Improved synthesis of three brominated analogues of the potent environmental mutagen 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX)

Article abstract of DOI:10.1016/S0040-4020(00)00258-1

The synthesis of three brominated analogues of the environmental mutagen 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX), namely, 3-chloro- 4-(bromochloromethyl)-5-hydroxy-2(5H)-furanone (BMX-1), 3-chloro-4- (dibromomethyl)-5-hydroxy-2(5H)-furanone (BMX-2) and 3-bromo-4- (dibromomethyl)-5-hydroxy-2(5H)-furanone (BMX-3), by employing a simple procedure from a common precursor, is described. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00258-1

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 3391±3397
Improved Synthesis of Three Brominated Analogues of the
Potent Environmental Mutagen
3-Chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX)
Maia Lloveras,a Isabel Ramos,a Elies Molinsb and Angel Messeguera,
*
aDepartment of Biological Organic Chemistry, IIQAB (CSIC), J. Girona, 18, 08034 Barcelona, Spain
Á
Institut de Ciencia de Materials de Barcelona (CSIC), Campus de la UAB, E. 08193 Bellaterra, Spain
b
Received 20 January 2000; accepted 29 March 2000
AbstractÐThe synthesis of three brominated analogues of the environmental mutagen 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-
furanone (MX), namely, 3-chloro-4-(bromochloromethyl)-5-hydroxy-2(5H)-furanone (BMX-1), 3-chloro-4-(dibromomethyl)-5-hydroxy-
2(5H)-furanone (BMX-2) and 3-bromo-4-(dibromomethyl)-5-hydroxy-2(5H)-furanone (BMX-3), by employing a simple procedure from
a common precursor, is described. q 2000 Elsevier Science Ltd. All rights reserved.
The occurrence of halogenated furanones in drinking water
is a matter of concern due to the potential toxicity of these
compounds. The most important representative of this
family of toxins is 3-chloro-4-(dichloromethyl)-5-hydroxy-
2(5H)-furanone (MX, Scheme 1). This compound is a
highly potent direct-acting mutagen in the Salmonella
typhimurium assay.1±3 The data available indicate that
MX is originated from the reactions of chlorine with the
humic substances not eliminated during the puri®cation of
water.4,5 It has been shown that MX induces DNA damage
in mammalian cells in vitro6±8 and in vivo,9,10 and that the
administration of this mutagen through drinking water
caused cancer effects in rats.11 More recently, adducts
from the reaction of MX with nucleosides and calf thymus
DNA have been identi®ed.12,13 These ®ndings, together with
the epidemiological studies showing a link between cancer
in humans and consumption of chlorinated drinking water,
suggest that MX should be considered as a health risk
factor.14,15
the generation of hypobromous acid or bromine by the
reaction of chlorine with bromide ion. Horth et al. described
the formation of three bromohydroxyfuranones, namely
BMX-1, BMX-2 and BMX-3 (Scheme 1) as result of the
chlorination of water containing bromide ions; furthermore,
mutagenic activities were reported for these compounds
although their rigorous chemical characterization was not
given.4,16 Suzuki et al. reported the occurrence of BMX-1,
BMX-2 and BMX-3, together with MX, in four different
samples of Japanese chlorinated drinking water.17
Finally, LaLonde and coworkers have recently published
the synthesis and mutagenicity in the Ames Salmonella
typhimurium TA-100 assay of two bromohydroxyfuranones,
namely BMX-2 and BMX-3.18 This work prompted us to
report our results on these MX analogues. In the present
contribution the synthesis of bromohydroxyfuranones
BMX-1, BMX-2 and BMX-3 is reported. For BMX-1,
this is the ®rst report on its synthesis and chemical
characterization. The preparation of these bromohydroxy-
furanones was accomplished by using a simple synthetic
pathway starting from methyl 3-methylbut-2-enoate as
common precursor.
The chlorine disinfection of drinking water with high
bromide content could also lead to the formation of bromi-
nated hydroxyfuranones. These compounds may arise from
Scheme 1. Structure of MX and of its brominated analogues BMX-1, BMX-2 and BMX-3.
Keywords: mutagen; synthesis; bromofuranones; polybrominated compounds.
* Corresponding author. Tel.: 139-93-4006121; fax: 134-93-2045904; e-mail: ampqob@iiqab.csic.es
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00258-1

3392
M. Lloveras et al. / Tetrahedron 56 (2000) 3391±3397
Results and Discussion
with base led to the bromo ester 2 in 77% yield. This ester
was treated with N-bromosuccinimide under irradiation to
give the pentabromo derivative 3 in 40% yield. Whereas the
®rst three bromine atoms were introduced in six hours in
almost quantitative yields (GC and GC/MS monitoring),
prolonged treatments with large excesses of the brominating
agent were required to accomplish the introduction of the
fourth one, and even then the conversion was not complete.
