Compartmental Analysis of Herbicide Efflux in Susceptible and Difenzoquat-Resistant Avena fatua L. Suspension Cells

Anthony J. Kern and William E. Dyer*

*Department of Plant, Soil, and Environmental Sciences
 Montana State University
 Bozeman, MT 59717-0312
This article was published in Pesticide Biochemistry and Physiology 61, 27-37(1998)

Abstract

The cellular localization of difenzoquat was investigated using suspension cell cultures derived from inbred wild oat (Avena fatua L.) lines that were susceptible (S) or resistant (R) to the herbicide. Compartmental analysis of S cell cultures resolved [3-¹⁴C]difenzoquat efflux from three cellular compartments which contained 75%, 19%, and 6%, respectively, of the total absorbed difenzoquat. Greater than 98% of the efflux from S cells occurred within 420 minutes. In contrast, compartmental analysis of R cell cultures differentiated only two cellular components, containing 73% and 27% of the absorbed difenzoquat. Further, difenzoquat efflux from R cell cultures occurred more slowly, requiring >2000 minutes for 98% efflux to be achieved. Washing cell cultures preloaded with [3-¹⁴C]difenzoquat in methanol:chloroform indicated that difenzoquat was preferentially bound in cell wall material of R cell cultures. In whole plants, a greater proportion of total leaf-absorbed [3-¹⁴C]difenzoquat was present in an insoluble residue in R plants than in S as determined by tissue homogenization and liquid scintillation counting. By 96 hours after treatment with [3-¹⁴C]difenzoquat, >90% of the absorbed radiolabel was extractable from S plants whereas <10% was extractable from R tissues. The results indicate that tight binding of difenzoquat in R cell walls may be responsible for difenzoquat resistance in wild oats.

Introduction

Difenzoquat (1,2-dimethyl-3,5-diphenyl-1H-pyrazolium) is a selective, postemergence herbicide used to control wild oats in wheat and barley. The precise mechanism of action of difenzoquat is not known, although it appears to have both bipyridinium contact-like activity similar to the chemically related herbicide paraquat, as well as growth inhibition activity which may be responsible for inhibition of meristematic cell division and expansion (1). Previous reports indicated that wild oat populations resistant to the chemically unrelated preemergence herbicide triallate (S-(2,3,3-trichloro-2-propenyl) bis(1-methylethyl)carbamothioate) were highly cross-resistant to difenzoquat, with R/S resistance ratios (ratio of 50% lethal doses for resistant (R) and susceptible (S) plant types) of 60 or more (2,3).

The goal of these studies was to use plant cell cultures to compare the physiological and cellular fate of difenzoquat in R and S cultured cells. Plant cell cultures are thought to qualitatively represent intact plants when investigating herbicide metabolism patterns (4), and their use avoids the confounding factors of herbicide translocation and presence in apoplastic spaces in intact plants (5). Cell cultures have been used to study the uptake kinetics of herbicides across the cell wall and plasmalemma. Atrazine (6), amitrole (7), diclofop-methyl (8), chlorsulfuron and clopyralid (9) have been shown to enter the plant cell by passive diffusion.

Compartmental analysis is used to study the uptake and cellular distribution of a radiolabeled compound once inside the cell. Typically, plant tissues or cell cultures are "loaded" in a medium containing a radioisotope for an appropriate amount of time. After steady state equilibrium is achieved, the solution is replaced with isotope-free medium and radiolabel efflux is monitored by frequently replacing the desorption medium. The loss of radiolabel has been described as a series of first-order reactions, corresponding to flux from the cell wall (the most rapidly effluxing cell component), followed by flux from the cytoplasm and finally the vacuole (the most slowly effluxing compartment) (10).

Compartmental analysis was originally developed to determine the intracellular movement and location of ions in giant algal cells (11), and the technique has since been used to study transmembrane ion and herbicide flux. Pierce and Higinbotham (12) used efflux data to differentiate ionic (Na⁺, Cl^-, and K⁺) fluxes between the plasmalemma and tonoplast, putatively identifying the location of ion pumps. Compartmental analysis of paraquat efflux showed that the herbicide was unevenly distributed in three cellular compartments: the cell wall, cytoplasm and vacuole (13). The authors concluded that paraquat was able to move bidirectionally across the plasmalemma and tonoplast, although uptake into the cytoplasm and vacuole was substantially slower than for other herbicides (6-9).

