M IIIIIIIIIIIIIIIIIIIIIIIIIIII

-——*-- m 1mm:(”Hill/NIWilli]!!! WIN/WWW

fiimigam S. tam
Umtverezty

 

THESIS

     
   

  

This is to certify that the

thesis entitled

ISOLATION OF HYPOXYIDN MAM’IATLM 'IOXINS

 

AND EFFECIS 0N POPULUS 'I‘RIMUIDIDES
presented by

 

Bruce Allen Stenner

has been accepted towards fulfillment
of the requirements for

Master of Science degree in Botany & Plant Pathology

 

 

7W” 4

Date 10 August 1981

0-7639

ovanouc FINES:
25¢ per day per item
{ffm‘fi f anumus LIBRARY MATERIALS:
m h

-‘ 51"":5 Place in book return to remove
~ “"” . charge from circulation records

 

_. n

 

{"x. , , L 2 -
”+41% s 45 ‘51:? ~ -
. ‘9’ ,'. .

3"

F, J
L: L1

(I.
U.

 

 

 

 

ISOLATION OF HYPOXYLON MAMMATUM TOXINS
AND EFFECTS ON POPULUS TREMULOIDES

 

By

Bruce Allen Stermer

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Botany and Plant Pathology

1981

G// :7/_.1'”

ABSTRACT

ISOLATION OF HYPOXYLON MAMMATUM TOXINS
AND EFFECTS ON POPULUS TREMULOIDES

 

By

Bruce Allen Stermer

Clones of E. tremuloides varied in sensitivity to toxins from fl.

mammatum, as shown by effects on leaves and young stems. A leaf assay,

sensitive enough to detect toxin in samples containing 10 ng dry weight,
was based on uptake of toxin solution by cut petioles. Leaves from a
sensitive clone showed black necrosis by 5 h after exposure; another
clone tolerated SOD-fold more toxin with no obvious damage. Sensitivity
of stems was correlated with sensitivity of leaves. Leaves of the
sensitive clone responded within 1 h by increased respiration (+36%); 02
uptake by the tolerant clone was not affected. SToxin caused an increase
(+500 to 600%) in leakage of electrolytes from tissues of the sensitive,
but not the tolerant, clone, beginning 6 h after exposure. Toxin was
partially purified by use of Bio-Gel P-2 and silicic acid columns.
Several host-specific toxic components were recovered. These and other
data suggest that the toxins are involved in development of Hypoxylon

canker.

ACKNOWLEDGEMENTS

Sincere thanks are expressed to Drs. Robert P. Scheffer and John H.
Hart for providing me with guidance and encouragement in these studies.
I also acknowledge with gratitude the financial support of Dr. J. H. Hart
which permitted me to make these studies.

Thanks are also expressed to Dr. Mark Lesney for his assistance and
encouragement in this work.

Lastly, I acknowledge Mr. Bob Livingston, Dr. Jerry Cohen, and

Mr. Steven Briggs for their aid with the many techniques used in these

studies.

ii

TABLE OF CONTENTS

LIST OF TABLES ................. ........ ............................
LIST OF FIGURES ....................................................
INTRODUCTION ....................................... ..... .... ...... .
LITERATURE REVIEW ..................................................

Host Range ....................................................
Mode of Infection .............................................
Histology of Pathogenesis .....................................
Effects of Environmental Factors on Development

of Hypoxylon Canker ........................................
Host Resistance to H. mammatum ................................
The H. mammatum Toxin .........................................

MATERIALS AND METHODS 9.0.0.0...OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Plant Material ................................................
Fungal Material ...............................................
Production of Toxin ...........................................
Bioassay 00.00000000000000000000000000000000000000IOOOOOOOOOOOO
Thin Layer Chromatography .....................................
High Pressure Liquid Chromatography ...........................
Absorbance Spectroscopy .......................................
High Voltage Paper Electrophoresis ............................
Ion Exchange ..................................................
Narburg Manometry .............................................
Electrolyte Leakage ...........................................

RESULTS 0.00....OOOOOOOOOOOOOOOOOOOO...OOOOOOOOOOOOOOOOOOOOOOOO0....

