alanine, (*) glycine, (䉫)α-alanine, and (䊐) taurine (each
0.055 M) in 0.2 M phosphate buffer at pH 8 and pH 6. Each
value represents one single UV measurement.

477
Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483 Properties of N,N-Dichlorotaurine
Figure 7. Bactericidal activity of each 0.055 M NCT and NDCT at pH 7. (䉭) Control, (䊐) NCT, (䊊) NDCT. Each value represents
the mean ±SD of three to five replicates; p<0.01 between NCT and NDCT for E. coli,P. aruginosa,and
P. mirabilis;p> 0.05 for staphylococci.
Sulfur-containing amino acids like cysteine, N-acetyl-cyst-
eine and methionine, however, reacted very fast with NDCT,
both at pH 4.1 and pH 8.9, under complete loss of oxi-
dation capacity.
Bactericidal activity
Since 55 mM NCT (1.0 %) proved to be well tolerated by
human tissue, bactericidal within minutes and therapeuti-
cally effective in previous studies [2, 11, 15⫺18], this con-
centration was chosen primarily for our experiments. The
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
results of a comparison between 55 mM NDCT and NCT
were not consistent in that the investigated gram-positive
strains S. aureus and S. epidermidis were equally resistant
to both compounds, while NDCT was significantly more
bactericidal against the gram-negative strains E. coli,
P. aeruginosa and P. mirabilis. This effect was the same at
both pH 7 and pH 5 (Figs. 7, 8). A significant increase in
the killing rate by lowering the pH from 7 to 5 is already
known for NCT [5, 11]. As can be seen in Figures 7 and 8,
this applied also to NDCT. For the same reduction in CFU/
mL, nearly a tenfold incubation time at pH 7 was needed

478
Gottardi et al. Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483
Figure 8. Bactericidal activity of each 0.055 M NCT and NDCT at pH 5. (䉭) Control, (䊐) NCT, (䊊) NDCT. Each value represents
the mean ±SD of three to five replicates; p< 0.01 between NCT and NDCT for E. coli,P. aruginosa,andP. mirabilis;p> 0.05
for staphylococci.
compared to pH 5. A similar pH effect was also verified
with a very low concentration of 30 µMat37°C (Figure 9),
conditions which occur at biological sites [1, 19].
Discussion
Synthesis of NCT and NDCT
Although NCT was discovered more than 30 years ago [7],
all attempts to isolate the compound from an aqueous sys-
tem were unsuccessful, a failure which roots in its incli-
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
nation to disproportionate to NDCT. In case of NDCT,
however, disproportionation turned out to be favorable for
the complete conversion into the dichloro derivative, which
therefore could be easily isolated.
Stability
The gradual acidification and simultaneous loss of oxi-
dation capacity gives evidence for a mechanism of de-
composition in aqueous solution. It can be attributed to
intramolecular redox and elimination reactions, for which
the following scheme can be established.

479
Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483 Properties of N,N-Dichlorotaurine
Figure 9. Bactericidal activity of 30 µM NDCT at pH 7 and pH 5. (*) Control pH 7, (䉭) control pH 5, (䊊)pH7,(䊐) pH 5. Each
value represents the mean ±SD of three to five replicates; p<0.01 between pH 7 and pH 5 (p< 0.05 for S. aureus 25923).
As a first step, an N-chloroaldimine is formed (Scheme 5)
which releases a second molecule of HCl or undergoes hy-
drolysis (Eqs. 7, 8), which finally leads to a nitril (Eq. 6)
or an aldehyde (Eqs. 8, 9). Both compounds were already
identified in the context of NCT decomposition [8, 20].