LaLonde and coworkers described the preparation of
BMX-2 and BMX-3 by using 4-(hydroxymethyl)-2(5H)-
furanone as common precursor for both bromohydroxy-
furanones.18 However, the synthetic procedures involved a
high number of steps and some of them proceeded in
moderate or low yields, which limit the availability of
these mutagens for toxicological studies. We designed an
alternative procedure which allows the preparation of all
three bromohydroxyfuranones in higher yields and in a
simple and shorter manner. This procedure resembles that
followed by Padmapriya et al. for the synthesis of MX.
These authors introduced the chlorine atoms into non-cyclic
intermediates and the cyclisation to the hydroxyfuranone
was performed in the last step.19 By contrast, the strategy
of LaLonde et al. consisted in the formation of the furanone
ring ®rst followed by its subsequent chlorination and
hydroxylation at the appropriate positions. Assays carried
out for the preparation of MX had indicated that the former
strategy was more convenient than that by the group of
LaLonde. Thus, in our procedure the readily available
methyl 3-methylbut-2-enoate (1) was used as common
precursor. As shown in Scheme 2, one route led to the
formation of BMX-1 and BMX-2 while the second afforded
BMX-3.
Interestingly, the 1H NMR spectrum of compound 3 showed
broad peaks at 7.21 and 6.79 ppm which were attributed to
the hydrogen atoms cis and trans with respect to the ester
moiety, respectively. In particular, the peak at 7.21 ppm was
the broadest. The 13C NMR spectrum of 3 showed two broad
peaks at 34.3 and 34.9 ppm assigned to both dibromomethyl
carbon atoms. These observations suggested the possibility
of a restricted rotation around the C-3/C-4 and C-3/C-5
bonds of this derivative.
To con®rm this hypothesis, the peaks shapes in the 1H NMR
spectrum of 3 were monitored at different temperatures in
deuterochloroform and in deuterated dimethyl sulfoxide
solutions (Fig. 1). The results obtained in deuterochloroform
indicated that the rotation was severely restricted at low
temperatures. Upon heating the rotation was possible but
slow enough to give very broad peaks, especially for H-5,
which suggests a more restricted rotation of the dibromo-
methyl moiety linked to this hydrogen atom. As expected,
these peaks became narrow by increasing the temperature.
While the shape of H-4 showed a similar pattern in the
spectra recorded at the same temperatures in both solvents,
the peak of H-5 was better resolved in deuterated dimethyl
sulfoxide. This suggests that the solvent plays a role in the
rotation of the corresponding dibromethyl group.
Synthesis of BMX-3
Our initial attempts were directed towards the allylic bromi-
nation of ester 1. This reaction gave the corresponding tetra-
bromo derivative in 13% yield. However, all efforts to
brominate or even chlorinate the conjugated double bond
of this intermediate were unsuccessful and only products
resulting from the rearrangement of the double bond to
the b,g-position were detected. This result suggested that
steric hindrance could be operating when both methyl
groups in 1 were dibrominated.20
This presumption was supported by the X-ray diffraction
analysis of this pentabromo derivative. As shown in Fig.
2, a hydrogen bond between H-5 and the carbonyl oxygen
Ê
of the ester moiety (bond distances: H5±O1 2.18(6) A; C5±
Ê
O1 2.831(10) A; bond angle 123(4)8) is present, which
Then we investigated the inverse sequence, i.e. ®rst bromi-
nation of the double bond which was followed by the intro-
duction of the allylic bromo substituents. Thus, the
bromination of ester 1 and subsequent dehydrobromination
explains the higher restriction of rotation observed for this
group in comparison with that of the other dibromomethyl
moiety.
Scheme 2. Synthetic route for the preparation of BMX-1, BMX-2 and BMX-3. Reaction conditions: (i) Br2, CCl4; (ii) Et3N; (iii) 4 equiv. NBS; (iv) 48% HBr;
(v) Cl2, CCl4; (vi) 3 equiv. NBS, hv; (vii) 70% CH3SO3H.

M. Lloveras et al. / Tetrahedron 56 (2000) 3391±3397
3393
1
Figure 1. Pro®le within the 6.5±7.8 ppm region of the H NMR spectra (300 MHz) of ester 3 at different temperatures in: (a) CDCl3; and (b) d6-DMSO
solutions.
