Biological Activity of the Isomeric Forms of Helminthosporiumsacchari Toxin and of Homologs Produced in Culture'
Received for publication June 17, 1983 and in revised form September 16, 1983
JONATHAN P. DuVICK, J. M. DALY*, Z. KRATKY, V. MACKO, W. ACKLIN, AND D. ARIGONIDepartment ofAgricultural Biochemistry, University ofNebraska, Lincoln, Nebraska 68583 (J. P. D., J.M. D); Boyce Thompson Institute at Cornell University, Ithaca, New York 14853 (Z. K., V. M.); andDepartment ofOrganic Chemistry, Swiss Federal Institute of Technology, CH-8092 Zurich, Switzerland(W A., D. A.)
ABSTRACT
The effect of Helminthosporium sacchari (HS) toxin isomers andrelated, pathogen-produced compounds on dark CO2 fixation in HS-susceptible sugar cane leaf slices was investigated. HS toxin consists ofa mixture of three isomeric bis-5-O-(6-galactofuranosyl)-$-galactofur-anosides (A, B, and C) differing in the position of one double bond in thesesquiterpene aglycone. Maximum inhibition of dark CO2 fixation insusceptible sugar cane (CP52-68) occurred within 30 to 40 minutes, andamounts necessary to reach 50% inhibition values typically were approx-imately 1.7 micromolar for natural toxin mixture (= 2:3:5 mixture ofisomers A:B:C) and 4, 6, and 0.7 micromolar for isomers A, B, and C,respectively. Other fractions from cultures of the pathogen consist ofcomparable mixtures of sesquiterpene isomers but have only 1, 2, or 3galactofuranose units (HS,, HS2, HS3) or two a-glucopyranose units aswell as four il-galactofuranose units (HS6). The lower toxin homologswere not toxic to clone CP52-68, but protected sugar cane from theeffects of toxin. Minimum ratios of protectant: toxin giving 95% protec-tion were approximately 50:1, 6:1, and 12:1 for HS,, HS2, and HS3,respectively. HS2 and HS3 protected when added up to 12 minutes aftertoxin as well as when added with or before toxin. Some common plantgalactopyranosides were not toxic and did not protect at 500:1 molarexcess. The sample of HS6 was toxic at 500 micromolar, and did notprotect apinst HS toxin. With the availability of purified, homogeneouspreparations ofHS toxin, homologs, and chemically modified or syntheticanalogs, the dark CO2 fixation assay should prove to be a useful tool forunderstanding the mode of action of HS toxin.
Independent efforts by two groups (10, 13-16) have recentlydemonstrated that the host-selective pathotoxin (HS toxin), pro-duced by Helminthosporium sacchari (Van Breda de Haan)Butler, and affecting sugar cane, consists of a sesquiterpenemoiety and several ,B-galactofuranose units. Livingston andScheffer (10) presented evidence that galactose was present as anoligomer of four to six (most probably five) ,-linked residueswith a terminal sesquiterpene aglycone. Macko et al. (15, 16),on the basis of high resolution nuclear magnetic resonance andmass spectra, concluded that HS toxin consists of a centralsesquiterpene aglycone linked to two residues of 5-O-(,B-galacto-furanosyl)-,B-galactofuranoside (Fig. 1). In addition, they reported
' Published as Paper No. 7192, Journal Series, Nebraska AgriculturalExperiment Station. Supported by United States Department of Agri-culture, Competitive Research Grants Office Grant Nos. 82-CRCR-1-1096 to J. M. D. and 59-2364-1-1-746-0 to V. M.
that the toxin is a mixture of three isomers differing in theposition of a double bond centered around carbon 4 of thecarbobicyclic ring system of the aglycone (Fig. 1). The isomers,separable by HPLC, are present in approximate percentages of20, 30, and 50 for A, B, and C, respectively.The existence of multiple species of a host-selected pathotoxin
is well documented (5). In some instances, for example HMTtoxin from H. maydis race T (18) and AAL toxin from Alternariaalternata f. sp. lycopersici (1), the individual species have equalinnate toxicity and selectivity. In others (AM toxin from A. mali[7]), there are significant differences in toxicity to susceptiblehosts which may be helpful in explaining the mode of action.
In the course of the isolation of HS toxin, the presence ofnontoxic lower homologs in culture filtrates has been observed(1 1-13, 17). These lower homologs possess 1, 2, or 3 galactoseunits and are designated in this paper as HS,, HS2, and HS3,respectively, in order to distinguish them from active toxin (HS4).Livingston and Scheffer recently reported a f3-galactofuranosi-dase from cultures of H. sacchari that was capable of convertingHS4 to lower homologs (12). Based on the structure of the threeHS toxin isomers (Fig. 1), successive loss ofgalactose units wouldgenerate 21 different lower homologs, all of which have nowbeen detected ( 13, 17). Livingston and Scheffer (10) also reportedthat the then incompletely characterized, lower mol wt homologswere not active in inducing ion leakage from susceptible clonesand, in a subsequent abstract (1 1), indicated that they protectedsusceptible clones from HS toxin-induced leakage.