The influence of buffers on the formation of NDCT
by disproportionation of NCT
Disproportionation of NCT represents a specific form of
chlorination, namely an auto-chlorination. The results pre-
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
R-CH
2
-NCl
2
씮R-CH⫽NCl ⫹HCl (5)
R-CH⫽NCl 씮R-C⬅N⫹HCl (6)
R-CH⫽NCl ⫹H
2
O↔R-CH⫽NH ⫹HOCl (7)
R-CH⫽NCl ⫹H
2
O씮R-CH⫽O⫹NH
2
Cl (8)
R-CH⫽NH ⫹H
2
O씮R-CH⫽O⫹NH
3
(9)
Scheme 5. Formation of N-chloroaldimine.
sented here (Figure 4) uncover, in addition to the influence
of pH, a major effect of the buffer concentration upon the
transhalogenation rate. They comply with kinetic studies

480
Gottardi et al. Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483
R-NHCl ⫹H
2
PO
4
⫺
↔R-NH
2
Cl
⫹
⫹HPO
42⫺
(10)
Scheme 6. Formation of the actual reactive species NCTH
+
.
R-NH
3
⫹
⫹HPO
42⫺
↔R-NH
2
⫹H
2
PO
4
⫺
(11)
Scheme 7. Equilibration of buffer with the amine.
gained in the presence of methoxyacetic acid/methoxyacet-
ate buffers [13]. A similar boosting effect has been observed
in a study on bacterial chlorine covers [12] and was ex-
plained by the proton-donating properties of buffers which
favor the formation of the actual reactive species NCTH
⫹
even at pH 7 [8] according to Scheme 6.
However, also an equilibration of buffer with the amine can
be considered, which would increase the concentration of
the non-protonated amine representing the actual nucleo-
philic agent (Scheme 7).
Figure 5 discloses that also in the case of the inverse reac-
tion, i.e. comproportionation, buffer concentration has a
crucial influence on the reaction rate. Since formation of the
protonated species (R-Cl
2
H
⫹
, see below) is less probable, a
substantial contribution of the equilibrium (Eq. 11) seems
plausible.
General readiness for transhalogenation of NCT
and NDCT
The term transhalogenation was defined for the transfer (ex-
change) of positive halogen between amine compounds.
Contrary to the substitution of a C-H bond, it occurs rather
fast already at RT and needs no catalyst, and it is not con-
nected with a loss of oxidation capacity [8]. The differing
behavior of NCT and NDCT can best be exposed by the
relation between disproportionation (Scheme 1) and com-
proportionation (Scheme 3). Both imply a transhalogen-
ation reaction whose equilibrium is highly pH dependent.
A simplified approach could be: disproportionation of NCT
occurs readily at pH <7, and comproportionation of
NDCT and taurine at pH >7.
Our results demonstrate that this concept is not limited to
the equilibria between taurine, NCT, and NDCT, but apply
generally to equilibria between unsubstituted amines,
mono-chloro amines, and/or dichloro amines (Scheme 8):
R-NHCl ⫹R⬘-NH
2
↔R-NH
2
⫹R⬘-NHCl
R-NHCl ⫹R⬘-NHCl ↔R-NH
2
⫹R⬘-NCl
2
Scheme 8. Equilibration between unsubstituted amines,
mono-chloro amines, and/or dichloro amines.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
Anyway, the readiness of NDCT for transhalogenation is
actually given at pH >7. The confirmation of a reaction of
NDCT with taurine at pH 7.1 needed a 400-fold surplus of
taurine, at least [13]. However, in the presence of a substan-
tial buffer concentration, the reaction could be proved even
at pH 6 (Figure 6).
Reactivity of NDCT against defined NH compounds
The observed rates at pH 8 (Figure 6) reveal the order taur-
ine >> α-alanine > glycine > β-alanine >> NH
4
⫹
, which is
rather bewildering. Since the reaction shown in Eq. 4
(Scheme 4) proceeds via a nucleophilic substitution (see
scheme below), the unprotonated amines act as nucleophilic
agents, and their pK
a
values should provide an indication
of their relative concentration at a given pH. However, the
pK
a
values of 9.19, 9.246, 9.91, 10.01 and 10.24 for taurine,
NH
4
⫹
, glycine, α-alanine, and β-alanine, respectively [21],
disclose that this concept does not coincide with the reac-
tion rates, most notably of taurine and NH
4
⫹
(Figure 6).
