|Table of Contents|
|1. General introduction|
|2. Comparative paraquat uptake...|
|3. Differential biochemical responses...|
|4. Differential photosynthetic...|
|5. Summary and conclusions|
|Table of Contents|
Table of Contents
1. General introduction
2. Comparative paraquat uptake and compartmentalization
3. Differential biochemical responses of resistant and sensitive biotypes of American black nightshade to bipyridinium herbicides
4. Differential photosynthetic electron transport and oxidative stress in resistant and sensitive biotypes of American black nightshade
5. Summary and conclusions
PHYSIOLOGICAL AND BIOCHEMICAL STUDIES OF
THE MECHANISM OF PARAQUAT RESISTANCE
IN AMERICAN BLACK NIGHTSHADE
CARLENE A. CHASE
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
I dedicate this work to my nieces and nephews Siobhan, Suraya, Stephen and
Matthew and to my brother Jonathan.
My stay at the University of Florida has been productive and happy. This has
largely been due to the helpfulness and the cooperation of the faculty, staff and
students of the Horticultural Sciences Department. I have been particularly fortunate
that Dr. Bewick and I chose to work together since his guidance and support have
been exemplary. His confidence in my abilities promoted a congenial and
professional working environment. I would also like to thank Dr. Bewick for
providing the financial support for my research, assistantship and attendance of
scientific meetings. I extend my thanks and appreciation to the other members of my
supervisory committee, Dr. J.A. Dusky, Dr. R.J. Ferl, Dr. C.L. Guy and Dr. D.G.
Shilling, for their participation in ushering me smoothly and successfully through my
program. Dr. Shilling, in particular, was always available for the meetings and
impromptu brain-storming sessions that yielded ideas that were critical to achieving
my objectives. He has been unselfish in allowing me the use of his laboratory,
equipment and supplies.
Others have also been generous in allowing me the use of their equipment: Dr.
C.E. Vallejos, Dr. L.C. Hannah, Dr. R.J. Ferl, J.K. Brecht, Dr. D.J. Huber, Dr.
D.J. Cantliffe, Dr. S.R. Kostewicz, Dr. S.J. Locasio, Dr. C.D. Chase and Dr. R.
Charudattan. I would like to thank Dr. K.C. Cline and his laboratory assistant Chang
Jiang Li for their help in optimizing my chloroplast isolation protocol. Dr. I.K. Vasil
and Dr. V. Vasil provided helpful information concerning the enzymic isolation of
cuticular membranes, cells and protoplasts. I am indebted to Dr. K.L. Smith who
was an invaluable resource person. The statistical consulting provided by Christy
Steible was greatly appreciated. I would like to thank Debbie Williams and Kathy
Bergsma for their instruction on the use of the gas chromatographs. I am grateful to
Greg MacDonald and Jim Gaffney for the preliminary fluorescence study they
conducted. I appreciated the cooperation, assistance and the camaraderie of the
laboratory personnel: Timothy McKechnie, Mark Shelby, Eric Bish, John Porter,
Bielinski Santos, Dr. Alan Miller and Dr. Ada Okoli. A number of undergraduate
student assistants made life a lot easier by doing some of the more routine and tedious
tasks for which they have my heart-felt thanks.
I would like to thank all my friends for their moral support and for their
assistance in getting me settled on arrival in Gainesville and every time I needed to
move house. Very special thanks are extended to Raymond Carthy who was always
considerate and supportive. Ray's contribution has been invaluable and has ranged
from assisting me in the laboratory to ensuring I got home safely after working late at
night. Finally, I would also like to thank my family whose confidence in me was
inspiring and whose recurring questions about when was I coming home kept me
focused on finishing.
TABLE OF CONTENTS
ACKNOWLEDGMENTS ................................ ....... iii
ABSTRACT .................................. .. ...... vii
1 GENERAL INTRODUCTION ... ......... ... .......... 1
Chloroplast Susceptibility to
Lipid Peroxidation ......
O yvon TnYiritv
Bipyridinium Herbicides .......
Chemical Structure and Herbicidal A
Mechanism of Action ......
Herbicide Resistance ......
Paraquat Resistance .......
Mechanisms of Paraquat Resistance
Altered Cuticular Membrane
Mutation at Site of Action . .
Paraquat Metabolism ......
Protective Enzymes .......
Paraquat Sequestration .....
Exclusion by Plasmalemma .
Translocational Mutants ...
. .. ,
. VI . .
. . .
. . . .
. . .
Paraquat-Resistant American Black Nightshade
2 COMPARATIVE PARAQUAT UPTAKE AND
Materials and Methods ........
Plant Material ..........
Paraquat Uptake .........
Paraquat Efflux . . . . .
Silanization of Glassware . ..
Results .. . .. . . . . .. .
Paraquat Uptake .........
Paraquat Efflux .........
3 DIFFERENTIAL BIOCHEMICAL RESPONSES OF RESISTANT
AND SENSITIVE BIOTYPES OF AMERICAN BLACK
NIGHTSHADE TO BIPYRIDINIUM HERBICIDES ........
M materials and M ethods .............................
Ethane Assay ................... .............
Ion Leakage .................................
Chlorophyll A Fluorescence .......................
Ethane Evolution ..............................
Electrolyte Leakage ............................
Chlorophyll A Fluorescence .......................
D discussion . . . . . . . . . . . . . . . . . .
4 DIFFERENTIAL PHOTOSYNTHETIC ELECTRON TRANSPORT
AND OXIDATIVE STRESS IN RESISTANT AND
SENSITIVE BIOTYPES OF AMERICAN BLACK
Materials and Methods .....................
Chloroplast and Thylakoid Membrane Isolation
Measurement of Superoxide Radicals
Results . . . . . . . . . . .
Electron Transport ............
Oxygen Radical Generation ......
D discussion ...................
5 SUMMARY AND CONCLUSIONS . .
BIOGRAPHICAL SKETCH ................
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . .
. . . . .
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the degree of Doctor of Philosophy
PHYSIOLOGICAL AND BIOCHEMICAL STUDIES OF
THE MECHANISM OF PARAQUAT RESISTANCE
IN AMERICAN BLACK NIGHTSHADE
Carlene A. Chase
Chairman: Thomas A. Bewick
Major Department: Horticultural Science
The primary mechanism of paraquat (1,l'-dimethyl-4,4'-bipyridinium ion)
resistance was investigated in American black nightshade (Solanum americanum
Mill.). Physiological and biochemical studies were conducted with a paraquat-
resistant (R) biotype and a paraquat-sensitive (S) biotype.
There was no difference in the amount of paraquat that penetrated to the
cytoplasm. The R biotype exhibited greater apoplastic adsorption of paraquat than the
S biotype; however, apoplastic sequestration is unlikely as a primary mechanism
because there was 3-fold more paraquat in the cytoplasm than was adsorbed in the
Paraquat treatment increased the rates of ethane production above control
levels in both biotypes, but to a greater extent in S biotypes. An enzyme inhibitor
ensured that this was not due to elevated levels of radical scavenging enzymes. A
larger difference between R and S biotypes in the rate of electrolyte leakage occurred
in light than in darkness. Concentration-dependent studies of paraquat-induced ion
leakage showed that the S biotype had higher ratios of I50(dark):Is(light) (I50 =
paraquat concentration that caused 50% ion leakage) than the R biotype. Diquat (6,7-
dihydrodipyrido[1,2-a:2', l'-c]pyrazinediium ion) and paraquat suppressed variable
fluorescence (F,) at the same rate, indicating rapid penetration to the site of action in
both biotypes. However, the magnitude of Fv reduction was smaller in the R biotype,
so that higher concentrations of diquat and paraquat were required to reduce Fv to a
level similar to that of the S biotype. The electrolyte leakage and fluorescence studies
support a light-dependent mechanism of resistance.
Electron transport rates, when measured as the photoreduction of ferricyanide
and NADP+ by isolated thylakoid membranes, were significantly higher in the S
biotype. The potential of the two biotypes to generate oxidative stress was also
compared. Nontreated thylakoids of the S biotype produced superoxide at a faster
rate than those of the R biotype. The increase in the rate of superoxide formation
resulting from paraquat treatment was larger in the thylakoids of S biotype than the R
biotype. The data imply that paraquat resistance in American black nightshade is due
to reduced photosynthetic electron flow.
Chloroplast Susceptibility to Oxygen Toxicity
Illuminated chloroplasts are subject to oxidative stress for a number of reasons
(Halliwell, 1987). The splitting of water to provide electrons to photosystem II (PSII)
creates a high ambient oxygen (02) concentration in the chloroplast. Therefore, 02 is
readily available to participate in reactions that generate toxic derivatives. Excitation
energy in illuminated chloroplasts that is not dissipated by electron transfer,
fluorescence or heat can cause the formation of triplet chlorophyll. This type of
chlorophyll can result in the formation of singlet oxygen, which can be extremely
damaging to chloroplast membranes (Halliwell, 1987). Also, inefficiencies in
photosynthetic electron transport lead to electrons being leaked to oxygen resulting in
its univalent reduction to superoxide (02-) (Asada et al., 1974). The reaction is
apparently associated with the autooxidation of electron acceptors on the reducing side
of photosystem I (PSI) and is considered to be a partial reaction of the Mehler
reaction (Asada et al., 1974). A small amount of 02- is also formed when oxygen is
reduced by ferredoxin (Allen, 1977) (Fig. 1). Chloroplast membrane lipids contain a
large amount of polyunsaturated fatty acids that predispose them to lipid peroxidation.
Therefore, the chloroplast is an organelle with inherent oxygen toxicity problems.
Protection against oxidative stress is provided by a battery of enzymes and
antioxidants that can alleviate normal oxidative stresses.
0 Reduced Ferredoxin
Route 1: Direct Ferredoxin 02
reduction of NADP+
oxygen by PSI reductase
Route 3: Formation Route 2: Reduction
of NADPH of oxygen by reduced
Figure 1. Alternative routes for electron flow beyond PSI. (Adapted from Halliwell and
Halliwell and Gutteridge (1989) defined free radicals as chemical species
capable of independent existence that possess one or more unpaired electrons. Such a
condition usually results in a species that is more reactive than the parent atom or
molecule. Radicals can potentially initiate a chain of reactions because the reaction of
a radical with a nonradical always results in the production of another radical
(Rabinowitch and Fridovich, 1983).
The superoxide radical can participate in both oxidation and reduction
reactions (Rabinowitch and Fridovich, 1983). Superoxide may also initiate and
propagate free radical chain reactions (Fridovich, 1988). Although 02- is not a very
reactive radical, Fridovich (1986, 1988) regards it as capable of direct damage to cell
constituents, and considers its innate toxicity to be underestimated while emphasis is
placed on its indirect effects through the formation of hydrogen peroxide (H202) and
hydroxyl radicals (OH').
The chloroplast is protected from 02- in a number of ways. The radicals
undergo spontaneous dismutation on collision with each other, yielding H202. This
reaction occurs at a much faster rate when catalyzed by superoxide dismutase (EC
184.108.40.206) (McCord and Fridovich, 1969).
O2- + 02- + 2H+ H202 + 02 (1)
Asada et al. (1973) demonstrated the presence of superoxide dismutase (SOD) in
spinach chloroplasts. Although most of the activity is due to the stromal Cu, Zn-
SOD, spinach chloroplasts also contain a membrane-bound SOD. Superoxide is also
scavenged by ascorbate (Nishikimi and Yagi, 1977).
Although H202 is not a radical, it is a substantial contributor to oxygen
toxicity in the chloroplast. A strong oxidant, H202 oxidizes several of the Calvin
cycle enzymes and inhibits carbon dioxide fixation (Kaiser, 1979; Law et al., 1983).
It inhibits SOD and can be reduced in an iron-catalyzed Haber-Weiss reaction to
produce the OH' (Halliwell and Gutteridge, 1989).
02- + H202 OH' + OH- + 02 (2)
The scavenging of H202 in the chloroplast was found to be catalyzed by a
peroxidase with ascorbate as the electron donor (Nakano and Asada, 1980, 1981).
Subsequently, it was shown that the addition of H202 to intact chloroplasts resulted in
a decrease in ascorbate content (Foyer et al, 1983; Law et al. (1983). Ascorbate
peroxidase (EC 220.127.116.11) is the first of a series of enzymes that comprise the
ascorbate-glutathione cycle or Halliwell-Asada pathway (Fig. 2). Dehydroascorbate
reductase (EC 18.104.22.168) is the next in the series and catalyzes the regeneration of
ascorbate from dehydroascorbate with reducing power provided by reduced
glutathione (GSH) (Nakano and Asada, 1981). This reaction also occurs
spontaneously at physiological pH (Foyer and Halliwell, 1976). Oxidized glutathione
(GSSG) is reduced to GSH in the presence of glutathione reductase (EC 22.214.171.124) and
NADPH (Foyer and Halliwell, 1976). Ascorbate and GSH have been detected in the
chloroplast at millimolar concentrations (Foyer et al., 1983; Law et al., 1983).
20"+2H-+ o2 + HO Ascorbate GSSG NADPH
2H0 "dehydro- 2GSH NADP
Figure 2. Ascorbate-glutathione cycle (adapted from Shaaltiel et al., 1988b). The
enzymes designated 1, 2, 3 and 4 are SOD, ascorbate peroxidase, dehydroascorbate
reductase and glutathione reductase, respectively.
As discussed previously, 02- and H202 can react in the presence of iron
complexes to give OH' (Rabinowitch and Fridovich, 1983). Equations 3 and 4 are
the partial reactions that result in the net reaction shown in Equation 2.
Fe(III) + 02- Fe(II) + 02 (3)
Fe(II) + H202 -- Fe(III) + OH- + OH' (4)
Equation 4 is known as the Fenton reaction. The source of the iron in vivo may be
phytoferritin (Dodge, 1989). Saito et al. (1985) have shown that O2- can cause the
release of iron from ferritin. The production of OH' has been demonstrated in
isolated thylakoids by detecting the formation of ethylene from methional (Elstner and
Hydroxyl radicals are extremely reactive. They react with almost all cell
constituents: phospholipids, nucleotides, amino acids, sugars and organic acids
(Halliwell and Gutteridge, 1989). The reactions are mainly hydrogen (H) abstraction,
addition, and electron transfer. Hydrogen abstraction is particularly important since it
is the means by which the peroxidation of cell membrane lipids is initiated (Halliwell
and Gutteridge, 1989). Hydroxyl radicals are scavenged by ascorbate, GSH and a-
tocopherol (Dodge, 1989).