Potential precursors of the HS toxins have recently beenisolated (13) and named HS6 (A, B, and C) because they consti-tute higher analogs differing from their HS4 counterparts throughthe additional presence of two a-glucosyl residues linked to the2 position of the terminal galactose units. Structural aspects ofHS toxin and its congeners have recently been reviewed (13).This paper examines the biological activities of these naturalproducts, using a quantitative assay based on inhibition of darkCO2 fixation in thin leaf slices (4).
MATERIALS AND METHODS
Plant Materials. Sugar cane clones CP57-603, C0453, andCP52-68 (all susceptible to H. sacchari [20]) and CP76- 1343 andH50-7209 (resistant) were obtained from Dr. J. L. Dean, USDA-ARS, Canal Pt., Fl. Plants were grown in 20-L plastic pots in agrowth chamber under a 12-h photoperiod (22°C ± 2 at benchtopheight). The plants were supplied with Hoagland solution twiceweekly.
Assays. Longitudinal strips (1.5 x 15 cm) were excised fromthe just-unrolled tissue on one side of the leaf midrib, using theyoungest leaf of plants at least 12 years old. Tissue was selectedon the basis of uniform color, such that neither very dark green,mature tissue nor very light green, immature tissue was included.
117 www.plantphysiol.orgon January 21, 2020 - Published by Downloaded from
to 3,ug Chl/slice. Chl was measured by the method of Winter-
mans and DeMots (26).Dark CO2 fixation by sugar cane leaf slices was assayed as
described by Daly and Barna for corn (4), with several modifi-
cations. Reaction vials containing slices in 0.475 ml assay buffer
were sealed with serum stoppers and incubated in a water bath
at 22 ±0.2C. For light pretreatments, the quantum flux (400-
700 nm) was 9-15 x103 MEm-2 s-'. Dark'4C02 fixation was
initiated, after a 5-min dark acclimatization (accomplished by
jacketing vials with aluminum foil sleeves), by adding 25 M1 of
60 mm NaHCO3 containing sufficient NaH'4CO3 to give 600 to
800 dpm/nmol NaHCO3. The reaction was terminated after a
fixed interval (10-20 min) by adding 0.25 ml of 1:2 (w/v)
TCA:H20 mixture. In preliminary experiments, the rate of dark
CO2 fixation was found to be essentially the same for intervals
of5 to 25 min. The vials were flushed with air for 20 to 30 minto drive off unreacted14CO2, and then 5.0 ml of scintillation
fluid (3a70; Research Products International, Mount Prospect,
IL) was added. Control vials containing assay buffer but no slices
retained insignificant amounts of14C label after aeration. Count-
ing rates were measured in a Packard Prias scintillation spectrom-
eter with an efficiency of 55 to 65%.Toxin was added to each vial at the start of the light preincu-
bation period, which was usually 1 h. Volumes of toxin solutionadded to vials ranged from 1 to 10 ul in most experiments; whenlarger volumes were applied (up to 50 u1), an equal volume ofH20 was added to control vials to minimize variations in volume.Each step in the assay was carefully timed, and vials were
handled sequentially with 15-s intervals between vials to preservethe absolute treatment interval for each vial. Three control vialswere used at the beginning, middle, and end of most experimentsin order to detect any change in fixation rates between vialshandled early and late in the assay. Replicate vials of toxin-treated slices, spaced at various intervals in the assay sequence,gave similar inhibition of dark C02 fixation, indicating that thesensitivity of the tissue did not change significantly during theassay. Each value given in an experiment is the average of atleast three replicates of 15 leaf slices each, and control values arethe average of six to nine replicates, except where noted. Allexperiments were repeated at least once.HS Toxin and Homolog Preparations. Toxin and toxin hom-
ologs were purified from H. sacchari culture filtrates as describedelsewhere (13, 16). Purified HS4 isomers A, B, and C, as well asnatural mixture of isomers of HS4 were used. HS,, HS2, HS3,and HS6 preparations all consisted of mixtures of A, B, and Cisomeric forms in roughly the same proportion as HS4 mixture(Macko, unpublished). In addition, since the aglycone is asym-metrical (Fig. 1), isomers of the lower homologs can be distin-guished based on which galactose unit(s) are missing. Thus, HS,and HS3 are each made up of six isomeric forms and HS2 nine(see Ref. 13 for details). These compounds were available inamounts ranging from a few mg (HS6) to approximately 20 mg(HS3). HS4 (mol wt = 884), HS3 (mol wt = 722), HS2 (mol wt =560), HS, (mol wt = 398) and HS6 (mol wt = 1208) weredissolved in distilled H20 to make 5 x 10-' to 5 x 10-2M stocksolutions, and 1/10 dilutions were prepared in H20. Duplicatestock solutions and dilutions of all compounds tested (exceptHS,) were prepared independently and used to confirm resultsobtained with the first set. All stock preparations and dilutionswere stored frozen at -20°C. The toxin preparations were quitestable, as no interconversion of isomers or decomposition toinactive products could be detected by HPLC methods, and theirrespective toxicities in the dark C02 fixation assay did not changeover a 9-month period.