Synthesis of BMX-1 and BMX-2
The synthesis of these mutagens was carried out as shown in
Scheme 2. The chlorination of ester 1 was problematic due
to the formation of products with different degree of chlori-
nation. The optimised procedure involved the slow addition
of a saturated solution of chlorine in carbon tetrachloride to
the ester 1. Even in this case, as shown in Scheme 3, the
addition of the ®rst equivalent of chlorine led to the forma-
tion of the chloro ester 8 (approx. 30% yield), while the
addition of a second equivalent of chlorine afforded a
mixture of products in which the dichloro ester 9 (30%)
and the diastereomeric mixture of trichloro esters 10
(16%) were the major components.
In view of the complexity and anticipated instability of this
reaction mixture, it was decided to perform the subsequent
dehydrochlorination without any further puri®cation. Thus,
Figure 2. X-Ray structure of the pentabromo ester 3 with atom labelling.
treatment of the above crude reaction mixture with base
afforded a 1.8:1 mixture of the unsaturated dichloro esters
4 and 5 in 55% yield.22 An attempt to purify this mixture by
When the hydrolysis and subsequent cyclisation of 3 was
attempted using the basic conditions as reported for MX, i.e.
lithium hydroxide in tetrahydrofuran/water followed by
treatment with potassium bicarbonate, BMX-3 was formed
in yields lower than 5%. Conversely, when this cyclisation
was performed in 48% hydrobromic acid,21 the expected
brominated mutagen was isolated in 48% yield, which
corresponds with 15% overall yield.
distillation led to the isolation of pure compound 5;
unfortunately, most of the dichloro ester 4 decomposed. In
addition, an E,Z mixture of the rearranged dichloro esters 11
(Scheme 4), which was not detected in the crude reaction
mixture, was also isolated.
When pure dichloro ester 5 was treated with an excess of
Scheme 3. Treatment of acrylate 1 with chlorine.

3394
M. Lloveras et al. / Tetrahedron 56 (2000) 3391±3397
overall yield obtained for BMX-2 (2%) was slightly higher
than that reported by the group of Lalonde (1.6%), and this
difference was considerably greater for BMX-3 in terms of
both yields (15% versus 0.7%) and simplicity. Moreover,
this procedure permitted the preparation and rigorous
chemical characterization of BMX-1 for the ®rst time. In
our case, the availability of these bromohydroxyfuranone
standards, especially of BMX-1, allows the study of their
cytotoxicity.24 This study revealed that they are potent direct
mutagens in the Ames test and that their eventual generation
in drinking water must be adequately monitored. In this
context, the reactivity of BMX-1, BMX-2 and BMX-3
with DNA model molecules was investigated and will be
reported elsewhere.
Scheme 4. Major by-products isolated in the distillation of the crude reac-
tion mixture containing dichloro esters 4 and 5.
N-bromosuccinimide under irradiation, a 1:1.3 mixture of
the tribromo esters 6 and 7 was formed. This result indicated
that an extensive isomerisation occurred during the bromi-
nation.23 This fact together with the problems encountered
in the puri®cation of the mixture of dichloro esters 4 and 5
led to the idea to brominate the crude reaction mixture
containing these dichloro esters. Accordingly, this mixture
was subjected to the allylic bromination under the same
conditions described above to render the tribromo esters 6
and 7 in 18% overall yield. Similarly to the allylic bromina-
tion during the synthesis of BMX-3, the introduction of the
third bromine atom was slow and dif®cult because of the
steric hindrance exerted by the two bromo substituents
already present and the chlorine atom attached to the double
bond. All efforts to separate both stereoisomers were
unsuccessful.
Experimental
Caution: MX and its brominated analogues (BMXs) are
hazardous due to their mutagenicity. Therefore, the use of
gloves, well ventilated fume cupboards and careful
destruction of residues with NaOCl is recommended for
the manipulation of these compounds.
