Phiirriiucology & Toricology 1992. 71, 165-172
Mutagenicity of Crude Senna and Senna Glycosides in Salmonella typhimurium Dagny Sandnes, Tori1 Johansen, Gerd Teien and Gunnar Ulsaker The Norwegian Medicines Control Authority, Sven Oftedals vei 6, N-0950 Oslo, Norway (Received April 13, 1991; Accepted February 13, 1992) The mutagenicity of' senna glycosides and extracts of senna folium and senna fructus was investigated in the Sulnionellu r.vp/iiniurium reversion assay. Senna glycosides were inactive in all strains, except for a slight, but significant increase in mutant frequency in TA102 in the absence and presence of liver microsomes. Extracts of senna fructus and senna folium demonstrated weak activity in TA9'7a. TAIOO and TA102 in the presence of liver microsomes, and in TA97a and TA102 in the absence of liver microsomes. A strong increase in mutant frequency (3- to 5-fold above background frequency) was observed with all extracts in TA98 in the presence of liver microsomes. This activity increased further following enzymatic hydrolysis with hesperidinase of extracts of senna fructus from one source, and could be correlated to the release of the flavonol aglycones kaempferol and quercetin. The weak or lacking activity of anthraquinone aglycones in the tested strains of Sulnionellii typhirizurium indicates that mutagenicity can not be attributed solely to the anthraquinone content of these plant materials. The chemical niiture of other mutagenic components has not been elucidated. Abstract:
Several stimulant laxatives are derived from plants, e.g. senna, rhubarb, aloe, cascara and frangula (van 0 s 1976; Brunton 1990). The anthraquinone glycosides found in these plants are considered as prodrugs, which are hydrolyzed by gut flora enzymes to release the active anthraquinone aglycones (Brunton 1990). Several of these anthraquinone aglycones, as well as the synthetic analogue danthron (l,8dihydroxyanthraquinone) are mutagenic in the Salmonella typhimurium reversion assay (Brown 1980; Tikkanen et a/. 1983; Westendorf et al. 1990). Some anthraquinone aglycones have also been shown to be active in tests for genotoxicity in mammalian cells (Westendorf et al. 1990), while others may have tumour-promoting activity (Wolfle et al. 1990). Danthron has been found to induce intestinal tumours in rats and mice and liver tumors in mice (Mori et al. 1985 & 1986). Following the voluntary withdrawal of medicinal products containing danthron, products containing senna or senna glycosides have become dominant among stimulant laxatives on the Norwegian market. Therefore, we found it of interest to examine the mutagenicity of these products. The principal constituents of senna are rhein diarithrone glucosides (sennosides A and B), which are metabolized in colon to sennidins, rhein-9-anthrone and rhein (Lemli & Lemmens 1980). Minor constituents are glycosides of aloeemodin dianthrone, heterodianthrones, flavonols and free aglycones (van 0 s 1976; Duke 1985). Morimoto et al. (1982) reported that a methanol extract of senna folium was mutagenic in Salmonrllu typhimurium. whereas senna glycosides were reported to be inactive (Tikkanen et al. 1983; Mengs 1988). This may indicate that crude senna preparations contain mutagenic components that are removed upon isolation of sennosides. Furthermore, these studies were performed with metabolic activation by rat liver microsomes. However, glycosides may be resistant to hydrolysis by liver microsomal enzymes (Brown & Dietrich 1979). As the prin-
cipal aglycone of senna, rhein, has been reported to be a weak mutagen in Salmonella typhimurium (Brown 1980; Westendorf et al. 1990) the reported lack of activity of senna glycosides might be due to resistance to hydrolysis. We have therefore compared the mutagenicity of pure sennosides and crude senna preparations and investigated the effect of enzymatic hydrolysis on the mutagenicity of these materials in the Salmonella typhimurium reversion assay. By comparison with anthraquinone and flavonol aglycones shown to be present in the preparations, we have attempted to identify the mutagenic components.