Toxin Isolation Procedures ....................................
Early stages in purification .............................
Gel filtration ...........................................
Thin layer chromatography ................................
Silicic acid column chromatography .......................
High pressure liquid chromatography ......................
Absorption spectra .......................................
Electrophoresis ..........................................
Ion EXChange 0.000...OOOOOOOOOOOOOOIOOOOOOO0000.00.00.00.0

Page

vi

o—It—H-u-I
NHHOflOto Q Noun UTWN N

12
12
13
13
13

15

15
15
16
17
17
20
21
21
21

Page

Toxin Bioassays ............................................... 23
Leaf puncture assay ...................................... 23
Stem assay ............................................... 24
Leaf petiole assay ....................................... 28

Effects of Toxin on Tissues.................................... 30
Toxin-induced changes in respiration ..................... 30
Toxin-induced changes in electrolyte leakage ............. 30
Isolated P, tremuloides protoplasts ...................... 32

Effects Of TOXin on Aspen Hybrids OOOOOOOOOOOOOOO0.00.0.0000... 34
Leaf assay 0f aspen hybrids OI.OOOOOOCOOOOOOOOOOOOOOOOOOOO 34

DISCUSSION C.0.00....O...COO...OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOIO 36
ISOIation 0f fl. mamatum TOXinS 0......OOOOOOOOOOOOOOOOOOOCOOOO 36
TOXin Bioassays 00......OOOOOOOOOOOOOOOOOO0.00000000000000COOOO 38
EffeCts Of TOXin O..0.0...O...COCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 41

BIBLIOGRAPHY 0.0.0......0.0.0..0...OOOOOOCOOOOOOOOOOOOO0.0.0.0000... 45

iv

LIST OF TABLES

Table Page

1 Light absorption and relative toxicity of fractions
from HPLC Offl. mamatum tOXTnS COOOOOOOOOOOOOOOOOOOOOO0.0... 22

2 Toxicity of H. mammatum toxin, NaCl, and methanol to
leaves from three clones of E. tremuloides .................. 25

 

3 Effect of H. mammatum toxin, methanol, and water on
stems of three clones of'E, tremuloides ..................... 26

 

4 Sensitivity of Populus hybrids to toxins from
E. mmatutfl 0.0.0....O..00...OOOOOOOOOOOOOOOOOO0.00......O... 35

LIST OF FIGURES

Figure Page

1 Effect of culture age on H. mammatum toxins recovered
from culture fluids. Stationary cultures were grown
on modified Fries medium at 24°C and harvested at 2, 4,
and 7 weeks. The toxins were chromatographed with a
Bio-Gel P-2 column (1.5 x 90 cm) ........................... 18

2 Chromatography of H. mammtum toxins on a silicic acid
column. Stationary cultures were grown on modified
Fries medium at 24°C and harvested at 7 weeks. The
toxins were chromatographed on a 1.5 x 90 cm column
packed with Bio-Sil A (100 to 200 mesh) and eluted with
n-hexane/isopropanol/acetone (6:4:1) ....................... 19

3 Effect of H. mammatum toxin on 0 consumption by
leaves of P. tremuloides clones 2 and 5 (C2 and C5).
Leaf disks (1 cm in diameter) were infiltrated at
reduced pressure with toxin solution or water. Leaf
disks (0.1 g) were placed in Warburg flasks and the
0 uptake measured manometrically. Each value is
tfie mean of two replicates ................................. 31

 

4 Effect of H. mammatum toxin on electrolyte leakage from
leaf disks of E. tremuloides clones 2 and 5 (C2 and C5).
Leaves were allowed to take up 10 ul of toxin solution
through the cut petioles. Leaf disks (1 cm diameter)
were then cut and 0.1 g of tissues were placed in vials
containing 2.0 ml of water. Conductivity of ambient
solutions was measured at invervals ........................ 33

vi

INTRODUCTION

Hypoxylon canker caused by Hypoxylon mammatum (Nahl.) Miller is one
of the most important diseases of quaking aspen, P0pulus tremuloides
Michx., in North America. It occurs throughout most of the range of
quaking aspen in the United States and Canada. In Michigan, Wisconsin,
and Minnesota, more than 12% of the live aspen are currently infected
with Hypoxylon canker, resulting in economic losses for this valuable
source of fiber in the millions of dollars per year (28). An estimated
300 million ft3 of aspen wood is lost each year in the Lake States as
the result of Hypoxylon canker, accumulation of these yearly losses until
harvest means the annual growth and utilization of aspen could be doubled
if Hypoxylon canker were eliminated (28). However, no effective control
measures for Hypoxylon canker are known.