Lipid peroxidation is the process in which unsaturated lipid molecules undergo
oxidative decomposition by means of a series of free-radical chain reactions. Two
types are recognized: enzymic and nonenzymic. Only the nonenzymic process is
pertinent and will be described briefly. Gutteridge (1988) and Halliwell and
Gutteridge (1989) may be consulted for a more comprehensive discussion. Chain
initiation occurs when a free radical with sufficient reactivity abstracts a H atom from
a methylene carbon of an unsaturated fatty acid. This results in a carbon radical that
undergoes stabilization to form a conjugated diene. Conjugated dienes react with
molecular oxygen giving rise to peroxyl radicals. Peroxyl radicals are important in
the propagation of the free-radical chain reaction because they stimulate the
peroxidation process by abstracting H atoms from adjacent fatty acid chains. Lipid
hydroperoxides are formed when the peroxyl radical combines with the abstracted H
atom. Termination reactions are also characteristic and these involve the reaction of
two radicals to form a nonradical (Gutteridge, 1988; Halliwell and Gutteridge, 1989).
Although OH' radicals can initiate lipid peroxidation, 02O radicals are not
sufficiently reactive and their charge precludes them from easily entering hydrophobic
lipid bilayers. The protonated form HO2' (the hydroperoxyl radical) is much more
reactive and, being neutral, easily penetrates membranes. It has been shown to
initiate peroxidation of fatty acids in vitro. However, at physiological pH the charged
species predominates and no evidence has yet been provided that either species
directly induces lipid peroxidation in vivo (Halliwell and Gutteridge, 1989).
Chemical Structure and Herbicidal Activity
The herbicidal properties of the bipyridiniums (bipyridyliums) were discovered
in 1955 (Brian et al., 1958). Currently only diquat (6,7-dihydrodipyrido[1,2-a:2',l'-
c]pyrazinediium ion) and paraquat (1,1l'-dimethyl-4,4'-bipyridinium ion) are in
commercial use as herbicides (Fig. 3). They are commonly synthesized as dibromide
and dichloride salts, respectively. However, only the cation is phytotoxic (Homer et
+ + CHi-N CN-CH
Figure 3. 6,7-dihydrodipyrido[1,2-a:2',1'-c]pyrazinediium ion (diquat) and 1,1'-
dimethyl-4,4'-bipyridinium ion (paraquat).
Summers (1980) reviewed the chemistry and other significant features of the
bipyridiniums. Paraquat is a diquaternary salt of 4,4'-bipyridine. It is comprised of
two pyridine rings, aromatic rings in which one carbon has been replaced by nitrogen
(N), linked at carbon 4. The term quaternary refers to the characteristic bonding of
the N atoms with the alkyl groups that results in a divalent cation (Halliwell and
Gutteridge, 1989). Diquat is a bridged diquaternary salt of 2,2'-bipyridine. The
pyridine rings are joined at carbon 2, and at the two N atoms with an ethylene bridge.
Paraquat salts were first synthesized in the late nineteenth century (Summers,
1980) and, as methyl viologen, have been used as oxidation-reduction indicators since
1932 (Michaelis and Hill, 1933). In the 1950s, it was determined that diquat was a
fast-acting, nonselective herbicide with desiccating action (Brian et al., 1958).
Primarily a contact herbicide, limited translocation was observed in the shoot.
Uptake was rapid, so that rain just after application did not reduce control. The
herbicide is inactivated in soil due to adsorption to soil colloids (Brian et al., 1958).
The bipyridiniums vary in herbicidal activity. The requirements for herbicidal
activity have been determined empirically. As compounds with low redox potentials,
they can undergo a single electron reduction (Michaelis and Hill, 1933).
BP2 + le- -- BP+ (5)
Only those compounds that can undergo a reversible univalent reduction to a stable
radical possess herbicidal activity (Homer et al., 1960). Radical stability is
characteristic of bipyridyls in which the pyridine rings are held in a coplanar
configuration. Stability is due to the ability of the radical to exist in a number of
resonance forms (Homer et al., 1960). The degree of activity has been shown to be
related to the redox potential of the compound. Compounds with lower potentials
required higher concentrations than those of higher potential to produce an equivalent
effect (Black, 1966; Homer et al., 1960; Kok et al., 1965). The highest herbicidal
activity occurs in bipyridyls with redox potentials in the range -0.3 to -0.5 V
(Summers, 1979). The redox potentials of paraquat and diquat are -0.35 and -0.45 V,
respectively (Homer et al., 1960). The size of the ion is also important because, only
compact molecules possess herbicidal activity (Summers, 1979).
Mechanism of Action
Mees (1960) showed that diquat had very slow herbicidal activity in darkness
but phytotoxicity in light was accelerated. He hypothesized that this pointed to the
involvement of the photosynthetic apparatus. Support for this hypothesis was
provided by Davenport (1963) who showed that several bipyridyls catalyzed
photophosphorylation in isolated chloroplasts. Zweig et al. (1965) observed that
diquat inhibited the reduction of NADP+ and proposed a site of action at the reducing
end of PSI. The site has been located more precisely at the terminal electron
acceptors FA/FB (Parrett et al., 1989). Fujii et al. (1990) have shown that Fg is the
main site of electron donation to paraquat and that FA has a lower affinity for the
herbicide (Fig. 4).
P700 A -- A -->FX-> A -- FERREDOXIN-+ NADP + NADP +
Figure 4. Scheme depicting the electron acceptors of PSI (Adapted from Dodge, 1989).
Harris and Dodge (1972a) found that continuous illumination of paraquat-
treated flax (Linum usitatissimum L.) cotyledons resulted in the rupture of the
tonoplast. This was followed by damage to mitochondrial and chloroplast
membranes. Membrane damage was also observed in dark-incubated tissue after 30
h. Further investigation provided more evidence of membrane injury in the presence
of paraquat (Harris and Dodge, 1972b). There was an increased rate of ion leakage
from tissue slices, enhanced accumulation of malondialdehyde and a progressive
change in lipid composition.
Baldwin (1969) concluded, NADPH deprivation due to diversion of electron
transport by paraquat, would result in a slow phytotoxicity similar to herbicides that
inhibit electron transport in PSII. Such deprivation does not explain the rapid wilting,
chlorosis and death of plants that commonly result from application of a bipyridyl
herbicide. Calderbank (1968) predicted the involvement of oxygen radicals in the
mode of action of bipyridyls. Farrington et al. (1973) felt that the membrane damage
observed by Harris and Dodge could be due to paraquat-derived toxic oxygen radicals
that were sufficiently long-lived to reach and react with membrane constituents. It
has now been established that the bipyridyl radicals react with 02 to produce 02-
(Asada and Kiso, 1973; Stancliffe and Pirie, 1971).
PQ+ + 02 pQ2+ + 02- (6)
Using pulse radiolysis, Farrington et al. (1973) determined the rate constant of the
reaction of the paraquat radical with 02 to be 7.7 x 108 M-1s- and proposed that 02O
was a good candidate for mediating paraquat toxicity. Evidence that 02- is generated
in paraquat-treated leaves was provided when a D-penicillamine-copper chelate was
shown to inhibit paraquat toxicity in flax cotyledons (Youngman et al., 1979). This
compound is capable of the dismutation of 02- (Youngman et al., 1979). It is
estimated that paraquat increases 02- production 10- to 20-fold, which is thought to be
sufficient to overwhelm the plant's protective mechanisms (Asada et al., 1977).
Carbon fixation in flax cotyledons was totally inhibited by paraquat by 5 h
after treatment (Harris and Dodge, 1972b). This is consistent with the diversion of
electrons from NADPH formation (Zweig et al., 1965) and the inactivation of Calvin
cycle enzymes by H202 (Kaiser, 1979). H202 is thought to be produced directly
during the autooxidation of bipyridinium radicals under aerobic conditions, as well as
indirectly via the spontaneous and catalytic dismutation of 02- (Black, 1966;
Davenport, 1963; Kok et al., 1965). The addition of paraquat to illuminated
chloroplasts resulted in oxidation of GSH to GSSG and a net loss of ascorbate (Foyer
et al., 1983; Law et al., 1983). This provided evidence that paraquat action produces
toxic oxygen species that require scavenging by these antioxidant molecules, and also
that the major function of the enzymes of the ascorbate-glutathione cycle was to
eliminate H202 in the chloroplast.
Paraquat-treated chloroplasts have also been shown to generate the OH'.
Initially, only the electron spin resonance signal of 02 was detected. On addition of
SOD, the signal changed to that of OH' (Dodge, 1989). Babbs et al. (1989) have
detected OH* in paraquat-treated plants at levels that were considered to be lethal. A
reaction proposed by Winterbour (1981) between the paraquat radical and H202
giving rise to the hydroxyl radical in the absence of a metal catalyst has since been
discounted (Halliwell and Gutteridge, 1989). Other highly reactive species have been
proposed to be generated in the presence of paraquat. These include 'crypto-OH'
(Youngman and Elstner, 1981; Youngman, 1984) and an iron (IV) species (Sutton et
In summary, paraquat applied to illuminated plants is reduced by PSI, which
decreases the NADPH concentration and forms paraquat radicals. Under the aerobic
conditions of the chloroplast, the paraquat radicals are instantly reoxidized by 02
which generates 02-. Superoxide is dismutated either spontaneously or by SOD to
form H202. The H202 not only blocks carbon fixation but also affects the ascorbate-
glutathione pathway since there is little reducing power for the functioning of
glutathione reductase. The enhanced production of 02 and H202 and their impaired
scavenging results in the formation of OH'. The indiscriminate reactivity of OH*
radicals leads to the destruction of cell constituents. One major consequence is the
peroxidation of cell membranes, including the tonoplast. There is a loss of osmotic
control and fluctuation in pH. The rupture of the tonoplast causes the release of the
toxic contents of the vacuole contributing to the general deterioration of the cell
With the increased occurrence of weed resistance to herbicides, reports on the
subject in the literature have become more frequent. There has been a problem of
inconsistency in the use of the terms tolerant and resistant. It was recently suggested
within a committee of the Weed Science Society of America that the term resistance
should supersede tolerance. The decision on terminology has been deferred until
other pest disciplines could be consulted on their terminology (Thill, 1993). The use
of the term resistance in this document is based on the definition by Warwick (1991).
Herbicide resistance may be defined as the condition whereby a plant
withstands the normal field dose of a herbicide, as a result of selection and
genetic response to repeated exposure to herbicides with a similar mode of
action. Susceptible plants are normally killed by recommended field doses.
(Warwick, 1991, page 95).
LeBaron and McFarland (1990) have compiled some characteristics of
herbicides that predispose weeds to develop resistance. Herbicides that target a single
site of action; those that effectively control a wide range of weed species; soil-applied
herbicides with residual activity that control germinating weeds throughout the growth
season, or are applied several times per season and; recurrent applications over
several seasons without rotating, alternating or combining with other types of
Although herbicide resistance has an economic cost because of a requirement
for alternative methods of control, there have also been benefits. The study of
herbicide resistant weeds has expanded our knowledge of many aspects of plant
biology and herbicide mode of action (LeBaron and McFarland, 1990). They are a
source of genes for breeding and genetically engineering herbicide resistant crops
(Mazur and Falco, 1989). In addition, there has been considerable interest in the use
of herbicide resistance as selectable markers in plant research (Mazur and Falco,
Herbicide resistant weeds tend to arise in situations where a single herbicide
exerts a considerable selection pressure. This is characteristic of herbicides with
prolonged residual activity. The residual action provides the selection pressure which
allows the mutant plants to thrive and reproduce. Paraquat is one of the exceptions.
It is a contact herbicide that is applied to the foliage. It is inactive in soil because it
becomes strongly adsorbed to clay colloids and organic matter (Brian et al., 1958).
Paraquat-resistant weeds have arisen at sites where there have been multiple
applications per year for many years or where paraquat was the only herbicide used
for a prolonged period (Table 1). Based on the criteria of LeBaron and McFarland
(1990) paraquat has a high risk for promoting the development of weed resistance.
Currently, eleven species of paraquat-resistant weeds are reported to have been
selected under such conditions. Although in some situations Lolium perenne L. is a
weed, it is also a forage crop. The resistant genotypes of this species, unlike the
other species listed in Table 1, were deliberately selected.
In vitro selection for paraquat resistance has also been accomplished in a few
plant species: Chenopodium rubrum L., Ceratopteris richardii Brongn. and Nicotiana
tabacum L. (Bhargava, 1993; Carroll et al., 1988; Furusawa et al., 1984; Hughes et
al., 1984). In addition, attempts have been made to genetically transform various
species to increase oxidative stress resistance. This was done by incorporating
additional genes encoding enzymes that scavenge oxygen radicals. Whereas the
enhanced production of SOD by transgenic Solanum tuberosum L. protected against
Table 1. Paraquat-resistant weeds (compiled from
LeBaron and McFarland, 1990).
Itoh and Matsunaka, 1990; and
paraquat toxicity, transformed N. tabacum and Lycopersicon esculentum MILL. plants
with enhanced SOD levels were as sensitive to paraquat as control plants (Perl et al.,
1993; Tepperman and Dunsmuir, 1990). N. tabacum plants transformed with
glutathione reductase genes from Escherichia coli were less susceptible to paraquat
than control plants (Aono et al., 1991, 1993).
Mechanisms of Paraquat Resistance
A review of the mechanisms of paraquat resistance was recently published
(Fuerst and Vaughn, 1990). The authors discussed five mechanisms that may impart
paraquat resistance. These included altered cuticular membrane; a change at the site
of action; metabolism of the herbicide to an inactive form; elevated levels of
protective enzymes; and rapid sequestration of the herbicide.
Altered Cuticular Membrane
A change in the composition or thickness in the cuticular membrane may limit
the penetration of the leaf surface by a xenobiotic. Although this hypothesis has been
investigated in a number of weeds and species, no evidence has been provided to
support this as a mechanism for resistance to paraquat. Leaves from resistant (R) and
sensitive (S) plants of Hordeum glaucum STEUD. gave the same kinetic response for
paraquat uptake, therefore, cuticular penetration was not involved (Bishop et al.,
1987). Fuerst et al. (1985) observed greater cuticular penetration by paraquat in the
R biotype than the S biotype of Conyza bonariensis (L.) CRONQ. This was
apparently due to the R biotype possessing a thinner cuticular membrane (Vaughn and
Fuerst, 1985). There was no difference in the short term uptake of L. perenne among
the three biotypes examined (Harvey et al., 1978).
Mutation at Site of Action
A change at the site of action in a paraquat-resistant biotype should have the
effect of reducing the efficiency of paraquat reduction and thus decreasing oxygen
radical formation and toxicity. Powles and Comic (1987) used isolated thylakoids of
H. glaucum to determine whether there were any differences in the affinity of PSI for
paraquat. They found no evidence to support this. Work with intact chloroplasts and
protoplasts indicated no difference in the inhibition of carboxylation. Electron
paramagnetic resonance was used in a direct analysis of the site of action in Conyza
(Norman et al., 1993). The spectra obtained did not appear to indicate any structural
differences at the site of action (Norman et al., 1993). These results are considered
to support earlier findings by Fuerst et al. (1985) that equivalent amounts of oxygen
uptake in both biotypes were induced by paraquat, diquat and triquat (7,8-dihydro-6H-
dipyrido[1,2-a:2',l'-c][1,4]diazepinediium ion) in PSI partial reactions.