Materials and Methods Materials. Quercetin, kaempferol, aloe-emodin and emodin of HPLC quality were obtained from Carl Roth KG. Karlsruhe, Germany. Rhein was a gift from Dr. Madaus GmbH & Co. Cologne, Germany. Standardized senna fructus concentrate powder, prepared from senna pods after removal of seeds, (46 mg sennosidedg) was obtained from Nycomed, Oslo, Norway. Standardized senna glycosides (sennosides A + B in the form of calcium salts) were obtained from Sandoz AG, Basle, Switzerland. Senna folium (36 mg sennosides/g) and senna fructus (24 mg sennosides/g) conforming to the European Pharmacopoeia were obtained from the Norwegian Medicinal Depot, Oslo, Norway. 4-Nitro-o-phenylene diamine and cumene hydroperoxide were from Fluka Chemie AG. Buchs, Switzerland. 2-Aminofluorene was from Aldrich Chemical Company, Steinheim, Germany. Sodium azide was from E. Merck. Darmstadt, Germany. Danthron and hesperidinase were obtained from Sigma Chemical Company, St. Louis, Mo.. U.S.A. Aroclor 1254 was from Supelco, Bellefonte, Pe.. U.S.A. All other chemicals were of analytical grade. Prepararion of test solutions. Senna fructus and senna foliurn were suspended in water at a concentration of 150 mg/ml and shaken overnight to dissolve glycosides. Following centrifugation, 300 pl of the supernatant was mixed with 500 p1 citric acid buffer pH 4.0 with or without 20 mg hesperidinase (Fukuoka et al. 1978), and incubated for 2 hr at 37". The volume was then adjusted to I ml with 0.07 N NaOH. Senna glycosides were dissolved in water (150
166
DAGNY SANDNES ET AZ. Rihle I
Mutagenicity of senna glycosides in Salmonellu typhimurium. Solutions of senna glycosides were tested in the plate incorporation assay (Maron & Ames 1983) in the absence and presence of rat liver microsomes (S9 fraction). The volume of S9 protein was 100 pliplate. Results are presented as mean number of revertant colonies/ pla1efS.D. of 3 4 separate experiments, each performed with duplicate plates. Control cultures were treated with 400 pl of the buffer solution prepared as described in Materials and Methods. P-values refer to the regression of revertant coloniesiplate upon dose, r is the correlation coefficient.
TA97a
0 50 100
200 400
TA98
106f5 120f 1 122+ I6 131 f 2 8 137f38
l76+ 10 185+ 15
194+7 188+9 l84+ I4
TA I02*
TAl00
-s 9
+s 9
-s 9
+s 9
-s 9
+s 9
31 + 3 35+8 41 + 9 40+ 15 50k 12
37f2 42f8 47f8 50f8 50+7
226f I4 258+ 12 261 +28 255+ 19 236k44
190f 19 208 27 228 f 30 224 f26 236k44
159k24 252f I6 265+21 271 +25 271 f 6 4
276k45 462+36 487+31 515+27 517+43
+s 9
-S9
Dose. pliplate
+
* TA102 -S9: slope. 0.48f0.15 revertantsipl. P=0.01 I. r=0.7029 (dose-range 0-200 pl) TA102 +S9: slope, 1.03+0.24 revertantsipl, P
100
a
U
0
I
6
26a
460
Dose. pl/plale
Dose,pl/plate
I L Z - 1
TA97a t S9 500
f a
400
300 a
5
200
b
a 100 0
-
0
200
Dose. pl/plale
400
0
200 400 Dose. pllplate
Fig. 1. Mutagenicity of senna extracts in Snlmonell~r typhimurirrm strain TA97a in the absence and presence of liver microsomal S9 fraction. Extracts were prepared as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pliplate. Control cultures were treated with 400 pl of buffer solution, prepared as described in Materials and Methods. Results are presented as mean number of revertant colonieslplate fS.D. of 3-4 separate experiments, each performed with duplicate plates. P-values refer to regression of revertant coloniesiplate upon dose, r is the correlation coefficient. Unless otherwise indicated, P was < 0.001. A-A-A Senna folium. In the absence of S9: slope=O 17f0.03 revertantsipl, r=0.7550. In the presencer of S9: slope=0.61 f0.06 revertantsipl, r=0.9284. 0 - E - 0 Senna fructus. In the absence of S9: slope=O 14k0.04 revertantsipl, P=0.003, r=0.7162. In the presence of S9: slope= 0.71 kO.07, revertantsipl. r=0.9482, 0-0-0 Senna fructuj concentrate. In the absence of S9: slope=0.17~0.03revertantsipl, r = 0.8240. In the presence of S9: slope=0.31 k0.06 revertanrsipl, r = 0.7622. -0 Hydrolyzed senna fructus concentrate. In the absence of S9: slope=0.19f0.04 revertants/pl. r=0.7555. In the presence of S9: Dose-response curve was linear up to 200 pl. Slope= I .05+0.28 revertantsifil. P=0.002, r=0.7099.