Galactose, a-methyl-galactopyranoside, and melibiose (allfrom Sigma) were dissolved in distilled H20 to make 0.5Msolutions; raffinose (Sigma) was made up to 0.25M.Under our growth conditions and using dark C02 fixation as
an assay for sensitivity, susceptible clone CP52-68 showed thegreatest sensitivity to HS toxin and was therefore used in mostof the subsequent experiments.
RESULTS
In early stages of this research, we encountered considerableexperiment-to-experiment variation in sensitivity of tissue to HStoxin. Some degree of variation is probably inherent in the tissue.Since sugar cane is a polyploid, clonally propagated species,homogeneity of genetic makeup of sister clones cannot be con-trolled as readily as in seed-propagated plants. Further, clonalpropagation does not allow ready selection of tissues which areontogenetically identical, as was the case with the fourth trueleaf of corn used in other similar studies (4). We were, however,able to identify two sources of experimental variation: (a) timeof harvest of leaves and (b) 'aging' time between cutting slicesand adding toxin. The first is likely to be a light/temperaturestress affecting uppermost leaves during the photoperiod. Areversible reduction in sensitivity of sugar cane to H. sacchariand HS toxin at -30° C and higher has been noted previously (2,8, 20, 23). Sensitivity was improved by harvesting tissue at thestart of the photoperiod before heat buildup had occurred nearthe chamber lights. It was further noted that toxin sensitivity andcontrol fixation rates were substantially improved if leaf slices
were allowed to sit for 60 to 90 min in assay buffer beforeproceeding with the assay. Control rates of dark CO2 fixation inexperiments reported here ranged from 7 to 19 nmol CO2 slice-'h-1 (0.2-0.5 jsmol mg Chl-' h-').Time Course of Toxin Effects on Susceptible and Resistant
Sugar Cane. The time course of inhibition of dark CO2 fixationby HS toxin mixture in susceptible clone CP52-68 and resistantclone CP76-1343 is shown in Figure 2. Onset of measurableinhibition of dark CO2 fixation in susceptible tissue occurred as
early as 20 min at high toxin concentrations. Maximum inhibi-tion was reached at 100 to 130 min at high concentrations and40 to 70 min at low concentrations of toxin. At toxin concentra-tions of 1.7 gM or less, toxin-induced inhibition of dark CO2fixation gradually decayed starting after 70 min incubation time.This decay was consistently observed at lower toxin concentra-tions: for example, in three separate experiments HS toxin (1,uM) resulted in 48.0, 37.6, and 40.0% inhibition at 70 min, and21.6, 21.3, and 18.4% inhibition at 130 min. No such decay wasdetected at higher toxin concentrations (3.3 gM and above) (Fig.2). In most experiments, the maximum percent inhibitionreached with saturating concentrations of HS4 was 85 to 90%.In some cases, up to 97% of dark CO2 fixation was inhibitableby HS toxin, but amounts required to reach 50% inhibition wereessentially the same (not shown).When tested over a range of concentrations from 10 to 500
Mm, resistant clones CP76-1343 (Fig. 2) and H50-7209 (notshown) showed little or no inhibition of dark CO2 fixation overa 120-min time period.Although the data reported in this paper were obtained under
conditions of light preincubation/dark CO2 fixation (light/dark),HS toxin inhibited dark CO2 fixation to a similar degree undercontinuous dark incubation and also inhibited photosyntheticCO2 fixation with either a light or dark preincubation (data notshown). In contrast, a preincubation period in the light is required
2z0
xL.
irIL0
z0
mIz
100* CP52-68 (S)A CP76-1343 (R) JopM 0 -
80 /
160
0040
20-0
20
__ __ __ _A -A
0 \i10O pM A
0 20 40 60 80 100 120
TIME (min)FIG. 2. Time course of inhibition of dark CO2 fixation in susceptible
sugar cane clone CP52-68 and resistant clone CP76-1343 by HS toxinmixture. HS toxin was added and vials were placed in the light at timezero, and at various intervals dark CO2 fixation was assayed over a 10-min time period following 5 min dark acclimatization. Control rateswere also determined for each interval, and data are expressed as a
percentage of control rate at that interval. In the figure, 'time' refers tothe total elapsed time from addition of toxin to end of CO2 fixation;thus, the earliest time point (10 min) represents addition of toxin and['4C]bicarbonate simultaneously followed by 10 min '4C02 fixation. Datafor CP52-68 are compiled from three separate experiments encompassing0.5 and 1.0 Mm, 1.7, 3.3, and 6.6 Mm, and 10.0 Mm HS toxin, respectively.Control values (which increased with increasing incubation time in thelight) for the three experiments were approximately 11 to 15, 13 to 19,and 7 to 13 nmol '4C02 fixed/slice-h, respectively. Control values forCP75-1 343 were 7 to 13 nmol/slice -h.