Metabolism of paraquat in plants has never been reported (Summers, 1980).
Very little work has been done in the area of paraquat metabolism in resistant species.
When the R biotype of Lolium was treated with a sublethal dose of 14C-paraquat, no
metabolites were recovered (Harvey et al., 1978). Norman et al. (1993) found no
evidence of metabolism in C. bonariensis.
Protective enzymes and their products scavenge toxic oxygen species and
prevent lipid peroxidation of cell membranes and oxidation of other cell constituents.
To serve as a primary mechanism of resistance, elevated levels of the enzymes need
to be constitutive. Because of the rapid action of paraquat, induction of protective
enzymes is not considered a feasible primary mechanism of resistance (Harvey and
Harper, 1982). In Lolium, elevated constitutive levels of SOD, catalase and
peroxidase were detected in crude extracts of some R genotypes. It was proposed that
resistance was conferred by the enhanced ability to scavenge 02- and H202 (Harper
and Harvey, 1978).
Reports on the constitutive levels of protective enzymes in C. bonariensis have
been in conflict. Whereas Shaaltiel and Gressel (1986) ascribe paraquat resistance in
this species to increased amounts of stromal SOD, ascorbate peroxidase and
glutathione reductase, other workers have not been successful in confirming these
results (Norman et al., 1993). In addition, Vaughn et al. (1989) argue that the lack
of cross-resistance to other compounds that sensitize the formation of oxygen radicals
indicates that resistance is not due to enzymic protection. However, Shaaltiel's group
has not only provided evidence of cross-resistance to environmental and xenobiotic
oxidants, but have also demonstrated that chelators of copper and zinc, which inhibit
SOD, had a synergistic effect on paraquat phytotoxicity in R and S C. bonariensis
(Shaaltiel and Gressel, 1987a; Shaaltiel et al., 1988b). The reason for the
discrepancy in the data from the two groups may lie in the growth stage of plants
used in their studies. The juvenile rosette stages of both R and S biotypes of C.
bonariensis are more susceptible to paraquat treatment than flowering plants
(Amsellem et al., 1993, Jansen et al., 1990).
Conflicting results have also been obtained for C. canadensis (L.) CRONQ. A
Hungarian group reported that the levels of protective enzymes in cell-free extracts
were lower in the resistant biotype (P616s et al., 1988). Conversely, oxygen
detoxifying activities of intact chloroplasts from C. canadensis were twice as high in
the R biotype as in the S biotype (Itoh and Matsunaka, 1990).
There was no evidence of elevation of constitutive levels of protective enzymes
in leaves of H. glaucum and Arctotheca calendula (L.) LEVYNS (Powles and Comic,
1987; Powles and Howat, 1990). Tepperman and Dunsmuir (1990) found that
transformed tobacco (N. tabacum) plants with as much as 50-fold normal chloroplastic
SOD activity were not resistant to oxidative stress. They concluded that elevation of
all three protective enzymes may be necessary to produce a resistant phenotype
(Tepperman and Dunsmuir, 1990). Amplification of glutathione reductase has also
been accomplished in tobacco (Aono et al., 1991, 1993). The transgenic plants were
less susceptible to oxidative stress produced by paraquat and sulfur dioxide.
The effect of a paraquat sequestration mechanism would be to restrict
penetration of the herbicide to the site of action. Through a process of elimination it
was concluded that resistance in H. glaucum must be due to apoplastic sequestration
(Bishop et al., 1987; Powles and Comic, 1987). Similarly, without direct evidence,
sequestration has been proposed as the mechanism of resistance in C. bonariensis
(Fuerst et al., 1985; Norman et al., 1993). H. leporinum LINK unlike H. glaucum
did not exhibit apoplastic sequestration (Preston et al., 1992).
Exclusion by Plasmalemma
A sixth mechanism not covered by Fuerst and Vaughn (1990) is the exclusion
of paraquat from the symplast due to a change in the plasma membrane. DiTomaso
et al. (1992b) reported decreased penetration of paraquat through the plasmalemma of
the R biotype of H. glaucum. This contradicts the work of Powles and Comic (1987)
who determined that there was no change in the permeability of the plasmalemma or
chloroplast envelope in intact protoplasts.
Some workers have demonstrated translocational differences between paraquat-
resistant and -sensitive biotypes (Bishop et al., 1987; Fuerst et al., 1985; Preston et
al., 1992; Tanaka et al., 1986). Carroll et al. (1988) have termed these
translocational mutants. Translocation in the R biotypes was limited or nonexistent,
while extensive distribution of paraquat was observed in the S biotypes. Bishop et al.
(1987) regarded these differences in translocation by H. glaucum biotypes to be
related to their proposed sequestration mechanism. Since the R biotype of H.
leporinum does not sequester paraquat in the cell wall, the resistance mechanism in
the two species is thought to be different (Preston et al., 1992). Fuerst and Vaughn
(1990) also consider restricted mobility of paraquat in R biotypes to be due to rapid
sequestration of the herbicide.
The pattern of translocation of 14C-paraquat fed through the petiole was
investigated in C. bonariensis. While there was uniform movement of paraquat
throughout the lamina of the S biotype, paraquat was restricted to regions along the
veins (Fuerst et al., 1985). In related species, E. philadelphicus L. and E. canadensis
L., Tanaka et al. (1986) found restricted uptake through the cut ends of excised leaves
and limited translocation in both R biotypes. They suggested that the difference in
movement of paraquat in S and R biotypes may be correlated with the resistance
Paraquat-Resistant American Black Nightshade
In Florida, an increase in tomato (L. esculentum) production and very effective
weed control on raised beds has been achieved due to a combination of polyethylene
mulch and soil fumigation. These measures have little effect on weeds between beds,
where additional means of control are required. Of the herbicides labelled for weed
control in tomatoes in Florida, paraquat is the most widely used for weed control in
row middles (Gilreath et al., 1987). Directed postemergence application minimizes
crop injury. One of the disadvantages of relying on paraquat is that at least three
applications per season are necessary for effective control (Gilreath and Stall, 1987).
American black nightshade (Solanum americanum MILL.) is a pernicious weed
of tomato production in Florida. Herbicide selectivity for nightshade (Solanum spp.)
in a tomato crop is difficult to obtain since they are both solanaceous plants with
similar physiology (Stall et al., 1987; Weaver et al., 1987). In the absence of an
effective control strategy, nightshade can competitively reduce tomato yields
(McGiffen et al., 1992; Weaver et al., 1987).
In 1987, Gilreath and Stall reported that growers in the Immokalee-Naples
area had observed a reduced efficacy of paraquat on American black nightshade.
Since Youngman et al. (1979) had found that a copper chelate of D-penicillamine,
with superoxide dismutase activity, inhibited paraquat toxicity in flax cotyledons; Stall
et al. (1987) hypothesized that the poor control of American black nightshade by
paraquat was due to copper-containing fungicides. Evidence was provided that
confirmed that some fungicides/bactericides decreased American black nightshade
control by paraquat (Bewick et al., 1990b). However, even in the absence of the
pesticides, a biotype of nightshade collected from the Naples area was shown to be
12-fold more resistant to paraquat than a sensitive biotype (Bewick et al., 1990a).
The purpose of this study was to elucidate the primary mechanism of paraquat
resistance in American black nightshade, using physiological and biochemical
techniques. To this end, the following null hypothesis was proposed. Paraquat
resistance in American black nightshade can be explained by a previously established
mechanism. The alternative hypothesis was: paraquat resistance in this species is due
to a mechanism that is unique and not previously reported.
The major objectives were two-fold. The first was to determine whether the R
biotype and a S biotype differed in their ability to deal with paraquat. Specifically,
the intention was to investigate whether (a) in the R biotype paraquat was prevented
from penetrating to the site of action; (b) it possessed an insensitive site of action or
(c) was capable of metabolizing the herbicide. The second major objective was to
compare the susceptibility of the two biotypes to oxidative stress. This area of
investigation was intended to (a) compare the ability of the two biotypes to generate
oxygen radicals and (b) compare their defenses against oxidative stress.
COMPARATIVE PARAQUAT UPTAKE
Paraquat is a postemergent herbicide that induces rapid desiccation (Summers,
1980). A member of the bipyridinium family, its site of action is at the FAFB iron-
sulfur protein in PSI where it oxidizes photosynthetic electron transport (Fujii et al.,
1990; Parrett et al., 1989). The paraquat cation is reduced to the radical form, which
is instantly reoxidized by molecular oxygen, giving rise to 02- (Farrington et al.,
1973). The accumulation of H202, resulting from the dismutation of 02-, inhibits the
fixation of carbon (Kaiser, 1976, 1979; Robinson et al., 1980). However, the
majority of cell damage is considered to result from the reaction of cell constituents
with the OH* (Dodge, 1989). The latter is generated by the iron-catalyzed Haber-
Weiss reaction between 02- and OH* (Halliwell and Gutteridge, 1989).
Six mechanisms of acquiring paraquat-resistance have been proposed:
(1) reduced cuticular penetration,
(2) sequestration of paraquat away from the site of action,
(3) exclusion by the plasmalemma,
(4) elevated levels of protective enzymes,
(5) mutation at the site of action and
(6) paraquat metabolism (Fuerst and Vaughn, 1990; DiTomaso et al., 1992b).
Although paraquat resistance has been reported in twelve weed species, evidence has
been provided in support of just three mechanisms. Elevated levels of superoxide
dismutase, ascorbate peroxidase and glutathione reductase have been measured in
paraquat-resistant L. perenne, E. philadelphicus and C. bonariensis (Harper and
Harvey, 1978; Itoh and Matsunaka, 1990; Shaaltiel and Gressel, 1986; Shaaltiel et
al., 1988a; Youngman and Dodge, 1981). These enzymes of the Halliwell-Asada
pathway are concerned with the scavenging of toxic oxygen species.
Apoplastic sequestration was considered a plausible explanation of resistance in
C. bonariensis and H. glaucum (Bishop et al., 1987; Fuerst and Vaughn, 1990;
Norman et al., 1993). This may involve adsorption of the herbicide in the apoplast.
Cytoplasmic accumulation in an organelle other than the chloroplast is also possible.
Gradual recovery of normal fluorescence, following treatment in only the R biotype,
led Lehoczki and Szigeti (1988) to propose that resistance in C. canadensis was due
to sequestration by an unknown mechanism.
Conflicting results were reported for paraquat penetration of the plasmalemma
in paraquat-resistant H. glaucum. Powles and Comic (1987) found no evidence of
reduced plasmalemma permeability in the R biotype. However, DiTomaso et al.
(1992b) found that paraquat resistance in H. glaucum was due to reduced herbicide
transport across the plasma membrane. In a more recent report, it was proposed that
resistance in this species was due to enhanced vacuolar sequestration of paraquat
(Lasat et al., 1994).
In 1990, Bewick et al. reported that a biotype of American black nightshade
was 12-fold more resistant to paraquat than a sensitive biotype. Although none of the
previous research has demonstrated the involvement of the cuticle, one of the
objectives of this work was to determine whether paraquat resistance in American
black nightshade was due to differential cuticular penetration. In addition, the
involvement of apoplastic sequestration and/or exclusion of the herbicide from the
symplast in conferring resistance was investigated.
Materials and Methods
Seeds were collected from a sensitive (S) population and a resistant (R)
population of American black nightshade in Alachua and Collier counties, respectively
(Stall et al., 1987). Seedlings at the cotyledon stage were transplanted into 10 cm
pots filled with a commercial potting mix (Metro-mix, Gracewood Horticultural
Products, Cambridge, MA). Fertilizer and water were provided as a 1/3 strength
Hoagland's solution (Hoagland and Arnon, 1950) via an automatic watering system
fitted with a fertilizer injector, three times per day.
One hundred plants of each biotype were scored for sensitivity to paraquat at
the three to four true leaf stage. A single fully expanded leaf of each paraquat-S plant
was painted with 48 1M paraquat (Ortho paraquat CL, Chevron Chemical Co., Ortho
Division, San Francisco, CA). The R biotype was painted with paraquat
concentrations from 0.7 mM to 6.3 mM. The leaves were assessed visually for injury
symptoms 24 h after painting and the most sensitive and the most resistant plants were
selected for experimental use.
Leaf disks 5 mm in diameter were used in all experiments. Disks were cut
from fully expanded leaves using a cork borer, taking care to avoid the midrib and
main veins. The disks were rinsed three times with distilled water to clean the
surfaces and to remove adhering debris from the cut edge. Disks were then allowed
to float on distilled water until needed.
The specific leaf weight was determined as follows. Twenty-five leaf disks
were blotted dry on filter paper and weighed. This was done on 42 sets of 25 leaf
disks for each biotype prior to use in the first paraquat uptake experiment. Based on
a total leaf area of 4.9 cm2, the specific leaf weight was expressed in units of g m2.
The buffer used in the assay of paraquat uptake consisted of 5 mM MES-TRIS
pH 5.8, 0.2 mM CaC12, 0.1% Tween 20 and 0.7 /iM (0.1 PCi) 14C-paraquat. The
ring-labeled paraquat was provided by ICI Agrochemicals (Jealott's Hill Research
Station, Bracknell, Berkshire, UK). Three milliliters of uptake buffer were added to
each 25 mL Erlenmeyer flask. Twenty-five leaf disks were blotted dry on filter paper
and placed in the buffer. The flasks were placed on a rotary shaker set at 100 rpm.
Time dependent uptake was performed for 0, 5, 10, 20, 30, 45, 60, 75, 90 and 180
minutes. A 50 jL aliquot of the buffer was sampled prior to addition of the disks and
at the end of the uptake period. Aliquots were placed in 20 mL liquid scintillation
vials. Ten milliliters of Aquasol universal liquid scintillation fluid (E.I. du Pont de
Nemours & Co., Boston, MA) were dispensed to each vial. The number of
disintegrations per minute (dpm) were determined by liquid scintillation spectrometry
(LSS) using a Beckman LS 7500 Liquid Scintillation System (Beckman Instruments
Inc., Fullerton, CA). Paraquat absorbed by the leaf disks was calculated as the
difference between the measurements taken before and after the uptake period.
The treatments were replicated 3 times and the experiment was conducted
three times. The data from the three experiments were combined and analyzed as a
randomized complete block design. The SAS System (SAS Institute Inc., Cary, NC)
was used for data analysis. Analysis of variance (ANOVA) was performed using the
general linear models (GLM) procedure. The nature of the response was determined
by regression analysis of the uptake means against time.