the exception of TA102, mutagenic activity increased in the presence of liver microsomes. Enzymatic hydrolysis with hesperidinase (Fukuoka et al. 1978) did not influence the mutagenic activity of senna glycosides, senna fructus or senna folium (data not shown). Following hydrolysis, mutant frequency induced by extracts of senna fructus concentrate increased further in TA97a, TA98 and TAlOO in the presence of liver microsomes and mutant frequency was significantly increased in TA98 in the absence of liver microsomes (table 2, fig. 1-3). The mutant frequency of TA102 in the presence of liver microsomes and TA97a, TAlOO and TA102 in the absence of liver microsomes (fig. 1 and 4, table 2) was not significantly altered by enzymatic hydrolysis. HPLC analysis of the solutions revealed peaks with retention times and UV-spectra corresponding to emodin, aloeemodin, quercetin and kaempferol (data not shown). Minor quantities of aloe-emodin (ranging from 1 to 5 pg/ml) and emodin (approximately 1 pg/ml) were found in all solutions, however, these did not increase upon hydrolysis. In extracts of senna fructus concentrate, the concentration of the flavonol aglycones kaempferol and quercetin increased from levels below the detection limit (1 pg/ml) before hydrolysis to approximately 9 pg/ml following hydrolysis. These flavonol aglycones were not detected in any of the other solutions. Rhein did not significantly increase mutant frequency in any strain of Salmonella typhimurium (tables 3 and 4). Emodin induced significant, dose-related increases in mutant
TA98tS9 300
300
Q
Q
3Q
I
5
5
200
p
100
.200 ul
Y) L
L
C c
5
a 0
0
0F-XF-ZDose. pl/plate
pl, both in the absence and presence of liver microsomes. Mutagenic activity was higher in the presence of liver microsomes. Extracts of senna folium, senna fructus and senna fructus concentrate induced significant dose-related increases in mutant frequency in all strains of Sar’monella typhimurium in the presence of liver microsomes (fig. 1 4 ) . The strongest increase in mutant frequency was observed in TA98 (fig. 2). In the absence of liver microsomes, weak, but significant dose-related increases in mutant frequency were observed with all extracts in TA97a (fig. 1) and stronger activity in TA102 (fig. 4). Weak activity in the absence of liver microsomes was also observed in TA98 with extracts of senna fructus and hydrolyzed senna fructus concentrate and in TAlOO with extracts of senna folium (table 2 ) . With
100
01
U 0)
0
200
400
Dose. pl/plate
Fig. 2. Mutagenicity of senna extracts in Salmonella typhimurium strain TA98 in the presence of liver microsomal S9 fraction. Extracts were preparated as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pl/plate. Control cultures were treated with 400 p1 of buffer solution, prepared as described in Materials and Methods. Results are presented as number of revertant colonies/platef S.D. of 3 4 separate experiments, each performed with duplicate plates. P-values refer to the regression of mutant colonies/plate upon dose, r is the regression coefficient. Unless otherwise indicated, P was < 0.001. A-A-A Senna folium. Slope=0.38+0.04 revertantsipl, r = 0.9160. 0-0-0Senna fructus. Slope=0.46+0.03 revertantsipl, r=0.9670, 0-0-0 Senna fructus concentrate. Slope =0.25 0.02 revertants/pl, r = 0.9529. H - 0 Hydrolyzed senna fructus concentrate. Dose-response curve was linear up to 200 pl, slope= I .lo+ 0.13 revertantsipl, r=0.939l.