We previously reported that wild oat lines resistant to triallate were highly cross-resistant to difenzoquat, a divalent cationic herbicide closely related to paraquat (3). We concluded that slight differences between R and S plants in uptake and translocation rates of difenzoquat were not sufficient to confer resistance (3). In addition, difenzoquat metabolites were not detected in either S or R wild oats, and preliminary studies suggested that translocated difenzoquat was more tightly bound in R wild oat leaf tissue (14).

The objectives of these studies were to: 1) analyze the cellular distribution and efflux kinetics of difenzoquat in R and S cell cultures using compartmental analysis; 2) determine the role cell wall material may play in difenzoquat binding; and 3) compare rates of difenzoquat binding in R and S tissues from whole plants.

Materials and Methods

Plant Material

S wild oat seeds were collected from greenhouse-grown populations of the nondormant inbred line SH430 (15). R wild oat seeds were originally collected from fields near Fairfield, MT in 1993 in which triallate had not provided acceptable control. Previous studies confirmed that these populations were resistant (12- to 20-fold) to triallate and were also cross-resistant to difenzoquat (2,3). One field collection designated FG93R22 contained about 80% R individuals and was used to develop the inbred difenzoquat-resistant line used in the experiments reported here. Field-collected seeds were planted in 55 by 35 by 10 cm flats using greenhouse soil mix (1:1:1 Bozeman silt loam:washed sand:peat moss, (v/v/v)). Flats were placed under greenhouse conditions, watered as needed, and plants at the four-leaf stage were treated with five times the recommended field rate of formulated difenzoquat (4.2 kg ha^-1) using a moving nozzle sprayer in 185 L ha^-1 water containing 0.25% (v/v) nonionic surfactant. The surviving individuals (about 40 out of 50 plants) were grown to maturity, self-pollinated, and their bulked progeny subjected to an identical difenzoquat treatment. All individuals survived, and about 100 plants were again grown, self-pollinated, and their progeny used for the experiments reported here. The inbred R line derived from FG93R22 was shown in previous experiments to be 60-fold more resistant to difenzoquat than S lines (3).

Tissue Culture

Suspension cell cultures of S and R wild oats were derived from mature embryos of greenhouse-grown SH430 and FG93R22 seeds, respectively. Nondormant seeds were peeled (lemma and palea removed by hand), the caryopses surface sterilized by shaking in 10% (v/v) commercial bleach solution for 15 min, followed by vigorous rinsing for 30 min in sterile water. Embryos were aseptically excised with a scalpel and cultured in the dark on MS salts (16) medium supplemented with 30 g L^-1 sucrose, 100 Fg L^-1 thiamine-HCl, 500 Fg L^-1 nicotinic acid, 500 Fg L^-1 pyridoxine-HCl, 100 mg L^-1 inositol, and 4 mg L^-1 2,4-dichlorophenoxyacetic acid (2,4-D) and solidified with 8 g L^-1 Difco agar in 15mm x 100mm plastic petri dishes (17). Four days after germination and as needed thereafter, shoots and roots were excised and callus tissue retained. Calli were transferred to the same medium as above every 2 weeks for 12 weeks, after which 1 g of highly friable callus was transferred to 50-ml Erlenmeyer flasks containing 5 ml liquid MS salts medium containing 2 mg L^-1 2,4-D. Cultures were shaken in the dark at 250 rpm at 24 ±3EC and subcultured every 2 weeks. Viability of cultured cells was determined using methylene blue vital stain (18). Cultures with $80% viability were used for experiments.