)F H. SACCHARI TOXIN 119
for inhibition of photosynthetic CO2 fixation by corn leaf slicestreated with HMT toxin (4).
Toxicity ofHS Toxin Isomers on Susceptible Sugar Cane. Theeffect of punfied HS toxin isomers on dark CO2 fixation in cloneCP52-68 was determined over a range of concentrations from0.1 to 18 gM (Fig. 3). In a typical assay, isomer C inhibited darkCO2 fixation by 50% at 0.7 ,M, and isomers B and A at 4 and 6,uM, respectively. Thus, isomer C was approximately 6 and 9times more toxic than B and A, respectively. Natural toxinmixture showed a level of toxicity intermediate between C andB, consistent with its composition (5:3:2, C:B:A) (Fig. 3). In fourindependent experiments (including the one shown in Fig. 3),HS toxin mixture inhibited dark CO2 fixation by 50% at approx-
?I100z0
480 / 7
z l _ /Z _ lIi
O ..01 02 O4 0-8 2 4 8 10
TOXIN (pM)FIG. 3. Inhibition ofdark CO2 fixation in clone CP52-68 as a function
of concentration of HS toxin isomers A, B, C, and natural mixture ofisomers. Dark CO2 fixation was measured for 15 min after 60 minpreincubation with toxin in the light at 22C. Points represent the averageof three replications. Vertical lines denote 50% inhibition of dark CO2fixation. Control rate was 10.6 ± 0.70 nmol/slice-h (average of ninereplications). Data for toxin mixture is from a separate experiment inwhich the control rate was 9.2 ± 0.80 nmol/slice-h (average of ninereplications).
Table I. Effect ofHS Toxin Homologs and HS Toxin on Dark CO2Fixation in LeafSlicesfrom Resistant and Susceptible Sugar CaneAssay conditions were as described in "Materials and Methods."
Numbers represent average oftwo or three vials per treatment, expressedas per cent of control rate (average of six vials). Data are compiled fromfour experiments, where control rates ranged from 8.4 to 12.5 nmol/slice. h.
imately 1.6, 1.8, 1.4, and 1.8 AM, respectively.The toxic activity of all isomers measured with 60 min incu-
bation tapered offvery gradually with increasing dilution oftoxin(Fig. 3). In other experiments, the lowest concentrations of toxinthat consistently gave detectable inhibition of dark CO2 fixationwere 0.075 AM (isomer C) and 0.1 AM (mixture). It should benoted that the small amount of inhibition measured at these lowconcentrations would have decayed to near zero with longerincubation times, as shown by Figure 2.
Dilutions prepared from duplicate independently preparedstock solutions of isomers A, B, C, and mixture gave results (datanot shown) similar to those shown in Figure 3.
Protection of Susceptible Tissue by Lower Homologs of HSToxin. HS,, HS,, and HS3 had no significant effect on dark CO,fixation either in susceptible (CP52-68) or resistant (CP76-1343)sugar cane at or below 500 ZlM (Table I). Two other susceptibleclones, CP57-603 and C0453, also were insensitive to the hom-ologs at 100 AM (not shown). However, when added with toxin,but in molar excess, HS,. HS,, and HS3 prevented toxin-inducedinhibition of dark CO2 fixation in clone CP52-68 in a concentra-tion-dependent manner (Fig. 4). HS, was the most effectiveprotectant with CP52-68, and preserved 95% of control CO,fixation rates at 6-fold molar excess over toxin. The same degreeof protection required approximately 15:1 and 50:1 molar excessof HS3 and HS,, respectively. Thus, HS, was more than twice aseffective as HS3 and about 10 times as effective as HS, inpreventing inhibition by HS4. Results for HS, and HS3 wereconfirmed by experiments with dilutions from independentlyprepared stock solutions (data not shown).
Order-of-addition experiments were carried out to determinewhether toxin homologs could protect sugar cane tissue whenadded after as well as before toxin. Leaf slices were incubated inthe light at 22°C, and protectant (sufficient to give 80-100% ofcontrol fixation rates) was added at various times before, simul-taneously with, or after toxin addition. After 120 min incubation(total time with toxin), dark CO2 fixation was assayed. HS2 andHS3 were both effective when added up to 12 min after toxin
& 100
0
080
z 2260/-
. 4.ComarisoHS2o HSp/a HS t
40-A
0
t Toxin only CelightwihSox(3.3 pM)
2 4 68 10 20 40 60
MOLAR RATIO (protectont toxin)
FiGi. 4. Comparison of protective ability of HS toxin homologs HS,,HS,, HS3. and HS6 as a function of logio (molar ratio of protectant to
toxin). Sugar cane slices (clone CP52-68) were incubated 120 min in the
light with HS toxin (3.3 AmM) and from 6.7 to 167 umM protectants and
rates of dark CO, fixation were determined. Corresponding molar ratios(protectant:toxin) ranged from 2:1 to 50:1. The control fixation rate was7.5 ± 0.4 nmol/slice.h; toxin alone resulted in 19.3% of control rate.Each data point represents the average of three replicate vials; controland toxin only values are derived from nine replicate vials. Data for HS6are from a separate experiment in which control rate was 10.1 ± 0.4nmol/slice.h, toxin alone gave 16.0% of control rate, and HS, + HS4(10:1 molar excess HS,) gave 96.2% of control rate (positive control forprotection).