Melting points were determined with a Kof¯er apparatus
and are uncorrected. The IR spectra were registered with a
MB model 120 Bomen apparatus and absorptions are given
in cm21. The 1H (300 MHz) and 13C (75 MHz) NMR
spectra were recorded with a Varian Unity 300 spectro-
meter. Unless stated otherwise, spectra were taken in
previously neutralized CDCl3 solution. Chemical shifts are
given in ppm related to tetramethylsilane for 1H and
deuterochloroform for 13C as internal standards. The
HPLC analyses were performed with a modular system
formed by two Waters 510 pumps and an Applied Bio-
systems detector-gradient controller (1000S Diode Array
Detector). The preparative puri®cations were performed
by employing a 30£0.78 cm ODS, 10 mm column, from
Tracer, S.A. Barcelona, Spain. The high resolution mass
spectra (HRMS) were obtained with a VG Autospec Q
instrument at the `Servei d'Espectrometria de Masses del
CID'. The GC/MS/EI spectra (70 eV) were obtained using a
Fisons model MD 800 mass spectrometer coupled to a
Fisons GC 8000 apparatus, which was equiped with a
25 m HP-5 capillary column; the halogen isotope pattern
observed and the postulated fragmentation are given in
parentheses. Elemental microanalyses were carried out at
1
The H NMR analysis of this mixture revealed features
similar to those observed for the pentabromo derivative 3.
Especially, the proton linked to the bromochloromethyl
moiety in 7 appeared as a broad peak. The peak width at
half height was 33.1 Hz at 21.58C in CDCl3 and this value
was reduced to 12.6 and 9.4 Hz at 30 and 408C, respectively.
This effect was considerably less pronounced for the
proton linked to the dibromomethyl moiety in 6 (3.75 and
2.35 Hz at 21.5 and 408C, respectively). These results
suggest that in 7 a strong hydrogen bond interaction is
present between the hydrogen atom linked to the bromo-
chloromethyl moiety and the carbonyl oxygen atom. For
both tribromo derivatives however, in contrast to 3, the
methine proton trans with respect to the ester moeity does
not exhibit this band broadening, which suggests that the
steric hindrance in these compounds is considerably
reduced.
Hydrolysis and subsequent cyclisation of these inter-
mediates in 70% methanesulphonic acid afforded the
mixture of BMX-1 and BMX-2, which was separated by
preparative reversed phase HPLC to give the pure
compounds in 25 and 20% yields, respectively. The use of
hydrobromic acid to promote the cyclisation was precluded
in this case since undesired halogen replacements were
observed for both BMX-1 and BMX-2. Although BMX-1
should be present as a diastereomer mixture, only one
peak was observed in the HPLC using different elution
conditions. In constrast, the derivatization of BMX-1
with bis(trimethylsilyl)tri¯uoroacetamide showed two
peaks in GC with mass spectra compatible with two
TMS esters. This derivatisation protocol may be of value
for the identi®cation of these diastereomers in water
samples.
Á
the `Servei de Microanalisi' of the CID by using a 1108
Carlo Erba analyser.
Unless stated otherwise, organic solutions obtained from the
treatment of crude reaction mixtures were dried over
MgSO4. Puri®cations by ¯ash chromatography were
performed by using 40±60 mm silicagel.
Synthesis of BMX-3
Methyl 2-bromo-3-methylbut-2-enoate (2). Bromine
(1.8 mL, 35.8 mmol) was added dropwise to a solution of
ester 1 (4.09 g, 35.8 mmol) in CCl4 (40 mL) and the mixture
was stirred for 1 h at 258C. The residue obtained after
removal of the solvent was redissolved in CH2Cl2
(30 mL). Then, a solution of Et3N (5.3 mL, 38.3 mmol) in
In summary, our procedure allows the preparation of all
three bromohydroxyfuranones in suf®cient amounts and
purities for future toxicological assays. In addition, the

M. Lloveras et al. / Tetrahedron 56 (2000) 3391±3397
3395
CH2Cl2 (6 mL) was added dropwise to the above solution,
Synthesis of BMX-1 and BMX-2
maintained at 08C. When the addition was completed, the
crude reaction mixture was stirred for 16 h at 258C (GC
monitoring) and ®ltered. The ®ltrate was washed with
0.3 N HCl, water, brine and dried. The residue obtained
from the elimination of solvents (5.9 g) was puri®ed by
¯ash chromatography on silicagel eluting with hexane±
EtOAc mixtures to give pure acrylate 2 as a pale yellow
Methyl (E,Z)-2,4-dichloro-3-methylbut-2-enoate (4 and
5). A satd. soln. of chlorine in CCl4 was added dropwise
and at room temperature to a solution of ester 1 (5 g,
44 mmol) in the same solvent (50 mL) until the reaction
was completed (GC monitoring) (72 mL, approx.
2.4 molar equivalents of chlorine were required). The
crude reaction mixture was washed with NaHCO3 satd.
solution, water, brine and dried. The elimination of the
solvent under vaccum rendered a pale yellow oil residue.