Difenzoquat dose response

Dose response assays were conducted by aseptically adding 3 ml of 14- to 28-day old R and S cell cultures to 5 ml MS medium in 50-ml Erlenmeyer flasks containing 0, 10, 100, 500, or 1000 FM technical grade difenzoquat. Cultures were shaken as described above and after 10 days packed cell volumes (PCV) were determined by centrifuging 1 ml of culture in 1.5-ml microcentrifuge tubes at 1500 x g for 1 min. There were two replications per treatment and the experiment was repeated once. Data are presented as the means of both experimental repetitions.

Difenzoquat efflux

R and S suspension cultures of similar viability were selected and 1-ml aliquots were treated with 1.67 x 10⁴ Bq [3-¹⁴C]difenzoquat (sp. act. 2.29 x 10⁶ Bq/mg) plus technical grade difenzoquat (96.8% pure) to a final concentration of 100 FM difenzoquat in MS salts medium for 2 hours. Uptake equilibrium was confirmed by sample oxidation and liquid scintillation counting (LSC) of cultures treated for an additional 10 hours, and R and S cultures contained similar amounts of ¹⁴C at the beginning of washout treatments (data not shown). After 2 hours of shaking, cells were placed on ice for 5 min, loosely packed by centrifuging for 3 sec at 4EC and the supernatant (-750 Fl) removed. The loose cell pellet was resuspended in 1 ml of ice-cold MS salts medium, flash spun, and pellet resuspended as before. The washing procedure was repeated twice and the samples placed at 25EC in a water bath. At various timepoints thereafter, cells were quickly centrifuged as above, the supernatant removed, and fresh room-temperature MS salts medium added. Efflux of ¹⁴C was determined by LSC of the supernatants and the experiments were continued until negligible amounts of ¹⁴C were liberated. The final cell pellet was obtained by centrifuging at 16,000 x g for 5 min, frozen in liquid nitrogen, and residual radioactivity determined by biological sample oxidation and LSC.

For some experiments, cells were incubated with radiolabeled and technical grade difenzoquat for 2 hr as above, centrifuged and resuspended in 750 Fl methanol:chloroform (2:1, v/v). Efflux of ¹⁴C from methanol:chloroform-treated cells and residual radioactivity were determined as above. For all efflux experiments, there were two replications of each timepoint and experiments were repeated once. Data are presented as the means of both experimental repetitions and are graphed as the natural log of the percent of total difenzoquat remaining in each sample versus efflux time.

Extractable difenzoquat in whole plants

To determine if altered herbicide binding to a subcellular component may be involved in difenzoquat resistance, R and S plants were treated with [3-¹⁴C]difenzoquat and leaf tissues were extracted at various timepoints after treatment. Greenhouse-grown R and S plants at the four-leaf stage were treated with 0.84 kg ha^-1 formulated difenzoquat as above. [3-¹⁴C]difenzoquat (1.67 x 10³ Bq) was dissolved in water containing 0.25% (v/v) nonionic surfactant and spiked with unlabeled formulated difenzoquat to achieve an identical herbicide concentration as above. The treatment mixture (3 µl) was spotted on the third fully expanded leaf 1 cm acropetal to the ligule. After 12, 24, 48, 72, or 96 hours, the 3-cm section of leaf tissue immediately acropetal to but not including the treated spot was excised and homogenized in 0.6 ml water using a Broeck tissue homogenizer. The crude homogenate was centrifuged for 5 min at 4EC and 16,000 x g, the supernatant removed and ¹⁴C present in the supernatant determined by LSC. The resulting pellet was sequentially extracted with ethanol, chloroform and benzene and liberated ¹⁴C determined by LSC. After the sequential extractions, the remaining pellet was oxidized and the residual bound ¹⁴C determined by LSC. There were three replications of R and S plants and the experiment was repeated twice. Data are presented as the means of the three experimental repetitions.

RESULTS AND DISCUSSION

Difenzoquat dose response

Figure 1.  Packed cell volumes of R (triangles) and S (circles) wild oat suspension cultures 10 days after treatment with difenzoquat.  Vertical bars are standard errors of means.

