- Email: [email protected]
Adsorption of a hydrophobic mutagen to dietary fibre from the skin and flesh of potato tubers
203
Mutation Research, 260 (1991) 203-213 © 1991 ElsevierSciencePublishers B.V. 0165-1218/91/$03.50 ADONIS 016512189100097P
MUTGEN 01660
Adsorption of a hydrophobic mutagen to dietary fibre from the skin and flesh of potato tubers Philip J. Harris 1, Anthony M. Roberton 2, H. John Hollands 1 and Lynnette R. Ferguson 3 Departments of 1 Botany and 2 Biochemistry and 3 Cancer Research LaboratorY, University of Aucklan~ Auckland (New Zealand)
(Received6 August 1990) (Revision received27 November1990) (Accepted 29 November 1990)
Keywords: Adsorption; Cell walls; Dietary fibre; 1,8-Dinitropyrene;Mutagen; Potato
Summary One of the theories to explain the protective action of some dietary fibres against colon cancer is that certain mutagens a n d / o r cancer promoters are adsorbed to these dietary fibres making the mutagens a n d / o r cancer promoters less available to gut mucosal cells. The abilities of 2 contrasting cell wall preparations (dietary fibre preparations) from potato tubers to adsorb in vitro the hydrophobic mutagen, 1,8-dinitropyrene (DNP), were studied using an incubation mixture containing D N P in phosphate-buffered saline (PBS). Walls from potato skins strongly adsorbed DNP and, at the highest wall concentration tested, only a small proportion of the D N P remained in solution. In marked contrast to the skin walls, potato flesh walls adsorbed only a small proportion of the DNP. Unexpectedly, the flesh walls also caused a large increase in the proportion of D N P found in solution. When flesh walls were pre-extracted with PBS, the ability of the extracted walls to bind D N P increased. The material extracted from the flesh walls was able to maintain D N P in solution, when added to the incubation medium in the absence of cell walls. Pectic polysaccharides appear to be the soluble component responsible for maintaining the D N P in solution. Competition between soluble and insoluble fibre components may have major implications for the availability and distribution of hydrophobic mutagens in the alimentary tract.
Much epidemiological evidence and many studies on experimental animals indicate that diets rich in dietary fibres reduce the risk of colon cancer (Burkitt, 1978; DeCosse et al., 1989). However, not all the epidemiological and experimental data support this protective role of dietary fibre
Correspondence: Dr. P.J. Harris, Department of Botany, University of Auckland, Private Bag, Auckland (New Zealand).
(Willet, 1989). Indeed, there are some studies which appear to show the converse effect (Jacobs and Lupton, 1986). A major reason for these conflicting data may be because the term dietary fibre includes a wide range of different materials. Dietary fibre consists mainly of plant cell walls and components obtained from these, and walls vary widely in structure and composition depending on cell type, plant species and stage of development (Bacic et al.,
204 1988). Thus, some cell walls may be much more effective in protecting against the development of colon cancer than others. One mechanism by which dietary fibre is thought to reduce the risk of colon cancer is that the dietary fibre may adsorb certain mutagens a n d / o r cancer promoters in the digestive tract. These are then carried out of the body adsorbed onto undigested dietary fibre. Thus the effective concentrations of these substances available to initiate or promote cancerous changes in the gut mucosal cells are lowered. Experiments in vitro have shown that certain mutagens and cancer promoters can adsorb to some dietary fibres (e.g., Smith-Barbaro et al., 1981; Barnes et al., 1983; Kada et al., 1984; Takeuchi et al., 1988; Roberton et al., 1990). Plant cell walls with contrasting chemical compositions and physical properties may adsorb particular mutagens to different extents. In this paper we report a study of the adsorption of the hydrophobic, environmental mutagen, 1,8-dinitropyrene (DNP) to 2 contrasting cell-wall preparations from potato. One preparation was from the skin of potato tubers and was rich in walls derived from cork cells of the periderm. The other preparation was from the flesh of potato tubers and contained mainly thin, unlignified walls derived from parenchyma cells.
out slowly (6 h per change) using 3:1, 1:1, and 1 : 3 v / v mixtures of ethanol and Technovit 7100 resin ( K y l z e r and Co. G m b H , D-6393 Wehrheim/Ts) with hardener I (benzyl peroxide) and finally in 100% Technovit 7100 resin with hardener I for 1 week. Polymerization was carried out at room temperature in fresh Technovit 7100 plus hardeners I and II (dimethyl sulphoxide (DMSO)). Sections (2-3 ~m) were cut with a Ralph knife on a LKB Historange microtome.
Microscopy
Materials and methods
Bright-field, Nomarski interference, and UV fluorescence microscopy were carried out using a Carl Zeiss research microscope (Oberkochen, F.R.G.) fitted with a 50-W tungsten lamp and a high-pressure mercury-vapour lamp (HBO 50). The microscope was also fitted with an incidence illuminator containing a G365 excitation filter, FT395 dichromatic beam splitter and LP420 bartier filter (Zeiss filter set 48 77 02). Unstained sections and isolated cell walls were examined by Nomarski interference and fluorescence microscopy. They were also examined by bright-field microscopy, either unstained or after staining with tohiidine blue (0.05% w / v ) in 20 m M sodium benzoate buffer (pH 4.4) (Feder and O'Brien, 1968). Starch was detected using a solution of iodine in potassium iodide (0.2 g iodine and 2 g potassium iodide in 100 ml of water) (Jensen, 1962).
Plant material
Isolation of cell walls from skins
Mature tubers of potato (Solanum tuberosum cv. Red King) were obtained from a local market.
Potato tubers were placed in a metal container, covered with liquid nitrogen, and removed when frozen. A stream of air was directed at them causing the skin to lift from the flesh. The skin was removed using forceps and stored at - 7 0 ° C until processed further. The skin (10 g) was then cut into pieces no larger than 5 m m x 5 mm. Batches of 0.5 g were homogenized for 3 rain in 20 mM sodium phosphate buffer (pH 7.0) containing 10 mM 2-mercaptoethanol at 4°C, using a specially constructed steel homogenizer (Harris, 1983). The homogenate was filtered on nylon mesh (pore size 10 /~m) and the residue washed with water until the filtrate was clear. The residue on the mesh was then resuspended in water and centrifuged (500 X g, 5 rain) and the supernatant dis-
Fixation, embedding and sectioning of potato tubers Fixation and embedding for microscopy were carried out at 4°C. Cubes (sides 2 mm long) of skin and adjacent flesh were fixed in a mixture of 2.5% (v/v) glutaraldehyde and 2% ( v / v ) formaldehyde in N a O H - P I P E S buffer (0.5 M, p H 7.3) (Salema and Brandao, 1973; Lawton and Harris, 1978) under vacuum for 20 rain, then left at atmospheric pressure for 20 h. They were then washed (3 x 30 min) in the buffer solution. Dehydration was carried out by washing with 2 changes (4 h each) of methyl ceUosolve, then with 2 changes (4 h each) of 100% ethanol. Infiltration was carried
205 carded. The pellet was washed by resuspending in water and recentrifuging as above until the supernatant was clear (8 × s). Starch grains contaminating the cell walls were removed in the following way. The pellet was divided into 2 equal parts. Each half of the pellet was suspended in water (200 nil) and heated in a boiling water bath for 5 rain. The suspension was then cooled rapidly to 37°C, porcine pancreatic a-amylase (282 units, type 1A, Sigma Chemical Co., St. Louis, MO) added, and the mixture incubated at 37°C for 60 rain. The 2 preparations were then sonicated (Model 380, Heat Systems-Ultrasonics, Farmingdale, NY) on ice for 1 min at full power using a 0.5 inch diameter probe. They were then centrifuged (500 × g, 5 rain) and the pellet washed onto nylon mesh (pore size 10 ~tm). The residue on the mesh was washed successively with water (2 1), 130 mM NaC1 (500 ml) and water (2 1). The residue was freeze-dried and stored over silica gel. Isolation of cell walls from flesh Tubers were washed, peeled, and the flesh cut into cubes (sides no longer than 2 mm). The cubes (100 g fresh weight) were homogenized at full speed for 12 min in 20 mM sodium phosphate buffer (300 ml, pH 7.0) containing 10 mM 2mercaptoethanol at 4°C using a Polytron homogenizer (Model PT10-35, Kinematica, CH-6010 Kriens-Luzern, Switzerland) fitted with a 12-mmdiameter generator. The homogenate was sonicated on ice for 5 min at full power and then divided into 4 equal volumes, each of which was filtered onto nylon mesh (pore size 350 #m) and washed with 2 1 of water. The residues were resuspended in water, combined and centrifuged (2000 × g, 30 min). The pellet was further homogenized (in batches of 7 g wet wt) for 2 rain using a specially constructed steel homogenizer in a mixer/mill (Model 8000, Spex, Edison, N J) (Harris, 1983). The homogenate was centrifuged (2000 × g, 20 rain), the pellet resuspended in 90% (v/v) DMSO (in water) and re-centrifuged (as above). The pellet was resuspended in fresh 90% (v/v) DMSO (200 ml), sonicated (5 min, full power) on ice, and stirred at room temperature for 30 min. The suspension was then centrifuged (as above), the pellet resuspended in fresh 90% DMSO (200 ml), and re-centrifuged (as above). The pellet was
