Chem. Res. Toxicol. 1994, 7, 779-783
779
Interactive Chlorine-by-Bromineand Hydrogen-by-HydroxylGroup Replacement Effects in 2(5H)-Furanone Mutagenicity Robert T. LaLonde*pT and Howard R. Leo* Department of Chemistry, College of Environmental Science & Forestry, State University of New York, Syracuse, New York 13210-2786, and Department of Physical Sciences, Onondaga County Community College, Syracuse, New York 13215 Received May 12, 1994@
Both bromine- and chlorine-substituted 2(5H)-furanones are produced by the chlorination of ligno-humic waters containing bromide ion. The molar mutagenicities of four bromineand chlorine-substituted 2(5H)-furanones were determined by the Salmonella typhimurium (TAlOO) assay to explore C1-by-Br and H-by-OH replacement effects on mutagenicity. Each of these two replacements was expected to enhance mutagenicity based on earlier work showing that lower LUMO energy levels and greater radical anion stability correlated with elevated TAlOO mutagenicity. The four compounds investigated were the following: 2,3-dibromo-5(reduced muhydroxy-2(5H)-furanone (mucobromic acid, MBA); 2,3-dibromo-2(5H)-furanone (mucochloric acid, MCA); and cobromic acid, RMBA); 2,3-dichloro-5-hydroxy-2(5H)-furanone 2,3-dichloro-2(5H)-furanone(reduced mucochloric acid, RMCA). Mean molar mutagenicities were found to be 5.54, 1.18, 2.92, and 0.105 revertantshmol for the four compounds in the order named. Mutagenicity enhancements resulting from C1-by-Br and H-by-OH replacements were analyzed by simple ratios of mean molar mutagenicity and by multiple regression analysis. The effect of the C1-by-Br replacement on mutagenicity amounted to a 1.9-fold enhancement in the presence of C-5 OH, but an 11-fold enhancement in the presence of C-5 H. This demonstrated that the two replacement effects were interactive. Higher mutagenicity values corresponded to lower AM1 computed LUMO energy levels and greater radical anion stabilities.
Introduction The chlorination of waters containing natural organic matter results in the formation of mutagenic, chlorinesubstituted 2(5m-furanones (Figure 1A) (1)and 4-methyl-2(5m-furanones (Figure 1B) (2-5). The latter are the more potent bacterial mutagens, and among this group, 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX,l Figure 1)is one of the most potent, with mutagenicity in the Salmonella typhimurium (TAlOO-Sg)assay ranging roughly from lo3 to lo4 revhmol (2, 3, 6). Recently however, the formation of bromine-, chlorine-, and mixed halogen-substituted 4-methyl-2(5m-furanones (Figure 1) has been reported to occur when waters containing naturally occurring organic matter and bromide ion are chlorinated (7). 3-Chloro-4-(dibromomethyl)-5-hydroxy-2(5H)-fixanone(BMX2) and 3-bromo4-(dibromomethyl)-5-hydroxy-2(5H)-furanone(BMX3) appear equally mutagenic, but slightly more so than MX (7).
This anticipated development prompts us to report our progress in an ongoing study of structure-activity relationships between halogen and hydroxyl group content +
State University of New York.
* Onondaga County Community College.
@Abstractpublished in Advance ACS Abstracts, October 15, 1994. Abbreviations: adjtd (adjusted);AM1 (Austin method one); BMX2 [3-chloro4(dibromomethyl)-5-hydroxy-2(5~furanonel; BMX3 [%bromo4-(dibromomethyl)-5-hydroxy-2(5H)-furanonel; EI-MS (electron impact mass spectrometry); LUMO (lowest unoccupied molecular orbital); MBA (mucobromic acid); MCA (mucochloric acid); MCk- (mucochloric acid radical anion); M , (molar mutagenicity); M , (mean molar mutpl mutagenicity); MX [3-chloro-4-(dichloromethyl)-2(5H)-furanonel; (multiple); rev (revertants); RMBA (reduced mucobromic acid); RMCA (reduced mucochloric acid); t (statistical distribution t ) ; TAlOO-Sg (tester strain TAlOO lacking S9 rat liver preparation).