In order to clarify whether the paraquat concentration being used was injuring
the leaf disks, an assay for ethane was performed on nontreated and paraquat-treated
disks. After the uptake period, the leaf disks were separated from the uptake solution
with a Buchner funnel and quickly rinsed with distilled water. The disks were blotted
dry and placed in 21 by 70 mm glass vials. The vials were sealed with serum
stoppers and illuminated for 5 min at 300 /E m2 s-1. A 1 mL sample of the
headspace was analyzed for ethane by gas chromatography using a Hewlett-Packard
5890 Series II gas chromatograph (Hewlett-Packard Company, Santa Clara, CA) fitted
with an alumina column. The temperature settings of the injector, oven and detector
were 2000C, 500C and 2500C, respectively.
For each biotype, 50 leaf disks were blotted dry and placed in a 25 mL
Erlenmeyer flask containing 3 mL uptake buffer. A 50 1L aliquot of the buffer was
removed before and after the 30 min loading period in order to determine uptake.
The flasks were allowed to shake during uptake and the subsequent efflux period at
100 rpm. The disks were separated from the uptake buffer by vacuum filtration and
transferred to a clean flask. Paraquat on the surface of the leaf disks was removed by
quickly rinsing with 2 mL of efflux buffer (uptake buffer minus paraquat). The flask
was evacuated using a water pump and a 3 mL volume of efflux buffer was added to
the flask, 1 mL of which was sampled at the end of the efflux interval. Paraquat
efflux was determined after 5, 10, 15, 30, 45, 60, 90 and 120 min. The amount of
14C-paraquat contained in the aliquot was determined by LSS and adjusted for a 3 mL
The treatments were replicated 3 times. The experiment was conducted twice
and the data were combined. A repeated-measures ANOVA was required since
sequential efflux measurements were done over time on the same leaf disks. The
analysis was performed using a split-plot approach. This allowed for a more stringent
hypothesis test of the main effect of biotype.
A difference between the biotypes in the efflux of paraquat could have been
due to adsorption within the apoplast or differential uptake into the symplast. To
distinguish between these possibilities an experiment was conducted to characterize
paraquat compartmentalization. Short-term efflux over a 15 min period was
determined in order to compare the amount of paraquat present in an unbound state in
the apoplast. The loading period was performed as described in the previous section.
Instead of rinsing, paraquat on the surface of the leaf disks was removed by blotting
with filter paper. Efflux was also performed similarly, except that the efflux period
was restricted to 15 min.
On completion of efflux of unbound apoplastic paraquat, another 3 mL volume
of efflux buffer was added. The plasma membranes of the leaf disks were then
disrupted by two cycles of freezing and thawing. The freezing was done by partially
immersing the flasks in liquid nitrogen. The buffer was thawed at room temperature.
The flasks were then allowed to shake for 15 min to promote maximum release of
cytoplasmic paraquat. A 1 mL aliquot was sampled for LSS and the flask evacuated.
Finally, the tissue was exposed to 3 mL of desorption buffer (5 mM MES-
TRIS pH 5.8 and 5 mM CaCl2) for 15 min in order to displace bound apoplastic
paraquat. A 1 mL aliquot was sampled and subjected to LSS.
Silanization of Glassware
During preliminary uptake experiments it became apparent that 14C-paraquat
was being lost from the uptake buffer which could not be explained by uptake into the
leaf tissue. A desorption procedure with 0.5 M NaCl solution revealed that paraquat
was being adsorbed to the glass surface of the Erlenmeyer flask. In order to prevent
this from occurring, the glassware was siliconized using 5% dichlorodimethylsilane in
toluene in a chemical hood. After evaporation of the toluene, the glassware was
baked for 2 h at 1800C or rinsed thoroughly with water before use (Sambrook et al.,
The paraquat uptake study was conducted in order to determine whether
differences in cuticular penetration was the mechanism of paraquat resistance in S.
americanum. The biotypes did not differ in their rate of paraquat absorption (Fig. 5).
Therefore, the cuticular membrane was not a barrier to uptake in the R biotype and
does not appear to be involved in conferring resistance. Time-dependent paraquat
uptake in S. americanum was best described by a quadratic relationship. Although
the S biotype had a greater mass per unit leaf area (Table 2), this did not appear to
influence paraquat uptake. Ethane evolution by the leaf disks was measured in order
to determine whether membrane damage occurred during the course of uptake. There
was no difference (p > 0.05) between the amount of ethane evolved by untreated and
paraquat-treated disks over the 3 h uptake period (Fig. 6). In addition, no biotypic
difference was observed. Therefore, the 0.7 MM 14C-paraquat did not cause
detectable membrane damage in either biotype over the course of the experiments.
30 60 90 120 150
Figure 5. Time-dependent 14C-paraquat uptake in S. americanum. The markers indicate
the means of 3 experiments. The paraquat-S biotype is designated by (A) and paraquat-R
biotype by (0).
Table 2. Specific leaf weight of paraquat-resistant and -sensitive American black
nightshade. Twenty-five 5 mm leaf disks were weighed. There were 42 replicates per
Biotype weight (g m-2)
Sensitive 157.6 (2.2)a
Resistant 148.8 (1.2)
aFigures in brackets are standard errors of the mean.
0 30 60 90 120 150 180
Figure 6. Comparison of ethane evolution from nontreated S biotype (A) and (0), and
paraquat-treated (0.7 /M) S biotype (A) and R biotype (*).
Paraquat efflux studies were used to characterize paraquat uptake and
distribution. Disks were allowed to load paraquat for 30 min prior to determining
efflux. Efflux data were expressed as a percentage of absorbed paraquat. Two
phases of efflux were observed: a rapid phase occurring during the first 10 min,
followed by a slower phase (Fig. 7). The rapid phase is thought to be due to efflux
from the apoplast, and the slower phase to be due to efflux from the symplast
(Nandihalli and Bhomik, 1991; Van Ellis and Shaner, 1988). Over the 2 h efflux
period, more paraquat effluxed from the S biotype than the R biotype. There was a
significant interaction between biotype and time (P <0.05) indicating that the rate of
paraquat efflux was not the same. This pattern of efflux could be due to greater
adsorption in the apoplast or increased uptake into the symplast by the R biotype.
Therefore, compartmentalization studies were conducted to determine the distribution
of paraquat after a loading period of 30 min.
5 10 15 30 45 60 90 120
Figure 7. Time-course comparing the efflux of 14C-paraquat from paraquat-sensitive (S)
and -resistant (R) biotypes of American black nightshade.
As for paraquat efflux, the data from the compartmentalization study also were
expressed as a percent of paraquat uptake (Table 3). The paraquat measured during
short-term efflux constituted the unbound apoplastic component. The total amount of
cell-free paraquat of the S biotype exceeded that of the R biotype. The rate of efflux
of paraquat was faster in the S biotype. Although more paraquat was adsorbed by the
R biotype than the S biotype, there was no difference in the percentage of absorbed
paraquat that penetrated into the cytoplasm. This indicated that it was unlikely that
paraquat resistance in American black nightshade was due to apoplastic sequestration
or to exclusion by a less permeable plasmalemma.
Table 3. Compartmentalization of 14C-paraquat. Free apoplastic paraquat was
determined by short-term efflux, bound apoplastic paraquat by desorption and
cytoplasmic paraquat by membrane disruption.
Treatment Sensitive Resistant LSD
Uptake 52284 (2567) 65429 (5801) 13815
Efflux (min) % of uptake
2.5 24.4 (0.7) 15.6 (1.3)
5.0 5.8 (0.3) 3.4 (0.2)
7.5 3.2 (0.2) 1.5(0.1)
10.0 1.8(0.1) 0.8(0.1)
15.0 1.8(0.2) 0.8(0.1)
Total efflux 37.0 (1.0) 22.2 (1.2) 3.6
Desorption 11.9 (0.6) 14.9 (0.8) 2.6
Freeze/thaw 40.0 (4.5) 47.6 (4.5) 9.6
Total 88.9 (5.0) 84.7 (5.3) 14.5
Initially, the intention was to study paraquat penetration through isolated
cuticular membranes. However, enzymic removal of the cuticular membranes could
not be consistently performed with S. americanum. When isolations were successful,
the cuticles proved to be too fragile to withstand subsequent procedures. The use of
leaf disks was a compromise. It allowed the assessment of paraquat uptake but did
not distinguish between cuticular penetration and plasmalemma penetration. Other
studies of paraquat uptake in resistant weeds have utilized intact attached leaves
(Bishop et al., 1987) and detached leaves, the petioles of which were maintained in
water (Fuerst et al., 1985). In the first case, differing rates of translocation of
paraquat away from the uptake site may have masked differences in cuticular
penetration. In both cases, the authors claimed to have determined the amount of
paraquat that partitioned into the cuticle by means of chloroform washes. This
technique will estimate only the fraction that is found in the epicuticular waxes and
not the entire cuticle (Bewick et al., 1993). Only negligible amounts of paraquat
were retrieved, and did not correlate with resistance (Bishop et al., 1987; Fuerst et
In concurrence with our findings, paraquat uptake through the cuticular
membrane was not correlated with paraquat resistance in C. bonariensis, H. glaucum
or L. perenne (Fuerst et al., 1985; Bishop et al., 1987; Harvey et al., 1978;
Youngman and Dodge, 1981). In H. glaucum and L. perenne, no difference in
uptake was observed (Bishop et al., 1987; Harvey et al. 1978). However, Fuerst et
al. (1985) reported greater cuticular penetration in the R biotype of hairy fleabane
while Youngman and Dodge (1981) found no biotypic difference in uptake in the
The concentration of paraquat in the solutions used in these studies was
restricted to a relatively low level because 8.5 /M paraquat was observed to cause
membrane damage in both biotypes within 2 h (Chapter 3). More extensive
characterization of uptake and compartmentalization has been accomplished with
maize roots (DiTomaso et al., 1993; Hart et al., 1992a). These authors had much
more flexibility in the concentrations and time courses they could utilize because root
tissue lacks chloroplasts and the phytotoxic action of paraquat is much slower. Such a
system was not practical in this situation since the effect of the cuticular membrane
was of interest.
Examination of the kinetics of paraquat uptake across the plasmalemma in
maize roots revealed a saturable component, which indicated a protein-mediated
system (Hart et al., 1992a). Further, uptake apparently occurred via the putrescine
membrane carrier system (Hart et al., 1992b). Paraquat uptake by plant cells is
thought to occur by passive transport (facilitated diffusion), the driving force being its
electrochemical gradient across the plasmalemma (Hart et al., 1993). This has been
corroborated by Zer et al. (1993), using a suspension of dark-grown Phaseolus
vulgaris L. cells, who found that spermine competitively inhibited the uptake of
paraquat. Uptake of paraquat was also reduced by low temperature and by the
addition of an uncoupler, carbonyl cyanide m-chlorophenylhydrazone (CCCP)1 (Zer
et al., 1993).
1CCCP is a proton-specific ionophore which is used to disrupt the pH gradient across
Reduced paraquat uptake in an R biotype of H. glaucum was interpreted as
being due to a change in the putrescine membrane carrier system (DiTomaso et al.,
1992a). The involvement of apoplastic sequestration or binding to the cell membranes
were eliminated as possibilities in these studies. This contradicted an earlier report
that showed no difference in carbon dioxide fixation by S and R protoplasts,
indicating that the resistance mechanism was not differential plasmalemma
permeability (Powles and Comic, 1987). More recently, based on a study of
paraquat uptake into plasmalemma vesicles isolated from the leaves of two H.
glaucum biotypes, Hart and DiTomaso (1994) have revised their earlier position.
Data derived from investigations of paraquat uptake by H. glaucum roots was
interpreted as showing no dramatic differences between the two biotypes. Instead, it
was proposed that paraquat resistance is due to enhanced sequestration of the
herbicide in the vacuole by the R biotype (Lasat et al., 1994).
By a process of elimination of other mechanisms, some groups have concluded
that the R biotype must be sequestering paraquat (Fuerst et al., 1985; Norman et al.,
1993; Powles and Comic, 1987; Lehoczki and Szigeti, 1988). No direct evidence
was provided to demonstrate paraquat binding in the apoplast or accumulation in an
organelle or vacuole. On the contrary, the lack of binding to insoluble cell wall
fragments observed by Fuerst et al. (1985) was later reinterpreted by Norman et al.
(1993) to perhaps be due to a requirement for a structurally intact cell wall.
In summary, neither the cuticle nor the plasmalemma was a barrier to the
penetration of paraquat into the cytoplasm of S. americanum. It is unlikely that the
primary mechanism of paraquat resistance is due to apoplastic sequestration, because
three-fold more paraquat was detected in the cytoplasm than was adsorbed in the
apoplast of the R biotype by the end of the 30 min loading period. In addition, there
was no biotypic difference in the percentage of cytoplasmic paraquat. These results
suggest that the resistance mechanism is located in the cytoplasm. These techniques
did not eliminate the involvement of symplastic sequestration. Therefore, the
possibility of paraquat resistance being due to restricted penetration to the site of
action will be addressed in the next chapter.
DIFFERENTIAL BIOCHEMICAL RESPONSES OF RESISTANT AND
SENSITIVE BIOTYPES OF AMERICAN BLACK NIGHTSHADE
TO BIPYRIDINIUM HERBICIDES
In the previous chapter, it was shown that the S and the R biotypes of
American black nightshade accumulated similar amounts of paraquat in the symplast.
A number of possibilities exist that may explain resistance occurring at the
cytoplasmic level. If the R biotype were capable of rapid sequestration of paraquat in
the vacuole or some other organelle, this could drastically reduce the amount that was
available for penetration into the chloroplast. This would result in reduced oxidative
stress. Rapid metabolism of the herbicide is unlikely since paraquat metabolism has
never been reported (Norman et al., 1993). In the absence of sequestration and
metabolism, it is also possible that the site of action in PSI may have undergone
mutation resulting in less efficient paraquat reduction. No evidence has thus far been
provided in support of this mechanism. Paraquat resistance can also be conferred by
elevated levels of protective enzymes (Harper and Harvey, 1978; Itoh and Matsunaka,
1990; Shaaltiel and Gressel, 1986).
Paraquat toxicity is mediated by oxygen radicals (Farrington et al., 1977).
Oxygen radicals are capable of initiating and propagating the peroxidation of cell
membrane lipids (Halliwell and Gutteridge, 1989). Consequently, techniques that can
quantify membrane damage occurring in response to paraquat application can be
useful indicators of whether paraquat is penetrating to the site of action and exhibiting
activity. Riely et al. (1974) found that the production of lipid peroxides and ethane
evolution by mammalian tissue were closely correlated and proposed that in vivo lipid
peroxidation could be assessed by measuring ethane evolution. Ethane evolution has
also been demonstrated in plant tissue exposed to conditions favoring peroxidation
(Konze and Elstner, 1978). The detection of ethane is regarded as a useful means of
characterizing the activities of some herbicides, if the results can be corroborated by
other methods of measuring lipid peroxidation (Elstner and Pils, 1979). Ethane
measurement has been used in the assessment of diphenyl ether and paraquat toxicity
as well as the effect of the fungal toxin cercosporin on tobacco membrane lipids
(Daub, 1982; Finckh and Kunert, 1985; Kunert and B6ger, 1984; Youngman et al.