1-
z0
x
"CP
A2s HS3 15:1 (A)
HS2 6:1 (0)
80
20 Toxin only (\3-3m)
-10 0 10 20 30 40 90TIME OF ADDITION OF PROTECTANT
(2ninbefore or after txi)
FIG. 5. Effect of time of addition ofHS toxin homologs HS, and HS3on their ability to protect sugar cane tissue from the effects of HS toxin.HS3 (A) and HS2 (0) were added at 15:1 and 6:1 molar excess, respec-tively. These ratios were previously shown (see Fig. 4) to protect 80 to100% of dark CO2 fixation. All slices were incubated 120 min with toxinin the light, and HS2 or HS3 was added at various times ranging from 10min before to 90 min after toxin. After 120 min in toxin, dark CO,fixation was measured for 20 min. Controls were incubated without toxinfor 120 min; control rate was 7.7 ± 0.26 nmol/slice-h (average of ninereplicates). Data points represent the average of three replicate vials. Thefirst part of the curve is a line of best fit for data from -10 to + 12 min,and the curving portion represents the trend of the data. Similar resultswere obtained when dark CO2 was measured at 60 instead of 120 min(not shown).
Table II. Effect ofD-Galactose and a-Galactopyranosides on InhibitionofDark CO, Fixation bi' HS Toxin
Assay conditions were as described in "Materials and Methods."
Dark CO, FixationTreatment Concn.
-Toxin +ToxinaM % ofcontrolb
None (100) 46.6D(+)galactose I0- 112.0 56.9Methyl a-D-galactopyranose 10-3 109.9 53.3Melibiose 10-3 100.4 41.9Raffinose 10-3 107.8 41.7a HS4 (mixture) at 2 ,AM.Number represents average rate of fixation of three samples, ex-
pressed as per cent of control rate (1 1.0 ± 0.5 nmol/slice- h; average ofnine samples).
(Fig. 5). After this time, protective ability of the homologsdeclined over a period of 30 to 40 min, and by 90 min additionof protectant had no effect on toxin-induced inhibition of darkCO2 fixation. HS, and HS3 were also effective when added 10min prior to addition oftoxin (Fig. 5). The 'window ofprotection'(i.e., period from time zero to drop off in protection) wassomewhat variable in duration from experiment to experiment,but consistently extended at least 5 min and as great as 15 minafter toxin addition; a much higher molar excess of protectantextended the 'window' only slightly (data not shown).
Effect of a-Galactopyranosides on Inhibition by HS Toxin.The protective ability of D(+)galactose, a-methyl-galactopyran-oside, and the naturally occurring a-galactopyranosides melibi-ose and raffinose, was tested under conditions similar to those
used to measure protection by toxin homologs. No significantdegree of protection against HS toxin was observed at up to500:1 molar excess of galactosides (Table II).
Effect of HS6 on Dark CO2 Fixation. HS6 had no effect ondark CO, fixation at concentrations betwen 1 and 50 AM (datanot shown). However, HS6 was substantially inhibitory at 500,M (70.6 ± 18% inhibition; average ofthree separate experiments[Table I]). Comparing this concentration to the concentration ofHS4 (mixture) required to give the same degree of inhibition (-3,gM) (Fig. 3), HS6 can be estimated to be approximately 150-foldless toxic than HS4.HS6 did not protect susceptible tissue against HS4 when added
at noninhibitory concentrations (Fig. 4). At higher concentra-tions, HS6 had a synergistic effect on inhibition of dark CO,fixation by HS4: with 100 jM HS6, rates were from 10 to 50%lower than toxin-only rates, based on three separate experiments.