MBA Y2 = Br,: 2 = OH
A y+j=(o
RMBA: Y2 = Br2;Z = H MCA: Y2 = C12; 2 = OH
A
2
RMCA: Y, = C1,; 2 = H
MX: W = X = Y = CI
B
H\O
BMX2: W = C1; X = Y = B r BMX3: W = X = Y = Br
Figure 1. (A) Structures of halogen-substituted 2(5H)-furanones. (B)Structures of halogen-substituted 4-methyl-2(5H)-
furanones.
and the level of mutagenicity for 2(5H)-furanones and their derivatives. Previous studies demonstrated correlation between chlorine-substituted 2(5H)-furanone TAlOO-S9 mutagenicity and computed electrophilicity indicators (8, 9). These indicators were the LUMO energy level and the stability (qd - AHf)that results from the addition of a n electron to the neutral compound. Both chlorinesubstituted 2(5H)-furanones and 4-methyl-2(5H)-furanones were included in the group of compounds. In contrast, the mutagenicity of the same group of compounds failed to correlate with another computed electrophilicity indicator: [AHFd(C-4) - AH4, the stability of the anion resulting from the attachment of a hydride ion to the C-4 position of halogen-substituted 2(5H)furanones. Similar, preliminary computations of LUMO energy levels of 2,3-dibromod-hydroxy-2(5H)-furanone(muco-
0893-228x/94/2707-0779$04.50/00 1994 American Chemical Society
780 Chem. Res. Toxicol., Vol. 7, No. 6, 1994
bromic acid, MBA) and 2,3-dibromo-2(5H)-furanone(reduced mucobromic acid, RMBA) suggested that these two would be more mutagenic than each of their chlorinesubstituted counterparts. We asked if the expected greater mutagenicity of bromine-substituted 2(5H)-furanones could be demonstrated, and if there was such a difference, might it be sufficiently large to expect mixed halogen-substituted 2(5H)-furanones, which might be discovered later, to possess varying degrees of mutagenicity depending on content and position of the halogen. Also of interest was the application of appropriate analysis of results and the methods for computing LUMO and (qd - AHf)energy levels. These questions were probed by examining four readily available bromine- and chlorine-substituted 2(5H)-furanones, which are as follows: MBA, RMBA, 2,3-dichloro-5-hydroxy-2(5H)-furanone (mucochloric acid, MCA), and 2,3-dichloro-2(5H)furanone (RMCA). Results are reported here.
Experimental Procedures Chemicals. 3,4-Dichloro-5-hydroxy-2(5H)-furanone(mucochloric acid, MCA), 3,4-dibromo-5-hydroxyy-2(5H)-furanone (mucobromic acid, MBA), and sodium borohydride were purchased from the Aldrich Chemical Co. (Milwaukee, WI). 3,4-Dichloro2(5H)-furanone (reduced mucochloric acid, RMCA) was prepared by reducing MCA a s previously described (10). 3,4-Dibromo2(5H)-furanone (reduced mucobromic acid, RMBA) was prepared from MBA by reducing it with sodium borohydride as previously described for the conversion of MCA t o RMCA (10). The properties of RMBA are: mp 85-87 "C [lit. 91 "C (1111;TLC Rf 0.87 (silica gel, CHzClz/methanol, 2:l); GC tR 15.00 min [30 m x 0.25 mm SPB-5 (film thickness 0.25 pm); initial temperature: 40 "C; final temperature: 120 "C, rate: 10 "C/min; initial hold time: 1 min; final hold time: 15 min; Nz carrier at 1.1 mumin]; lH NMR (300 MHz, CDC13) 6 4.86 (s, C-5 H) (TMS 6 0.00 ppm); 13CNMR (75.45 MHz, CDC13) 6 74.13 (t, C-5), 114.58 (s, (2-31, 143.45 (s, (3-41, 166 (s, C-2) (TMS 6 0.00 ppm); EI-MS (solid probe) mlz 240 ( C 4 H ~ ~ ~ B r z 19.09, 0 2 , M+), 242 (C4Hz79's1BrzOz, 38.98, M+), 244 (C4Hzs1Brz02, 18.27, M+), 211 (C3H79BrzO, 8.83), 213 (C3H79,81Brz0, 17.74), 215 (C3HB1Brz0,8-36], 183 (CzH79Brz,0.54),185 ( C Z H ~ ~ ~1.63),187 ~ ~ B ~ (CzHs1Br2, Z, 0.641, 161 ( C d H ~ ~ ~ B r1001, 0 2 , 163 (C4Hzs1Br0zr92.72). All compounds were a t least 97% pure a s determined by GC and lH NMR. Caution:MBA, RMBA, MCA, and RMCA have tested positive in the Ames mutagenicity assay with S. typhimurium (TAlOO) without metabolic activation (see Mutagenisis Assay). Although none of these compounds are as yet known carcinogens, caution should be exercised in their handling and disposal. Mutagenesis Assay. The histidine-requiring (his-) S. typhimurium tester strain TAlOO supplied by Dr. Bruce Ames (University of California, Berkeley) was used in the standard plate incorporation mutagenesis assay performed according to Maron and Ames (12). The assay period was 72 h. The mutagenic activities of the four compounds, added in freshly prepared (CH3)zS0solutions to the top agar, were