The measurement of electrolyte leakage is another common method of
detecting lipid membrane damage (Whitlow et al., 1992). When plant tissue is
exposed to conditions that cause membrane injury, the ability of the plasma membrane
to regulate the transport of ions is impaired. Leakage of electrolytes from the tissue
occurs, which can be detected by measuring the conductance of the bathing solution.
Prior to Munday and Govindjee (1969), penetration of intact cells by paraquat
had not been demonstrated. These authors showed that after adding paraquat to algal
cells, there was a progressive quenching of peak fluorescence (Fp) over 15 minutes,
which proved that the herbicide penetrated to the site of action. In paraquat-resistant
plants chlorophyll fluorescence induction measurement has been used to study the
presence and activity of paraquat at its site of action in the chloroplast (Fuerst et al.,
1985; Lehoczki et al., 1992; Vaughn et al., 1989).
The objectives of this study were to investigate whether paraquat can penetrate
to the site of action in the R biotype, and to evaluate the biochemical responses of a
paraquat-S and -R biotypes of American black nightshade to diquat and paraquat.
This was accomplished indirectly through the detection of ethane and ion leakage and
directly by measurement of chlorophyll a fluorescence.
Materials and Methods
Leaf disks, 13 mm in diameter, were cut with a cork borer and held in water
to prevent wilting. The disks were rinsed three times with deionized water before
use. Ten leaf disks were placed in 25 ml Erlenmeyer flask containing 8 ml of
incubation solution. The incubation solutions contained 5 mM MES pH 5.2, 1 mM
KCN and paraquat at one of three rates: 0, 8.5 and 85 iM. The KCN was included
to inhibit the protective enzymes of the Halliwell-Asada pathway. The flasks were
foil-wrapped in order to exclude light. The leaf disks were vacuum infiltrated with
the incubation solution and the flasks were sealed with rubber septum stoppers.
The leaf disks were allowed to absorb paraquat in darkness for 2 h. The foil
was then removed and the disks exposed to light at 300-400 /tE m-2s-1. The light was
provided by 3 incandescent fixtures and was filtered through a 0.2 M CuSO4 solution
to reduce heat. The disks were illuminated for 2, 4, 6, and 8 h. Each combination
of paraquat and period of illumination was replicated 3 times.
The headspace of each flask was mixed by repeatedly filling and releasing a 1
ml tuberculin syringe equipped with a sub-Q 26G 16 mm needle (outside diameter =
0.46 mm). A 1 mL aliquot of the headspace gas was removed and assayed for ethane
using a Hewlett-Packard HP5710A gas chromatograph equipped with a flame
ionization detector and a packed alumina column. The oven temperature was set at
100C. Ethane was quantified by comparison with a 10 AL L-1 standard (Scott
Specialty Gases, Inc., Plumsteadville, PA). The experiment was repeated and the
data from 2 experiments were combined. The data were analyzed as a randomized
complete block design. The procedure used was analysis of variance (ANOVA).
Means were generated and are presented with standard errors.
Plants of the S and R biotypes were scored (as described previously) in order
to select plants that were most sensitive and most resistant to paraquat. Several
different groups of plants were used in the performance of these studies and the
relative resistance and sensitivity varied from group to group. Leaf disks were cut to
give a diameter of 10 mm. The disks were washed three times with distilled water
and floated on distilled water until required.
These experiments were performed using scintillation vials that were silanized
to prevent adsorption of paraquat to the glass. A preliminary experiment was
conducted to compare the time-dependent effect of a lethal dose of paraquat on ion
leakage from leaf disks incubated in light (300 1AE m-2s-1) and in darkness. Time-
courses of 36 h for the illuminated incubations and 48 h for the dark incubations were
used. The dark treatments were done at room temperature (26 C). The temperature
of the illuminated treatments was maintained at 26 C, by immersing the vials in a
water bath and by using a water filter to reduce infrared radiation from the lamps.
Five leaf disks were placed in 5 mL aliquots of 0.3 M paraquat solution.
The initial conductivity of the paraquat solutions was determined prior to the
addition of the leaf disks using a YSI Model 31 Conductivity Bridge (Yellow Springs
Instrument Co. Inc., Yellow Springs, Ohio). Final conductivity was measured at the
end of the incubation period, and total conductivity measured after 2 cycles of
alternate freezing and thawing. Ion leakage was expressed as a percentage of the total
Concentration-dependent assays were also conducted. In the first experiment
the disks were allowed to incubate in light (300 pE m-2s-1) for 8 h or darkness for 12
h at 260C. The paraquat concentrations ranged from 0 to 0.43 mM for the S biotype
and 0 to 8.5 mM for the R biotype. In the next 2 experiments, the concentration
ranges were reduced to 0 to 17 jM for the S biotype and the upper limit for the R
biotype was 2.1 mM. The incubation times were extended to 12 h in the light and to
24 h for dark incubation. In all 3 experiments the treatments were replicated 3 times.
The data were sorted by illumination type and biotype. The raw data as well as
treatment means were analyzed using the REG procedure of SAS. 150 values (the
paraquat concentration required to cause loss of 50% of the electrolytes) were
determined. 150 values and 95 % confidence limits were calculated using inverse
regression for a linear model. Where the model was not linear, the data were first
transformed using a square root or natural log transformation.
Chlorophyll A Fluorescence
Chlorophyll a fluorescence was measured using a Plant Productivity
Fluorometer Model SF-20 (Richard Brancker Research Ltd, Ottawa, Canada). The
experiments were done with 10 mm leaf disks prepared as described previously. The
leaf disks were floated on a dish of distilled water and held in darkness for 1 h. The
experiments were conducted in a darkroom using a green safelight. After dark-
adaption, the leaf disks were blotted dry with filter paper. Disks were placed adaxial
side down in 15 mL of 20 pM diquat or 1.7 mM paraquat in plastic Petri plates. The
latter were covered with foil to exclude light. Leaf disks were allowed to incubate for
10 to 60 min. At the end of the incubation period, disks were removed from the
herbicide solution and placed adaxial side down on the sensing probe of the
fluorometer. Initial, peak and terminal fluorescence were measured. These terms
correspond to Fo, Fp and Fs using the nomenclature of van Kooten and Snel (1990).
Fs was measured 50 s after F0. Variable fluorescence (F,) was determined by taking
the difference between Fp and Fo. Data for a single diquat experiment and 2 paraquat
experiments were used in the respective analyses. The experiments were analyzed
using the ANOVA procedure of SAS. The experimental designs were randomized
Concentration-dependent fluorescence was also measured. The dark-adaptation
was combined with herbicide incubation. In the preliminary experiment, which
included both diquat and paraquat, the dark-adaptation/incubation period was 30 min.
The herbicide concentrations ranged from 0 to 20 mM for diquat and 0 to 17 mM for
paraquat. It was intended that the same concentrations be used for both herbicides,
however, the hydration of paraquat was inadvertently not considered when the
solutions were prepared. Concentration-dependent experiments were repeated 2 more
times using paraquat only and an incubation period of 20 min. An additional
concentration of 170 mM was added for the R biotype only. The experiments were
analyzed as randomized complete block designs using the GLM procedure of SAS.
The analysis was adjusted to accommodate the unequal number of treatments used in
the last concentration-dependent experiment. This required using the type IV sums of
squares for hypothesis testing (Littel et al., 1992).
Biotypes differed in the amount of ethane they produced (Fig. 8). However,
the response was not the same over all levels of time and paraquat concentration.
Since paraquat was expected to cause membrane injury, the data were sorted by
S 1.00 1.00
2 4 6 8 2 4 6 8
Time (h) Time (h)
Z 0.50 L 0.50
2 4 6 8 0 25 50 75 100
Time (h) Paraqual (uM)
Figure 8. Ethane evolution over time by leaf disks treated with 0 (A), 8.5 (B) and 85
pM (C). Paraquat-dependent ethane production by the S biotype (A) and the R biotype
(0) is shown in (D).
paraquat concentration and reanalyzed. The nontreated (0 paraquat) leaf disks
evolved ethane over time (Fig. 8a). This was expected since inefficiencies in
photosynthetic electron transport chain result in the reduction of 02 to form 02-
(Asada et al., 1974). The addition of KCN was intended to inhibit the scavenging of
02-, so that possible differences in constitutive levels of protective enzymes would not
confound the results. Although there was a tendency for the S biotype to produce
more ethane than the R biotype, this was not significant (Fig. 8a). The addition of
paraquat resulted in a considerable increase in ethane production. There was a
significant interaction (P<0.05) between biotype and time for the 8.5 and the 85 /M
concentrations. This was due to a more rapid rate of ethane evolution by the S
biotype (Fig. 8b,c). Increasing the concentration of paraquat from 8.5 to 85 /M
resulted in a small but significant (P <0.05) increase of ethane production by both
biotypes (Fig. 8d). These results indicate that paraquat was capable of penetrating the
chloroplasts of the R biotype and generating oxidative stress. However, membrane
damage occurred more rapidly in the S biotype than in the R biotypes.
There was no biotypic difference in the amount of ion leakage in the absence
of paraquat (Fig. 9a). Ion leakage increased with time of incubation in both darkness
and light. However, in the former case there was a biotype by time interaction
(p <0.05) which indicated that the slope of the response was unique for each biotype.
The time-dependent response was the same for both biotypes in the light. The
addition of paraquat increased ion leakage significantly (Fig 9b). Whereas untreated
leaf disks leaked more ions in the darkness than in the light, the reverse was true for
paraquat-treated disks. Ion leakage occurred more rapidly under illuminated
conditions when paraquat was present. This was expected since bipyridyl toxicity is
accelerated due to oxygen radical production resulting from paraquat photoreduction
in the chloroplast (Mees, 1960). There was a significant interaction (p <0.05)
between biotype and time which was indicative of differences in biotypic response
over time. Further analysis indicated that the interaction was due to a lower amount
of ion leakage recorded for the R biotype at 24 h than for the S biotype (Fig. 9b).
The S biotype had a more rapid rate of ion leakage than the R biotype when incubated
12 24 36
Figure 9. Time-course of ion leakage from untreated leaf disks (A) or leaf disks
incubated in 0.3 mM paraquat (B). R and S biotypes incubated in darkness are
depicted by 0 and A, and in light o and A, respectively.
Concentration-dependent experiments were also performed. The data for these
experiments (Fig. 10, 11 and 12) were not combined since incubation times of the
first experiment were not consistent with those of the other two experiments, and the
paraquat concentrations used for the light-incubations of the S biotype were different
in each experiment. The regressions of ion leakage on paraquat concentration were
performed using means of the three replicate treatments. The equations and response
curves for the effect of paraquat concentration on ion leakage are illustrated in
Figures 10, 11 and 12. Within the concentration ranges used, both biotypes showed
an increase in ion leakage with increasing rates of paraquat in darkness and in light.
In experiments 1 and 2, dark-incubated leaf disks of the R biotype had lower levels of
ion leakage than those of the S biotype (Fig. 10a, 11a). There was no biotypic
difference in ion leakage in experiment 3 (Fig. 12a). Whereas the difference in ion
leakage between biotypes in darkness was small or absent, in light, the difference was
very large. This necessitated the use of lower paraquat concentrations in order to
characterize the response in the S biotype and precluded the use of ANOVA to
compare the 2 biotypes in light. However, the greater sensitivity of the S biotype in
the light was clearly apparent when the 50 values for the two biotypes were compared
(Table 4). The parameter estimates that were used to calculate the 150 values and
95% confidence limits were derived from regression analysis of the raw data rather
than means since the resulting confidence limits are more sensitive. The parameter
estimates and the R2 for the regression model were tabulated (Table 5). Ratios of 50
(dark):150 (light) were calculated for each biotype (Table 4). The ratios ranged from
0 450 900 1350 1800 2250
0 900 1800 2700 3600 4500
R -0.72 T
0 90 180 270 360 450
Figure 10. Experiment 1: Concentration-dependent ion leakage in darkness for 12 h by
leaf disks of the R [*] and S [A] biotypes (A) and in light for 8 h by the R biotype [o]
(B) and the S biotype [A] (C).
0 450 900 1350 1800 2250
0 450 900 1350 1800 2250
0 5 10 15 20 25 30 35 40 45
Figure 11. Experiment 2: Concentration-dependent ion leakage in darkness for 24 h by
leaf disks of the R [*] and S [A] biotypes (A), and in light for 12 h by the R biotype [o]
(B) and the S biotype [A] (C).
0 250 500 750 1000 1250 1500 1750
0 450 900 1350 1800 2250
0.45 0.90 1.35
Figure 12. Experiment 3: Concentration-dependent ion leakage in darkness for 24 h by
leaf disks of the R [*] and S [A] biotypes (A), and in light for 12 h by the R biotype [o]
(B) and the S biotype [A] (C).
Table 4. Estimates of I50a values and ratios of I50(dark):I50(light) derived from the
concentration-dependent electrolyte leakage studies.
Dark Light Dark Light
Expt 1 -M
150 1230 31.3 1610 1079
Lower 95% CI 1100 14.4 1449 971
Upper 95% CI 1391 50.3 1824 1198
Dark/Light 39.3 1.5
150 490 4.1 1071 700
Lower 95% CI 410 0.7 991 586
Upper 95% CI 562 8.1 1157 827
Dark/Light 119.5 1.5
150 353 0.31 534 111
Lower 95% CI 247 0.18 454 98
Upper 95% CI 460 0.42 619 121
Dark/Light 1138.7 4.8
a 150 refers to the concentration of paraquat required to cause the leakage of 50%
of the electrolytes from the leaf disks.
1.5 to 4.8 for the R biotype and from 39 to 1139 for the S biotype. The discrepancy
in the 150 values between the biotypes can be explained by a mechanism of paraquat
resistance that is light-dependent. There are three possible explanations for the
variability of the 150 values within biotypes, particularly the S biotype. The most
Table 5. Parameter estimates used in the determination of 150 values and confidence
limits, derived from regression of percent conductivity (Y) against paraquat concentration
Biotype Illumination Intercept X In X sqrt X R2
S Dark 12.7 0.03 0.85
S Light 27.0 6.7 0.57
R Dark 9.8 0.03 0.87
R Light 7.3 1.3 0.95
S Dark 10.9 2.1 0.93
S Light 21.9 13.8 0.82
R Dark 8.0 0.04 0.92
R Light 12.0 1.4 0.85
S Dark 8.0 2.2 0.79
S Light 83.5 28.5 0.82
R Dark 11.1 0.07 0.83
R Light -26.6 16.3 0.76
sensitive response to paraquat by the S biotype apparently occurs after an incubation
period greater than the 8 h used for the light incubation in experiment 1. In addition,
different groups of plants were used for each experiment and so different growth
conditions may have influenced the susceptibility to paraquat. It is also possible that
the inherent genetic variability of the S biotype may be greater than that of the R
biotype since the S biotype had not been subjected to the selection pressure of
multiple paraquat sprays per season for several years. Both the time-dependent and
concentration-dependent experiments gave evidence of paraquat activity in darkness
and in light. However, the biotypic difference in ion leakage was greater in the light
than in darkness.