To confirm that cell suspension cultures derived from the inbred R line retained their herbicide response levels, dose response assays were conducted on both R and S suspension cultures (Figure 1). In the absence of difenzoquat, R and S wild oat cultures grew rapidly, increasing from 100 Fl to 280 Fl PCV in 10 days. Growth of S cells was inhibited 38% by 10 FM difenzoquat and growth was severely inhibited at 100 FM difenzoquat and above. From these data we estimated a GR₅₀ value (treatment rate at which growth was inhibited 50%) of 25 FM for S cells. In contrast, growth of R cells was inhibited only 13% at the 100 FM treatment rate, and some growth occurred even at 1000 FM difenzoquat. The estimated GR₅₀ value for R cells was 700 FM difenzoquat, representing a 28-fold increase in resistance levels in R over S suspension cells. This level of resistance is somewhat lower than the 60-fold increase we observed in R plants (3), a difference that may be explained by the physiology of cultured cells. In addition to difenzoquat’s proposed mechanism of plant toxicity by interrupting meristematic cell division and/or expansion, the herbicide is also thought to act as an electron acceptor from Photosystem I, generating oxygen radicals and causing lipid peroxidation (19). Since the suspension cultured cells were not photosynthetically active, the second mechanism of toxicity probably did not contribute to the herbicidal effect.

Difenzoquat efflux

Figure 2.  Efflux of [14C] difenzoquat from S wild oat suspension cultures in aqueous desorption medium after (A) 1-420 min, (B)1-120 min, and (C) 1-10 min.  Regression lines (r2 > 0.93) were determined as under Results and Discussion. Vertical bars are standard errors of means.  Efflux half-times (t1/2) were calculated from slopes of the regression lines.

Difenzoquat efflux from S suspension cells occurred rapidly during the first 10 min after adding desorption medium, and slowed dramatically thereafter (Figure 2). After 420 min (the last timepoint tested), >98% of the difenzoquat present in S cells at t=0 had effluxed. Analysis of the efflux kinetics indicated that difenzoquat efflux occurred in three distinct linear components (cell compartments) in S cells, each with different kinetics. Using regression analysis and extrapolating regression lines of each linear component to the y axis, the total percentages of difenzoquat present in each compartment were determined. The y-intercept calculated from the most slowly effluxing compartment (linear from 180 to 420 min, r²=0.93, Figure 2A) was 0.61, corresponding to approximately 4% of the total difenzoquat present in the cells before desorption. To calculate the percentages of radioactivity present in the other two compartments, extrapolated values for the most slowly effluxing compartment were first subtracted for each corresponding timepoint (1 to 120 min), the data graphed and regression analysis done on the resulting next linear component as above (15 to 120 min, r²=0.99, Figure 2B). The corresponding y-intercept of 1.28 indicated that 19% of the absorbed difenzoquat effluxed from the second-slowest compartment. The extrapolated values in Figure 2B were likewise subtracted to generate a line for the most quickly-effluxing compartment (<15 min, r²=0.96, Figure 2C) and the y-intercept indicated that it contained approximately 75% of the total radioactivity.

Efflux of [3-¹⁴C]difenzoquat from R suspension cells was similar to S cells for the first 30 min, with >60% of the difenzoquat effluxing from the cells during this time (Figure 3). However, efflux of the remaining fraction of difenzoquat from R cells occurred much slower than in S cells, requiring >2000 min for 98% efflux. The linear component of Figure 3A was determined as above (45 to 2180 min, r²=0.97) and the corresponding y-intercept value of 1.43 indicates that 27% of the absorbed difenzoquat was present in this fraction. Similar calculations showed that efflux occurring from R cells 1 to 30 min after loading contained 73% of the absorbed difenzoquat (r²=0.98, Figure 3B). While three distinct linear components of the efflux curve could be differentiated in S cells, only two linear components could be identified in the efflux pattern from R cells with a high degree of probability (regression lines with r² >0.90).

Figure 3. Efflux of [14C]difenzoquat from R wild oat suspension cultures in aqueous desorption medium from (A) 1-2180 minutes and (B) 1-30 minutes. Regression lines (r2$0.95) were determined as in Results and Discussion. Vertical bars are standard errors of means.