Schipper (40) reported that a toxin produced by H. mammatum was
necessary for the infection of aspen by the fungus. Other investigators
(13,18) have attempted to screen for resistance in aspen to Hypoxylon
canker by selecting trees tolerant of the toxin from H, mammatum. The
purpose of my study was to further purify the toxin, to learn more about
the character of the toxin, and to examine some of the effects of the

toxin on aspen tissues.

LITERATURE REVIEW

Hypoxylon canker was first found in 1920 in Essex County, New York,
and Oakland County, Michigan, and described by Povah (32) in 1924.
However, the fungus itself was described by Klotsch (26) in 1833 as
Sphaeria pruinata. Later in 1883 Cooke (14) transferred it to the genus
Hypoxylon and described the fungus as H, pruinatum (Klot.) Cooke.
Recently, Miller (29) has combined H, pruinatum with H. mammatum (Wahl.)
Miller. Koch's postulates were completed for Hypoxylon canker by Bier
(9) in 1940; the general characteristics of the disease have been

reviewed by Gruenhagen (20).

Host Range

The most common host of H. mammatum is Populus tremuloides. Other

 

hosts in decreasing order of susceptibility are P. grandidentata Michx.,

 

.P. balsamifera L., and P, trichocarpa Torr & Grey (21). Naturally
occurring cankers have also been described on P. tremula L. (3), and on
Acer rubrum L. and A. saccharum Marsh. (8). Miller reported H, mammatum

on wood of Acer, Alnus, Betula, Carpinus, Fagus, Picea, Populus, Pyrus,

 

Salix, Sorbus, and Hlflfléfi however, he does not discuss pathogenicity to
these trees (29).

Attempts at inducing Hypoxylon cankers by artificial inoculation
have given various results. Generally, efforts have been unsuccessful in

producing cankers with H. mammatum by inoculating aspen buds, leaves,

petioles, or leaf axils with mycelia or with suspensions of ascospores or
conidia (9,20,42,43). However, cankers developed readily on stems and
branches inoculated with mycelia (9,20,42,43). H, mammatum has a wider
host range on excised branches; cankers are produced on cuttings of H.

deltoides, H, deltoides x.H. balsamifera (7), Acer macrophyllum Pursch.,

 

 

and Salix sp. (10).

Mode of Infection

 

Transmission and the mode of infection for H, mammatum is uncertain.
Researchers have been unable to infect aspen by inoculating the unwounded
periderm and agree that the fungus is a wound parasite on branches and
trunks (9,20). The type of wounds which serve as infection courts is not
clear, but insect injuries are commonly associated with cankers.

Anderson et al. (2) found that out of 169 natural cankers in an aspen
plantation in Minnesota, 95% originated in galls induced by the insect
Saperda inornata, although only 1% of the total galls were infected.
Similar observations in Michigan showed that an important relationship
exists between Hypoxylon canker and insects of aspen (19). More studies
of the aspen-insect-fungus interaction are needed before the role of
insects in Hypoxylon canker is understood.

The major inoculum for natural infections is thought to be
ascosoores (7), but efforts to infect with the spores have been negative
(9,42). Ascospores are discharged from stromata throughout the year
following precipitation, either rain or snow, even when the ambient
temperature is below 0 C (47). Hypoxylon cankers bearing perithecia are
present in most, if not all quaking aspen stands within the range of the
disease, and cankers are obtained easily by wound inoculation with

mycelia derived directly from ascospores. Therefore, Berbee and Rogers

(7) suggested that certain requirements are necessary for gerimination of
ascospores or for the initial mycelial growth and that these requirements
must be met for producing cankers. Reports of ascospore germination in
the sapwood (1) and the xylem (5) of aspen indicates that infection RY.fl-
mammatum depends on conditions for mycelial establishment and not on
ascospore germination.

Bagga and Smalley (5) consistently produced cankers on aspen inocu—
lated with ascospores when cell-free fungal extracts from H, mammatum
cultures were applied tapically over the inoculated wounds. No cankers
developed when ascospores were inoculated with 1% glucose, 2% malt
extract, or water. The materials produced in culture were phytotoxic and

caused extensive bark necrosis when applied to H, tremuloides stems.

 

Bagga and Smalley concluded that a heat-resistant, water-soluble product
from the culture filtrate was necessary for the production of cankers by
ascospores.