The strikingly low reactivity of NH
4
⫹
can be ascribed to
the absence of an inductive effect. In case of attached alkyl
groups, the C-H functions act as electron donors and in-
crease the nucleophilicity of the amine function. However,
this approach conflicts to a certain degree with the rates of
glycine, α- and β-alanine.
The lower reactivity at pH 6, on the other hand, can easily
be explained with the actual concentration of the non-pro-
tonated amine compounds, which is significantly lower than
at pH 8. From the pK
a
values it can be derived that, e.g. for
taurine (β-alanine), the concentration of the free amine
form is by a factor of 94.0 (99.5) higher at pH 8 compared
to pH 6.
General reactivity of NDCT compared to other
active chlorine compounds
The ability to produce chlorine covers on protein matrices
allowed the ranking of active chlorine compounds accord-
ing to their chlorinating potency regarding transhalogen-
ation. In case of chlorine covers on human skin, the order
HOCl/OCl
⫺
> Chloramine T > NH
2
Cl > NCT was found
[22], and in case of bacterial covers, it was DCI-Na >
Chloramine T > NCT > NDCT [12]. Combining both rank-
ings evidences the very last position of NDCT. This is rather
surprising because it possesses two chlorine atoms on the
same nitrogen, suggesting a higher reactivity in terms of
transfer of positive chlorine. The lesser potency of NDCT
[12] can be attributed to mechanistic effects. As pointed out
in [8], the halogen transfer between NCT and an amine
compound implies an attack of the free electron pair of the
amino group of the acceptor molecule R
1
-NH
2
towards the
chlorine atom of the protonated donor molecule R
2
-
NH
2
Cl
⫹
according to a nucleophilic substitution:

481
Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483 Properties of N,N-Dichlorotaurine
Scheme 9. Halogen transfer between NCT and an amine
compound.
In case of NDCT, it seems plausible that two chlorine atoms
at the same nitrogen cause a positive partial charge high
enough to impede the intermediary formation of the pro-
tonated donor molecule R-NCl
2
H
⫹
. This could explain why
NDCT, contrary to NCT, is not suited in such a rate for a
chlorine transfer and why comparatively weak chlorine co-
vers are formed [12]. For the same reason, also hydrolysis
according to Eq. 2 (Scheme 2) does not take place.
In spite of these considerations a transhalogenation is con-
ceivable also with the non-protonated NDCT molecule,
where the second chlorine atom induces a positive partial
charge at the chlorine reacting with the nucleophilic agent.
However, this partial charge is less than at the protonated
NCT molecule which bears a real charge. NDCT, therefore,
reacts markedly only in a weakly alkaline milieu where the
portion of the free amines is high. Among the N-chloro
species of taurine the following order in terms of electro-
philicity seems plausible: R-NClH
2
⫹
> R-NCl
2
> R-NHCl
resp. NCTH
⫹
> NDCT > NCT.
Bactericidal activity
Regarding the experiments conducted with the same molar
concentration of 55 mM (Figs. 7, 8), one would expect a
stronger activity of NDCT because its oxidation capacity is
twice as much as that of NCT. This was in fact observed
both at pH 5 and 7; however, only with gram-negative bac-
teria. The lower sensitivity of gram-positive bacteria to
NDCT cannot be explained until now. The throughout
higher bactericidal activity of NDCT at pH 5 (Figs. 6, 7)
complies with the behavior of NCT [5, 11], but also with
other active chlorine compounds like hypochlorite [23] and
Chloramine T [24]. For the latter compounds, a pH shift
from 7 to 5 implies a substantial change in the equilibrium
concentrations of the active species: in hypochlorite solu-
tions, e.g., an increase of the ratio [HOCl]/[OCl
⫺
] from 3.4
to 344. Accordingly, the higher bactericidal effect of hypo-
chlorite at pH 5 was explained [23]. Since the molecule
NDCT does not undergo any change this approach does
not work. An influence of pH upon the susceptibility of
bacteria towards chlorinating agents is conceivable. This
applies also to NCT whose increase of killing activity ⫺
based on the results presented here ⫺cannot be explained
by disproportionation.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
A possible explanation could rely on a change of electric
charges that occurs if the pH changes from 7 to 5. Since
bacterial membranes are bearing negative charges, the ap-
proach of the anions NCT and NDCT is hampered in a
neutral environment. If the solution becomes acidified, an
increase of non-ionized molecules and a decrease of nega-
tive charges on the bacterial surface take place, resulting in
a better access of the chlorinating agents. Concerning the
mechanism of bacterial kill, it can be stated that transhal-
ogenation is not the main cause. If this was true, NDCT
with its significantly minor potency to form chlorine covers
[12] would be less active than NCT, which was not observed.