The significantly longer time period required for efflux from R than from S cells may be related to these changed kinetics. We hypothesize that increased difenzoquat binding in R cells (most likely to the cell wall) resulted in slower desorption which in turn interfered with efflux from intracellular spaces. Such a phenomenon has been previously documented for Ca²⁺ efflux in compartmental analysis of algal cells (11). A significant amount of Ca²⁺ was proposed to be bound to negatively charged sites in the cell wall, and the resulting slow desorption interfered with vacuolar efflux, thus preventing a three-compartment analysis. For difenzoquat, as binding sites slowly became available in R cells, continued binding of difenzoquat effluxing from the cytoplasm or vacuole may have occurred. Based on these data we hypothesize that such a binding mechanism was active in R suspension cells and caused the dramatically increased times needed for difenzoquat efflux.

Previous reports (10, 12, 13) attributed efflux from the most quickly effluxing compartment as representing the "surface film" of the cells, or the compound associated with the water layer surrounding cells prior to efflux analysis. The next two slower components are thought to represent efflux from the cytoplasm and the vacuole, respectively. Based on this model, the slowly effluxing vacuole should contain the highest amount of radioactivity, due to its high proportion of the total cellular volume (10). However, other authors have reported experimental results that do not support such a model for compounds that only slowly accumulate in a cellular compartment. Using intact roots from maize seedlings, Hart et al. (13) found that only 7.6% of radiolabeled paraquat was present in the most slowly effluxing fraction, supporting the idea of a saturable tonoplast carrier system.

We attribute the three-component efflux pattern from S cells as representing difenzoquat from the cytoplasm (the most rapidly effluxing compartment), the vacuole (the next slower effluxing compartment), and the cell wall (the slowest effluxing compartment). It seems unlikely that >75% of the effluxed radioactivity could have originated from a "surface film", since it should have been removed by the three ice-cold washes done prior to starting efflux measurements. Also, our data showing that a much higher proportion of difenzoquat was bound in the leaves of R plants than in S leaves (14, also see below) supports the idea that slowly effluxing difenzoquat is gradually being released from binding sites, most likely in the cell wall.

Figure 4. Efflux of [14C]difenzoquat from S wild oat suspension cultures in methanol:chloroform desorption medium from (A) 1-75 minutes and (B) 1-10 minutes. Regression lines (r2$0.95) were determined as in Results and Discussion. Vertical bars are standard errors of means.

To determine if the most slowly-effluxing fraction of difenzoquat from R cells was from cell walls or from vacuolar accumulation, we repeated the efflux experiments as described above except using methanol:chloroform (2:1, v/v) as the desorption medium. Previous efflux studies used this treatment to remove all cellular contents while leaving the apoplast and cell wall material undamaged (13). This protocol also allowed living cell suspensions to equilibrate with the [3-¹⁴C]difenzoquat medium prior to desorption in methanol:chloroform, thus allowing any possible enzymatic sequestration processes to occur. Slowly effluxing difenzoquat could only originate from cell walls in this desorption medium. Similar efflux kinetics from S cells were observed as in the aqueous medium, although efflux occurred more rapidly (Figure 4). By 10 min after immersion, S cells effluxed >90% of the absorbed difenzoquat. The slowest-effluxing component (Figure 4A) comprised 6% of the total absorbed difenzoquat as determined by regression analysis (r²=0.98) and nearly all absorbed difenzoquat had effluxed by 75 min. In contrast, R cells effluxed difenzoquat more slowly than S cells, with a y-intercept value of 1.68 for the slowest-effluxing compartment, corresponding to nearly 50% of the absorbed difenzoquat (Figure 5A; r²=0.95). A second linear efflux component from R cells was identified from 1 to 7.5 min after treatment (Figure 5B;r²=0.95), representing the remaining 50% of absorbed difenzoquat. Because of the rapid initial efflux of intracellular difenzoquat from both R and S cells in this desorption medium, we attribute the most slowly effluxing component in S cells (6% of the total absorbed difenzoquat) and R cells (50% of the total absorbed difenzoquat) as difenzoquat bound to the cell wall.