washed with water onto nylon mesh (pore size 10 /xm). The residue on the mesh was washed with water (2 1), 130 mM sodium chloride (1 1), water (2 1) and then freeze-dried. Extraction of flesh walls with PBS at 100°C Walls (200 mg) were refluxed for 1 h in a solution of phosphate-buffered saline (PBS) (40 ml) containing 20 mM sodium phosphate buffer (pH 6.5) and 130 mM sodium chloride. The suspension was cooled rapidly to ambient temperature, centrifuged (1000 x g, 20 rain) at 4 ° C and the supernatant filtered through a 0.2-/tin-pore-size filter (AcroCap, Gelman Sciences, Ann Arbor, MI, U.S.A.) and then stored at - 2 0 ° C . The pellet was resuspended in fresh PBS (40 ml) and refluxed for 30 min. The suspension was cooled rapidly, centrifuged (as above), and the supernatant discarded. The pellet was again refluxed with PBS (40 ml), cooled and centrifuged (as above). The final pellet was washed (3 x s) by resuspension in water followed by centrifugation (as above), freeze dried, and stored over silica gel. Extraction of flesh walls with PBS at 4°C and 37°C Flesh walls (200 mg) were incubated for 1 h in PBS (40 ml) at 4°C. Another preparation was incubated at 37°C. Each suspension was centrifuged at 4°C (1000 x g, 20 rain) and the supernatant removed and filtered as described for the IO0°C extract. Incubation of walls with DNP in PBS All incubations were done in acid-washed, conical glass centrifuge tubes (capacity 12 ml). Wall preparations were weighed into the tubes and 1.96 ml of PBS was added 30 rain before the beginning of the experiment. At time zero, 40 /~1 of a solution of DNP in DMSO (100 ng DNP, final DMSO concentration 2% v/v) was added and the contents of the tubes mixed using a vortex mixer. The tubes were shaken (120 rev/min) on an orbital shaker for 55 min at 37°C, centrifuged (2500 x g, 2.5 rain), and aliquots (3 × 50 /~1) of the supernatants taken for the mutagenicity assay. Control incubations were also done with no walls or with a-cellulose (10 mg) (Roberton et al., 1990). The DNP adsorbed to the cell walls, a-cellulose and
206 test tube walls was recovered and the mutagenicity measured as previously described (Ferguson et al., 1990). The results were calculated as a percentage of the mutagenicity of the D N P added initially. The latter was determined by adding a solution of D N P (40 /~1 containing 100 ng) in D M S O to D M S O (1.96 ml), mixing and assaying the mutagenicity of aliquots (3 × 50 /tl). Data from a minimum of 2 experiments were averaged.
Incubation of wall extracts with DNP in PBS Wall extracts (1.96 ml) were used undiluted and after diluting (1 : 1, 1 : 3 and 1 : 7 v / v ) with PBS. D N P (100 ng) in D M S O (40/xl) was added to the wall extract in PBS at zero time and incubated as above. After 55 rain the tube was centrifuged (2500 × g, 2.5 rain) and allquots (3 × 50 ~1) removed for mutagenicity assay. D N P adsorbed to the test tube walls was measured as previously described (Ferguson et al., 1990) and the results calculated as described above. Chemical characterization of the IO0°C extract of flesh walls The extract was dialysed for 16 h at 4°C against water and freeze dried. The residue from 0.5 ml of extract (0.74 mg) was hydrolysed with 2 M trifluoroacetic acid (TFA) (0.5 ml) for 1 h at 121°C under argon (Albersheim et al., 1967). The hydrolysate was evaporated to dryness with a stream of air and the neutral monosaccharides were reduced and acetylated by a modification of the method of Blakeney et al. (1983) as described by Rae et al. (1985). Alditol acetates were separated and quantiffed on a 25-m-long, 0.33-mm-inner-diameter, BPX70, wall-coated open tubular (WCOT) fused silica column (film thickness 0.25 Fm) (SGE Pty., Melbourne, Vic., Australia) in a Hewlett-Packard (Avondale, PA, U.S.A.) 5890A gas chromatograph fitted with a flame-ionization detector and a dedicated cool on-column capillary inlet. Zero-grade helium (New Zealand Industrial Gases Ltd., Penrose, Auckland, New Zealand) was used as the carrier gas at a column head pressure of 40 kP a. The initial oven temperature, 38°C, was maintained for 30 sec following injection, then raised to 170°C at 50°C/rain, then to 230°C at 2 ° C / r a i n and kept at 230°C for 5 rnin. The detector temperature was held at 250°C.
Uronic acids in a hydrolysate of the extract were detected in the following way. Freeze-dried extract (2 mg) was hydrolysed with 2M T F A (0.5 ml) and evaporated to dryness as described above. Water (0.1 ml) was added to the dry hydrolysate and an aliquot (10 #1) subjected to thin-layer chromatography using cellulose plates (Merck, Darmstadt, F.R.G.) with b u t a n - l - o l : acetic acid: water ( 3 : 1 : 1 v / v ) followed by ethyl acetate: pyridine : water (10 : 4 : 3 v / v ) in the same direction. M o n o s a c c h a r i d e s were detected with aniline-hydrogen phthalate (Fry, 1988).
Bacterial mutagenicity assay The Salmonella typhimurium plate incorporation assay used strain TA98 as described by M a r o n and Ames (1983). In order to minimize day-to-day variability in tester strain sensitivity to mutagen, which is largely due to the exact growth phase of the culture, we grew cells for quantitative work as follows. A 1-ml vial (2 × 10 s cells) was removed from storage at - 8 0 ° C and inoculated into bacterial complete medium (20 ml). It was grown until a 1 in 10 dilution in fresh medium gave an absorbance of 0.11-0.12 at 654 nm, which took approximately 3 h. D N P levels in experiments were measured by assaying mutagenicity and each experimental point was performed in triplicate on at least 2 separate occasions. There is a linear relationship between
3000' 0.
e-.
o
2000"
o c
>
1000O J~
E z
0
Amount of DNP (ng/plate)
Fig. 1. The relationship between the amount of DNP (in DMSO) added to each plate of the bacterial mutagenicity assay and the number of revertant colonies. 3 separate calibration experiments were done and in each the experimental points were performed in triplicate. The bars represent standard errors between experiments.
207 the amount of D N P added to each plate and the number of revertant colonies over the dose range 0 - 5 ng (Fig. 1). Using this relationship we have converted the numbers of revertant colonies into amounts of DNP. The maximum amount of D N P added to each plate in experiments with cell walls was 2.5 ng. The background number of revertant colonies was approximately 65. Positive controls c o n t a i n i n g 4 - n i t r o - o - p h e n y l e n e d i a m i n e (100 /tg/plate) were included in all experiments to test the bacterial strain response. Results of experiments in which D N P was incubated with cell
Fig. 2. Transverse section of the skin and adjacent flesh of a potato tuber stained with toluidine blue and photographed using bright-field microscopy. The skin (periderm) consists mainly of cork cells (co). The flesh is composed of thin-walled parenchyma cells. Those in the outer cortex (oc), which is just beneath the skin, are smaller than those in the inner cortex (it). Bar = 100/~rn.
walls, a-cellulose or wail extracts were expressed as a percentage of the mutagenicity of the D N P added initially (see above). Results
Anatomy of the potato tuber and microscopy of the cell-wall preparations from the flesh and skin The majority of the tuber is composed of thinwalled parenchyma cells containing starch granules (Figs. 2 and 3A). The parenchyma cells just beneath the skin (outer cortex) are smaller and contain fewer starch granules than those further in (inner cortex and pith) (Fig. 2). The outer region of the tuber constitutes the skin or periderm tissue which consists mainly of cork cells (Fig. 2). These cells are flattened radially and are arranged in compact radial rows. Their walls fluoresced blue in UV radiation because of the presence of the polymer, suberin, which contains phenolic domains (Fig. 3B) (O'Brien and McCully, 1981; Kolattukudy and Espelie, 1985). The walls of the parenchyma cells did not fluoresce indicating the absence of lignin or suberin (Harris and Hartley, 1976; O'Brien and McCully, 1981). The cell-wall preparation from the flesh was composed mainly of wall fragments derived from parenchyma cells (Fig. 3C). These stained bright pink with toluidine blue which is consistent with the presence of polyanions, such as those present in the acidic pectic polysaccharides, rhanmogalacturonans (O'Brien and McCully, 1981). They did not fluoresce in UV radiation. Very occasional wall fragments derived from tracheary elements were also observed which stained blue with toluidine blue and fluoresced blue in UV radiation indicating the presence of lignin (Harris and Hartley, 1976). The cell-wall preparation from the skin was rich in wall fragments that fluoresced blue in UV radiation and were derived from the cork ceils of the periderm (Fig. 3D). These fragments stained weakly blue or purple with toluidine blue. The preparation also contained wall fragments that did not fluoresce in UV radiation and stained bright pink with toluidine blue. These were probably derived from the parenchyma cells beneath the periderm. N o starch granules were detected in either cell-wail preparation.