The main cause for bacterial kill might be found in the oxi-
dation of thiols which occurs in the whole pH range with
a high rate and which causes irreversible mutations in the
protein structure.
The consequence of NDCT formation by dispro-
portionation at biological sites
Three features can be specified:
(1) Based on the fact that two molecules of NCT yield one
of NDCT, our results suggest rather a decrease of bacteri-
cidal activity of NCT caused by disproportionation, at least
in the case of gram-positive bacteria.
(2) Because NDCT produces weaker chlorine covers on bac-
teria than NCT [12], disproportionation might also induce
a weakening of the post-antibiotic effect, which is beneficial
in overcoming infections by the immune system [5, 15].
(3) Another drawback might originate in the minor stability
of NDCT in solution (Figure 2), as NDCT, in contrast to
NCT, decomposes under acidification. This favors the rate
of disproportionation even more.
In summary, it can be postulated that appreciable concen-
trations of NDCT will usually be unfavorable for the per-
formance of NCT. That is why conditions that promote dis-
proportionation, e.g. pH <7, should be avoided when apply-
ing NCT as an antiseptic in human medicine. For the same
reasons, NDCT-Na, despite of its surprising stability, seems
not suitable as a substitute for NCT-Na.
Acknowledgments
This study was supported by the Austrian Science Fund
(grant P15240-MED). We thank Prof. J. Schantl, Institute
for Organic Chemistry (Innsbruck Medical University,
Innsbruck, Austria) for valuable discussions and Eva-Maria
Lemberger for excellent technical assistance.

482
Gottardi et al. Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483
Experimental
Chemical experiments
Reagents
Trichloroisocyanuric acid (TCI) was purchased from Fluka AG and
contained 44.3 % Cl
⫹
(assayed iodometrically), which corresponds
to M
r
(eff) ⫽240.0. All other chemicals were from Merck and of
the highest available purity.
Analytic methods
Oxidation capacity was assayed by iodometric titration keeping the
sequence acidification (50% acetic acid) of the analyte solution be-
fore addition of potassium iodide. The reverse order resulted in a
value ~30% too low for NDCT, while it had no influence on the
assessment of NCT. The origin of this effect is not quite clear. Since
the reaction of an N-chloro compound with iodide is connected
with an initial strong alkalization (>N-Cl ⫹2I
⫺
⫹H
2
O씮>N-H
⫹I
2
⫹OH
⫺
) an intramolecular redox reaction is conceivable (see
Scheme 5, Eqs. 5, 6) that is connected with a loss of oxidation
capacity.
Preparation of an aqueous solution of NDCT-H
To a stirred ice-cooled solution of 1.25 g taurine (0.01 mol) in
100 mL water was gradually added 1.6 g trichloroisocyanuric acid
(0.02 mol Cl
⫹
). After 0.5 h of stirring at 1⫺2°C, the precipitated
cyanuric acid was removed by filtration. On the basis of the UV
spectrum (A
302.4
⫽0.360, A
251.5
⫽0.254) and the known molar
absorption coefficients [8], the resulting strongly acidic solution
contained 0.109 M NDCT-H. Iodometric titration yielded
0.207 ±0.002 M Cl
⫹
, which also indicates a quantitative reaction.