Triethylamine (3.1 mL, 22 mmol) was added dropwise to a
solution of this residue in CH2Cl2, maintained at 0±58C, and
the mixture was stirred for 1 h at this temperature and for
20 h at 208C (GC monitoring). The crude reaction mixture
was washed with 0.5 N HCl, water, brine and dried. The
dark brown residue obtained from the elimination of
solvents was puri®ed by ®ltration through silicagel eluting
with hexane±ethyl acetate (98:2) to give a 1.8:1 mixture of
dichloro esters 4 and 5 (5.8 g, 76% purity, 55% overall
oil (5.50 g, 77% yield). 2:25 IR (®lm): 1716; H NMR d:
1
3.79 (s, 3 H), 2.14 (s, 3 H, Z-CH3), 2.05 (s, 3 H, E-CH3); 13
C
NMR d: 164.5 (C-1), 149.1 (C-3), 108.2 (C-2), 52.7 (CH3),
27.2 (E-CH3), 23.2 (Z-CH3); MS (EI): m/z, 192, 194 (M, Br);
177, 179 (M2Me, Br); 161, 163 (M2OMe, Br); 160, 162
(M2[OMe1H], Br); 133, 135 (M2[OMe1CO], Br); 132,
134 (M2[OMe1CO1H], Br); 53 (base peak,
M2[OMe1CO1Br1H]).
Methyl 2,4,4-tribromo-3-(dibromomethyl)but-2-enoate
(3). A soln. of the bromo ester 2 (3.42 g, 17.7 mmol) in
CCl4 (150 mL) was treated with NBS (13.2 g, 74.2 mmol)
and the mixture was stirred while irradiating with a 300 W
mercury lamp for 15 h (GC monitoring). The crude reaction
mixture was ®ltered, the ®ltrate was washed with CCl4,
concentrated and treated with additional NBS (4.09 g,
23 mmol) and irradiation for 3 more days. The crude
reaction mixture was ®ltered and the precipitate was washed
thoroughly with CCl4. The residue obtained from the
elimination of the solvent was puri®ed by ¯ash chromato-
graphy on silicagel eluting with 30:1 hexane:EtOAc solvent
mixture, to give compound 3 as a solid (3.45 g, 40% yield).
3: mp103±1048C (hexane:CHCl3); IR (KCl): 1710, 734; 1H
NMR d: 7.21 (br, 1 H, Z-CHBr2), 6.79 (br, 1 H, E-CHBr2),
3.94 (s, 3 H, CH3); 13C NMR d: 163.0 (C-1), 142.5 (C-3),
120.4 (C-2), 54.2 (CH3), 34.9 (E-CHBr2), 34.3 (Z-CHBr2);
MS (EI): m/z, 473, 475, 477, 479, 481, 483 (Br5, M-OCH3),
445, 447, 449, 451, 453, 455 (Br5, M-CO2CH3), 425,
427, 429, 431, 433 (Br4, M-Br), 397, 399,401, 403, 405
(Br4), 50 (base peak). Elemental analysis for C6H5Br5O2:
C, 14.12; H, 0.98; Br, 78.56. Found: C, 13.95; H, 0.91,
Br, 78.30.
1
yield). IR (®lm): 1726; H NMR d: 4.57 (s, 2 H, CH2Cl,
4), 4.30 (s, 2 H, CH2Cl, 5), 3.85 (s, 6 H, CO2Me, 4 and 5),
2.25 (s, 3 H, CH3, 5), 2.15 (s, 3H, CH3, 4); 13C NMR d:
163.3 (C-1, 5), 162.9 (C-1, 4), 144.4 (C-3, 4), 143.1 (C-3, 5),
123.2 (C-2, 4), 121.3 (C-2, 5), 53.0 (CO2CH3, 4), 52.9
(CO2CH3, 5), 45.9 (CH2Cl, 5), 44.1 (CH2Cl, 4), 20.7
(CH3, 4), 19.1(CH3, 5); MS (EI): m/z, 182, 184, 186 (M1,
Cl2), 147, 149 (M2Cl), 146, 148 (M2HCl), 87, 89 (base
peak). Elemental analysis for C6H8Cl2O2. Calcd: C, 39.37;
H, 4.41; Cl, 38.74. Found: C, 39.44; H, 4.41; Cl, 39.16.
HRMS: calcd, 181.9901. Found, 181.9903.