determined without activation by r a t liver homogenate fraction S9. Three plates per dose level were prepared, and mutagenic responses were determined for each dose. Additionally, prior assays had been carried out to establish toxic dose levels in relation to the lower nontoxic dose-response levels. Values for mutagenicity a s revlpg were obtained a s the positive linear regression slopes from the linear portion of the dose-response plots. The spontaneous TAlOO mutants, observed from the (CH3)zSO controls, were taken as the zero-dose points. Each assay included positive (CH&SO, crystal violet, ampicillin, and sodium azide controls. Bacterial and (CH&DO controls were conducted in quintuplicate. Slopes in revlpg were multiplied by the molecular weight in units of pglpmol x to obtain the molar mutagenicities (M,) in units of rev/nmol for comparison with the mutagenicities of compounds whose mutagenicities had
LaLonde and
Leo
been determined pre$ously by this same method. The mean molar mutagenicity (M,)was obtained from values of M. Computation Method. Electronic properties were calculated at the AM12 level using MOPAC version 6.0 (13)running on a VAX 3100. All bond lengths, bond angles, and dihedral angles were optimized using the default BFGS method with the keyword PRECISE. A self-consistentfield was achieved for each calculation, and the Herberts test was satisfied in BFGS for the calculations seeking the lowest heats of formation. For both MBA and MCA the hydroxyl group was rotated 360" in 10-deg increments to ensure that the structures were not trapped in a localized minimum and to determine the effect on the LUMO energy.
Results, Analysis, and Discussion Comparison of Mutagenicities. Molar mutagenicities (M,)and related statistical indicators for the assays are given in Table 1. The dose-response curves obtained in Assay 1 are displayed in Figure 2 and are typical of those for the six different assays that yielded the data given in Table 1. Differences between the mean molar were statistically significant at the mutagenicities 95% confidence level by the two-tailed t test for the following four pairs: MBA and MCA (t = 3.221, RMBA and RMCA (t = 6.84), MCA and RMCA (t = 8.361, and MBA and RMBA(t = 5.67). The four compounds MBA, MCA, RMBA, and RMCA represent two classes of substituent replacements: chlorine-by-bromine (Cl-by-Br)and hydrogen-by-hydroxyl(Hby-OH). In the presence of the C-5 OH, the effect of the C1-by-Br replacement on mutagenicity is assessed simply which is 1.9. In the presence as the ratio MmmAlfimMCA, of C-5 H, the replacement effect is given by the ratio fimmBAlfimRMCA, which is 11. Or, instead, looking at the H-by-OH replacement in the presence of Br or C1, the amounts to 4.7, while the ratio ratio MmMBAIMmRMBA MmMCAIMmRMCAequals 28. The overall net effect of both C1-by-Br and OH-by-H effects mutually influencing one another is given by the ratio MmmlMmRMcA,which equals 53. An earlier report from this laboratory showed that the H-by-OH replacement amounted to a 46-fold mutagenicity enhancement for the MCA-RMCA pair (10). For another group of 2(5H)-furanones, the chlorine-substituted 4-methyl-2(5H)-furanones,the H-by-OH replacements resulted in a mutagenicity enhancement factor of about 10 (14, 15). However, in an exceptional case, the factor was of the order of 100. This was the case for 44chloromethyl)-5-hydroxy-2(5H)-furanone (15). The mutual influence of the two classes of replacements on Mm could be assessed more rigorously and directly from the eighteen M , values for all four compounds. This alternative analysis was achieved by regressing the eighteen log M , values on predictor variables Y and 2 and an interaction term (YZ)that are given in the regression model:3
(um)
log M, = constant
+ aY + j3Z + y ( Y Z ) + E
(1)
A value of 0 was assigned to the predictor variables Y The A M 1 method was chosen over the PM3 method since the latter failed to reach a plausible stationary point for the radical anions of MBA and RMBA. We are currently applying AM1 to the compounds which we investigated previously using PM3. Equation 1 is the formal expression of the regression model. This equation includes 6, the random deviation from the population regression surface. The presence of in the model accounts for observed points falling above or below this surface.