209
1 O0 ,
100" A
0__.__0/O~C -o o n 7 cl
/
6O
0
o Q.. :7
40
60
.A
zx~~ _~__(
40
20 o~
0
10
20
30
:
L
40
50
Cell wall concn. ( r n g / 2 m l ) Fig. 4 The effect of increasing concentrations of potato skin walls on the proportion of D N P in solution. D N P (100 n g / 2 ml) was incubated in PBS with various concentrations of skin walls. Samples were taken after 55 min for mutagenicity assays. Values represent total numbers of revertant colonies (minus those for negative controls) expressed as a percentage of the n u m b e r s seen for a comparable sample of D N P diluted into D M S O and plated immediately. Mutagenic activity present in the supernatant (A), associated with the potato skin walls (o), and adsorbed to the tube walls (I-q).
Effect of increasing concentrations of skin walls on the distribution of DNP Fig. 4 shows the DNP present in the supernatant after incubating DNP (50 ng/ml) in PBS with a range of concentrations (1-50 m g/ 2 ml) of skin walls. The supernatant contained only a small proportion of the DNP added. In the absence of walls, 35% of the DNP added was in the supernatant after 1 h. The presence of 1 mg of walls decreased this to less than 25 %, and this decreased to only 2% in the presence of 50 mg of skin walls. The proportion of the DNP associated with the skin walls increased with the amount of wall added (Fig. 4), and accounted for approximately
/
H
0
t 10
0(I)
~ I 20
N
~ ,30
L 40
[ 50
Cell wall c o n c n ( m g / 2 m l ) Fig. 5. The effect of increasing concentrations of potato flesh walls on the proportion of D N P in solution. The experiment was carried out similarly to that described in Fig. 4, except flesh walls were used instead of the skin walls. Mutagenic activity present in the supernatant (A), associated with the potato flesh walls (o), and adsorbed to the tube walls ([2).
78% of the DNP added in the presence of 10 mg of walls and approximately 90% with 50 mg of walls. Only in the absence of walls was a significant proportion of the DNP associated with the glass walls of the tubes (Fig. 4). The results of a parallel experiment carried out with a-cellulose showed that 10 mg of a-cellulose adsorbed 67% of the DNP and 50 mg of a-cdlulose adsorbed 75% of the DNP (data not shown).
Effect of increasing concentrations of flesh walls on the distribution of DNP In marked contrast to the results of the above experiment with the skin walls and with a-cellulose (used as a positive control), most of the DNP added remained in the supernatant with flesh walls. Fig. 5 shows the effects of incubating a
Fig. 3. (A) Transverse section of the inner cortex region of the flesh of a potato tuber. The section was unstained a n d photographed using Nomarski interference optics. It shows thin-walled parenchyma cells containing large starch granules (arrow). Bar = 50 lam. (B) Transverse section of the skin and adjacent flesh of a potato tuber photographed using U V fluorescence microscopy. The walls of the cork cells (cc) of the skin (periderm) fluoresce, whereas the walls of the parenchyma cells of the flesh just beneath the skin (outer cortex) (oc) do not. Bar = 50 #m. (C) Bright-field mierograph of the cell-wail preparation from the flesh of a potato tuber stained with tohiidine blue. The preparation consists mainly of wall fragments derived from parenchyma cells. Bar = 50 # m . (D) UV fluorescence micrograph of the cell-wall preparation from the skin of a potato tuber. The preparation was rich in wall fragments derived from cork cells which fluoresce in UV radiation. Bar = 50 # m .
210 range of concentrations (1-50 m g / 2 ml) of flesh walls with D N P (50 n g / m l ) in PBS. After incubation with 10 mg walls, more than 80% of the D N P added was recovered in the supernatant, whereas the control (in the absence of walls) contained only approximately 20% of the D N P added. Incubation with increasing concentrations of walls led to the progressive loss of D N P from the supernatant. However, even in the presence of 50 m g / 2 ml walls, more than 40% of the D N P added was recovered in the supernatant. Some D N P was associated with the walls (Fig. 5) and this increased with increasing concentrations of walls, but accounted for only approximately 40% of the D N P added, even at the highest concentration of walls. Even smaller proportions of the D N P added were associated with the glass walls of the tubes (Fig. 5).
Effect of extracted flesh walls on the distribution of DNP In contrast to the results with the unextracted flesh walls, a smaller proportion of the D N P added remained in the supernatant with flesh walls that had been extracted with PBS at 100°C. Fig. 6A shows the effects on D N P content of incubating a range of concentrations (5-30 m g / 2 ml) of extracted flesh walls with PBS containing D N P (50 ng/ml). After incubation with 5 mg of extracted walls, the proportion of the D N P added that was recovered in the supernatant was similar to the control without walls (approximately 20%). Incubation with increasing concentrations of extracted walls led to the progressive loss of D N P in the supernatant. More D N P was associated with the extracted than the unextracted flesh walls (Fig. 6B). At all cell-wall concentrations, only small amounts of D N P were also associated with the glass walls of the tubes (Fig. 6C).
100
8O "o
o O_ Z
"6
60 40 20, 0
I 10
0
I 20
,30
Cell wall concn ( r n g / 2 m l )
I O0
8O "o "o
"o 60 ¸ o CL z a
,._
40
0
20, o
OI 0
O
I
I
10
20
30
Cell wall concn ( m g / 2 m l )
I°°|C
"o
60'
(3_ Z
40" Fig. 6. The effect of extracting potato flesh walls at 100°C on their ability to adsorb DNP. The experiment was carried out similarly to that described in Fig. 5. Mutagenic activity (A) present in the supernatant (a, z~), (B) associated with flesh walls (e, o), and (C) adsorbed to the tube walls (11, O), when extracted walls were used (closed symbols) as compared to unextracted wails (open symbols).
20'
0
10
20
Cell wall concn ( m g / 2 m l )
3O
211 100
galacturonic acid (Ring and Selvendran, 1978; Jarvis et al., 1981). Discussion
"~6 40.
A~ ~
20. /
0
20
40
60
80
1O0
extract
Fig. 7. The effect of the 100°C extract from potato flesh walls on the concentration of DNP in solution. DNP (100 ng/2 nil) was incubated in PBS with various dilutions of the extract. Samples were taken after 55 rain for mutagenicity assays. Mutagenic activity present in the supernatant (-), and adsorbed to the tube walls (I).