Methyl (E,Z)-4-bromo-3-dibromomethyl-2,4-dichloro-
but-2-enoate (6 and 7). A solution of the dichloro esters
4 and 5 (1.9 g, 10 mmol) in CCl4 (70 mL) was treated with
N-bromosuccinimide (6.5 g, 37 mmol) and the mixture was
irradiated with a 300 W mercury lamp, under stirring, for
110 h divided in three different periods (GC monitoring). At
the end of the ®rst period (30 h), the crude reaction mixture
was ®ltered and the precipitate was washed with CCl4. The
residue obtained after the elimination of the solvent was
puri®ed by ¯ash chromatography (95:5 hexane:EtOAc) to
give a fraction which was dissolved in CCl4 (30 mL) and
irradiated in the presence of NBS (2.2 g, 12 mmol) for 30 h.
The whole process was repeated with a new addition of NBS
(1.2 g, 7 mmol) and a ®nal irradiation period of 50 h. The
crude reaction mixture was ®ltered and the precipitate was
washed thoroughly with CCl4. The residue obtained from
the elimination of the solvent was puri®ed by ®ltration
through silicagel eluting with hexane±EtOAc (from 99:1
to 95:5) to give the expected tribromo derivatives which
were isolated after crystallization from CHCl3±hexane as
a 1:1.3 mixture of 6 and 7, respectively (0.79 g, 18%
yield). All attempts to separate this mixture by further
fractional crystallization were unsuccessful. Mp 78±808C
Cyclisation of the pentabromo ester 3. A suspension of 3
(0.51 g, 1 mmol) in 48% HBr (5 mL) was heated for 8 h at
gentle re¯ux (TLC monitoring). The crude reaction mixture
was diluted with water and extracted with EtOAc. The
organic fraction was washed with brine and dried. The
brownish residue obtained after the elimination of the
solvent (0.33 g) was puri®ed by ¯ash chromatography on
silicagel, eluting with a hexane:EtOAc solvent mixture to
give an enriched sample of BMX-3. A ®nal puri®cation by
cristallyzation (CHCl3) afforded the pure bromohydroxy-
furanone as a colorless solid (170 mg, 48% yield).
BMX-3:18 mp 71±738C (CHCl3); IR (KBr): 1797, 1780;
1H NMR d: 6.47 (s, 1 H, CHBr2), 6.42 (br, 1H, CHOH),
4.97 (s, 1 H, OH); 13C NMR d: 164.9 (C-1), 155.7 (C-3),
114.1 (C-2), 98.6 (C-5), 26.4 (C-6). Elemental analysis for
C5H3Br3O3: C, 17.11, H, 0.85, Br, 68.35. Found: C, 17.12,
H, 0.76, Br, 68.25. Silylation of BMX-3 using bis(trimethyl-
silyl)tri¯uoroacetamide led to the TMS ester. MS (EI): m/z,
419, 421, 423, 425 (Br3, M-1), 405, 407, 409, 411 (Br3,
M2Me), 223, 225, 227 (Br2, base peak, M2[Br1CHO1O-
SiMe3]), 195, 197, 199 (Br2, M2[Br1CHO1OSi-
Me31CO]).
1
(hexane±CHCl3); IR (KBr): 1716; H NMR (CDCl3), d:
7.57 (br, 1 H, CHBrCl, 7), 7.51 (br, 1 H, CHBr2, 6), 6.81
(s, 1 H, CHBrCl, 6), 6.78 (s, 1 H, CHBr2, 7), 3.94 (s, 6 H,
CH3, 6 and 7); 13C NMR (CDCl3, 258C), d: 162.1 (C-1),
142.5 (C-3), 128.5 (C-2), 54.1 (CH3), 51.0 (CHBrCl), 50.8
(CHBrCl), 34.0 (CHBr2), 30.8 (CHBr2); MS (EI): m/z, 385,
387, 389, 391, 393 (Cl2Br3, M2CH3O), 380, 382, 384, 386,
388 (ClBr3, M2HCl), 337, 339, 341, 343 (Cl2Br2, M2Br),

3396
M. Lloveras et al. / Tetrahedron 56 (2000) 3391±3397
258, 260, 262 (base peak, Cl2Br, M22 Br). Elemental
analysis for C6H6Br3Cl2O2. Calcd: C, 17.16; H, 1.19; Br,
57.12; Cl, 16.98. Found: C, 17.26; H, 1.08; Br, 57.15; Cl,
17.00.