Isolation of free acid
The aqueous solution containing NDCT-H (see above) was
vacuum-evaporated (water-jet vacuum pump), whereby the tem-
perature of the water bath was increased to 90°C at the end. The
resulting pungently smelling, micro-crystalline material-containing
oil was suspended in 15 mL 1,2-dichloroethane and once more
vacuum-evaporated after filtration. The remaining hygroscopic clear
yellowish oil (1.7 g; calculated 1.94 g) was free of cyanuric acid and
represented for the most part the free acid NDCT-H. By iodometric
titration, we found 32.4 ±0.3 % Cl
⫹
(calculated 36.54 % Cl
⫹
). The
UV spectrum showed the NCl
2
band at 302.4 nm. It was neither
possible to purify (i.e. to remove water) the very aggressive, i.e.
strongly acidic and highly oxidizing, substance nor to record an
IR spectrum.
General procedure for preparing alkali salts by neutralization of
NDCT-H
To the ice-cooled aqueous solution of NDCT-H was added the cal-
culated amount of NaOH or KOH up to pH 艐5. The resulting
solution was then vacuum-evaporated as described above. The
crystalline residue containing the alkali salt was purified by re-crys-
tallization, resulting in 1 g of the potassium salt from 13 mL boiling
methanol and 1 g of the sodium salt from 15 mL boiling ethanol/
methanol (4 :1). NDCT-Na: found 31.77 ±0.30 % Cl
⫹
(N ⫽3);
calculated: 32.82 % Cl
⫹
(semi-hydrate: 31.51 % Cl
⫹
). The sodium
salt contained crystal water (bands at 3565 and 3470 cm
⫺1
in the
IR spectrum) which disappeared in the vacuum over P
2
O
5
. NDCT-
K: found 30.00 ±0.072% Cl
⫹
(N ⫽3); calculated: 30.55% Cl
⫹
.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
Monitoring the reactions of NDCT
Equilibria between NCT, NDCT and taurine (disproportionation
and comproportionation) were monitored photometrically using the
absorption maxima of NCT (λ
max
⫽250.9 nm, ε⫽397.4 L mol
⫺1
cm
⫺1
) and NDCT (λ
max
⫽302.4 nm, ε⫽332.9 L mol
⫺1
cm
⫺1
)as
set forth in [8]. In case of reactions with NH
4
⫹
/NH
3
(ammonium
chloride), glycine, α- and β-alanine, the absorption at 302.4 nm was
used, which allowed to assess the temporal decrease of NDCT.
Microbiological experiments
Bacteria
Bacterial strains (Staphylococcus aureus ATCC 25923, Staphylo-
coccus epidermidis ATCC 12228, Proteus mirabilis ATCC 14153,
Escherichia coli ATCC 11229, Pseudomonas aeruginosa ATCC
27853 deep-frozen for storage were grown overnight on tryptic soy
agar (Merck, Darmstadt, Germany). Colonies from this agar were
grown in tryptic soy broth (Merck, Darmstadt, Germany) at 37°C
overnight and twice washed with saline.
Killing tests
Bacteria were diluted in buffered NDCT (NCT) solution to
0.6 ⫻10
7
⫺3.5 ⫻10
7
CFU/mL. Immediately and subsequent to
different incubation times at RT, aliquots were removed and NDCT
(NCT) was inactivated with sodium thiosulfate. Undiluted aliquots
(50 µL) as well as 100-fold dilutions in saline (50 µL) were spread
onto tryptic soy agar plates with an automatic spiral plater (Don
Whitley Scientific Limited, West Yorkshire, UK) in duplicates, al-
lowing a detection limit of 10 CFU/mL. Plates were grown at 37°C,
and CFU were counted after 24 and 48 h. Controls without NCT
were treated in the same way.