The effect of 100°C extract of flesh walls on the retention of DNP in solution In the absence of extract, less than 20% of the D N P remained in the PBS after incubation for 1 h (Fig. 7). However, after incubation for 1 h in the presence of the 100°C extract, the proportion of D N P maintained in solution increased with increasing amount of extract (Fig. 7). In the presence of undiluted extract, more than 60% of the D N P added remained in solution. The proportion of D N P associated with the glass walls of the tubes decreased with increasing amount of extract (Fig. 7). Wall extracts obtained with PBS at 4 and 37°C had a similar, but lesser effect (data not shown). Chemical characterization of the IO0°C extract of flesh walls The neutral monosaccharide composition of hydrolysates of the extract was (average % w / w of 2 analyses): rhamnose, 4.8%; arabinose, 8.1%; xylose, 1.0%; mannose, 1.3%; galactose, 84.1%; glucose, 0.8%. Thin-layer chromatography of a hydrolysate of the extract showed a spot with the same mobility as galacturonic and glucuronic acids which co-chromatograph. From previous work on the polysaccharides of the walls of potato flesh, it is likely that the spot was composed mainly of
The ability of different plant cell walls (dietary fibres) to adsorb mutagens (as tumour initiators) needs to be quantified in order to predict their effectiveness in removing mutagens from the digestive tract and thus protecting against colon cancer. In a previous in vitro study (Roberton et al., 1990), we showed that the hydrophobic, environmental mutagen, 1,8-dinitropyrene (DNP), adsorbed to a model fibre preparation (a-cellulose). In the present study we have examined the adsorptive properties of 2 contrasting cell-wall preparations from the skin and flesh of potatoes. Each preparation was rich in a cell-wall type that is common in Western diets. The skin walls reduced the proportion of D N P in solution as was previously observed for the model fibre, a-cellulose (Roberton et al., 1990). However, to our surprise, we found that the flesh walls increased the proportion of D N P in solution relative to the control without cell walls. The walls of the cork cells of potato skins contain the complex hydrophobic polymer, suberin, and associated waxes (Espelie et al., 1980). Suberin consists of aliphatic polyester domains covalently attached to phenolic domains (Kolattukudy and Espelie, 1985). Because of the presence of these hydrophobic components, we would predict that the hydrophobic mutagen, DNP, would strongly adsorb to a preparation rich in these walls. This proved to be the case. We found that a co-mutagenic effect was not responsible for the increased mutagenic activity of the D N P solution in the presence of flesh walls. In fact, the effect was caused by a PBS-soluble fraction released from the walls maintaining the D N P in solution. This soluble fraction could be extracted from the walls and when incubated with D N P in the absence of walls, it maintained the D N P in solution. The flesh cell walls of potato tubers are known to contain a high proportion of pectic polysaceharides (Ring and Selvendran, 1978; Jarvis et al., 1981). Pectic polysaccharides are a complex class of wall polysaccharides, comprising a family of acidic polysaccharides (rham-
212 nogalacturonans) and several neutral polysaccharides (arabinans, galactans, and arabinogalactans) (Bacic et al., 1988). They can be partially extracted from walls with water and our analyses of the 100°C extract of flesh cell walls with PBS are consistent with the presence of rhamnogalacturonans and neutral pectic polysaccharides, mainly consisting of galactans. This predominance of galactans in the neutral pectic polysaccharides of the flesh cell walls of potato tubers has been reported previously (Ring and Selvendran, 1978; Jarvis et al., 1981). In our experiments, the pectic polysaccharides in the PBS extract appear to be responsible for maintaining the D N P in solution. The finding that a hydrophobic molecule, such as DNP, can be maintained in solution by hydrophilic molecules, such as pectic polysaccharides, is surprising. It would appear to break a fundamental physico-chemical rule. One possible explanation is the mechanism proposed by Wiggins and van Ryn (1990) to account for how a hydrophilic molecule, provided that it is highly charged, can maintain hydrophobic molecules in solution. These authors have shown that in the water surrounding highly charged molecules the water-water hydrogen bonding is weaker than in normal water and hydrophobic molecules selectively accumulate into that environment. The acidic rhamnogalacturonans could theoretically act in this way to maintain D N P in solution. Even after removing the PBS-soluble components of the potato flesh walls, their ability to adsorb DNP was much less than the skin walls or the model fibre, a-cellulose. These flesh walls are derived from parenchyma cells and walls of this cell type from dicotyledons that have been chemically analysed have been found to have broadly similar compositions (Bacic et al., 1988). As this cell type is very common in many dicotyledonous fruits and vegetables in Western diets, the effect which we have described may be widespread. Potato tubers are usually cooked in one of a variety of ways before they are eaten, however, cooking makes the isolation of cell walls much more difficult and therefore we used cell walls isolated from raw tubers. We would expect that boiling peeled potatoes would result in the loss of some pectic polysaccharides into the cooking water, especially from the surface parenchyma
cells. However, within the flesh, although pectic polysaccharides may be partially released from the walls, the presence of surrounding cells should prevent their loss into the cooking water. Refluxing of isolated flesh cell wails in PBS for 1 h, as described in the present study, is a much more drastic treatment than would be encountered in the regular boiling of potatoes and even after this PBS treatment the walls adsorbed D N P only weakly. Boiling will probably have httle effect on the ability of skin cell wails to adsorb D N P as we found that refluxing isolated skin walls in PBS for 1 h increased their ability to adsorb D N P by only 9% (data not shown). The release of the soluble fraction from parenchyma cell walls, the adsorptive properties of the insoluble components of the cell walls, as well as the digestion of both of these by bacteria in the colon will all affect the distribution of hydrophobic components, such as DNP, in the intestines. Whether the adsorption of hydrophobic mutagens by the soluble fraction from the parenchyma walls will increase or decrease the availability of the mutagens to the gut mucosal cells is unknown. Therefore, parenchyma walls of dicotyledons, as exemplified by the potato, may not necessarily protect against colon cancer. These findings may have important consequences for protection against colon cancer in Western populations.
Acknowledgements This work was supported by a grant from the Auckland Medical Research Foundation. We thank the Auckland University Research Committee for equipment grants to P.J.H. to purchase the sonicator and mixer/mill, and the Provincial Grand Lodge (Irish Constitution) for an equipment grant to A.M.R. to purchase a shaking water bath. Jude Antonimuttu, Marlene F. Kendon, Robert J. McKenzie, and Dr. Julie B. White provided valuable assistance in this work.
References Albersheim, P., D.J. Nevins, P.D. English and A. Karr (1967) A method for the analysis of sugars in plant cell-wall polysaccharidesby gas-liquid chromatography,Carbohydr. Res., 5, 340-345.
213 Bacic, A., P.J. Harris and B.A. Stone (1988) Structure and function of plant cell walls, in: J. Preiss (Ed.), The Biochemistry of Plants, Vol. 14, Carbohydrates, Academic Press, San Diego, CA, pp. 297-371. Barnes, W.S., J. Maiello and J.H. Weisburger (1983) In vitro binding of the food mutagen 2-amino-3-methylimidazo[4,5-f]quinoline to dietary fibers, J. Natl. Cancer Inst., 70, 757-760. Blakeney, A.B., P.J. Harris, R.J. Henry and B.A. Stone (1983) A simple and rapid preparation of alditol acetates for monosaccharide analysis, Carbohydr. Res., 113, 291-299. Burkitt, D.P. (1978) Colonic-rectal cancer: tiber and other dietary factors, Am. J. Clin. Nutr., 31, $58-$64. DeCosse, J.J., H.M. Miller and M.L. Lesser (1989) Effect of wheat fiber and vitamins C and E on rectal polyps in patients with familial adenomatous polyposis, J. Natl. Cancer Inst. 81, 1290-1297. Espelie, K.E., N.Z. Sadek and P.E. Kolattukudy (1980) Composition of suberin-associated waxes from the subterranean storage organs of seven plants - parsnip, carrot, rutabaga, turnip, red beet, sweet potato and potato, Planta, 148, 468-476. Feder, N., and T.P. O'Brien (1968) Plant microtechniques: some principles and new methods, Am. J. Bot., 55, 123-142. Ferguson, L.R., P.J. Harris, H.J. Hollands and A.M. Roberton (1990) Effects of bile salts on the adsorption of a hydrophobic mutagen to dietary fiber, Mutation Res., 245, 111117. Fry, S.C. (1988) The growing plant cell wall: chemical and metabolic analysis. Longman Scientific and Technical, Harlow. Harris, P.J. (1983) Cell walls, in: J.L. Hall and A.L. Moore (Eds.), Isolation of Membranes and Organdies from Plant Cells, Academic Press, London, pp. 25-53. Harris, P.J., and R.D. Hartley (1976) Detection of bound ferulic acid in cell walls of the Gramineae by ultraviolet fluorescence microscopy, Nature (London), 259, 508-510. Jacobs, L.R., and J.R. Lupton (1986) Relationship between colonic luminal pH, cell proliferation, and colon carcinogenesis in 1,2-dimethylhydrazine treated rats fed high fiber diets, Cancer Res., 46, 1727-1734. Jarvis, M.C., M.A. Hall, D.R. Threlfall and J. Friend (1981) The polysaccharide structure of potato cell walls: chemical fractionation, Planta, 152, 93-100.
Jensen, W.A. (1962) Botanical Histochemistry, Freeman, San Francisco, CA. Kada, T., M. Kato, K. Aikawa and S. Kiriyama (1984) Adsorption of pyrolysate mutagens by vegetable fibers, Mutation Res., 141, 149-152. Kolattukudy, P.E., and K.E. Espelie (1985) Biosynthesis of cutin, suberin, and associated waxes, in: T. Higuchi (Ed.), Biosynthesis and Biodegradation of Wood Components, Academic Press, Orlando, CA, pp. 161-207. Lawton, J.R., and P.J. Harris (1978) Fixation of senescing plant tissues: sclerenchymatous fibre cells from the flowering stem of a grass, J. Microsc., 112, 307-318. Maron, D.M., and B.N. Ames (1983) Revised methods for the Salmonella mutagenicity test, Mutation Res., 113, 173-215. O'Brien, T.P., and M.E. McCully (1981) The study of plant structure: principles and selected methods, Termarcarphi Pty., Melbourne. Rae, A.L., P.J. Harris, A. Bacic and A.E. Clarke (1985) Composition of the cell walls of Nicotiana alata Link et Otto pollen tubes, Planta, 166, 128-133. Ring, S.G., and R.R. Selvendran (1978) Purification and methylation analysis of cell wall material from Solanum tuberosum, Phytochemistry, 17, 745-752. Roberton, A.M., P.J. Harris, H.J. Hollands and L.R. Ferguson (1990) A model system for studying the adsorption of a hydrophobic mutagen to dietary fiber, Mutation Res., 244, 173-178. Salema, R., and I. Brandao (1973) The use of PIPES buffer in the fixation of plant cells for electron microscopy, J. Submicrosc. Cytol., 5, 79-96. Smith-Barbaro, P., D. Hanson and B.S. Reddy (1981) Carcinogen binding to various types of dietary fiber, J. Natl. Cancer Inst., 67, 495-497. Takeuchi, M., M. Hara, T. Inoue and T. Kada (1988) Adsorption of mutagens by refined corn bran, Mutation Res., 204, 263-267. Wiggins, P.M., and R.T. van Ryn (1990) Changes in ion selectivity with changes in density of water in gels and cells, Biophysica, in press. Willett, W. (1989) The search for the causes of breast and colon cancer, Nature (London), 338, 389-394.