Chem. Res. Toxicol., Vol. 7, No. 6,1994 781
Halogen Replacement Effects on Mutagenicity
Table 1. A Summary of Results: The S. typhimurium (TAlOO) Mutagenicities of Four 3,4-Dihalo-2(5H)-furanones y2
substituents: Y, Z
assay
spontaneous revertants
nc
r2 d
YZ= Br2, Z = OH (MBA)
1
2 4 5 6
156 153 130 119 98
29.6 185 15.4 110 21.4 114 14.5 100 26.3 84 fimf = 5.54, SD = 1.70, n = 5
14 20 17 20 13
0.770 0.870 0.943 0.946 0.888
7.63 3.99 5.52 3.75 6.79
Y2 = Br2, Z = H (RMBA)
1 2 4 5 6
156 153 130 119 98
23 20 20 20 10
0.964 0.892 0.952 0.949 0.953
1.09 0.850 1.39 1.07 1.50
fimf=
4.50 151 3.51 152 5.74 121 4.42 108 6.22 97 1.18, SD = 0.262, n = 5
Yz = Cl2, Z = OH (MCA)
1 2 3 4 5
156 153 153 130 119
20 23 20 20 14
0.942 0.808 0.919 0.800 0.921
3.58 2.13 2.71 2.84 3.32
fimf=
21.2 167 12.6 182 16.0 169 16.8 136 19.6 122 2.92, SD = 0.563, n = 5
23 14 20
0.705 0.855 0.884
0.0659 0.119 0.129
156 130 119 a
slopea
interceptb
0.431 150 0.779 128 0.841 118 fimf= 0.105, SD = 0.0339, n = 3
Mm'
In units of rev/pg. b In units of redplate. The number of observations. The correlation coefficient squared. e In units of revhmol. molar mutagenicities for n cases.
f Mean
Table 2. Multiple Linear Regression of Log M m on the Representing the Predictor Variables Y,2,and (n), Substituent Classes (Br or C1 or OH or H) Attached to 2(5H)-Furanonea YP I
4 PI
a
c 400
t
/
I
3 4
-
variable class
coeff
std error
t (DF = 15) probability partial r2
4
PI
0
p:
300
200
Y
-
z
(YZ) const
1.06 (a) 0.0861 1.46 (p) 0.0861 -0.792 ( y ) 0.1140 -0.998
12.3 16.9 - 6.95
0.000 0.000 0.000
0.915 0.953 0.775
, -
a
Std error of est = 0.118; adjtd rz = 0.962; r2 = 0.961; mutpl r2
= 0.984.
100
indicators for this regression as well. As the t values of Table 2 demonstrate, all coefficients are statistically significant. The appreciable nonzero coefficient ( y ) of Figure 2. Dose-response plots for Assay 1 for compounds (YZ) implies an interaction of two factors Y and 2. MBA (v),MCA (O),RMBA (A),and RMCA (m).Mean values of Consequently, these coefficients have been distributed for the revertantdplate and error bars, representing the standard the contributions of C1-by-Br and H-by-OH replacements, deviation (SD),are plotted at each symbol. Error bars are concealed when the SD is less than the height of the symbol. as depicted in Figure 3, to illustrate the extent of the Actual doses of RMCA and RMBA were multiplied by a scaling interactive effects of Br, C1, OH, and H on mutagenicity. factor of 0.1 for the purpose of displaying together the doseThe values of the distributed coefficients show that the response plots of RMCA and RMBA with those of MCA and C1-by-Br replacement has a greater effect on log M for Z MBA, which required smaller doses for response. The scaling factor for MCA and MBA is 1. = H (a= 1.06) than Z = OH (a = 0.268). Looked at another way, the H-by-OH replacement is more effective and Z when they represented C1 or H substituents, for Y = C1 (p = 1.46) than Y = Br (p = 0.668). In general, respectively, whereas a value of 1 was assigned when the negative y indicates the two replacements together these two variables represented Br or OH, respe~tively.~ tend to diminish the individual enhancements of a single The regression of the eighteen log M , values on Y, 2, replacement. and YZ yielded the regression coefficients that are given The outcomes from analyses by multiple regression and in Table 2, which includes the pertinent statistical simple ratios agree. For example, from the multiple regression analysis, log MmMBA = 0.73 and log &lmRMCA = 4 Note that the regression analysis technique as used here is -0.998. Subtraction gives the following: log equivalentto the classical analysis of variance applied to a 2 by 2 crosslog MmRMCA = log(MmMBA/MmRMCA) = 1.73, whose antilog classification experimental design. 0
2
4 6 8 10 Done in 1(g x factor
12
14
16
amMBA