Methyl 2,4,4-tribromo-3-(dibromomethyl)but-2-enoate (3).
empirical formula: C6H5Br5O2, Mˆ508.65, triclinic, space
Ê
Ê
Ê
group P-1, aˆ6.236 A, bˆ8.480 A, cˆ11.502 A,
Ê 3
aˆ93.038, bˆ101.738, gˆ94.258, Vˆ592.5 A , Zˆ2,
Dcˆ2.851 g/cm3, mˆ16.925 mm21
,
3793 re¯ections
Cyclization of the mixture of halo esters 6 and 7. A
suspension of this mixture of esters (0.38 g, 0.9 mmol) in
70% methanesulfonic acid (30 mL) was heated for 6 h at
1408C (TLC monitoring). The crude reaction mixture was
cooled, poured into water and extracted with ethyl acetate.
The organic fraction was washed with brine and dried. The
residue obtained after the elimination of the solvent
(0.230 g) was prepuri®ed by ¯ash chromatography on
silicagel eluting with a 70:20:10 hexane±EtOAc±AcOH
mixture, to give an enriched mixture of BMX-1 and
BMX-2. A 50 mg fraction of this mixture was redissolved
in MeOH (400 mL) and separated by preparative HPLC
(30£0.78 cm ODS-2, 10 mm column). The elution condi-
tions were 20:80 MeOH: 50 mM HCOOH±Et3N buffer
(pH 3.16) at 2.5 mL/min. The collected eluates correspond-
ing to the bromohydroxyfuranones were independently
extracted with EtOAc and the organic fractions were
washed with brine and dried. Final puri®cation of each
sample by preparative TLC on silicagel, eluting with a 1:1
hexane±EtOAc solvent mixture led to the isolation of pure
BMX-1 and BMX-2.
collected (u-rangeˆ1.81±29.988), 3453 independent re¯ec-
tions (Rintˆ0.0380), 129 parameters, R1ˆ0.0543,
wR2ˆ0.1239 (all re¯ections), residual electron density
Ê 23
1.210 and 21.049 e A
.
Acknowledgements
Financial support from CICYT-Generalitat de Catalunya
(Grant QFN 95-4717) and CICYT (Grant SAF989-0059),
Â
and fellowships from the `Direccio General de Recerca' to
two of us (I.R. and M.Ll.) are acknowledged. We thank Dr
Â
F. Sanchez-Baeza for helpful discussions on the NMR
spectra.
References
1. Tikkanen, L.; Kronberg, L. Mutat. Res. 1990, 240, 109±116.
2. Meier, J. R.; Blazak, W. F.; Knohl, R. B. Environ. Mol.
Mutagen. 1987, 10, 411±424.
3-Chloro-4-(bromochloromethyl)-5-hydroxy-2(5H)-
furanone (BMX-1). 11 mg (25%); H NMR, d: 6.63, 6.62
3. Kronberg, L.; Holmbom, B.; Reunanen, M.; Tikkanen, L.
Environ. Sci. Technol. 1988, 22, 1097±1103.
1
(s, 1 H, CHBrCl), 6.43 (br, 1 H, H-5), 4.75 (0.7 H, OH), 4.32
(0.3 H, OH); 13C NMR d: 163.8 (CO), 151.0 (C-4), 124.4
(C-3), 96.5 (CHOH), 43.73 (CHBrCl); HRMS for
C5H2BrCl2O2: calcd, 242.8615. Found: 242.8618. Silylation
of BMX-1 using bis(trimethylsilyl)tri¯uoroacetamide led to
a mixture of TMS esters in a 1:2 diastereomeric ratio (GC/
MS monitoring). MS (EI): m/z, 317, 319, 321, 323 (BrCl2,
M-15), 239, 241, 243 (Cl2, M-CHBr), 179, 181, 183
[BrCl,M2(Cl1Me3SiO1CHO)], 73, 75 [base peak, Cl,
M2(Br1Cl1Me3SiO12CO)].
4. Horth, H. J. Fr. Hydrol. 1990, 21, 135±145.
5. Hemming, J.; Holmbom, B.; Reunanen, M.; Kronberg, L.
Chemosphere 1986, 15, 549±556.
6. Harrington-Brock, K.; Doerr, C. L.; Moore, M. M. Mutat. Res.
1995, 348, 105±110.
7. Jansson, K.; Hyttinen, J. M. T. Mutat. Res. 1994, 322,
129±132.
8. Chang, L. W.; Daniel, F. B.; DeAngelo, A. B. Teratog.
Carcinog., Mutagen. 1991, 11, 103±114.
9. Furihata, C.; Yamashita, M.; Kinae, N.; Matsushima, T. Water
Sci. Technol. 1992, 25, 341±345.
3-Chloro-4-(dibromomethyl)-5-hydroxy-2(5H)-fura-
10. Jansson, K.; Maki-Paakkanen, J.; Vaittinen, S. L.; Vartiainen,
T.; Komulainen, H.; Tuomisto, J. Mutat. Res. 1993, 299, 25±28.