Mutation Research, 260 (1991) 203-213 © 1991 ElsevierSciencePublishers B.V. 0165-1218/91/$03.50 ADONIS 016512189100097P
MUTGEN 01660
Adsorption of a hydrophobic mutagen to dietary fibre from the skin and flesh of potato tubers Philip J. Harris 1, Anthony M. Roberton 2, H. John Hollands 1 and Lynnette R. Ferguson 3 Departments of 1 Botany and 2 Biochemistry and 3 Cancer Research LaboratorY, University of Aucklan~ Auckland (New Zealand)
(Received6 August 1990) (Revision received27 November1990) (Accepted 29 November 1990)
Keywords: Adsorption; Cell walls; Dietary fibre; 1,8-Dinitropyrene;Mutagen; Potato
Summary One of the theories to explain the protective action of some dietary fibres against colon cancer is that certain mutagens a n d / o r cancer promoters are adsorbed to these dietary fibres making the mutagens a n d / o r cancer promoters less available to gut mucosal cells. The abilities of 2 contrasting cell wall preparations (dietary fibre preparations) from potato tubers to adsorb in vitro the hydrophobic mutagen, 1,8-dinitropyrene (DNP), were studied using an incubation mixture containing D N P in phosphate-buffered saline (PBS). Walls from potato skins strongly adsorbed DNP and, at the highest wall concentration tested, only a small proportion of the D N P remained in solution. In marked contrast to the skin walls, potato flesh walls adsorbed only a small proportion of the DNP. Unexpectedly, the flesh walls also caused a large increase in the proportion of D N P found in solution. When flesh walls were pre-extracted with PBS, the ability of the extracted walls to bind D N P increased. The material extracted from the flesh walls was able to maintain D N P in solution, when added to the incubation medium in the absence of cell walls. Pectic polysaccharides appear to be the soluble component responsible for maintaining the D N P in solution. Competition between soluble and insoluble fibre components may have major implications for the availability and distribution of hydrophobic mutagens in the alimentary tract.
Much epidemiological evidence and many studies on experimental animals indicate that diets rich in dietary fibres reduce the risk of colon cancer (Burkitt, 1978; DeCosse et al., 1989). However, not all the epidemiological and experimental data support this protective role of dietary fibre
Correspondence: Dr. P.J. Harris, Department of Botany, University of Auckland, Private Bag, Auckland (New Zealand).
(Willet, 1989). Indeed, there are some studies which appear to show the converse effect (Jacobs and Lupton, 1986). A major reason for these conflicting data may be because the term dietary fibre includes a wide range of different materials. Dietary fibre consists mainly of plant cell walls and components obtained from these, and walls vary widely in structure and composition depending on cell type, plant species and stage of development (Bacic et al.,
204 1988). Thus, some cell walls may be much more effective in protecting against the development of colon cancer than others. One mechanism by which dietary fibre is thought to reduce the risk of colon cancer is that the dietary fibre may adsorb certain mutagens a n d / o r cancer promoters in the digestive tract. These are then carried out of the body adsorbed onto undigested dietary fibre. Thus the effective concentrations of these substances available to initiate or promote cancerous changes in the gut mucosal cells are lowered. Experiments in vitro have shown that certain mutagens and cancer promoters can adsorb to some dietary fibres (e.g., Smith-Barbaro et al., 1981; Barnes et al., 1983; Kada et al., 1984; Takeuchi et al., 1988; Roberton et al., 1990). Plant cell walls with contrasting chemical compositions and physical properties may adsorb particular mutagens to different extents. In this paper we report a study of the adsorption of the hydrophobic, environmental mutagen, 1,8-dinitropyrene (DNP) to 2 contrasting cell-wall preparations from potato. One preparation was from the skin of potato tubers and was rich in walls derived from cork cells of the periderm. The other preparation was from the flesh of potato tubers and contained mainly thin, unlignified walls derived from parenchyma cells.
out slowly (6 h per change) using 3:1, 1:1, and 1 : 3 v / v mixtures of ethanol and Technovit 7100 resin ( K y l z e r and Co. G m b H , D-6393 Wehrheim/Ts) with hardener I (benzyl peroxide) and finally in 100% Technovit 7100 resin with hardener I for 1 week. Polymerization was carried out at room temperature in fresh Technovit 7100 plus hardeners I and II (dimethyl sulphoxide (DMSO)). Sections (2-3 ~m) were cut with a Ralph knife on a LKB Historange microtome.
Microscopy
Materials and methods
Bright-field, Nomarski interference, and UV fluorescence microscopy were carried out using a Carl Zeiss research microscope (Oberkochen, F.R.G.) fitted with a 50-W tungsten lamp and a high-pressure mercury-vapour lamp (HBO 50). The microscope was also fitted with an incidence illuminator containing a G365 excitation filter, FT395 dichromatic beam splitter and LP420 bartier filter (Zeiss filter set 48 77 02). Unstained sections and isolated cell walls were examined by Nomarski interference and fluorescence microscopy. They were also examined by bright-field microscopy, either unstained or after staining with tohiidine blue (0.05% w / v ) in 20 m M sodium benzoate buffer (pH 4.4) (Feder and O'Brien, 1968). Starch was detected using a solution of iodine in potassium iodide (0.2 g iodine and 2 g potassium iodide in 100 ml of water) (Jensen, 1962).
Plant material
Isolation of cell walls from skins
Mature tubers of potato (Solanum tuberosum cv. Red King) were obtained from a local market.
Potato tubers were placed in a metal container, covered with liquid nitrogen, and removed when frozen. A stream of air was directed at them causing the skin to lift from the flesh. The skin was removed using forceps and stored at - 7 0 ° C until processed further. The skin (10 g) was then cut into pieces no larger than 5 m m x 5 mm. Batches of 0.5 g were homogenized for 3 rain in 20 mM sodium phosphate buffer (pH 7.0) containing 10 mM 2-mercaptoethanol at 4°C, using a specially constructed steel homogenizer (Harris, 1983). The homogenate was filtered on nylon mesh (pore size 10 /~m) and the residue washed with water until the filtrate was clear. The residue on the mesh was then resuspended in water and centrifuged (500 X g, 5 rain) and the supernatant dis-
Fixation, embedding and sectioning of potato tubers Fixation and embedding for microscopy were carried out at 4°C. Cubes (sides 2 mm long) of skin and adjacent flesh were fixed in a mixture of 2.5% (v/v) glutaraldehyde and 2% ( v / v ) formaldehyde in N a O H - P I P E S buffer (0.5 M, p H 7.3) (Salema and Brandao, 1973; Lawton and Harris, 1978) under vacuum for 20 rain, then left at atmospheric pressure for 20 h. They were then washed (3 x 30 min) in the buffer solution. Dehydration was carried out by washing with 2 changes (4 h each) of methyl ceUosolve, then with 2 changes (4 h each) of 100% ethanol. Infiltration was carried
205 carded. The pellet was washed by resuspending in water and recentrifuging as above until the supernatant was clear (8 × s). Starch grains contaminating the cell walls were removed in the following way. The pellet was divided into 2 equal parts. Each half of the pellet was suspended in water (200 nil) and heated in a boiling water bath for 5 rain. The suspension was then cooled rapidly to 37°C, porcine pancreatic a-amylase (282 units, type 1A, Sigma Chemical Co., St. Louis, MO) added, and the mixture incubated at 37°C for 60 rain. The 2 preparations were then sonicated (Model 380, Heat Systems-Ultrasonics, Farmingdale, NY) on ice for 1 min at full power using a 0.5 inch diameter probe. They were then centrifuged (500 × g, 5 rain) and the pellet washed onto nylon mesh (pore size 10 ~tm). The residue on the mesh was washed successively with water (2 1), 130 mM NaC1 (500 ml) and water (2 1). The residue was freeze-dried and stored over silica gel. Isolation of cell walls from flesh Tubers were washed, peeled, and the flesh cut into cubes (sides no longer than 2 mm). The cubes (100 g fresh weight) were homogenized at full speed for 12 min in 20 mM sodium phosphate buffer (300 ml, pH 7.0) containing 10 mM 2mercaptoethanol at 4°C using a Polytron homogenizer (Model PT10-35, Kinematica, CH-6010 Kriens-Luzern, Switzerland) fitted with a 12-mmdiameter generator. The homogenate was sonicated on ice for 5 min at full power and then divided into 4 equal volumes, each of which was filtered onto nylon mesh (pore size 350 #m) and washed with 2 1 of water. The residues were resuspended in water, combined and centrifuged (2000 × g, 30 min). The pellet was further homogenized (in batches of 7 g wet wt) for 2 rain using a specially constructed steel homogenizer in a mixer/mill (Model 8000, Spex, Edison, N J) (Harris, 1983). The homogenate was centrifuged (2000 × g, 20 rain), the pellet resuspended in 90% (v/v) DMSO (in water) and re-centrifuged (as above). The pellet was resuspended in fresh 90% (v/v) DMSO (200 ml), sonicated (5 min, full power) on ice, and stirred at room temperature for 30 min. The suspension was then centrifuged (as above), the pellet resuspended in fresh 90% DMSO (200 ml), and re-centrifuged (as above). The pellet was