782 Chem. Res. Toxicol., Vol. 7, No.6, 1994
LaLonde and Leo Table 4. Heat of Formation of Neutral Compounds (AHd and Radical Anions (qd), and Differences in Heats of Formation - AHr) in kcaYmol
(ed
compd MBA MCA
RMBA RMCA
OH 10,ll = [CI,OHl
1
;
J
[Br,OH] = [ 1,l I
,461 + ,268Y
,461
.73
log w n
1%
u,
Figure 3. The distribution of a and ,!I,the coefficients of Y and 2,respectively, in the multiple regression equation [log M = const + aY + ,!I2 + @'Z) + E ] for C1-by-Br (horizontal)and H-byOH (vertical) replacements. Substituent designator values of 0 or 1 for C1, Br, H, and OH are indicated at each corner region of the rectangle along with log 2, values for each of the four compounds. The distributed coefficients of Y and Z are given in conjunction with the equation associated with each of the vertical and horizontal sides of the rectangle. Table 3. Calculated (AM11LUMO Coefficients at Atoms (2-2, C-3, C-4, and 0 - 7 and LUMO Energies (em compd C-2 C-3 C-4 0-7 LUMOenergy MBA MCA RMBA RMCA
-0.24 -0.27 -0.26 -0.28
-0.60 -0.59 -0.58 -0.56
0.66 0.65 0.67 0.66
0.25 0.26 0.25 0.26
-1.53 -1.31 -1.23 -1.02
is 53, the value noted previously from the simple ratio M"*/M"C*.
In summarizing the replacement results, each of the two bromine-substituted compounds is more mutagenic than its chlorine-substituted counterpart. Also, the presence of the hydroxyl group lessens the C1-by-Br replacement effect so that mutagenic differences among mixed halogen-substituted 2(5H)-furanones are more likely to be found significant for the compounds lacking the C-5 hydroxyl group than those possessing it. Computational Results and Correlations. The LUMO energies of MCA and MBA proved to be somewhat dependent on the angle of rotation of the hydroxyl group. For these compounds, the lowest LUMO energy was not coincidental with the lowest heat of formation. In both cases the lowest heat of formation occurred when the HO-C5-C4 dihedral angle was approximately 65". In contrast, the lowest LUMO energy occurred when the dihedral angle was in the range of 5-10'. The energy barrier for the rotation of the hydroxyl group in both MBA and MCA was 5.1 kcal/mol, with the maximum occurring a t approximately -75'. The lowest LUMO energies and the coefficients of the basis set orbitals for the four compounds are reported in Table 3. The coefficients of the orbitals indicate that the LUMO of each compound was constructed primarily from the pporbitals of C-3 and C-4. The pp orbitals of C-2 and the carbonyl oxygen provided minor contributions to the LUMO's. A node occurred between C-3 and C-4 and again between C-2 and the carbonyl oxygen for each structure. The coefficients varied little from one compound to another, and no significant correlation could be established between any coefficient and the mutagenic activities of the compounds. The contribution of the halogens to the LUMO makeup was very small and varied from 2.0% to 2.8%. For each compound, the
AHf
A l p
A H y d - AHf
- 91.44 -114.11 - 42.83 - 65.22
-141.56 -157.85 - 86.44 -102.44
-50.12 -43.74 -43.61 -37.22
coefficient of the halogen at C-2 was slightly smaller than that at C-3. The properties of the radical anion form of the compounds were also calculated. For all of the compounds the calculated heats of formation of the radical anions were lower than those of the corresponding neutral compounds (Table 4). The difference between the calculated heat of formation of the radical anion and the corresponding neutral compound gives an approximate measure of the relative stability of the radical anion to the neutral compound, as well as its electrophilicity. Inspection of the mutagenicity values of Table 1along with AM1 computed values for LUMO and (qd AHf) energies of Tables 3 and 4 reveals a general correlation: greater mutagenicity corresponds to both a lower LUMO energy level and a greater stability of the radical anion resulting