DISCUSSION
The data clearly demonstrate that an early response of suscep-tible sugar cane tissue incubated with purified HS toxin is arapid, substantial (80-90%) inhibition of dark CO, fixation. Thebasis for HS toxin-induced inhibition of dark and photosyntheticCO, fixation is at present not known. Dark CO2 fixation mayserve as a cytoplasmic 'pH stat', responding to proton fluxes (6).Any perturbation of intracellular pH, as might result from mem-brane damage, would thus in turn affect CO2 fixation. Recentresults from Schroter et al. (21; personal communication) lendsupport to this interpretation. They used microelectrode tech-niques to study the effect of HS toxin on the E,,,2 of the sugarcane leaf cell plasmalemma. Changes in E,,, (depolarization) wereobserved with toxin concentrations as low as 0.05 gM. At hightoxin concentrations, the effect was measurable as early as 4 minafter addition of toxin. Thus, an early effect of HS toxin onindividual cells appears to be perturbation of a proton gradientacross the plasmalemma. This effect may precede the onset ofinhibition of dark CO2 fixation, which in leaf slices was notapparent until 15 to 20 min incubation (Fig. 2). Electrolyteleakage from leaf discs also was first detected after 15 minexposure to HS toxin (20), although this was in response toconsiderably higher concentrations of toxin (50 gg/ml) than inthe present study. We cannot, based on present data, concludewhich effect actually occurs earliest, since whole tissue responsemay lag behind that of individual cells due to diffusion barriers.While all three responses (depolarization, inhibition of dark CO2fixation, and ion leakage) are consistent with a primary site fortoxin action on the plasmalemma, our data do not rule out thepossibility of a primary site of action other than the plasma-lemma.The partial recovery of dark CO2 fixation rates observed at
low toxin concentrations (Fig. 2) may be related to the partialrecovery of E,,, that was observed by Schroter et al. (21) with lowHS toxin concentrations under continuous light incubation. Norecovery of E,,, was observed with high concentrations of toxinor under continuous dark incubation (21). Thus at least two ofthe toxin's effects on sugar cane tissue apparently are partiallyreversible.The lowest concentration of HS toxin mixture giving a detect-
able inhibition of dark CO2 fixation in CP52-68 (0.1 Mm) iswithin the same range as that reported as the lower limit fordetection of ion leakage in the same clone by Scheffer andLivingston (20) (0.15 sg/ml or 0.17 uM) and for membranedepolarization by Schroter et al. (21) (0.05 uM). On the otherhand, the extremely high sensitivity reported for another clone(C0453), in terms of ion leakage (0.01 ,ug/ml or 0.012 gM) (20),is difficult to reconcile with the fact that, under our conditions,
clone C0453 is less sensitive to inhibition of dark CO2 fixationthan is CP52-68.The existence of three closely related isomers provided an
interesting first step in determining structure/activity relation-ships of HS toxin. HS toxin isomer C, which possesses a doublebond between carbons 3 and 4 of the aglycone (Fig. 1) wasconsiderably more active than isomers B or A in inhibiting darkCO2 fixation in susceptible clone CP52-68. Isomer C probablyaccounts for most of the activity of natural mixtures, ofwhich itconstitutes approximately 50%. The same relative and absolutetoxicities were obtained from independently isolated and purifiedpreparations, making it unlikely that dilution errors can accountfor the difference. Clearly, the toxicity of the HS toxin moleculeis significantly influenced by differences in three-dimensionalstructure related to an altered point of unsaturation in theaglycone. Chemical modification of HS toxin functional groupsand synthesis of toxin analogs, as has been done with HMT toxin(24), should provide additional information on structure/activityrelationships.
Several interpretations ofthe slight toxicity ofHS, are possible:(a) HS6 itself is marginally toxic; (b) the HS, preparation maycontain a small amount (0.5-1%) of active toxin (HS,) whichaccounts for its activity; (c) small amounts of HS, are convertedto HS4 by the action of glucosidases in sugar cane tissue. Exper-iments are underway to distinguish among these possibilities.Another aspect of the relationship between structure and ac-
tivity is apparent in the comparison between the biologicalactivity of HS4 and that of its homologs. HS, HS,, and HS,, werenot toxic to three susceptible sugar cane clones; this indicates aminimum requirement of four galactose units per HS toxinmolecule for toxicity. Addition of glucopyranosyl groups to bothtermini (i.e., HS() also reduced toxicity, a further indication ofthe structural constraints on toxic activity. These homologs do,however, possess biological activity, as demonstrated by theireffect on inhibition of dark CO2 fixation by HS4. The protectionof susceptible sugar cane by HS,, HS,, and HS3 confirms theearlier report by Livingston and Scheffer ( 11) in which nontoxiccompounds from H. sacchari inhibited toxin-induced loss ofelectrolytes from leaf discs. The most effective compound re-ported by them inhibited 95% of ion loss at 40-fold (by weight)excess. Recently, Livingston and Scheffer (personal communi-cation) have determined that HS3 was the most effective protec-tant (80% at 5:1 excess wt) in electrolyte leakage assays involvingseveral clones other than CP52-68. In the present study, HS, wasthe most effective protectant, with 95 to 100% protection at a 6-to 7-fold molar excess. Possible explanations for the conflictingresults obtained in the two laboratories include differences in:(a) the protective response of different susceptible sugar caneclones; (b) environmental or cultural conditions affecting theresponse; (c) the way in which the two assay systems measurethe protective response; and (d) the isomeric composition of thehomolog preparations used (since the isomeric forms of a hom-olog such as HS2 may be present in different relative amounts indifferent preparations, and may not all be equally active asprotectants). Regarding point d, we have preliminary evidenceindicating that the three sugar positional isomers of HS, havesignificantly different relative protective activities. The relativeactivities of the sesquiterpene isomers of the homologs are notknown. More detailed protection studies employing purified,individual isomers of HS2 and HS3 are currently underway.The lack of protection by D-galactose or a-D-galactopyrano-
sides is not consistent with an earlier report that a, but not 13,galactopyranosides protect sugar cane tissue from runner lesionsinduced by HS toxin preparations (22). Since in the earlierexperiments leaf strips were incubated in a-galactoside solutionsfor 24 h prior to assaying for sensitivity to HS toxin (leaf punctureassay), it is possible that the observed protection was an indirect
result of uptake and metabolism of test sugars rather than adirect competition at a common binding site as was claimed.(No experiments in which toxin and tests sugars were appliedtogether to sugar cane tissue were reported). In addition, sincethe galactosyl units of HS toxin and its homologs are now knownto be of the ,B-linked furanose form, there is little reason to expectthat a-galactopyranosides would be active as competitive inhib-itors of toxin action. Preliminary evidence (unpublished) indi-cates that several synthetic l-D-galactofuranoside compounds doin fact have protective activity in the dark CO2 fixation bioassay.