washed with water onto nylon mesh (pore size 10 /xm). The residue on the mesh was washed with water (2 1), 130 mM sodium chloride (1 1), water (2 1) and then freeze-dried. Extraction of flesh walls with PBS at 100°C Walls (200 mg) were refluxed for 1 h in a solution of phosphate-buffered saline (PBS) (40 ml) containing 20 mM sodium phosphate buffer (pH 6.5) and 130 mM sodium chloride. The suspension was cooled rapidly to ambient temperature, centrifuged (1000 x g, 20 rain) at 4 ° C and the supernatant filtered through a 0.2-/tin-pore-size filter (AcroCap, Gelman Sciences, Ann Arbor, MI, U.S.A.) and then stored at - 2 0 ° C . The pellet was resuspended in fresh PBS (40 ml) and refluxed for 30 min. The suspension was cooled rapidly, centrifuged (as above), and the supernatant discarded. The pellet was again refluxed with PBS (40 ml), cooled and centrifuged (as above). The final pellet was washed (3 x s) by resuspension in water followed by centrifugation (as above), freeze dried, and stored over silica gel. Extraction of flesh walls with PBS at 4°C and 37°C Flesh walls (200 mg) were incubated for 1 h in PBS (40 ml) at 4°C. Another preparation was incubated at 37°C. Each suspension was centrifuged at 4°C (1000 x g, 20 rain) and the supernatant removed and filtered as described for the IO0°C extract. Incubation of walls with DNP in PBS All incubations were done in acid-washed, conical glass centrifuge tubes (capacity 12 ml). Wall preparations were weighed into the tubes and 1.96 ml of PBS was added 30 rain before the beginning of the experiment. At time zero, 40 /~1 of a solution of DNP in DMSO (100 ng DNP, final DMSO concentration 2% v/v) was added and the contents of the tubes mixed using a vortex mixer. The tubes were shaken (120 rev/min) on an orbital shaker for 55 min at 37°C, centrifuged (2500 x g, 2.5 rain), and aliquots (3 × 50 /~1) of the supernatants taken for the mutagenicity assay. Control incubations were also done with no walls or with a-cellulose (10 mg) (Roberton et al., 1990). The DNP adsorbed to the cell walls, a-cellulose and
206 test tube walls was recovered and the mutagenicity measured as previously described (Ferguson et al., 1990). The results were calculated as a percentage of the mutagenicity of the D N P added initially. The latter was determined by adding a solution of D N P (40 /~1 containing 100 ng) in D M S O to D M S O (1.96 ml), mixing and assaying the mutagenicity of aliquots (3 × 50 /tl). Data from a minimum of 2 experiments were averaged.
Incubation of wall extracts with DNP in PBS Wall extracts (1.96 ml) were used undiluted and after diluting (1 : 1, 1 : 3 and 1 : 7 v / v ) with PBS. D N P (100 ng) in D M S O (40/xl) was added to the wall extract in PBS at zero time and incubated as above. After 55 rain the tube was centrifuged (2500 × g, 2.5 rain) and allquots (3 × 50 ~1) removed for mutagenicity assay. D N P adsorbed to the test tube walls was measured as previously described (Ferguson et al., 1990) and the results calculated as described above. Chemical characterization of the IO0°C extract of flesh walls The extract was dialysed for 16 h at 4°C against water and freeze dried. The residue from 0.5 ml of extract (0.74 mg) was hydrolysed with 2 M trifluoroacetic acid (TFA) (0.5 ml) for 1 h at 121°C under argon (Albersheim et al., 1967). The hydrolysate was evaporated to dryness with a stream of air and the neutral monosaccharides were reduced and acetylated by a modification of the method of Blakeney et al. (1983) as described by Rae et al. (1985). Alditol acetates were separated and quantiffed on a 25-m-long, 0.33-mm-inner-diameter, BPX70, wall-coated open tubular (WCOT) fused silica column (film thickness 0.25 Fm) (SGE Pty., Melbourne, Vic., Australia) in a Hewlett-Packard (Avondale, PA, U.S.A.) 5890A gas chromatograph fitted with a flame-ionization detector and a dedicated cool on-column capillary inlet. Zero-grade helium (New Zealand Industrial Gases Ltd., Penrose, Auckland, New Zealand) was used as the carrier gas at a column head pressure of 40 kP a. The initial oven temperature, 38°C, was maintained for 30 sec following injection, then raised to 170°C at 50°C/rain, then to 230°C at 2 ° C / r a i n and kept at 230°C for 5 rnin. The detector temperature was held at 250°C.
Uronic acids in a hydrolysate of the extract were detected in the following way. Freeze-dried extract (2 mg) was hydrolysed with 2M T F A (0.5 ml) and evaporated to dryness as described above. Water (0.1 ml) was added to the dry hydrolysate and an aliquot (10 #1) subjected to thin-layer chromatography using cellulose plates (Merck, Darmstadt, F.R.G.) with b u t a n - l - o l : acetic acid: water ( 3 : 1 : 1 v / v ) followed by ethyl acetate: pyridine : water (10 : 4 : 3 v / v ) in the same direction. M o n o s a c c h a r i d e s were detected with aniline-hydrogen phthalate (Fry, 1988).
Bacterial mutagenicity assay The Salmonella typhimurium plate incorporation assay used strain TA98 as described by M a r o n and Ames (1983). In order to minimize day-to-day variability in tester strain sensitivity to mutagen, which is largely due to the exact growth phase of the culture, we grew cells for quantitative work as follows. A 1-ml vial (2 × 10 s cells) was removed from storage at - 8 0 ° C and inoculated into bacterial complete medium (20 ml). It was grown until a 1 in 10 dilution in fresh medium gave an absorbance of 0.11-0.12 at 654 nm, which took approximately 3 h. D N P levels in experiments were measured by assaying mutagenicity and each experimental point was performed in triplicate on at least 2 separate occasions. There is a linear relationship between
3000' 0.
e-.
o
2000"
o c
>
1000O J~
E z
0
Amount of DNP (ng/plate)
Fig. 1. The relationship between the amount of DNP (in DMSO) added to each plate of the bacterial mutagenicity assay and the number of revertant colonies. 3 separate calibration experiments were done and in each the experimental points were performed in triplicate. The bars represent standard errors between experiments.
207 the amount of D N P added to each plate and the number of revertant colonies over the dose range 0 - 5 ng (Fig. 1). Using this relationship we have converted the numbers of revertant colonies into amounts of DNP. The maximum amount of D N P added to each plate in experiments with cell walls was 2.5 ng. The background number of revertant colonies was approximately 65. Positive controls c o n t a i n i n g 4 - n i t r o - o - p h e n y l e n e d i a m i n e (100 /tg/plate) were included in all experiments to test the bacterial strain response. Results of experiments in which D N P was incubated with cell
Fig. 2. Transverse section of the skin and adjacent flesh of a potato tuber stained with toluidine blue and photographed using bright-field microscopy. The skin (periderm) consists mainly of cork cells (co). The flesh is composed of thin-walled parenchyma cells. Those in the outer cortex (oc), which is just beneath the skin, are smaller than those in the inner cortex (it). Bar = 100/~rn.
walls, a-cellulose or wail extracts were expressed as a percentage of the mutagenicity of the D N P added initially (see above). Results
Anatomy of the potato tuber and microscopy of the cell-wall preparations from the flesh and skin The majority of the tuber is composed of thinwalled parenchyma cells containing starch granules (Figs. 2 and 3A). The parenchyma cells just beneath the skin (outer cortex) are smaller and contain fewer starch granules than those further in (inner cortex and pith) (Fig. 2). The outer region of the tuber constitutes the skin or periderm tissue which consists mainly of cork cells (Fig. 2). These cells are flattened radially and are arranged in compact radial rows. Their walls fluoresced blue in UV radiation because of the presence of the polymer, suberin, which contains phenolic domains (Fig. 3B) (O'Brien and McCully, 1981; Kolattukudy and Espelie, 1985). The walls of the parenchyma cells did not fluoresce indicating the absence of lignin or suberin (Harris and Hartley, 1976; O'Brien and McCully, 1981). The cell-wall preparation from the flesh was composed mainly of wall fragments derived from parenchyma cells (Fig. 3C). These stained bright pink with toluidine blue which is consistent with the presence of polyanions, such as those present in the acidic pectic polysaccharides, rhanmogalacturonans (O'Brien and McCully, 1981). They did not fluoresce in UV radiation. Very occasional wall fragments derived from tracheary elements were also observed which stained blue with toluidine blue and fluoresced blue in UV radiation indicating the presence of lignin (Harris and Hartley, 1976). The cell-wall preparation from the skin was rich in wall fragments that fluoresced blue in UV radiation and were derived from the cork ceils of the periderm (Fig. 3D). These fragments stained weakly blue or purple with toluidine blue. The preparation also contained wall fragments that did not fluoresce in UV radiation and stained bright pink with toluidine blue. These were probably derived from the parenchyma cells beneath the periderm. N o starch granules were detected in either cell-wail preparation.