Figure 5.  Efflux of [14C]difenzoquat from R wild oat suspension cultures in methanol:chloroform desorption medium from (A) 1-600 minutes and (B) 1-7.5 minutes. Regression lines (r2$0.95) were determined as in Results and Discussion. Vertical bars are standard errors of means.

These results support the hypothesis that a significant amount of difenzoquat is bound to cell wall material in R wild oat suspension cultures, and that this fraction exchanges very slowly with the external medium. Similar results were noted with the divalent cation paraquat, where 12% of the total paraquat absorbed by intact maize roots was tightly bound to cell wall material (13). Interestingly, the authors attributed the most slowly effluxing fraction of paraquat to be efflux from the vacuole, not the cell wall. From those studies as well as the data presented here, it appears that the strong cationic nature of paraquat and difenzoquat causes a significant amount of these herbicides to be bound to the cell wall, preventing their entry into the site of action (the chloroplast). The additional levels of difenzoquat binding in R wild oat cell walls documented here represent a reasonable explanation for a resistance mechanism.

Extractable difenzoquat from whole plants

To confirm the altered difenzoquat binding levels in whole R plants, we treated R and S plants with [3-¹⁴C]difenzoquat and quantified binding and metabolism patterns (Table 1). In S plants, >90% of the ¹⁴C present in the 3-cm section of leaf acropetal to the treated leaf spot was extractable in water at all times after treatment. In contrast, the amount of water-extractable difenzoquat from R plants declined dramatically after treatment. By 12 hours after treatment, <60% of the radioactivity in R tissues was extractable and amounts steadily decreased to <10% by 96 hours after treatment. Difenzoquat uptake and translocation occurred more rapidly in R leaves than in S leaves, until at 96 hours after treatment 41% more ¹⁴C had accumulated in R than S. These results agree with our earlier findings that showed more difenzoquat uptake in R than S plants, possibly due to the severe herbicidal injury exhibited by S plants by 24 hours after treatment (3). We were unable to detect any metabolites of difenzoquat in R or S plants at any time after treatment using high-performance liquid chromatography (HPLC) protocols (data not shown). Similar results were reported by Sharma et al. (20) in barley and wild oats, suggesting that difenzoquat is not metabolized in plant systems. Although all extractable radioactivity from R tissues was unmetabolized difenzoquat, the chemical nature of the remaining unextractable fraction was not determined. Subsequent extractions of the remaining pellet with ethanol, chloroform, and benzene failed to liberate additional radioactivity (data not shown), indicating that the unextractable radioactivity in R tissues was very tightly bound. We are currently conducting additional experiments to determine the cellular location and form of this bound fraction.

Compartmental analysis of difenzoquat efflux from cell cultures indicates that there are fundamental differences in herbicide localization between R and S wild oats and suggests that difenzoquat is selectively bound to cell wall material in R cells. Based on our findings, we attribute the most slowly-effluxing difenzoquat (4% of absorbed) in S cultures to be bound to cell wall material and hypothesize that the increased binding we observed in R cultures is involved in the mechanism of difenzoquat resistance. Binding of the related herbicide paraquat to negatively-charged sites in the cell wall has been proposed as a mechanism of herbicide resistance in Hordeum glaucum, where uptake into isolated protoplasts did not differ between R and S biotypes although translocation was dramatically reduced in R leaves (21). Altered paraquat binding could be due to an increased number of binding sites in R cell walls or a mutation that reduces herbicide affinity of a putative transport protein on the plasmalemma, thus increasing the likelihood of paraquat’s interaction with and subsequent binding to cell wall material. We do not know which of these scenarios may cause increased binding of difenzoquat in R wild oats. Nonetheless, it appears that an alteration in the uptake and binding patterns of this herbicide plays a major role in the mechanism of difenzoquat resistance in wild oats.

Acknowledgements

This work was partially supported by the Montana Agricultural Experiment Station and the Montana Noxious Weed Trust Fund. The authors wish to thank the American Cyanamid Company for furnishing financial support and ¹⁴C labeled difenzoquat.

References

1.  K.M. Thai, J. Sakti, and L.C. Fowke, Cell membrane permeability and ultrastructural effects of difenzoquat on wild oats (Avena fatua), Weed Sci. 37, 98 (1989).