168
DAGNY SANDNES ET A L . TA 1OO+S9
300
; 100 0
400 pilplate
200 DOSE,
0
200 400 Dose, pl/plale
Fig. 3. Mutagenicity of senna extracts in Su/rilonc,//tr /,vp/iiniuriuni strain TAl00 in the presence of liver microsomal S9 fraction. Extracts were prepared as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pliplate. Control cultures were treated with 400 pI of buffer, prepared as described in Materials and Methods. Results are presented as mean number of revertant coloniesiplatef S.D. of 3 4 experiments, each performed with duplicate plates. P-values refer to the regression of mutant colonies/ plate upon dose, r is the correlation coefficient. Unless otherwise indicated, P was < 0.001. A -A - A Senna folium: Slope=0.44f0.05 revertantsipl, r = 0.9022. U - L - U Senna fructus: slope=0.32+0.04 revertants/pl. r=0.9077. 0-0-0 Senna fructus concentrate: P=0.40 by analysis of variance. regression analysis: slope = 0.20 f0.08 revertants/pl, Hydrolyzed senna fructus concentrate: P=0.034, r = 0.5486.0-0P=0.063 by analysis of variance, dose-response curve was linear up to 100 pl, slope= l.OOf0.39 revertantsipl, P=0.038. r=0.6935.
strain, whereas activity in the absence of liver microsomes was weaker (table I , fig. 4). In contrast to senna glycosides, extracts of senna fructus and senna folium demonstrated significant activity in all strains of Sulmonellu typhimurium, indicating that these plant materials contain mutagenic components that are removed upon isolation of sennosides (fig. 1 4 , table 2). Apart from activity in TA98, the mutagenic activity of senna extracts was weak, with no more than a doubling of the background frequency at the highest dose. However, the significant dose-related increase in mutant frequency indicates an effect of biological significance. Due to limitations in the volume of plant extract that could be applied to the plates, higher doses were not tested. This would require
600
1001
, ,
O J ,
0
frequency in TA97a and TA102 in the presence of liver microsomes (fig. 5). N o activity was observed in the absence of liver microsomes and in the other test strains, except for a slight increase in mutant frequency in TAIOO in the presence of liver microsomes (tables 3 and 4). Aloe-emodin increased mutant frequency in TA97a and TA98 in the absence of liver microsomes, and in all strains in the presence of liver microsomes, although the increase was slight in TAIOO (fig. 5, table 4). In TA97a and TA98 activity was reduced in the presence of liver microsomes. The response of test strains to anthraquinone aglycones was not improved by the preincubation procedure (Maron & Ames 1983) or by varying the concentration of S9 per plate (data not shown). Quercetin and kaempferol significantly increased mutant frequency in all strains in the presence of liver microsomes (fig. 6). In the absence of liver microsomes, kaempferol was inactive in TA97a and TA98, whereas quercetin was inactive in TA102 (tables 3 and 4). Activity was higher in all strains in the presence of liver microsomes. Discussion
In the present study, we confirmed the lack of activity of senna glycosides in Sulmonellu typhimurium in previously tested strains (Tikkanen et ul. 1983; Mengs 1988). In TA102 (table I ) . which was not tested previously, we observed a weak, but significant, dose-related increase in mutant frequency, both in the absence and presence of liver microsomes. Activity in the presence of liver microsomes was similar to the activity of other senna preparations in this
TA 102.S9
TA10249 600
I
1
0
200 400 DOSE.pl/plate
I
TAlO2+S9
-
1
500
-
m
m
5 c
TA102rS9
600-
800500
200 460 Dose. pllplale
'
400-
5n
300-
E 300-
\
400-
v)
a
w
f;
3 U
5
200-
a
100-
0
2w 400 Dose. p p i a t e
200-
100-
O J , 0
,
,
200 4w Dose. pl/plaie
Fig. 4. Mutagenicity of senna extracts in Scr/morie//uryphimuriimi strain TA102 in the absence and presence of liver microsomal S9 fraction. Extracts were preparated as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pliplate. Control cultures were treated with 400 pI of buffer solution, prepared as described in Materials and Methods. Results are presented as mean number of revertant coloniesiplatef S.D. of 3 4 separate experiments, each performed with duplicate plates. P-values refer to the regression of revertant colonies/plate upon dose, r is the correlation coefficient. Unless otherwise indicated, P was