when the neutral compound assumes a single electron. Both lower LUMO energy levels and greater radical anion stability are indicated by larger negative values. At least for the limited comparison allowed by this study, each bromide is more mutagenic than its chloride counterpart and has lower computed LUMO energy levels and greater radical anion stabilities. An analysis involving comparison by leastsquares, straight-line plots shows the eighteen values of M, or log M, correlate with the computed LUMO and - AHf) values. However, because the total span of mutagenicity values is only 50-fold, that is, ranging between one and two powers of 10, it is not surprising that there is so little difference in the strength of M, and log M, correlations. The squared correlation coefficients (1.2) for these four correlations are: LUMO vs Mm, 0.80; LUMO vs log Mm,0.84; (qd - AHf)vs M,0.73; (qd - AHd vs log M,, 0.80. The LUMO and (qd AHf) vs log M, appear to be stronger, but there is no statistically significant difference. However, the outcome of the comparisons made here for bromides and chlorides agrees with the earlier correlations of log M, with LUMO - AHf) for a larger set of mostly chlorineand (qd substituted 4-methyl 2(5H)-furanones (8). The present and previous correlations between mutagenicity and both LUMO energies and radical anion stabilities point to mutagenicity resulting in a process involving a radical anion. How might such a radical anion be generated, and how might it be involved in chemical processes manifested ultimately as mutagenesis? Recently, MCA radical anion (MCk-) was observed as an intermediate when MCA was treated with GSH (16).The stable reaction products were the GSH-MCA conjugates and GSSG. Conceivably, MCk-, or a neutral protonated form thereof, could react with biomolecules directly involved in DNA replication. For example, the radical-initiated degradation of the deoxyribose moiety of deoxyribonucleotides can occur when a free radical abstracts one of the several types of hydrogens from the deoxyribose ring (17).Thus, the foregoing leads one to propose that cellular glutathione potentiates the halogensubstituted 2(5H)-furanones. However, solutions of chlo-
(wd
Halogen Replacement Effects on Mutagenicity
rine-substituted 2(5H)-furanones are at least partially inactivated by thiols, such as glutathione ( 6 ) and Nacetylcysteine (18). The thiol conjugate reaction products have tested nonmutagenic in the TAlOO-Sg assay (18, 19). Furthermore, the efficiency of conjugate formation may be greater in cells where the presence of glutathione transferase could enhance a conversion of halogenated 2(5H)-furanone to a nonmutagenic conjugate, this occurring a t the expense of radical anion production. Nevertheless, the convergent structure-activity and MCA radical anion findings provide a substantial platform for further inquiry regarding the mechanism of halogensubstituted 2(5H)-furanone mutagenesis.
Chem. Res. Toxicol., Vol. 7, No. 6, 1994 703
butenoic acid in chlorine-treated humic water and drinking water extracts. Environ. Sci. Technol. 22, 1097-1103. (6) Ishiguro, Y., LaLonde, R. T., Dence, C. W., and Santodonato, J. (1987) Mutagenicity of chlorine-substituted furanones and their inactivation by reaction with nucleophiles. Enuiron. Toxicol. Chem. 6,935-946. (7) Horth, H., Fielding, M., James, C. P., and Gwilliam, R. D. (1991) Identification of Mutagens in Drinking Water (EC 9105 SLD). PRD 2038-M. Water Research Centre, Marlow, Bricks, U.K. (8) LaLonde, R. T., Leo, H., Perakyla, H., Dence, C. W., and Farrell, R. P. (1992) Associations of the bacterial mutagenicity of halogenated 2(5H)-furanones with their MNDO-PM3 computed properties and mode of reactivity with sodium borohydride. C h m . Res. TOX~COZ. 6, 392-400. (9) Tuppurainen, K., Latjonen, S.,Laatikainen, R., Vartianen, T., Maran, U., Standberg, M., and