Chlorophyll A Fluorescence
Time-dependent effects of diquat and paraquat
No biotypic difference in initial fluorescence (Fo) was observed when leaf
disks were incubated in 20 tIM diquat (Table 6). Although Fo appeared to decrease
with time, this was not significant. Peak fluorescence (Fp) was rapidly quenched in
the presence of diquat so that, after an hour, Fp had been reduced to about one half
the level of untreated plants. Quenching of Fp occurred to the same extent in both
biotypes. Similarly, there was no biotypic difference in the time-dependent response
of terminal fluorescence (Fs). An equivalent decrease in Fs was observed in both S
and R resistant biotypes (Table 6).
Variable fluorescence (F,) is a frequently used parameter in the early detection
of stress in plants (Lichtenthaler, 1988). The level of Fo is directly related to the
chlorophyll concentration per unit leaf area (Lichtenthaler, 1988). Therefore, F, is
more useful than Fp since it adjusts for differences in chlorophyll content. The time-
dependent effect of diquat on F, was summarized in Figure 13. The longer the
incubation period in the diquat solution, the greater was the reduction of Fv in both
biotypes. F, was reduced to a lower level in the S biotype (Table 7). Although there
was no significant difference in Fp and Fo, the small differences that did occur
combined to cause the biotypic difference in Fv. The absence of an interaction
between biotype and time means that the suppression of fluorescence occurred at the
same rate and provides evidence of similar rates of diquat penetration to the site of
Table 6. Time-dependent effect of 20 JM diquat on initial, peak and terminal
fluorescence of leaf disks from paraquat-sensitive and -resistant biotypes.
Relative Chlorophyll Fluorescence
Fp Fs F
126.04.6 72.74.4 57.71.9 1:
92.73.0 70.01.5 51.35.9
69.33.9 63.71.8 47.76.4 7
67.32.2 61.72.4 52.01.2
57.0+1.7 56.02.6 49.71.8
65.74.1 62.72.8 53.01.0
65.72.7 65.31.3 50.32.0
When the effect of incubating leaf disks in 1.7 mM paraquat was examined
over a 60 minute time-course, there was no biotypic difference in Fo (Table 8).
However, there was a decrease in Fp as time of exposure to the herbicide increased.
Unlike diquat, paraquat suppressed Fp to a greater extent in the S biotype than the R
biotype. Fs was observed to decrease over the time-course with no biotypic
difference (Table 8). There was greater inhibition of fluorescence in the S biotype
10 20 30 40 50 60
Figure 13. Effect of 20 AM diquat on the Fv of paraquat-R (o) and -S (A) American
Table 7. Comparison of chlorophyll fluorescence a in leaf disks of the paraquat-S and -
R biotypes incubated in 20 AiM diquat.
Fluorescence Sensitive Resistant P> F
Fo 689.9 87.9 0.0620
Fp 53.7 57.8 0.0601
F, 24.9 34.1 0.0028
Fs 86.2 86.5 0.8283
Table 8. Time-dependent effect of 1.7 mM paraquat on chlorophyll a fluorescence of
leaf disks from a S and an R biotype.
Relative Chlorophyll Fluorescence
Time Sensitive Resistant
min Fo Fp Fs Fo Fp Fs
0 64.84.7 119.84.0 82.85.6 67.84.6 138.53.8 83.73.6
10 62.73.4 94.33.6 75.04.8 64.23.6 119.55.8 79.53.4
20 64.23.7 84.75.0 75.24.7 65.25.2 103.05.4 81.25.1
30 65.53.0 80.82.5 73.23.2 65.24.6 102.73.4 78.73.9
40 65.24.1 80.33.7 75.73.7 65.02.5 102.83.9 79.72.7
50 61.73.7 76.03.3 73.33.4 62.03.7 90.33.1 74.74.1
60 63.04.0 80.03.9 73.73.6 64.73.3 96.34.1 77.04.3
(Fig. 14). However, this appeared to be due to inherently higher fluorescence by
disks from untreated R plants (0 time). The possibility of a modification of the site of
action cannot yet be discounted, but such a change is likely to result in a greater
degree of resistance than was observed. Like diquat, there was no interaction
between biotype and time, indicating that paraquat caused Fv to decrease at the same
rate in both biotypes (Fig. 14). This indicated that paraquat as well as diquat are
capable of rapid penetration to the site of action in both biotypes. In addition, these
two experiments provided evidence of bipyridinium reduction at the site of action.
Concentration-dependent effects of diquat and paraquat
The concentration-dependent effects of diquat and paraquat were also
investigated. Data from preliminary experiments are summarized in Tables 9 and 10,
respectively. Increasing the rates of diquat and paraquat in the bathing solution had
no effect on Fo indicating that in these experiments both biotypes had similar levels of
chlorophyll per unit leaf area. In both biotypes, Fp decreased as the herbicide rates
increased, but the rate of decrease was not the same. Fp decrease was more rapid in
the S biotype. Fs decreased with increasing concentrations of diquat and paraquat
with no biotypic difference. Figures 15 and 16 illustrate the concentration-dependent
effect of diquat and paraquat on F,. Diquat and paraquat were more effective in
suppressing F, in the S biotype than in the R biotype. The concentration-dependent
experiment was repeated 2 more times. A more convenient incubation period of 20
minutes was used and an additional concentration was added for the R biotype in
order to determine whether fluorescence could be suppressed to the same degree as
the S biotype. Fo, Fp, Fv and Fs were all higher in the R biotype than the S biotype
(Table 11). Fo, Fp and Fs decreased in a similar manner as paraquat concentration
was increased (Table 12). When Fv was expressed as a percentage of the control,
there was an interaction (P < 0.05) between biotype and concentration (Fig. 17).
Higher concentrations of paraquat were required by the R biotype to produce an
equivalent decrease by the S biotype. Since the time-dependent effect of diquat and
paraquat revealed that the rate of bipyridyl penetration to the site of action was the
same, the effect of concentration can be interpreted to mean that the resistance
mechanism is related to chloroplast photochemistry rather than exclusion or
Figure 14. Effect of 1.7 mM paraquat on Fv of a R biotype (o) and a S biotype (A) of
American black nightshade.
Table 9. Concentration-dependent effects of diquat on chlorophyll a fluorescence by
leaf disks of the paraquat-S and -R biotypes dark-incubated in the diquat solution for
Relative Chlorophyll Fluorescence
Conc. Sensitive Resistant
(M) F F F Fo Fp Fs
0 63.02.4 127.76.3 86.03.3 62.52.2 135.8 3.0 82.01.9
2E-8 58.51.0 119.52.3 79.00.9 58.72.2 136.0 3.6 78.71.7
2E-7 60.01.5 129.05.0 81.72.4 57.51.6 127.0 5.0 75.02.1
2E-6 59.23.2 97.55.4 76.02.6 59.02.3 129.3+10.3 81.05.0
2E-5 62.25.5 75.73.7 71.52.3 56.02.5 105.7 8.8 71.23.8
2E-4 57.02.2 71.52.5 69.22.1 57.21.0 91.5 2.2 71.70.8
2E-3 55.51.3 67.21.3 67.21.3 58.52.7 88.5 2.7 70.72.8
25 ---A-- A
0 2E-7 2E-6 2E-5 2E-4 2E-3 2E-2
Figure 15. Concentration-dependent effects of diquat on Fv of leaf disks from paraquat-
R (*) and -S (A) biotypes of American black nightshade incubated in darkness for 30
Table 10. Concentration-dependent effect of paraquat on chlorophyll a fluorescence
when leaf disks were dark-incubated in the paraquat solution for 30 min.
Relative Chlorophyll Fluorescence
Conc. Sensitive Resistant
(M) Fo F F F Fp Fs
S- __ o ___ l P ____ s____l o ____ P ___ L s
0 I ---
0 2E-7 2E-6 2E-5 2E-4
Figure 16. Concentration-dependent effects of paraquat on Fv of leaf disks of a R
(o) and an S (A) biotype incubated in darkness for 30 min.
Table 11. Effect of paraquat on chlorophyll a fluorescence in leaf disks of a S and R
biotype of American black nightshade.
Fluorescence Sensitive Resistant P> F
Fo 60.7 63.2 0.0001
Fp 110.5 133.2 0.0001
F, 49.8 70.0 0.0001
Fs 75.6 81.7 0.0001
Table 12. Concentration-dependent effect of paraquat on chlorophyll a fluorescence
when leaf disks were incubated in darkness over a 20 min period.
Relative Chlorophyll Fluorescence
Figure 17. Concentration-dependent effects of paraquat on Fv of leaf disks from R (o)
and S (A) biotypes incubated in darkness for 20 min.
2E-6 2E-5 2E-4 2E-3 2E-2
The measurement of ethane production was used as an indirect assay for
paraquat penetration to and activity at the site of action. The stimulation of lipid
peroxidation in both biotypes indicated that paraquat penetrated to the site of action
and was capable of inducing phytotoxicity in the R biotype in a similar manner to the
S biotype. Ethane evolution was also used to compare susceptibility to lipid
peroxidation in S and R C. canadensis (Lehoczki et al., 1992). In accordance with
our results, they found that in the absence of paraquat there was no difference in
ethane production between biotypes. With the addition of paraquat there was an 8-
fold increase in ethane production by the S biotype of C. canadensis compared with
only a 3-fold increase in the R biotype. Similar results were reported by P616s et al.
(1988) who found increased ethane and malondialdehyde (MDA) production in the
paraquat-S biotype of C. canadensis when leaves were treated with 100 tM paraquat.
The paraquat-R biotype, which is also resistant to atrazine, was unaffected.
Similarly, no ethane was produced by paraquat-R C. bonariensis (Youngman and
Electrolyte leakage, MDA formation and chlorophyll bleaching have also been
used to compare paraquat activity in R and S C. bonariensis (Vaughn and Fuerst,
1985). While there was only a small change in these components by the R biotype
over control levels when leaf disks were treated with 10 pM, considerably more
damage occurred in S biotype. Paraquat activity has also been examined in isolated
chloroplasts of C. bonariensis (Jansen et al., 1990; Shaaltiel and Gressel, 1987b).
Untreated chloroplasts of the R biotype evolved less ethane than the S biotype. The
addition of paraquat caused a greater increase in ethane in the S biotype. The use of
isolated chloroplasts eliminated the possibility of cytoplasmic sequestration. The
limited lipid peroxidation of the R biotype was coincident with constitutively elevated
protective enzymes, which were thought to be the primary mechanism of resistance
(Jansen et al., 1990; Shaaltiel and Gressel, 1987b). Another group working with this
species has been unable to corroborate these findings (Norman et al., 1993).
Elevation of constitutive levels of protective enzymes is not a plausible explanation of
paraquat resistance in American black nightshade since in our assays these enzymes
were inhibited by KCN.
Norman et al. (1993) indirectly investigated the ability of paraquat to interact
with the site of action by measuring MDA production, and directly by EPR analysis
and polarographically (partial PSI assay of oxygen consumption) in isolated
thylakoids. They found no evidence of a change in the site of action. This supported
evidence derived from a PSI partial reaction with isolated chloroplasts that there was
no change at the site of action (Fuerst et al., 1985). This technique has also been
used to assess paraquat activity in H. glaucum (Powles and Comic, 1987). No
evidence has as yet been provided that implicates mutation of the site of action in
conferring paraquat resistance.
The rapid phytotoxic activity of paraquat requires the photolysis of water
(Homer et al., 1960; Mees, 1960). Light is the energy source that drives the release
of electrons from water molecules, providing the reducing power to generate the
radical form of the herbicide. The bipyridinium herbicides are capable of killing non-
photosynthetic tissue as well as dark-incubated green tissue (Summers, 1980). The
process is much slower and respiration is thought to be the source of the reducing
power (Homer et al., 1960; Mees, 1960). The electrolyte leakage data from
American black nightshade demonstrated that the difference in paraquat activity
between biotypes in the dark was small. This indicated that a sufficient quantity of
paraquat must have remained unsequestered in both biotypes, resulting in the observed
membrane damage in the dark. The biotypic difference in paraquat activity under
illumination was considerably greater. This suggested that the resistance mechanism
was light-dependent and not sequestration or metabolism.
Photosynthetic electron transport can be assessed with great sensitivity in a
nondestructive manner by measuring chlorophyll fluorescence (Smillie and Gibbons,
1981). Fo depends to a large degree on the chlorophyll content of the leaf per unit
leaf area (Lichtenthaler, 1988). Fv is defined as the increase in fluorescence above
the Fo to Fp. Its variable nature arises from its responsiveness to changes in electron
flow through PSII (Smillie and Gibbons, 1981). Thus, the determination of Fv is a
useful technique for measuring the effect of various stresses on photosynthetic
electron transport. Conditions that inhibit electron flow on the photoreducing side of
PSII increase F,, while those that inhibit electron flow on the photooxidizing side
reduce Fv (Smillie and Gibbons, 1981). The measurement of F, has been used as a
direct assay of the presence of paraquat at the site of action in PSI (Munday and
Govindjee, 1969) and is a useful tool in investigations of the mechanism of paraquat
The analysis of chlorophyll a fluorescence in leaf disks of American black
nightshade confirmed that paraquat does penetrate to and is reduced at the site of
action of the R biotype. However, both diquat and paraquat were more effective in
suppressing Fv in the S biotype. Chlorophyll fluorescence has also been used to
assess paraquat activity in paraquat-R C. bonariensis and C. canadensis. Although
these results were comparable to those with American black nightshade, the
interpretations were not always in accord. In C. bonariensis, fluorescence quenching
was observed in both biotypes with much higher concentrations of paraquat required
to effect the response in the R biotype (Fuerst et al., 1985; Vaughn et al., 1989).
The inability of the R biotype to quench fluorescence to the same degree as the S
biotype was regarded as direct evidence that the herbicide did not reach the site of
Decrease in Fv of the R biotype of C. canadensis occurred within 7 minutes
(Lehoczki et al., 1992). Even though more fluorescence suppression occurred in the
S biotype, the workers felt that suppression in the R biotype was a definite indication
of paraquat penetration of the chloroplasts of the R plants and interaction with the site
of action. This was consistent with our interpretation of similar data in American
black nightshade. However, they concluded that the data supported the protective
enzyme mechanism in combination with a mechanism for the elimination of paraquat
activity (Lehoczki et al., 1992). Isolated chloroplasts from an S biotype and an R
biotype of C. canadensis co-resistant to paraquat and atrazine quenched maximum
fluorescence (Fm) to the same extent (Poils et al., 1988). This also was true for
short-term measurement on intact leaves. However, when measured at 24 h after
paraquat treatment, less quenching was apparent in the R biotype. The quenching of
chlorophyll fluorescence in the R biotype was a transient effect, but was not
reversible in the S biotype (Lehoczki and Szigeti, 1988; Lehoczki et al., 1992; P616s
et al., 1988; Szigeti et al., 1988). The gradual recovery of chlorophyll fluorescence
levels was thought to indicate the progressive elimination of paraquat activity.