Mutagenicity of Crude Senna and Senna Glycosides in Salmonella typhimurium Dagny Sandnes, Tori1 Johansen, Gerd Teien and Gunnar Ulsaker The Norwegian Medicines Control Authority, Sven Oftedals vei 6, N-0950 Oslo, Norway (Received April 13, 1991; Accepted February 13, 1992) The mutagenicity of' senna glycosides and extracts of senna folium and senna fructus was investigated in the Sulnionellu r.vp/iiniurium reversion assay. Senna glycosides were inactive in all strains, except for a slight, but significant increase in mutant frequency in TA102 in the absence and presence of liver microsomes. Extracts of senna fructus and senna folium demonstrated weak activity in TA9'7a. TAIOO and TA102 in the presence of liver microsomes, and in TA97a and TA102 in the absence of liver microsomes. A strong increase in mutant frequency (3- to 5-fold above background frequency) was observed with all extracts in TA98 in the presence of liver microsomes. This activity increased further following enzymatic hydrolysis with hesperidinase of extracts of senna fructus from one source, and could be correlated to the release of the flavonol aglycones kaempferol and quercetin. The weak or lacking activity of anthraquinone aglycones in the tested strains of Sulnionellii typhirizurium indicates that mutagenicity can not be attributed solely to the anthraquinone content of these plant materials. The chemical niiture of other mutagenic components has not been elucidated. Abstract:
Several stimulant laxatives are derived from plants, e.g. senna, rhubarb, aloe, cascara and frangula (van 0 s 1976; Brunton 1990). The anthraquinone glycosides found in these plants are considered as prodrugs, which are hydrolyzed by gut flora enzymes to release the active anthraquinone aglycones (Brunton 1990). Several of these anthraquinone aglycones, as well as the synthetic analogue danthron (l,8dihydroxyanthraquinone) are mutagenic in the Salmonella typhimurium reversion assay (Brown 1980; Tikkanen et a/. 1983; Westendorf et al. 1990). Some anthraquinone aglycones have also been shown to be active in tests for genotoxicity in mammalian cells (Westendorf et al. 1990), while others may have tumour-promoting activity (Wolfle et al. 1990). Danthron has been found to induce intestinal tumours in rats and mice and liver tumors in mice (Mori et al. 1985 & 1986). Following the voluntary withdrawal of medicinal products containing danthron, products containing senna or senna glycosides have become dominant among stimulant laxatives on the Norwegian market. Therefore, we found it of interest to examine the mutagenicity of these products. The principal constituents of senna are rhein diarithrone glucosides (sennosides A and B), which are metabolized in colon to sennidins, rhein-9-anthrone and rhein (Lemli & Lemmens 1980). Minor constituents are glycosides of aloeemodin dianthrone, heterodianthrones, flavonols and free aglycones (van 0 s 1976; Duke 1985). Morimoto et al. (1982) reported that a methanol extract of senna folium was mutagenic in Salmonrllu typhimurium. whereas senna glycosides were reported to be inactive (Tikkanen et al. 1983; Mengs 1988). This may indicate that crude senna preparations contain mutagenic components that are removed upon isolation of sennosides. Furthermore, these studies were performed with metabolic activation by rat liver microsomes. However, glycosides may be resistant to hydrolysis by liver microsomal enzymes (Brown & Dietrich 1979). As the prin-
cipal aglycone of senna, rhein, has been reported to be a weak mutagen in Salmonella typhimurium (Brown 1980; Westendorf et al. 1990) the reported lack of activity of senna glycosides might be due to resistance to hydrolysis. We have therefore compared the mutagenicity of pure sennosides and crude senna preparations and investigated the effect of enzymatic hydrolysis on the mutagenicity of these materials in the Salmonella typhimurium reversion assay. By comparison with anthraquinone and flavonol aglycones shown to be present in the preparations, we have attempted to identify the mutagenic components.