Tamm, T. (1991) About the mutagenicity of chlorine-substituted furanones and halopropenals. A QSAR study using molecular orbital indices. Mutat. Res. Acknowledgment. This research was supported by 247,97-102. the U.S. Geological Survey, Department of the Interior, (10) LaLonde, R. T., Perakyla, H., Cook, G. P., and Dence, C. W. (1990) under USGS Award 14-08-0001-G1912. The views and Contribution of the 5-hydroxyl group to the mutagenicity of conclusions contained in this document are those of the mucochloric acid. Enuiron. Toxicol. Chem. 9, 687-691. authors and should not be interpreted as necessarily (11) Dupont, G., Dulou, R., and Lefebvre, G. (1954) Some mono representing the official policies, either expressed or derivatives of 2-butyne-1, 4-diol. Bull. SOC. Chim Fr., 816-820. (12) Maron, D. M., and Ames, B. N. (1983) Revised methods for the implied, of the US.Government. The assistance of David Salmonella mutagenicity test. Mutat. Res. 113, 173-215. J . Kiemle in determining the mass and NMR spectra, (13) Stewart, J. J. P. (1990) MOPAC: A general molecular orbital Jeffrey S. Johnson for preparing RMBA, and LeMoyne package (version 6.0). QCPE Bull. 10, 86. College, Syracuse, NY,for VAX 3100 computational time (14) LaLonde, R. T., Cook, G. P., Perakyla, H., Dence, C. W., and is also gratefully acknowledged. Babish, J. G. (1991) Salmonella typhimurium (TA100) mutagenicity of 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone References and its open- and closed-ring analogs. Enuiron. Mol. Mutagen. 17,40-48. Kronberg, L., and Franzen, R. (1993)Determination of chlorinated (15) LaLonde, R. T., Cook, G. P., Perakyla, H., and Dence, C. W. (1991) furanones, hydroxyfuranones, and butenedioic acids in chlorineEffect on mutagenicity of the stepwise removal of hydroxyl treated water in pulp bleaching liquor. Environ. Sci. Technol. 27, group and chlorine atoms from 3-chloro-4-(dichloromethyl)-51811-1818. hydroxy-2(51.n-furanone: 13C NMR chemical shifts as determiHolmbom, B. R., Voss, R. H., Mortimer, R. D., and Wong, A. (1981) nants of mutagenicity. Chem. Res. Toxicol. 4, 35-40. Isolation and identification of an Ames-mutagenic compound (16) LaLonde, R. T., Xie, S., Chamulitrat, W., and Mason, R. P. (1994) present in kraft chlorination effluents. Tappi 64, 172-174. Oxidation and radical intermediates associated with the gluMeier, J. R., Knohl, R. B., Coleman, W. E., Ringhand, H. P., tathione conjugate of mucochloric acid. Chem. Res. Toxicol. 7, Munch, J. W., Kaylor, W. H., Streicher, R. P., and Kopfler, F. C. 482-486. (1987) Studies on the potent mutagen, 3-chloro-4-(dichloro(17) Dedon, P. C., and Goldberg, I. H. (1992) Free-radical mechanisms methyl)-5-hydroxyl-2(5H)-furanone; aqueous stability, XAD rein the formation of sequence-dependent bistranded DNA lesions covery and analytical determination in drinking water and in by the antitumor antibiotics bleomycin, neocarzinostatin, and chlorinated humic acid solutions. Mutat. Res. 189, 363-373. calicheamicin. Chem. Res. Toxicol. 6, 311-332. Holmbom, B. R., Voss, R. H., Mortimer, R. D., and Wong, A. (1984) (18) LaLonde, R. T., and Xie, S. (1992)A study of inactivation reactions Fractionation, isolation and chlorination of Ames mutagenic of N-acetylcysteine with mucochloric acid, a mutagenic product compounds in kraft chlorination effluents. Enuiron. Sci. Technol. of the chlorination of humic substances in water. Chem. Res. 18,333-337. TOX~CO~. 6, 618-624. Kronberg, L., Holmbom, B., Reunanen, M., and Tikkanen, L. (1988) Identification and quantification of the Ames mutagenic (19) LaLonde, R. T., and Xie, S. (1993) Glutathione and N-acetylcyscompound 3-chloro-4-~dichloromethyl)-5-hydroxy-2(5H)-f~anone teine inactivations of mutagenic 2(5H)-furanones from the chloand its geometric isomer (E)-2-chloro-3-(dichloromethyl)-4-oxorination of humics in water. Chem. Res. Toxicol. 6, 445-451.