There is an inverse relationship between fluorescence and the rate of
photosynthesis. When photosynthesis is optimized, most of the absorbed light energy
is utilized and there is very little fluorescence (Lichtenthaler, 1988). It is thus
conceivable that the higher Fp and Fv values obtained for the R biotype of American
black nightshade may be due to differences in photosynthetic electron transport. If
photosynthetic electron transfer is impaired in paraquat-R American black nightshade,
the higher levels of Fv imply that the causal mutation is located between PSII and
PSI. Thus it is likely that a PSI partial reaction will not detect it. Midmore and
Prange (1992) have found that plants grown under conditions of low temperature and
low irradiance had low rates of photosynthesis and correspondingly higher Fp than
similar plants grown at higher temperature. The increased Fp and Fv was felt to be
due to less efficient movement of electrons through PSII. Evidence of reduced rate of
electron transfer between QA and Qg and of a 20% reduction in the relative rate of
electron transport between water and plastoquinone has been reported for the R
biotype of C. canadensis (Lehoczki et al., 1992). However, it was apparently not
considered important enough to warrant mention in the conclusions. There was no
difference in the rate of electron transfer between QA and Qg when an S and an R
biotype of C. canadensis with atrazine co-resistance were compared (P616s et al.,
Based on the data collected from the measurement of ethane, electrolyte
leakage and fluorescence in the presence of paraquat it can be concluded that paraquat
does enter the chloroplast of the R biotype. The involvement of constitutive
protective enzymes was eliminated. Paraquat was capable of diverting electrons from
PSI and exhibited phytotoxic action, albeit at a lower level than the S biotype. The
results suggest that the resistance mechanism is chloroplastic and may be related to
differences in electron transport or a change at the site of action. However,
cytoplasmic mechanisms such as vacuolar sequestration or paraquat metabolism that
may reduce the amount of paraquat that penetrates to the site of action could not be
ruled out unequivocally.
DIFFERENTIAL PHOTOSYNTHETIC ELECTRON TRANSPORT AND
OXIDATIVE STRESS IN RESISTANT AND SENSITIVE BIOTYPES
OF AMERICAN BLACK NIGHTSHADE
Phytotoxicity of bipyridiniums is caused by the formation of toxic oxygen
radicals resulting from the autoxidation of the bipyridyl radical. Superoxide (02-) is
the first product of the reaction between the bipyridyl radical and oxygen. In vitro
production of 02- from the autooxidation of diquat and paraquat radicals has been
demonstrated (Giannopolitis and Ries, 1977; Stancliffe and Pirie, 1971). The
involvement of 02- in mediating paraquat toxicity was proposed because it was
determined that the radical was sufficiently long-lived to promote the observed
phytotoxic effect (Farrington et al., 1973). Several bipyridyl compounds were shown
to stimulate the reduction of ferricytochrome c by isolated thylakoids, a reaction that
was inhibited by superoxide dismutase (Asada et al., 1974). Youngman et al. (1979)
also provided in vivo evidence that 02- was involved in promoting paraquat toxicity.
Superoxide can participate directly in oxidation and reduction reactions with
cell components leading to toxic effects. Superoxide also reacts with H202 through
the iron-catalyzed Haber-Weiss reaction giving rise to the more reactive hydroxyl
radical (OH'). The OH* reacts indiscriminately with almost all cell components and
initiates and propagates the peroxidation of cell membranes. Hydroxyl radicals have
been detected in paraquat-treated plants (Babbs et al., 1989). The dismutation of 02-
spontaneously or by catalytic action of SOD results in the formation of H202. A
strong oxidant, H202 inhibits CO2 fixation through oxidation of sulfhydryl groups of
the enzymes that catalyze the process (Kaiser, 1979). The formation of 02- is an
unavoidable consequence of photosynthetic electron transport due to the leakage of a
small amount of electrons to 02 in a process known as the Mehler reaction (Asada
and Kiso, 1973). The resultant active oxygen species are scavenged by SOD followed
by the enzymes of the Halliwell-Asada pathway (Halliwell, 1987). The enzymes of
this pathway are ascorbate peroxidase, dehydroascorbate reductase and glutathione
reductase. Paraquat stimulates radical production to levels that overwhelm the
The results presented in chapter three strongly implicated the involvement of
chloroplast photochemistry in American black nightshade paraquat resistance. In light
of evidence suggesting that electron transport was reduced in the R biotype, the
possibility existed that the paraquat-R and -S biotypes of American black nightshade
differed in their ability to generate toxic oxygen species. Therefore, the objectives of
these experiments were two-fold: (1) to confirm whether there were inherent
differences between the rates of photosynthetic electron transport of the paraquat-R
and the -S biotypes, and (2) to determine whether the biotypes differed in their ability
to generate oxygen stress.
Materials and Methods
Chloroplast and Thylakoid Membrane Isolation
Each group of R and S plants used in these experiments were scored for
resistance or sensitivity to paraquat and held in darkness for 24-48 h in order to
reduce the size of starch granules in the chloroplasts. Different individuals were used
for each assay. Two or three of the most recently expanded leaves per plant were
detached, washed and blotted dry. Leaves were sliced thinly and placed in 150 mL of
partially frozen grinding (GR) buffer. The grinding buffer consisted of 0.4 M
sorbitol, 50 mM MOPS-KOH pH 7.5, 10 mM MgCl2 and 1% bovine serum albumin
(BSA). This grinding buffer was chosen because it was used successfully in the
isolation of chloroplasts from a related species, S. nigrum (Siegenthaler and Mayor,
1992). Leaf material was homogenized with a Waring blender in two 5 s bursts. The
homogenate was filtered through 4 layers of cheesecloth and 2 layers of nylon screen
moistened with GR buffer. Filtrate was poured into chilled 50 mL polypropylene
centrifuge tubes and centrifuged at 2800 rpm using an IEC clinical centrifuge with a
swing-out head rotor (International Equipment Company, Needham, MA). All
centrifugation, unless otherwise indicated, was done at 2800 rpm with the IEC clinical
centrifuge. Centrifuge tubes filled with ice were placed in the rotor cups between
centrifugations so that all operations were performed at low temperature. The filtrate
was centrifuged for 3 min. The supernatant was discarded and the pellet resuspended
in GR buffer. For the determination of 02- formation it was necessary to reduce the
possibility of mitochondrial contamination of the chloroplast preparation because
mitochondrial cytochrome c oxidase reverses the reaction of interest (Asada, 1984).
Therefore, a Percoll gradient was prepared by combining 50% 2X GR buffer with
50% Percoll (Sigma Chemical Company, St Louis, Mo). The gradient was generated
by centrifugation of 25 mL of the Percoll mixture at 20,000 rpm for 40 min, using a
Beckman model J2-21 refrigerated centrifuge (Beckman Instruments Inc., Fullerton,
CA). The resuspended chloroplasts were gently layered onto the gradient.
Centrifugation for 15 min in the Beckman centrifuge at 4C with a fixed angle JA 20
rotor did not produce very distinct layers, and so the IEC clinical centrifuge with its
swing-out head rotor was employed in the second experiment. The top layers were
aspirated off and the lower green band retrieved with a wide bore Pasteur pipet. This
was diluted to 50 mL with chloroplast wash buffer and centrifuged for 3 min. The
wash buffer consisted of GR buffer minus the BSA. The pellet was resuspended and
washed again. After resuspension in 5 mL chloroplast wash buffer, the chloroplasts
were examined by means of phase contrast microscopy to confirm integrity of the
Chloroplasts were lyzed by incubating them for 2 min in 5 mL chloroplast
wash buffer plus 45 mL 20 mM potassium phosphate buffer, pH 7.8. Thylakoids
were collected by centrifugation for 5 min. Thylakoids were resuspended in a 20 mM
potassium phosphate buffer, pH 7.8, containing 1 mM ethylenediaminetetraacetic acid
(EDTA). This buffer was used to deplete the thylakoids of stromal SOD. After 1 h
at 4C, the thylakoids were collected by centrifugation for 5 min and the SOD
removal step repeated with fresh buffer. On completion of the second wash step the
thylakoids were collected by centrifugation as previously described and resuspended in
a small quantity of 50 mM potassium phosphate buffer, pH 7.8, containing 10 mM
For the electron transport experiments, instead of a continuous gradient, a
Percoll cushion was used to purify the chloroplasts. It consisted of 50% GR buffer
and 40% Percoll. The resuspended chloroplasts were layered onto the Percoll
cushion and centrifuged for 3 min. The pellet was resuspended in 5 mL chloroplast
wash, which was then diluted to 50 mL with cold distilled water. Chloroplasts were
allowed to lyse over the next 5 min, then the thylakoids were collected by
centrifugation. They were resuspended in 1 mL of chloroplast wash buffer.
Chlorophyll was determined after Bruinsma (1961).
The photoreduction of ferricyanide to ferrocyanide was one of two methods
used to compare photosynthetic electron transport in the 2 biotypes. The
photoreduction was conducted as per Siegenthaler and Mayor (1992) and ferrocyanide
was detected after Avron and Shavit (1963). The 1 mL reaction mixture consisted of
0.3 M sorbitol, 50 mM MOPS-KOH pH 7.2, 5 mM MgCl2, 5 mM NH4CI and 0.5
mM K3Fe(CN)6. Because of differences in yield of chloroplasts, 20 j/g chlorophyll
mL-1 were used in experiment 1 and 10 1/g chlorophyll mL-1 were used in the second
experiment. Thylakoids were illuminated at 350 /xE m-2 s-1 with a 300 W
incandescent lamp. A 3.5 cm deep water filter was used to reduce heating by IR
radiation and red light was provided by means of a red filter that eliminated
wavelengths below 650 nm. At intervals of 0.5, 1, 1.5, 2 and 2.5 min, the reaction
was stopped by the addition of 100 jxL of 30% trichloroacetic acid (TCA). Each
treatment was replicated 3 times. An appropriate control was prepared for each
biotype by addition of the TCA without illumination.
Accumulation of ferrocyanide was monitored spectrophotometrically. The
reaction mixture was transferred to micro-centrifuge tubes and a clear supernatant
obtained by centrifugation at 12000 rpm for 1 min, using a Fisher micro-centrifuge
model 59A (Fisher Scientific, Pittsburgh, PA). An aliquot of the supernatant was
sampled: 300 /L in the first experiment, 600 tL in the second. To the aliquot, 1.8
mL of distilled water were added, followed by 0.9 mL of reagent mixture. The stock
reagents were 3 M sodium acetate (pH 6.2), 0.2 M citric acid, 3.3 mM FeCl3
dissolved in 0.1 M acetic acid and 4,7-diphenyl-l,10-phenantroline, sulfonated. The
concentration of the last reagent was 100 mg per 30 mL water. The reagents were
mixed in the ratio 2:2:1:1 just prior to use. After addition of the reagent mixture, the
solutions were exposed to ambient room light for 5 minutes for color development,
then they were covered with foil until absorbance was determined. This was done at
535 nm using a Shimadzu UV-Visible recording spectrophotometer, model UV 160
(Shimadzu Corporation Spectrophotometric Instruments Plant, Kyoto, Japan).
The data were expressed as /mol ferrocyanide mg-1 chlorophyll using a molar
absorption coefficient of 20,500 (Avron and Shavit, 1963). The data were analyzed
as a randomized complete block design. ANOVA was performed and the means were
regressed against time.
NADP+ reduction was successfully accomplished in only 2 of 4 studies. The
procedure employed in the two successful studies was as follows. The first
experiment was performed in 1 mL of reaction mixture containing: 40 mM tricine-
NaOH (pH 7.8), 5 mM MgCl2, 1 mM ADP, 10 mM Na2HPO4, 0.2 mM NADP+, 50
jg ferredoxin (Fluka Chemical, Ronkonkoma, NY) and thylakoids at 15 Jtg m-1. In
addition to an untreated control, thylakoids were also exposed to 100 JM paraquat.
Illumination was provided by a projector equipped with a 300 W lamp. Light was
filtered through 6 cm of water (flat glass bottle containing deionized water) to reduce
heating due to IR radiation and a red filter was used to exclude wavelengths of less
than 650 nm. The light intensity was 250 /E m-2 s-1. The reaction mixture was
exposed to light for 2.5, 5.0, 7.5, 10, 15 and 20 min. NADPH formation was
detected spectrophotometrically by absorbance increase at 340 nm. In order to
compensate for the gradual decrease in the absorbance of the blank over the course of
the assay, three replicate blanks were prepared with and without paraquat for each
biotype. The absorbance of these were determined without illumination and used to
adjust sample absorbances. Data were expressed as lsmol NADPH mg-1 chlorophyll
using a molar absorptivity coefficient of 6,200. Each treatment was replicated 3
times and the experiment was analyzed as a randomized complete block design.
For the second experiment, several changes were made in the procedure. The
chloroplasts were isolated on the day prior to use in the assay and so were more than
24 h old when used. The chlorophyll concentration was 5 tg mL1 in a 3 mL
reaction volume. The reaction volume was changed in order to make use of glass
cuvettes that gave more precise readings than the disposable cuvettes used previously.
Complete blanks were prepared with and without paraquat. Atrazine, a
photosynthesis inhibitor was included at 50 /iM in an attempt to stabilize the blank.
Three replicate cuvettes were exposed to light and repeated absorbance determinations
were done at 3, 6, 9 and 12 min. As a result, a more stringent analysis of the data
was required. A repeated measures analysis was done using a split-plot design. One
replicate with thylakoids from the R biotype consistently gave a negative absorbance
that did not change over time. The data for this replicate were not included in the
analysis. The means for the nontreated R biotype were based on two replicates.
Measurement of Superoxide Radicals
The procedure for measuring 02" generation was taken from Asada (1984).
The assay involved the spectrophotometric determination, at 550 nm, of
ferricytochrome c reduction to ferrocytochrome c by 02-.
Fe3 cyt c + 02- Fe2+cyt c + 02
The experiment was conducted in a 2 mL reaction mixture, which consisted of 50
mM potassium phosphate, pH 7.8, 10 mM NaCI, 20 1M ferricytochrome c and 100
MM paraquat. An untreated control was included for comparison. For the first
experiment, the chlorophyll content was 5 ig mL-1 and 10 ltg mL'1 for the second
experiment. The reaction mixture was illuminated as described previously for the
NADP+ reduction. The illumination periods were 1, 2.5, 5, 7.5 and 10 min for
experiment 1 and the small yield of chloroplasts restricted the number of replicates to
two. In experiment 2, some additional illumination periods were used: 15 and 20 for
the S biotype and 15, 20 and 40 min for the R biotype. The treatments were
replicated 3 times in experiment 2.