Materials and Methods Materials. Quercetin, kaempferol, aloe-emodin and emodin of HPLC quality were obtained from Carl Roth KG. Karlsruhe, Germany. Rhein was a gift from Dr. Madaus GmbH & Co. Cologne, Germany. Standardized senna fructus concentrate powder, prepared from senna pods after removal of seeds, (46 mg sennosidedg) was obtained from Nycomed, Oslo, Norway. Standardized senna glycosides (sennosides A + B in the form of calcium salts) were obtained from Sandoz AG, Basle, Switzerland. Senna folium (36 mg sennosides/g) and senna fructus (24 mg sennosides/g) conforming to the European Pharmacopoeia were obtained from the Norwegian Medicinal Depot, Oslo, Norway. 4-Nitro-o-phenylene diamine and cumene hydroperoxide were from Fluka Chemie AG. Buchs, Switzerland. 2-Aminofluorene was from Aldrich Chemical Company, Steinheim, Germany. Sodium azide was from E. Merck. Darmstadt, Germany. Danthron and hesperidinase were obtained from Sigma Chemical Company, St. Louis, Mo.. U.S.A. Aroclor 1254 was from Supelco, Bellefonte, Pe.. U.S.A. All other chemicals were of analytical grade. Prepararion of test solutions. Senna fructus and senna foliurn were suspended in water at a concentration of 150 mg/ml and shaken overnight to dissolve glycosides. Following centrifugation, 300 pl of the supernatant was mixed with 500 p1 citric acid buffer pH 4.0 with or without 20 mg hesperidinase (Fukuoka et al. 1978), and incubated for 2 hr at 37". The volume was then adjusted to I ml with 0.07 N NaOH. Senna glycosides were dissolved in water (150
166
DAGNY SANDNES ET AZ. Rihle I
Mutagenicity of senna glycosides in Salmonellu typhimurium. Solutions of senna glycosides were tested in the plate incorporation assay (Maron & Ames 1983) in the absence and presence of rat liver microsomes (S9 fraction). The volume of S9 protein was 100 pliplate. Results are presented as mean number of revertant colonies/ pla1efS.D. of 3 4 separate experiments, each performed with duplicate plates. Control cultures were treated with 400 pl of the buffer solution prepared as described in Materials and Methods. P-values refer to the regression of revertant coloniesiplate upon dose, r is the correlation coefficient.
TA97a
0 50 100
200 400
TA98
106f5 120f 1 122+ I6 131 f 2 8 137f38
l76+ 10 185+ 15
194+7 188+9 l84+ I4
TA I02*
TAl00
-s 9
+s 9
-s 9
+s 9
-s 9
+s 9
31 + 3 35+8 41 + 9 40+ 15 50k 12
37f2 42f8 47f8 50f8 50+7
226f I4 258+ 12 261 +28 255+ 19 236k44
190f 19 208 27 228 f 30 224 f26 236k44
159k24 252f I6 265+21 271 +25 271 f 6 4
276k45 462+36 487+31 515+27 517+43
+s 9
-S9
Dose. pliplate
+
* TA102 -S9: slope. 0.48f0.15 revertantsipl. P=0.01 I. r=0.7029 (dose-range 0-200 pl) TA102 +S9: slope, 1.03+0.24 revertantsipl, P
100
a
U
0
I
6
26a
460
Dose. pl/plale
Dose,pl/plate
I L Z - 1
TA97a t S9 500
f a
400
300 a
5
200
b
a 100 0
-
0
200
Dose. pl/plale
400
0
200 400 Dose. pllplate
Fig. 1. Mutagenicity of senna extracts in Snlmonell~r typhimurirrm strain TA97a in the absence and presence of liver microsomal S9 fraction. Extracts were prepared as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pliplate. Control cultures were treated with 400 pl of buffer solution, prepared as described in Materials and Methods. Results are presented as mean number of revertant colonieslplate fS.D. of 3-4 separate experiments, each performed with duplicate plates. P-values refer to regression of revertant coloniesiplate upon dose, r is the correlation coefficient. Unless otherwise indicated, P was < 0.001. A-A-A Senna folium. In the absence of S9: slope=O 17f0.03 revertantsipl, r=0.7550. In the presencer of S9: slope=0.61 f0.06 revertantsipl, r=0.9284. 0 - E - 0 Senna fructus. In the absence of S9: slope=O 14k0.04 revertantsipl, P=0.003, r=0.7162. In the presence of S9: slope= 0.71 kO.07, revertantsipl. r=0.9482, 0-0-0 Senna fructuj concentrate. In the absence of S9: slope=0.17~0.03revertantsipl, r = 0.8240. In the presence of S9: slope=0.31 k0.06 revertanrsipl, r = 0.7622. -0 Hydrolyzed senna fructus concentrate. In the absence of S9: slope=0.19f0.04 revertants/pl. r=0.7555. In the presence of S9: Dose-response curve was linear up to 200 pl. Slope= I .05+0.28 revertantsifil. P=0.002, r=0.7099.