Statistics
One-way analysis of variance (ANOVA) and Dunnett’s Multiple
Comparison test (GraphPad Software Inc., CA, USA) was applied;
pvalues <0.05 were considered significant.
References
[1] M. B. Grisham, M. M. Jefferson, D. F. Melton, E. L. Thomas,
J. Biol. Chem. 1984,259, 10404⫺10413.
[2] M. Nagl, W. Gottardi, Hyg. Med. 1996,21, 597⫺605.
[3] S. J. Weiss, N. Engl. J. Med. 1989,320, 365⫺376.
[4] J. Marcinkiewicz, Immunol. Today 1997,18, 577⫺580.
[5] M. Nagl, M. Hess, K. Pfaller, P. Hengster, W. Gottardi, Anti-
microb. Agents Chemother. 2000,44, 2507⫺2513.
[6] E. Park, J. Jia, M. R. Quinn, G. Schuller-Levis, Clin. Immunol.
2002,102, 179⫺184.
[7] J. M. Zgliczynski, T. Stelmaszynska, J. Domanski, W. Ostrow-
ski, Biochim. Biophys. Acta 1971,235, 419⫺424.
[8] W. Gottardi, M. Nagl, Arch. Pharm. Pharm. Med. Chem. 2002,
335, 411⫺421.
[9] R. W. Light, Clin. Chest. Med. 1985,6,55⫺62.
[10] H. P. Simmen, J. Blaser, Am. J. Surg. 1993,166,24⫺27.
[11] M. Nagl, W. Gottardi, Hyg. Med. 1992,17, 431⫺439.
[12] W. Gottardi, M. Nagl, J. Antimicrob. Chemother. 2005,55,
475⫺482.
[13] F. H. Kayser, Medizinische Mikrobiologie (Eds.: F. H. Kayser,
K. A. Bienz, J. Eckert, R. M. Zinkernagel), Thieme, Stuttgart
⫺New York, 1998, chapter 1.

483
Arch. Pharm. Chem. Life Sci. 2005, 338, 473−483 Properties of N,N-Dichlorotaurine
[14] W. Gottardi, V. Bock, Fourth Conference on Progress in Chemi-
cal Disinfection 1988,35⫺60.
[15] M. Nagl, P. Hengster, E. Semenitz, W. Gottardi, J. Antimicrob.
Chemother. 1999,43, 805⫺809.
[16] M. Nagl, V. A. Nguyen, W. Gottardi, H. Ulmer, R. Höpfl, Br.
J. Dermatol. 2003,149, 590⫺597.
[17] A. Neher, M. Nagl, E. Appenroth, M. Gstottner, M.
Wischatta, F. Reisigl, M. Schindler, H. Ulmer, K. Stephan,
Laryngoscope 2004,114, 850⫺854.
[18] B. Teuchner, M. Nagl, A. Schidlbauer, H. Ishiko, E. Dragosits,
H. Ulmer, K. Aoki et al.,J. Ocular Pharmacol. Ther. 2005,
21, 157⫺165.
[19] E. L. Thomas, Infect. Immun. 1979,25, 110⫺116.
2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.archpharm.de
[20] C. Cunningham, K. F. Tipton, H. B. Dixon, Biochem. J. 1998,
330, 939⫺945.
[21] A. Albert, E. P. Serjeant in The Determination of Ionization
Constants, Chapman and Hall Ltd., London, 1971, chapter 3.
[22] W. Gottardi, A. Karl, Zent. bl. Hyg. Umweltmed. 1991,191,
478⫺493.
[23] G. R. Dychdala in Disinfection, Sterilization and Preservation
(Ed.: S. S. Block), Lippincott Williams Wilkins, Philadelphia,
2001, chapter 7.
[24] G. R. Weber, U. S. Public Health Reports 1950,65, 503⫺512.
[25] J. M. Antelo, F. Arce, P. Calvo, J. Crugeiras, A. Rios, J. Chem.
Soc. Perkin Trans. 2 2000,2, 2109⫺2114.