2. J.T. O'Donovan, M.P. Sharma, K.N. Harker, D. Maurice, M.N. Baig, and R.E. Blackshaw, Wild oat (Avena fatua) populations resistant to triallate are also resistant to difenzoquat, Weed Sci. 42, 195 (1994).

3. A.J. Kern, C.T. Colliver, B.D. Maxwell, P.K. Fay, and W.E. Dyer, Characterization of wild oat (Avena fatua L.) populations and an inbred line with multiple herbicide resistance, Weed Sci. 44, 847 (1996).

4. T.M. Sterling, N.E. Balke, and D.S. Silverman, Uptake and accumulation of the herbicide bentazon by cultured plant cells, Plant Physiol. 92, 1121 (1990).

5. T.M. Sterling, Mechanisms of herbicide absorption across plant membranes and accumulation in plant cells, Weed Sci. 42, 263 (1994).

6. W.B. McCloskey and D.E. Bayer, Atrazine absorption and effects on photosynthesis in wheat leaf proroplasts, Pestic. Biochem. Physiol. 37, 227 (1990).

7. S.R. Singer and C.N. McDaniel, Transport of the herbicide 3-amino-1,2,4-triazole by cultured tobacco cells and leaf protoplasts, Plant Physiol. 69, 1382 (1982).

8. S.A. Tritter, F.B. Holl, and B.G. Todd, Diclofop-methyl and diclofop uptake in oat (Avena sativa L.) protoplasts, Can. J. Plant Sci. 67, 215 (1987).

9. M.D. Devine, H.D. Bestman, and W.H. Vanden Born, Uptake and accumulation of the herbicides chlorsulfuron and clopyralid in excised pea root tissue, Plant Physiol. 85, 82 (1987).

10. K.T. Hodges, Ion absorption by plant roots, Adv. Agron. 25, 163 (1973).

11. R.M. Spanswick and E.J. Williams, Ca fluxes and membrane potentials in Nitella translucens, J. Exp. Bot. 16, 463 (1965).

12. W.S. Pierce and N. Higinbotham, Compartments and fluxes of K+, Na+ and Cl in Avena coleoptile cells, Plant Physiol. 46, 666 (1970).

13. J.J. Hart, J.M. DiTomaso, D.L. Linscott, and L.V. Kochian, Characterization of the transport and cellular compartmentation of paraquat in roots of intact maize seedlings, Pestic. Biochem. Physiol. 43, 212 (1992).

14. A.J. Kern and W.E. Dyer, Compartmental analysis of difenzoquat efflux in resistant and susceptible Avena fatua L. suspension cells, Proc. West. Soc. Weed Sci. 50, 79 (1997).

15. S. Jana and J.M. Naylor, Genetic adaptation for seed dormancy in Avena fatua, Can. J. Bot. 60, 1611 (1976).

16. T. Murashige and F. Skoog, A revised medium for rapid growth and bioassays with tobacco tissue cultures, Physiol. Plant. 15, 473 (1963).

17. C.M. Stiff, Barley tissue culture methodology: media, culture techniques, and approaches to biolistic transformation, Lab Manual, Program in Genetics and Cell Biology, Washington State University (1991).

18. C.N. Huang, M.J. Cornejo, D.S. Bush, and R.L. Jones, Estimating viability of plant protoplasts using double and single staining, Protoplasma 135, 80 (1986).

19. J.H. Goldbeck and J.M. Cornelius, Photosystem I charge separation in the absence of centers A and B. I. Optical characterization of center A2 and evidence for its association with a 64-kDa protein, Biochim. Biophys. Acta 849, 16 (1986).

20. M.P. Sharma, W.H. Vanden Born, H.A. Friesen, and D.K. McBeath, Penetration, translocation, and metabolism of 14C-difenzoquat in wild oat and barley, Weed Sci. 24, 379 (1976).

21. S.B. Powles and G. Cornic, Mechanism of paraquat resistance in Hordeum glaucum. I. Studies with isolated organelles and enzymes. Aust. J. Plant Physiol. 14, 81 (1987).