Strobel (22) reported the isolation from susceptible sugar caneleaves of a membrane protein which was capable of binding `4C-labeled HS toxin as well as a-galactoside sugars and proposedthat this protein was the initial site of toxin activity. However,Lesney et al. (9) were unable to reproduce experiments in which'4C-labeled HS toxin purportedly was bound preferentially bymembrane vesicles from susceptible sugar cane (22). This, to-gether with the lack of structural similarity between HS toxinand a-galactoside sugars, makes the current status of the a-galactoside-binding protein uncertain.The protection against HS toxin by the lower homologs is not
inconsistent with the notion that these molecules bind to acommon receptor site(s) in susceptible sugar cane. Such a recep-tor site(s) could either be involved in transport oftoxin into cells,or could represent the primary site causing damage, or both. Ourdata do not distinguish among these possibilities. In either case,the relative protective activities of the homologs might reflecttheir relative affinities for the receptor in competition with toxin.The ability of HS2 and HS3 to protect when added severalminutes after toxin (Fig. 5) is more difficult to interpret, butcould be due to more rapid transport of protectant to active sitesor displacement of toxin from active sites before toxicity ismanifested. Alternatively, protectants might 'repair' or reverseinitial damage by the toxic species by a mechanism not neces-sarily involving a common site (3, 25). Ultimately, a more directapproach using purified, labeled toxin or a toxic derivative as aprobe will likely be necessary to elucidate the nature and locationof the active site(s) for HS toxin.The presence of nontoxic compounds that antagonize toxin
action has important implications for the search for toxins inculture filtrates of plant pathogens. HS toxin homologs arepresent in considerable quantity in older cultures of the fungus,apparently as a result of 13-galactofuranosidase cleavage of activetoxin ( 12). Antagonism by HS2 is detectable at 2:1 molar excessand is virtually complete at 8:1 molar excess; thus, crude prepa-rations from older cultures could contain inherent, but 'masked',toxin activity. Similarly, toxins produced by other pathogens inculture could conceivably be overlooked during screening ifantagonistic compounds were present in culture fluids or extracts.
Ultimately, the significance of the toxic, nontoxic, and antag-onistic compounds produced by H. sacchari in culture must alsobe considered from the standpoint of how they affect diseasedevelopment in the sugar cane host. Important, unansweredquestions include the relative and absolute amounts of thesecompounds produced in situt, and the role, if any, played by hostenzymes such as galactosidases or glucosidases in modifying thetoxin and homologues in susceptible or resistant tissue.
In summary, the data provide some potentially valuable in-sights into some of the structural feature that are important forthe biological activity of HS toxin and its congeners: (a) alterationof a single point of unsaturation at carbon-4 of the aglyconesignificantly affects toxicity; (b) loss of one or more galactofura-nose units from the HS toxin molecule renders the moleculenontoxic to all susceptible clones thus far tested, but still biolog-ically active as an antagonist of toxin; (c) the number of galac-tofuranose units is important in determining the relative protec-
tive activity of the lower homologs; (d) addition of two gluco-pyranose units (HS6) to active toxin greatly reduces toxicity butdoes not yield an active protectant.
Additional information on structure/function relationshipscould be derived from a comparison of protection by individualisomers of the homologs as well as from studies of the protectiveor toxic properties of chemically altered toxin and natural orsynthetic analogs.