The difference absorption coefficient between ferri- and ferrocytochrome c,
used for expressing the data in terms of nmol, was 19,000 M-1 cm-1 (Asada, 1984).
A complete blank plus 50 /g mL1 SOD was prepared for each biotype. There was a
gradual decrease in the absorbance of the blank resulting in a total decrease in
absorbance of 0.049 for the S biotype and 0.031 for the R biotype over the course of
the experiment. In the second experiment an attempt was made to adjust for the
changes that were occurring in the blank. For the S biotype, the reading of the blank
was noted when the sample absorbance was recorded so that it was possible to adjust
for the changes from the original reading of the blank. For the R biotype the
absorbance of the blank against water was read. This absorbance was used to adjust
the sample readings which were also blanked against water. The data from the first
experiment were analyzed as a randomized complete block, blocking on replicate.
The ANOVA procedure of SAS was used and standard errors were calculated. For
the second experiment the GLM procedure was used for analyzing the raw data and
the REG and NLIN procedures were used to regress the means.
Photosynthetic electron transport was compared using an artificial electron
acceptor, ferricyanide, and the natural electron acceptor, NADP+. The reduction of
ferricyanide occurred more rapidly in the S biotype than the R biotype (Fig. 18),
indicating that the rate of electron transfer was faster in the S biotype.
(A) and -R
Reduction of ferricyanide to ferrocyanide in isolated thylakoids of paraquat-S
(o) biotypes of American black nightshade.
Corroboration of these results was provided by the data from the experiments
measuring NADP+ reduction. Again, more rapid formation of NADPH by the S
biotype provided evidence that electron transport was slower in the R biotype (Fig.
19). While the addition of paraquat reduced NADPH production in the S biotype, no
effect was observed in the R biotype in the first experiment (Fig. 19). In the second
experiment, there was a significant (P <0.05) biotype by time interaction. Further
analysis indicated that the effect of time was not significant in the R biotype, whereas
NADPH accumulated with time in the S biotype (Fig. 20). The inclusion of 100 pM
paraquat completely eliminated NADP+ reduction in both biotypes (data not shown).
ZA-- A A
0 I f I
0 5 10 15 20
Figure 19. Production of NADPH in isolated thylakoids of paraquat-S (A) and -R (o)
biotypes as measured by absorbance at 340 nm. Closed symbols indicate NADPH
production in the presence of at 100 /M paraquat.
0.00 1 1 1 I
0 3 6 9 12 15
Figure 20. Production of NADPH in isolated thylakoids of paraquat-S (A) and -R (o)
as measured by absorbance at 340.
The magnitude of the response in these two experiments varied because the
chloroplasts used in the second experiment were slightly aged because they had been
isolated the previous day. This tends to reduce the activity of the chloroplasts. Since
the chloroplasts were isolated from two different groups of plants, this may also
account for some of the observed variability in the response.
Oxygen Radical Generation
Similar results were obtained in both experiments in which 02- generation was
measured. The data were not combined since the experiments were not performed
over the same time period, and there were differences in the way in which the
experiments were conducted. Figure 21 illustrates the data collected from the first
0 1 2.5 5 7.5 10
Figure 21. Comparison of superoxide production in the S (A) and R (o) biotypes.
Closed symbols indicate paraquat treatment at 100 MM. (Experiment 1).
assay. Superoxide was generated by thylakoids of both biotypes in the absence of
paraquat. The interaction between biotype and time (P<0.05) showed that the rate of
radical production in the S biotype exceeded that of the R biotype. The addition of
paraquat elevated 02- production in both biotypes with greater stimulation occurring
in the S biotype (Fig. 21). The trends of the second experiment were similar to those
of the first experiment. In the absence of paraquat the S biotype had a high rate of
02- production that was enhanced by the addition of paraquat (Fig. 22a). The R
biotype had a lower rate of radical formation than the S biotype when no paraquat
was present. Addition of paraquat stimulated the rate of 02- generation (Fig. 22b). It
was clear from both Figures 21 and 23 that in the presence of 100 MM paraquat,
Y= 1.8X-0.03X +1.9
5 10 15
Y=0.97X-0.0 IX+6.3 R =0.88
10 20 30
Figure 22. Comparison of superoxide production in: [A] the S biotype with 100 tM
paraquat (A) or without (A), and [B] the R biotype with 100 iM paraquat (0) or
without (0). (Experiment 2).
0 5 10 15 20
Figure 23. Comparison of superoxide production in the S (A) and R (o) biotypes.
Closed symbols indicate paraquat treatment at 100 UM. (Experiment 2).
isolated thylakoids from the R biotype generated levels of 02- similar to those of the
S biotype in the absence of paraquat. This suggests that the addition of paraquat to
the R biotype produces a level of oxygen stress that can be withstood by the plant
without irreparable damage occurring. The similarity in response when electron
transport and radical generation were measured seem to indicate that the restricted
oxygen stress in the R biotype is due to reduced electron transport.
The results presented herein provide evidence of a slower rate of electron transport in
the paraquat-R biotype of S. americanum. This impairment results not only in a
lower rate of the Mehler reaction but also in reduced oxygen stress in the presence of
paraquat when measured as 02-. The data of Yanase et al. (1992) support the
argument that reduced photosynthesis translates into reduced paraquat toxicity. They
showed, by measuring electrolyte leakage, that paraquat toxicity can be alleviated by
photosynthesis inhibitors such as DCMU. Toxicity is reduced because the inhibitors
reduce the electron supply from the electron transport system to paraquat.
In the absence of a direct measurement of the activity of the protective
enzymes (SOD and the Halliwell-Asada enzymes) in the R biotype of S. americanum,
it is assumed that either the constitutive levels are adequate to cope with the oxidative
stress or that constitutive levels provide sufficient protection to maintain cell integrity
until existing levels are augmented. The latter is plausible since 50 tM paraquat has
been shown to cause enhanced levels of SOD mRNAs in illuminated Nicotiana
plumbaginifolia (Tsang et al., 1991). The increase was 40-fold in the chloroplasts,
30-fold in the mitochondria, and 15-fold in the cytoplasm. Unpublished data from
this laboratory (T.A. Bewick and K.L. Smith) also supports this possibility. They
found that constitutive levels of the antioxidant ascorbate were the same in both
biotypes of S. americanum. However, while application of paraquat caused a net loss
of ascorbate in the S biotype, ascorbate depletion was transient in the R biotype.
Subsequently, ascorbate was observed to accumulate in excess of the constitutive level
(Bewick and Smith, unpublished).
Paraquat resistance due to impaired electron transport as observed here in S.
americanum is unique. Lower rates of electron transport have been reported in leaf
disks and protoplasts of a paraquat-R biotype of C. canadensis. However, there were
no differences at the chloroplast level and the mechanism of resistance was attributed
to enhanced constitutive levels of protective enzymes (Itoh and Matsunaka, 1990).
Photosynthesis has also been assessed in other paraquat-R biotypes. Szigeti and
Lehoczki (1992) compared photosynthetic electron transport between PSII and PSI in
two R biotypes and one S biotype of C. canadensis by evaluation of P700 (reaction
center of PSI) absorbance changes. They found no evidence to support the
occurrence of a mutation along the electron transport chain from plastoquinone to the
cytochrome b6f. In another report on C. canadensis, CO2 fixation rates at time zero
in the R and S biotypes for the same species as read from a figure were 1.75 and
2.25 nmol cm-2 s-1, respectively. However, there was no indication of whether the
rates were significantly different (Po61s et al, 1988). In general, the focus of
investigations of photosynthetic electron transport and CO2 fixation has been to
determine if target site mutation has occurred or to compare the extent to which
paraquat inhibited CO2 fixation in an R and an S biotype. Such experiments with C.
bonariensis, H. glaucum, H. leporinum yielded no evidence of differences in electron
transport that could be implicated in the resistance phenomenon (Fuerst et al., 1985;
Norman et al., 1993; Powles and Comic, 1987; Preston et al., 1992).
In summary, a unique mechanism is responsible for paraquat resistance in S.
americanum. Photosynthetic electron transport is impeded in the R biotype. This
causes a level of oxidative stress, in the presence of 100 AM paraquat, similar to that
of the S biotype in the absence of paraquat. This magnitude of stress is probably well
within the protective capacity of the enzyme systems of R biotype and so
phytotoxicity is not observed until sufficient paraquat is applied, to raise the oxidative
stress to a level with which the plant cannot cope.
As long as paraquat remains in the chloroplast, oxygen toxicity will be
promoted. As a result, a continuous allocation of NADPH would be required to drive
the scavenging of toxic oxygen species by the Halliwell-Asada enzymes. It may be
that sublethal doses of paraquat improve the efficiency of electron transport in the R
biotype so that both NADPH and paraquat radicals are formed. However, lethal
doses of paraquat would be expected to totally eliminate NADPH formation so that all
scavenging of toxic oxygen ceases. Alternatively, the gradual elimination of paraquat
from the chloroplast may occur by vacuolar sequestration.
SUMMARY AND CONCLUSIONS
Reports by Collier County tomato farmers of reduced control of S.
americanum by paraquat (Gilreath and Stall, 1987) encouraged efforts to determine
the cause of the problem. The possibility of an interaction with copper-containing
pesticides was investigated (Bewick et al., 1990a,b). While some of the pesticides did
reduce the efficacy of paraquat, in the absence of copper pesticides the progeny of
plants collected from Collier County were 12-fold more resistant to paraquat than
plants from an unselected Gainesville population (Bewick et al., 1990a). This
suggested that as for other paraquat-resistant weeds, the high selection pressure of
multiple sprays of paraquat per season had permitted the establishment of a resistant
population in Collier county.
The purpose of this study was to determine the mechanism of paraquat
resistance in S. americanum. A number of mechanisms have been previously
proposed to explain paraquat resistance in other weed species. Therefore, the
hypothesis tested in this study was: paraquat resistance in S. americanum can be
explained by a previously established mechanism. There were two main objectives:
(1) to determine whether paraquat could penetrate to the site of action and undergo
normal reduction in the R biotype, and (2) to compare the ability of the R biotype and
the S biotype to generate and defend against oxidative stress.
Mechanisms that would restrict the penetration of paraquat into the cytoplasm
include: (1) reduced cuticular penetration, (2) apoplastic sequestration, and (3)
exclusion by the plasmalemma. Experiments were conducted to determine whether
resistance was due to reduced cuticular absorption. The cuticle did not appear to be a
barrier to paraquat uptake by leaf disks. The biotypes did not differ in the amount of
paraquat that was absorbed, and paraquat uptake in S. americanum was characterized
by a quadratic relationship. Examination of efflux kinetics showed that efflux from
both biotypes occurred at a rapid rate within the first 10 min, then tapered off.
However, the rate of paraquat efflux was significantly greater in the S biotype. This
pattern of efflux could arise from either greater apoplastic sequestration or greater
absorption of paraquat into the cytoplasm by the R biotype. To distinguish between
these two possibilities a study of paraquat compartmentalization was conducted.
There was more unbound paraquat in the apoplast of the S biotype. More paraquat
was adsorbed in the apoplast of the R biotype than the S biotype. However, it is
unlikely that rapid sequestration of paraquat by apoplastic adsorption accounts for
resistance in S. americanum because the level of paraquat in the cytoplasm was
similar in both biotypes. In the R biotype, this was 3-fold more paraquat than was
bound in the apoplast. The biotypes had a similar level of cytoplasmic paraquat.
Thus, the data did not appear to support either apoplastic sequestration or exclusion
by a less permeable plasmalemma as a possible primary mechanism of resistance.
In order to determine whether paraquat could penetrate to the site of action in
the R biotype and was capable of undergoing reduction, a number of biochemical
responses to paraquat treatment were compared in the two biotypes. Ethane evolution
was measured as an index of lipid peroxidation. In the absence of paraquat there was
no difference in ethane production by the two biotypes and ethane was produced at
the same rate by both biotypes. The addition of paraquat caused increased ethane
production in both biotypes. The rate of ethane production was greater in the S
biotype. The incremental increase in ethane that occurred on raising the paraquat
concentration from 8.5 $M to 85 iM was smaller than the increase measured on
raising the concentration from 0 to 8.5 IM. Since the protective enzymes, superoxide
dismutase and ascorbate peroxidase, were inhibited by inclusion of KCN in the
incubation solution, it is unlikely that the differential response observed was due to
enhanced radical scavenging.
In darkness, time-dependent electrolyte leakage from paraquat-treated leaf
disks was the same for the first 12 h. After 12 h, electrolyte leakage increased at a
more rapid rate in the S biotype than the R biotype. Illuminated paraquat-treated leaf
disks of the S biotype leaked ions faster than those of the R biotypes, so that 100%
leakage occurred within 12 h for the S biotype and by 24 h in the R biotype.
Concentration-dependent ion leakage was also investigated in darkness and in light.
Whereas the differences in ion leakage between biotypes were small in the darkness,
they were very large under illumination. These data suggested that the resistance
mechanism may be light-dependent.
The effect of diquat and paraquat on chlorophyll a fluorescence was also
investigated. The ability of the bipyridyls to accept electrons at PSI keeps the
electron transfer chain oxidized and thus suppresses fluorescence. Thus, fluorescence
is a direct measure of the presence of bipyridyl cations at the site of action. With
both paraquat and diquat, chlorophyll fluorescence was reduced at the same rate in
both biotypes. This indicated that both paraquat and diquat penetrated to the site of
action at the same rate. However, the magnitude of the fluorescence quenching was
smaller in the R biotype. Concentration-dependent chlorophyll fluorescence was also
examined. Greater concentrations of diquat and paraquat were required by the R
biotype in order to achieve similar levels of fluorescence quenching. Because control
levels of fluorescence tended to be higher in the R biotype, the data were interpreted
to indicate that paraquat resistance was probably due to reduced electron transport or
reduced affinity for bipyridinium ions at the site of action.
Taken together, ethane production, ion leakage and fluorescence showed that
paraquat penetrated to the site of action, that the active site was capable of reducing
both diquat and paraquat and that sufficient oxidative stress was generated in the R
biotype to induce phytoxicity, although at a lower level than the S biotype. However,
they do not unequivocally eliminate the possibility that less paraquat may penetrate to
the target site or that a target site mutation may have occurred. The ethane assays did
eliminate elevated levels of protective enzymes from being the primary mechanism
because KCN was used to inhibit SOD and ascorbate peroxidase and a differential
response was still obtained.
In order to test whether the primary mechanism was chloroplastic or
cytoplasmic, assays for differential electron transport and 02- production were