In further studies, Schipper (40) found that an avirulent or a viru-
lent isolate of H, mammatum successfully infected aSpen when exogenous
"mammatoxin" from culture filtrates was applied to the inoculated wounds;
there were no successful infections by either isolate when toxin was
withheld. Schipper noted that infection of aspen by H. mammatum has been
successful only when mycelium growing on agar or grain was introduced
into wounds, and that both agar and grain upon which H. mammatum has
grown contains measurable amounts of the toxin. He suggested that
successful natural infection occurred only when mycelium and toxin were
introduced into wounds on aspen concurrently, perhaps by a vector
carrying a piece of cankered host tissue into a wound on a healthy aspen

tree.

Histology of Pathogenesis

.H. mammatum grows primarily in the sapwood of infected trees and
invades the bark from within (35). The fungus apparently invades only
dead bark, presumably because aspen bark contains fungitoxic compounds
(41). The pathogen has been isolated from discolored wood up to 5 cm

beyond the visible canker limit, and under sound bark of H. tremuloides

 

(25). Macrosc0pic and microsc0pic examination of inoculated wounds
revealed that bark collapse preceded the hyphae of H. mammatum in the
bark (38). The toxin found in culture filtrates of the fungus may be
responsible for the bark necrosis seen in Hypoxylon cankers (41). The
pathogen was isolated by Gruenhagen (20) from wood up to 0.8 cm beneath
the cambium. Bagga (4) isolated H, mammatum from a naturally infected
stem (9 cm dbh) throughout the entire area of discolored wood, up to the

pith.

Effects of Environmental Factors on Devel0pment of Hypoxylon Canker

The effect of the environment on Hypoxylon canker has been studied
extensively by many workers. Any factor which contributed to moisture
stress in the host increased susceptibility to H. mammatum (6). Relative
humidities higher than 50% and soils with high water-holding capacities
were unfavorable for canker develOpment. Non-fertilized plants were more
susceptible than plants fertilized regularly.

A model accounting for 92% of the variation in canker incidence was
constructed by Bruck and Manion (12) using 56 measured parameters. Highly
significant negative correlations were observed between canker frequency
and soil moisture. They summarized their results by stating that physical
and chemical parameters of the soil which relate most directly to moisture

were highly correlated with Hypoxylon canker incidence.

 

Host Resistance to H. mammatum

 

Factors responsible for resistance to Hypoxylon canker have not been
established. Berbee and Rogers (7) postulated that rapid formation of
callus around wounds was important for resistance. They observed rapid
production of large amounts of wound callus in species of Pogulus
resistant to H. mammatum infection, but little callus in susceptible
species. Constitutent bark materials may also be involved in resistance

to infection (17). Aqueous extracts from the outer bark of H. tremuloides

 

stimulated ascospore germination of H. mammatum, while extracts from the
underlying green layer inhibited their germination. Hubbes (23) reported
that extracts of aspen bark strongly inhibited the fungus in culture, and
attributed the inhibition to pyrocatechol, two glycosides, and possibly
other substances in the bark.

Recent work has shown that phytoalexins are produced when H. mammatum
is placed on debarked aspen stem disks (15,16). The significance of this
is questionable; although mycelial plugs of H. mammatum caused production
of phytoalexins, the phytoalexins were not capable of inhibiting mycelial
growth of H. mammatum. Germination of the ascospores of H. mammatum were
inhibited markedly by the aspen phytoalexins, but the spores did not
induce the stem disks to produce phytoalexins. Other mechanisms of
resistance to Hypoxylon canker that have been suggested include microbial
competition (11) and rapid death of infected branches (46).

Genetic control of resistance to H. mammatum in aspen is complex. In
inoculation tests of aspen species and their hybrids there was a lack of
dominance for resistance or susceptibility to infection (27). The hybrids

were intermediate to their parents and varied in susceptibility to the

fungus. In contrast, a study by Berbee and Rogers (7) suggested that
susceptibility of aspen to Hypoxylon canker was dominant. Overall,
evidence indiciates that the reaction of aspen to Hypoxylon canker is

controlled by many genes (27,45,46).

The H. mammatum Toxin

 

In 1964 Hubbes was the first to report that H. mammatum produced
substances that were toxic to H. tremuloides tissues, and that the
diffusible substances prevented wound callus formation (24,25). In order
to have successful initiation of canker develOpment with H. mammatum
ascospores, application of toxic materials produced in culture by the
fungus was necessary (5). The fungal metabolite was most toxic on the

tree species most susceptible to Hypoxylon canker, H. tremuloides, and

 

much less toxic on other Papulus species (38). The fungal metabolite is
a host-specific toxin that causes bark collapse around aspen stem wounds
identical with that caused by H, mammatum infection. Cell-free extracts
of H. mammatum increased the activity of peroxidase enzymes in treated
aspen seedlings (39).