209
1 O0 ,
100" A
0__.__0/O~C -o o n 7 cl
/
6O
0
o Q.. :7
40
60
.A
zx~~ _~__(
40
20 o~
0
10
20
30
:
L
40
50
Cell wall concn. ( r n g / 2 m l ) Fig. 4 The effect of increasing concentrations of potato skin walls on the proportion of D N P in solution. D N P (100 n g / 2 ml) was incubated in PBS with various concentrations of skin walls. Samples were taken after 55 min for mutagenicity assays. Values represent total numbers of revertant colonies (minus those for negative controls) expressed as a percentage of the n u m b e r s seen for a comparable sample of D N P diluted into D M S O and plated immediately. Mutagenic activity present in the supernatant (A), associated with the potato skin walls (o), and adsorbed to the tube walls (I-q).
Effect of increasing concentrations of skin walls on the distribution of DNP Fig. 4 shows the DNP present in the supernatant after incubating DNP (50 ng/ml) in PBS with a range of concentrations (1-50 m g/ 2 ml) of skin walls. The supernatant contained only a small proportion of the DNP added. In the absence of walls, 35% of the DNP added was in the supernatant after 1 h. The presence of 1 mg of walls decreased this to less than 25 %, and this decreased to only 2% in the presence of 50 mg of skin walls. The proportion of the DNP associated with the skin walls increased with the amount of wall added (Fig. 4), and accounted for approximately
/
H
0
t 10
0(I)
~ I 20
N
~ ,30
L 40
[ 50
Cell wall c o n c n ( m g / 2 m l ) Fig. 5. The effect of increasing concentrations of potato flesh walls on the proportion of D N P in solution. The experiment was carried out similarly to that described in Fig. 4, except flesh walls were used instead of the skin walls. Mutagenic activity present in the supernatant (A), associated with the potato flesh walls (o), and adsorbed to the tube walls ([2).
78% of the DNP added in the presence of 10 mg of walls and approximately 90% with 50 mg of walls. Only in the absence of walls was a significant proportion of the DNP associated with the glass walls of the tubes (Fig. 4). The results of a parallel experiment carried out with a-cellulose showed that 10 mg of a-cellulose adsorbed 67% of the DNP and 50 mg of a-cdlulose adsorbed 75% of the DNP (data not shown).
Effect of increasing concentrations of flesh walls on the distribution of DNP In marked contrast to the results of the above experiment with the skin walls and with a-cellulose (used as a positive control), most of the DNP added remained in the supernatant with flesh walls. Fig. 5 shows the effects of incubating a
Fig. 3. (A) Transverse section of the inner cortex region of the flesh of a potato tuber. The section was unstained a n d photographed using Nomarski interference optics. It shows thin-walled parenchyma cells containing large starch granules (arrow). Bar = 50 lam. (B) Transverse section of the skin and adjacent flesh of a potato tuber photographed using U V fluorescence microscopy. The walls of the cork cells (cc) of the skin (periderm) fluoresce, whereas the walls of the parenchyma cells of the flesh just beneath the skin (outer cortex) (oc) do not. Bar = 50 #m. (C) Bright-field mierograph of the cell-wail preparation from the flesh of a potato tuber stained with tohiidine blue. The preparation consists mainly of wall fragments derived from parenchyma cells. Bar = 50 # m . (D) UV fluorescence micrograph of the cell-wall preparation from the skin of a potato tuber. The preparation was rich in wall fragments derived from cork cells which fluoresce in UV radiation. Bar = 50 # m .
210 range of concentrations (1-50 m g / 2 ml) of flesh walls with D N P (50 n g / m l ) in PBS. After incubation with 10 mg walls, more than 80% of the D N P added was recovered in the supernatant, whereas the control (in the absence of walls) contained only approximately 20% of the D N P added. Incubation with increasing concentrations of walls led to the progressive loss of D N P from the supernatant. However, even in the presence of 50 m g / 2 ml walls, more than 40% of the D N P added was recovered in the supernatant. Some D N P was associated with the walls (Fig. 5) and this increased with increasing concentrations of walls, but accounted for only approximately 40% of the D N P added, even at the highest concentration of walls. Even smaller proportions of the D N P added were associated with the glass walls of the tubes (Fig. 5).
Effect of extracted flesh walls on the distribution of DNP In contrast to the results with the unextracted flesh walls, a smaller proportion of the D N P added remained in the supernatant with flesh walls that had been extracted with PBS at 100°C. Fig. 6A shows the effects on D N P content of incubating a range of concentrations (5-30 m g / 2 ml) of extracted flesh walls with PBS containing D N P (50 ng/ml). After incubation with 5 mg of extracted walls, the proportion of the D N P added that was recovered in the supernatant was similar to the control without walls (approximately 20%). Incubation with increasing concentrations of extracted walls led to the progressive loss of D N P in the supernatant. More D N P was associated with the extracted than the unextracted flesh walls (Fig. 6B). At all cell-wall concentrations, only small amounts of D N P were also associated with the glass walls of the tubes (Fig. 6C).
100
8O "o
o O_ Z
"6
60 40 20, 0
I 10
0
I 20
,30
Cell wall concn ( r n g / 2 m l )
I O0
8O "o "o
"o 60 ¸ o CL z a
,._
40
0
20, o
OI 0
O
I
I
10
20
30
Cell wall concn ( m g / 2 m l )
I°°|C
"o
60'
(3_ Z
40" Fig. 6. The effect of extracting potato flesh walls at 100°C on their ability to adsorb DNP. The experiment was carried out similarly to that described in Fig. 5. Mutagenic activity (A) present in the supernatant (a, z~), (B) associated with flesh walls (e, o), and (C) adsorbed to the tube walls (11, O), when extracted walls were used (closed symbols) as compared to unextracted wails (open symbols).
20'
0
10
20
Cell wall concn ( m g / 2 m l )
3O
211 100
galacturonic acid (Ring and Selvendran, 1978; Jarvis et al., 1981). Discussion
"~6 40.
A~ ~
20. /
0
20
40
60
80
1O0
extract
Fig. 7. The effect of the 100°C extract from potato flesh walls on the concentration of DNP in solution. DNP (100 ng/2 nil) was incubated in PBS with various dilutions of the extract. Samples were taken after 55 rain for mutagenicity assays. Mutagenic activity present in the supernatant (-), and adsorbed to the tube walls (I).