the exception of TA102, mutagenic activity increased in the presence of liver microsomes. Enzymatic hydrolysis with hesperidinase (Fukuoka et al. 1978) did not influence the mutagenic activity of senna glycosides, senna fructus or senna folium (data not shown). Following hydrolysis, mutant frequency induced by extracts of senna fructus concentrate increased further in TA97a, TA98 and TAlOO in the presence of liver microsomes and mutant frequency was significantly increased in TA98 in the absence of liver microsomes (table 2, fig. 1-3). The mutant frequency of TA102 in the presence of liver microsomes and TA97a, TAlOO and TA102 in the absence of liver microsomes (fig. 1 and 4, table 2) was not significantly altered by enzymatic hydrolysis. HPLC analysis of the solutions revealed peaks with retention times and UV-spectra corresponding to emodin, aloeemodin, quercetin and kaempferol (data not shown). Minor quantities of aloe-emodin (ranging from 1 to 5 pg/ml) and emodin (approximately 1 pg/ml) were found in all solutions, however, these did not increase upon hydrolysis. In extracts of senna fructus concentrate, the concentration of the flavonol aglycones kaempferol and quercetin increased from levels below the detection limit (1 pg/ml) before hydrolysis to approximately 9 pg/ml following hydrolysis. These flavonol aglycones were not detected in any of the other solutions. Rhein did not significantly increase mutant frequency in any strain of Salmonella typhimurium (tables 3 and 4). Emodin induced significant, dose-related increases in mutant
TA98tS9 300
300
Q
Q
3Q
I
5
5
200
p
100
.200 ul
Y) L
L
C c
5
a 0
0
0F-XF-ZDose. pl/plate
pl, both in the absence and presence of liver microsomes. Mutagenic activity was higher in the presence of liver microsomes. Extracts of senna folium, senna fructus and senna fructus concentrate induced significant dose-related increases in mutant frequency in all strains of Sar’monella typhimurium in the presence of liver microsomes (fig. 1 4 ) . The strongest increase in mutant frequency was observed in TA98 (fig. 2). In the absence of liver microsomes, weak, but significant dose-related increases in mutant frequency were observed with all extracts in TA97a (fig. 1) and stronger activity in TA102 (fig. 4). Weak activity in the absence of liver microsomes was also observed in TA98 with extracts of senna fructus and hydrolyzed senna fructus concentrate and in TAlOO with extracts of senna folium (table 2 ) . With
100
01
U 0)
0
200
400
Dose. pl/plate
Fig. 2. Mutagenicity of senna extracts in Salmonella typhimurium strain TA98 in the presence of liver microsomal S9 fraction. Extracts were preparated as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pl/plate. Control cultures were treated with 400 p1 of buffer solution, prepared as described in Materials and Methods. Results are presented as number of revertant colonies/platef S.D. of 3 4 separate experiments, each performed with duplicate plates. P-values refer to the regression of mutant colonies/plate upon dose, r is the regression coefficient. Unless otherwise indicated, P was < 0.001. A-A-A Senna folium. Slope=0.38+0.04 revertantsipl, r = 0.9160. 0-0-0Senna fructus. Slope=0.46+0.03 revertantsipl, r=0.9670, 0-0-0 Senna fructus concentrate. Slope =0.25 0.02 revertants/pl, r = 0.9529. H - 0 Hydrolyzed senna fructus concentrate. Dose-response curve was linear up to 200 pl, slope= I .lo+ 0.13 revertantsipl, r=0.939l.