LITERATURE CITED
1. BoTTINI AT, JR BOWEN, DG GILCHRIST 1981 Phytotoxins. II. Characterizationof phytotoxic fractions of Alternaria alternata lvcopersici. Tetrahedron Lett22: 2723-2726
2. BYTHER RS, GW STEINER 1975 Heat-induced resistance of sugar-cane toHelminthosporium sacchari and helminthosporoside. Plant Physiol 56: 415-419
3. DALY JM 1981 Mechanisms of action. In RD Durbin, ed, Toxins in PlantDisease. Academic Press, New York, pp 331-394
4. DALY JM, B BARNA 1980 A differential effect of race T toxin on dark andphotosynthetic CO2 fixation by thin leaf slices from susceptible corn. PlantPhysiol 66: 580-583
5. DALY JM, HW KNOCHE 1982 The chemistry and biology of pathotoxinsexhibiting host-selectivity. Adv Plant Pathology 1: 83-138
6. DAVIES DD 1973 Control of and by pH. Symp Soc Exp Biol 27: 513-5297. KOHMOTO K, ID KHAN, Y RENBUTSU, T TANIGUCHI, S NisHIMURA 1976
Multiple host-specific toxins of Alternaria mali and their effect on thepermeability of host cells. Physiol Plant Pathol 8: 141-153
8. LARKIN PJ, WR SCOWCROFT 1981 Eyespot disease of sugarcane. Induction ofhost-specific toxin and its interaction with leaf cells. Plant Physiol 67: 408-414
9. LESNEY MS, RS LIVINGSTON, RP SCHEFFER 1982 Effects of toxin from Hel-minthosporium sacchari on nongreen tissues and a reexamination of toxinbinding. Phytopathology 72: 844-849
10. LIVINGSTON RS, RP SCHEFFER 1981 Isolation and characterization of host-selective toxin from Helminthosporium sacchari. J Biol Chem 256: 1705-1710
11. LIVINGSTON RS, RP SCHEFFER 1981 Fungal products chemically related toHelminthosporium saccari toxin protect sugar cane tissues from the toxin.Phytopathology 71: 891 (abstr)
12. LIVINGSTON RS, RP SCHEFFER 1983 Enzymic conversion ofHelminthosporiumsacchari toxin to toxoids by ,B-galactofuranosidase from Helminthosporiuim.Plant Physiol 72: 530-534
13. MACKO V 1983 Structural aspects of toxins. In JM Daly, BJ Deverall, eds,Toxins and Plant Pathogenesis. Academic Press, Sydney, Australia, pp 41-80
14. MACKO V, W ACKLIN, D ARIGONI 1982 Structure of host-specific toxinsproduced by Helminthosporium sacchari. Abstr 252, Am Chem Soc 183rdNational Meeting
15. MACKOV,W ACKLIN, C HILDENBRAND, F WEIBEL, D ARIGONI 1983 Structureof three isomeric host-specific toxins from Helminthosporium sacchari.Experientia 39: 343-347
16. MACKO V, K GOODFRIEND, T WACHS, JAA RENWICK, W ACKLIN, D ARIGONI1981 Characterization ofthe host-specific toxins produced by Helminthospo-rium sacchari, the causal organism of eyespot disease of sugarcane. Exper-ientia 37: 923-924
17. MACKO V, C GRINNALDS, J GOLAY, D ARIGONI, W ACKLIN, F WEIBEL, CHILDENBRAND 1982 Characterization of lower homologues of host-specifictoxins from Helminthosporium sacchari. Phytopathology 72: 942 (abstr)
18. PAYNEG, HW KNOCHE, Y KONO, JM DALY 1980 Biological activity ofpurifiedhost-specific pathotoxin produced by Bipolaris (Helminthosporium) maydis,race T. Physiol Plant Pathol 16:227-239
19. RATHNAM CKM 1978 Metabolic regulation of carbon flux during photosyn-thesis. I. Evidence for parallel CO2 fixation by mesophyll and bundle sheathcells in situ. Z Pflanzenphysiol 87: 65-84
20. SCHEFFER RP, RA LIVINGSTON 1980 Sensitivity of sugarcane clones to toxinfrom Helminthosporium sacchari as determined by electrolyte leakage. Phy-topathology 70: 400-404
2 1. SCHROTER H, A NOVACKY, V MACKO 1983 Helminthosporium sacchari-toxinaffects the energy-dependent component ofmembrane potential. Plant Phys-iol 7 1: s-24
22. STROBEL GA 1973 The helminthosporoside-binding protein of sugarcane. Itsproperties and relationship to susceptibility to the eye spot disease. J BiolChem 248: 1321-1328
23. STROBEL GA 1979 The relationship between membrane ATPase activity insugarcane and heat-induced resistance to helminthosporoside. Biochim Bio-phys Acta 554: 460-468
24. SUZUKI Y, KJ TEGTMEIER, JM DALY, HW KNOCHE 1982 Analogs of host-specific phytotoxin produced by Helminthosporium maydis, race T. II.Biological activities. Bioorg Chem I1: 313-321
25. WHEELER H 1969 In Symposium on Potential Crop Protection. Cornell Uni-versity, Ithaca, pp 9-13
26. WINTERMANSJFGM, A DEMOTS 1965 Spectrophotometric characteristics ofchlorophylls a and b and their phaeophytins in ethanol. Biochim BiophysActa 109:448-453
Terence M. Murphy, Gerald B. Matson, and Steven L. Morrison.Ultraviolet-Stimulated KHCO3 Efflux from Rose Cells. Reg-ulation ofCytoplasmic pH.
Table I, column 2, row 4, should read: 1.07 ± 0.09
Jonathan P. Duvick, J. M. Daly, Z. Kratky, V. Macko, W.Acklin, and D. Arigoni. Biological Activity of the IsomericForms ofHelminthosporium sacchari Toxin and of HomologsProduced in Culture.
Page 1 7, column 2, line 3 from bottom, should read: .... atleast 12 weeks old....