11. Komulainen, H.; Kosma, V. M.; Vaittinen, S. L.; Vartianen,
1
none (BMX-2).18 18 mg (20%); H NMR, d: 6.49 (s, 1 H,
CHBr2), 6.43 (br, 1 H, H-5), 4.41 (s, 1 H, OH); 13C NMR, d:
163.6 (CO), 151.6 (C-4), 123.8 (C-3), 97.1 (CHOH), 24.80
(CHBr2); HRMS for C5H2Br2ClO2: calcd, 286.811001.
Found: 286.810829. Silylation of BMX-2 using bis-
(trimethylsilyl)tri¯uoroacetamide led to the TMS ester.
MS (EI): m/z, 375, 377, 379, 381 (ClBr2, M21), 361,
363, 365, 367 (ClBr2, M2Me), 223, 225, 227 (Br2,
M2[OSiMe31CHO1Cl]), 179, 181, 183 (ClBr,
M2[OSiMe31CHO1Br]), 151, 153, 155 (ClBr,
M2[OSiMe31CHO1Br1CO]), 73, 75 (Cl, base peak,
M2[2Br1OSiMe312CO]).
È È
T.; Kaliste-Korhonen, E.; Lotjonen, S.; Tuominen, R. K.;
Tuomisto, J. J. Natl. Cancer Inst. 1997, 89, 848±856.
12. Le Curieux, F.; Munter, T.; Kronberg, L. Chem. Res. Toxicol.
1997, 10, 1180±1185.
È
13. Munter, T.; Le Curieux, F.; Sjoholm, R.; Kronberg, L. Chem.
Res. Toxicol. 1998, 11, 226±233.
14. Melnick J. Natl. Cancer Inst. 1997, 89, 832±833.
15. Koivusalo, M.; Jaakkola, J. K.; Vartianinen, T.; Hakulinen, T.;
Karjalainen, S.; Pukkala, E.; Tuomisto, J. Am. J. Public Health
1994, 84, 1223±1228.
16. Horth, H.; Fielding, M.; James, H. A.; Thomas, M. J.; Gibson,
T.; Wilcox, P. Water Chlorination 1990, 6, 107±124.
17. Suzuki, N.; Nakanishi, J. Chemosphere 1995, 30, 1557±1564.
18. LaLonde, R. T.; Bu, L.; Henwood, A.; Fiumano, J.; Zhang, L.
Chem. Res. Toxicol. 1997, 10, 1427±1436.
Single crystal X-ray diffraction analysis
All data were collected using an Enraf-Nonius CAD4
Ê
diffractometer with MoKa radiation (lˆ0.71073 A) and
19. Padmapriya, A. A.; Just, G.; Lewis, N. G.; Can J. Chem. 1985,
63, 828±832.
were processed with the program WINGX.26 The structures
were solved by direct methods using the shelx97 program27
and re®ned by full-matrix least squares methods on F 2 using
the shelx97 program.28
20. For previous reports on the reaction of bromine with crowded
ole®ns, see Bellucci, G.; Bianchini, R.; Chiappe, C.; Lenoir, D.;
Attar, A. J. Am. Chem. Soc. 1995, 117, 6243±6248.

M. Lloveras et al. / Tetrahedron 56 (2000) 3391±3397
3397
21. Martin, R.; Chapleo, C. B.; Svanholt, K. L.; Dreidig, A. S.
Helv. Chim. Acta 1976, 59, 2724±2727.
24. Ramos, I.; Lloveras, M.; Solans, X.; Huici, A.; Messeguer, A.
Submitted for publication.
22. It is worth noting that a similar sequence (halogenation/
dehydrohalogenation), but performed with bromine as in the
synthesis of BMX-3, did not lead to the formation of the
corresponding unsaturated dibromoesters.
25. Owen, L. N. J. Chem. Soc. 1949, 236±239.
26. WingxÐA Windows Program for Crystal Structure Analysis.
Farrugia, L. J. University of Glasgow, Glasgow, 1998.
27. Sheldrick, G. M. Acta Crystallogr. A 1990, 46, 467±473.
28. Sheldrick, G. M. Shelx198; A computer program for the
re®nement of crystal structures, University of Gottingen, 1997.
23. For previous reports on isomerisations involved in allylic
È
brominations, see Loffer, A.; Norris, F.; Taub, W.; Svanholt, K.
L.; Dreidig, A. S. Helv. Chim. Acta. 1970, 53, 403±417.