168
DAGNY SANDNES ET A L . TA 1OO+S9
300
; 100 0
400 pilplate
200 DOSE,
0
200 400 Dose, pl/plale
Fig. 3. Mutagenicity of senna extracts in Su/rilonc,//tr /,vp/iiniuriuni strain TAl00 in the presence of liver microsomal S9 fraction. Extracts were prepared as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pliplate. Control cultures were treated with 400 pI of buffer, prepared as described in Materials and Methods. Results are presented as mean number of revertant coloniesiplatef S.D. of 3 4 experiments, each performed with duplicate plates. P-values refer to the regression of mutant colonies/ plate upon dose, r is the correlation coefficient. Unless otherwise indicated, P was < 0.001. A -A - A Senna folium: Slope=0.44f0.05 revertantsipl, r = 0.9022. U - L - U Senna fructus: slope=0.32+0.04 revertants/pl. r=0.9077. 0-0-0 Senna fructus concentrate: P=0.40 by analysis of variance. regression analysis: slope = 0.20 f0.08 revertants/pl, Hydrolyzed senna fructus concentrate: P=0.034, r = 0.5486.0-0P=0.063 by analysis of variance, dose-response curve was linear up to 100 pl, slope= l.OOf0.39 revertantsipl, P=0.038. r=0.6935.
strain, whereas activity in the absence of liver microsomes was weaker (table I , fig. 4). In contrast to senna glycosides, extracts of senna fructus and senna folium demonstrated significant activity in all strains of Sulmonellu typhimurium, indicating that these plant materials contain mutagenic components that are removed upon isolation of sennosides (fig. 1 4 , table 2). Apart from activity in TA98, the mutagenic activity of senna extracts was weak, with no more than a doubling of the background frequency at the highest dose. However, the significant dose-related increase in mutant frequency indicates an effect of biological significance. Due to limitations in the volume of plant extract that could be applied to the plates, higher doses were not tested. This would require
600
1001
, ,
O J ,
0
frequency in TA97a and TA102 in the presence of liver microsomes (fig. 5). N o activity was observed in the absence of liver microsomes and in the other test strains, except for a slight increase in mutant frequency in TAIOO in the presence of liver microsomes (tables 3 and 4). Aloe-emodin increased mutant frequency in TA97a and TA98 in the absence of liver microsomes, and in all strains in the presence of liver microsomes, although the increase was slight in TAIOO (fig. 5, table 4). In TA97a and TA98 activity was reduced in the presence of liver microsomes. The response of test strains to anthraquinone aglycones was not improved by the preincubation procedure (Maron & Ames 1983) or by varying the concentration of S9 per plate (data not shown). Quercetin and kaempferol significantly increased mutant frequency in all strains in the presence of liver microsomes (fig. 6). In the absence of liver microsomes, kaempferol was inactive in TA97a and TA98, whereas quercetin was inactive in TA102 (tables 3 and 4). Activity was higher in all strains in the presence of liver microsomes. Discussion
In the present study, we confirmed the lack of activity of senna glycosides in Sulmonellu typhimurium in previously tested strains (Tikkanen et ul. 1983; Mengs 1988). In TA102 (table I ) . which was not tested previously, we observed a weak, but significant, dose-related increase in mutant frequency, both in the absence and presence of liver microsomes. Activity in the presence of liver microsomes was similar to the activity of other senna preparations in this
TA 102.S9
TA10249 600
I
1
0
200 400 DOSE.pl/plate
I
TAlO2+S9
-
1
500
-
m
m
5 c
TA102rS9
600-
800500
200 460 Dose. pllplale
'
400-
5n
300-
E 300-
\
400-
v)
a
w
f;
3 U
5
200-
a
100-
0
2w 400 Dose. p p i a t e
200-
100-
O J , 0
,
,
200 4w Dose. pl/plaie
Fig. 4. Mutagenicity of senna extracts in Scr/morie//uryphimuriimi strain TA102 in the absence and presence of liver microsomal S9 fraction. Extracts were preparated as described in Materials and Methods and tested in the plate incorporation assay (Maron & Ames 1983). The volume of S9 protein was 100 pliplate. Control cultures were treated with 400 pI of buffer solution, prepared as described in Materials and Methods. Results are presented as mean number of revertant coloniesiplatef S.D. of 3 4 separate experiments, each performed with duplicate plates. P-values refer to the regression of revertant colonies/plate upon dose, r is the correlation coefficient. Unless otherwise indicated, P was