The effect of 100°C extract of flesh walls on the retention of DNP in solution In the absence of extract, less than 20% of the D N P remained in the PBS after incubation for 1 h (Fig. 7). However, after incubation for 1 h in the presence of the 100°C extract, the proportion of D N P maintained in solution increased with increasing amount of extract (Fig. 7). In the presence of undiluted extract, more than 60% of the D N P added remained in solution. The proportion of D N P associated with the glass walls of the tubes decreased with increasing amount of extract (Fig. 7). Wall extracts obtained with PBS at 4 and 37°C had a similar, but lesser effect (data not shown). Chemical characterization of the IO0°C extract of flesh walls The neutral monosaccharide composition of hydrolysates of the extract was (average % w / w of 2 analyses): rhamnose, 4.8%; arabinose, 8.1%; xylose, 1.0%; mannose, 1.3%; galactose, 84.1%; glucose, 0.8%. Thin-layer chromatography of a hydrolysate of the extract showed a spot with the same mobility as galacturonic and glucuronic acids which co-chromatograph. From previous work on the polysaccharides of the walls of potato flesh, it is likely that the spot was composed mainly of
The ability of different plant cell walls (dietary fibres) to adsorb mutagens (as tumour initiators) needs to be quantified in order to predict their effectiveness in removing mutagens from the digestive tract and thus protecting against colon cancer. In a previous in vitro study (Roberton et al., 1990), we showed that the hydrophobic, environmental mutagen, 1,8-dinitropyrene (DNP), adsorbed to a model fibre preparation (a-cellulose). In the present study we have examined the adsorptive properties of 2 contrasting cell-wall preparations from the skin and flesh of potatoes. Each preparation was rich in a cell-wall type that is common in Western diets. The skin walls reduced the proportion of D N P in solution as was previously observed for the model fibre, a-cellulose (Roberton et al., 1990). However, to our surprise, we found that the flesh walls increased the proportion of D N P in solution relative to the control without cell walls. The walls of the cork cells of potato skins contain the complex hydrophobic polymer, suberin, and associated waxes (Espelie et al., 1980). Suberin consists of aliphatic polyester domains covalently attached to phenolic domains (Kolattukudy and Espelie, 1985). Because of the presence of these hydrophobic components, we would predict that the hydrophobic mutagen, DNP, would strongly adsorb to a preparation rich in these walls. This proved to be the case. We found that a co-mutagenic effect was not responsible for the increased mutagenic activity of the D N P solution in the presence of flesh walls. In fact, the effect was caused by a PBS-soluble fraction released from the walls maintaining the D N P in solution. This soluble fraction could be extracted from the walls and when incubated with D N P in the absence of walls, it maintained the D N P in solution. The flesh cell walls of potato tubers are known to contain a high proportion of pectic polysaceharides (Ring and Selvendran, 1978; Jarvis et al., 1981). Pectic polysaccharides are a complex class of wall polysaccharides, comprising a family of acidic polysaccharides (rham-
212 nogalacturonans) and several neutral polysaccharides (arabinans, galactans, and arabinogalactans) (Bacic et al., 1988). They can be partially extracted from walls with water and our analyses of the 100°C extract of flesh cell walls with PBS are consistent with the presence of rhamnogalacturonans and neutral pectic polysaccharides, mainly consisting of galactans. This predominance of galactans in the neutral pectic polysaccharides of the flesh cell walls of potato tubers has been reported previously (Ring and Selvendran, 1978; Jarvis et al., 1981). In our experiments, the pectic polysaccharides in the PBS extract appear to be responsible for maintaining the D N P in solution. The finding that a hydrophobic molecule, such as DNP, can be maintained in solution by hydrophilic molecules, such as pectic polysaccharides, is surprising. It would appear to break a fundamental physico-chemical rule. One possible explanation is the mechanism proposed by Wiggins and van Ryn (1990) to account for how a hydrophilic molecule, provided that it is highly charged, can maintain hydrophobic molecules in solution. These authors have shown that in the water surrounding highly charged molecules the water-water hydrogen bonding is weaker than in normal water and hydrophobic molecules selectively accumulate into that environment. The acidic rhamnogalacturonans could theoretically act in this way to maintain D N P in solution. Even after removing the PBS-soluble components of the potato flesh walls, their ability to adsorb DNP was much less than the skin walls or the model fibre, a-cellulose. These flesh walls are derived from parenchyma cells and walls of this cell type from dicotyledons that have been chemically analysed have been found to have broadly similar compositions (Bacic et al., 1988). As this cell type is very common in many dicotyledonous fruits and vegetables in Western diets, the effect which we have described may be widespread. Potato tubers are usually cooked in one of a variety of ways before they are eaten, however, cooking makes the isolation of cell walls much more difficult and therefore we used cell walls isolated from raw tubers. We would expect that boiling peeled potatoes would result in the loss of some pectic polysaccharides into the cooking water, especially from the surface parenchyma
cells. However, within the flesh, although pectic polysaccharides may be partially released from the walls, the presence of surrounding cells should prevent their loss into the cooking water. Refluxing of isolated flesh cell wails in PBS for 1 h, as described in the present study, is a much more drastic treatment than would be encountered in the regular boiling of potatoes and even after this PBS treatment the walls adsorbed D N P only weakly. Boiling will probably have httle effect on the ability of skin cell wails to adsorb D N P as we found that refluxing isolated skin walls in PBS for 1 h increased their ability to adsorb D N P by only 9% (data not shown). The release of the soluble fraction from parenchyma cell walls, the adsorptive properties of the insoluble components of the cell walls, as well as the digestion of both of these by bacteria in the colon will all affect the distribution of hydrophobic components, such as DNP, in the intestines. Whether the adsorption of hydrophobic mutagens by the soluble fraction from the parenchyma walls will increase or decrease the availability of the mutagens to the gut mucosal cells is unknown. Therefore, parenchyma walls of dicotyledons, as exemplified by the potato, may not necessarily protect against colon cancer. These findings may have important consequences for protection against colon cancer in Western populations.
Acknowledgements This work was supported by a grant from the Auckland Medical Research Foundation. We thank the Auckland University Research Committee for equipment grants to P.J.H. to purchase the sonicator and mixer/mill, and the Provincial Grand Lodge (Irish Constitution) for an equipment grant to A.M.R. to purchase a shaking water bath. Jude Antonimuttu, Marlene F. Kendon, Robert J. McKenzie, and Dr. Julie B. White provided valuable assistance in this work.
References Albersheim, P., D.J. Nevins, P.D. English and A. Karr (1967) A method for the analysis of sugars in plant cell-wall polysaccharidesby gas-liquid chromatography,Carbohydr. Res., 5, 340-345.
213 Bacic, A., P.J. Harris and B.A. Stone (1988) Structure and function of plant cell walls, in: J. Preiss (Ed.), The Biochemistry of Plants, Vol. 14, Carbohydrates, Academic Press, San Diego, CA, pp. 297-371. Barnes, W.S., J. Maiello and J.H. Weisburger (1983) In vitro binding of the food mutagen 2-amino-3-methylimidazo[4,5-f]quinoline to dietary fibers, J. Natl. Cancer Inst., 70, 757-760. Blakeney, A.B., P.J. Harris, R.J. Henry and B.A. Stone (1983) A simple and rapid preparation of alditol acetates for monosaccharide analysis, Carbohydr. Res., 113, 291-299. Burkitt, D.P. (1978) Colonic-rectal cancer: tiber and other dietary factors, Am. J. Clin. Nutr., 31, $58-$64. DeCosse, J.J., H.M. Miller and M.L. Lesser (1989) Effect of wheat fiber and vitamins C and E on rectal polyps in patients with familial adenomatous polyposis, J. Natl. Cancer Inst. 81, 1290-1297. Espelie, K.E., N.Z. Sadek and P.E. Kolattukudy (1980) Composition of suberin-associated waxes from the subterranean storage organs of seven plants - parsnip, carrot, rutabaga, turnip, red beet, sweet potato and potato, Planta, 148, 468-476. Feder, N., and T.P. O'Brien (1968) Plant microtechniques: some principles and new methods, Am. J. Bot., 55, 123-142. Ferguson, L.R., P.J. Harris, H.J. Hollands and A.M. Roberton (1990) Effects of bile salts on the adsorption of a hydrophobic mutagen to dietary fiber, Mutation Res., 245, 111117. Fry, S.C. (1988) The growing plant cell wall: chemical and metabolic analysis. Longman Scientific and Technical, Harlow. Harris, P.J. (1983) Cell walls, in: J.L. Hall and A.L. Moore (Eds.), Isolation of Membranes and Organdies from Plant Cells, Academic Press, London, pp. 25-53. Harris, P.J., and R.D. Hartley (1976) Detection of bound ferulic acid in cell walls of the Gramineae by ultraviolet fluorescence microscopy, Nature (London), 259, 508-510. Jacobs, L.R., and J.R. Lupton (1986) Relationship between colonic luminal pH, cell proliferation, and colon carcinogenesis in 1,2-dimethylhydrazine treated rats fed high fiber diets, Cancer Res., 46, 1727-1734. Jarvis, M.C., M.A. Hall, D.R. Threlfall and J. Friend (1981) The polysaccharide structure of potato cell walls: chemical fractionation, Planta, 152, 93-100.
Jensen, W.A. (1962) Botanical Histochemistry, Freeman, San Francisco, CA. Kada, T., M. Kato, K. Aikawa and S. Kiriyama (1984) Adsorption of pyrolysate mutagens by vegetable fibers, Mutation Res., 141, 149-152. Kolattukudy, P.E., and K.E. Espelie (1985) Biosynthesis of cutin, suberin, and associated waxes, in: T. Higuchi (Ed.), Biosynthesis and Biodegradation of Wood Components, Academic Press, Orlando, CA, pp. 161-207. Lawton, J.R., and P.J. Harris (1978) Fixation of senescing plant tissues: sclerenchymatous fibre cells from the flowering stem of a grass, J. Microsc., 112, 307-318. Maron, D.M., and B.N. Ames (1983) Revised methods for the Salmonella mutagenicity test, Mutation Res., 113, 173-215. O'Brien, T.P., and M.E. McCully (1981) The study of plant structure: principles and selected methods, Termarcarphi Pty., Melbourne. Rae, A.L., P.J. Harris, A. Bacic and A.E. Clarke (1985) Composition of the cell walls of Nicotiana alata Link et Otto pollen tubes, Planta, 166, 128-133. Ring, S.G., and R.R. Selvendran (1978) Purification and methylation analysis of cell wall material from Solanum tuberosum, Phytochemistry, 17, 745-752. Roberton, A.M., P.J. Harris, H.J. Hollands and L.R. Ferguson (1990) A model system for studying the adsorption of a hydrophobic mutagen to dietary fiber, Mutation Res., 244, 173-178. Salema, R., and I. Brandao (1973) The use of PIPES buffer in the fixation of plant cells for electron microscopy, J. Submicrosc. Cytol., 5, 79-96. Smith-Barbaro, P., D. Hanson and B.S. Reddy (1981) Carcinogen binding to various types of dietary fiber, J. Natl. Cancer Inst., 67, 495-497. Takeuchi, M., M. Hara, T. Inoue and T. Kada (1988) Adsorption of mutagens by refined corn bran, Mutation Res., 204, 263-267. Wiggins, P.M., and R.T. van Ryn (1990) Changes in ion selectivity with changes in density of water in gels and cells, Biophysica, in press. Willett, W. (1989) The search for the causes of breast and colon cancer, Nature (London), 338, 389-394.