The necrotic effect of the toxin on tissues of H. tremuloides is the

 

basis of a bioassay. Schipper (38,41) placed fully expanded young aspen
leaves in 27 ml vials that contained enough of the test solution to
immerse the cut end of the petiole. The leaves were examined for
necrosis after 16 to 96 hours of exposure. In another leaf bioassay,
excised leaves were punctured with a 16-gauge hypodermic needle and a 20
pl drop of test solution was immediately placed on the puncture (18).
Leaves maintained in humid chambers developed a discrete black necrotic
spot which radiated from the point of toxin application. Aspen stem

bioassays have been tested, but they have not proven to be useful (41).

Only 3, tremuloides was highly sensitive to the toxin from H,

mammatum, as shown by leaf bioassays of numerous plants in 24 different

genera. French (18) reported that clones of_H. tremuloides varied

 

greatly in sensitivity of the toxin; many clones growing in the field were
significantly tolerant of the toxin. French was not able to correlate
tolerance to the toxin with resistance to infection by H. mammatum.

All isolates of the fungus tested by French produced toxin in
culture, but the non-pathogenic isolates or those with low virulence
produced smaller amounts of toxin (18). The fungus produced the toxic
metabolites under all tested conditions which allowed growth of the
fungus. The toxin has been isolated from Hypoxylon cankers (41).

The toxin produced in culture by H. mammatum was partially purified
and was described as a heat-stable, non-protein compound with a mol wt
of 700 to 1,100 (18, 41). The toxin was highly soluble in acetone,
alcohols, and water, but was insoluble in chloroform, ether, or n-hexane.
As many as 7 toxic forms were detected when the toxin was subjected to

thin layer or silicic acid column chromatography (41).

 

MATERIALS AND METHODS

Plant Material

 

Young aspen trees were obtained from root cuttings of naturally

occurring clones of_H. tremuloides located near East Lansing, Michigan.

 

The clones were identified by French (18) as highly sensitive to H.
mammatum toxin (clone 5), highly tolerant to the toxin (clone 2), or
intermediate in reaction to the toxin (clone 3). Subsurface runner roots
2.5 cm or less in diameter were collected, scrubbed in soap and water,
cut into sections 10 to 15 cm long, and planted in a 1:1 sand-vermiculite
mixture. Root sprouts were excised from root sections when 3 to 5 cm
tall, and planted in the same mixture in flats covered with saran wrap.
Rooted sprouts were then transplanted into 4 inch pots. When the young
trees were 75 to 100 cm tall they were trimmed to encourage branching

and planted in 10 inch pots. The plants were potted in a soil-less mix
(Sphagnum peat moss, vermiculite, and perlite) and fertilized biweekly
with 20-19-18 soluble fertilizer. Spraying was done as needed for

mite control with Plictran or Vendex at a rate of 3/4 teaspoon per
gallon. A photoperiod of at least 16 h was maintained with supplemental

illumination.

Fungal Material
The culture of H. mammatum used for toxin production was a single

spore isolate previously described by French as isolate RL5A-7 (18).

10

Production of Toxin

 

One liter Roux bottles containing 200 ml of a modified Fries medium
(33) were inoculated with 1 ml of blended mycelium. The bottles were
kept stationary at 24 C under low light. After 2 to 7 weeks the culture
fluid was harvested by filtration through 2 layers of cheesecloth, followed
by filtration through a layer of miracloth in a Buchner funnel. The
crude filtrate was concentrated under reduced pressure at 30 to 40 C to
approximately 10% of its original volume. An equal volume of methanol was
added and the solution was stored overnight at 4 C. Precipitates were
removed by filtration through Hhatman GF/C filter paper and the solution
reduced to approximately 2.5% of the original volume. The concentrated,
deproteinized, aqueous culture filtrate was partitioned 3 times against
equal volumes of chloroform and then 3 times against equal volumes of
water-saturated n-butanol. The butanol extracts, which contained most of
the toxic activity, were combined, reduced to near dryness, and resuspended
in 300 ml of distilled water. Activated charcoal (HCl-washed, from Sigma
Chemical Co.) was added (4 to 5 g per liter of original filtrate) and the
slurry was stirred overnight at 4 C. The charcoal slurry was centrifuged,
the supernatant discarded, and the pellet washed by resuspending in water
and recentrifuging. Next the toxin was desorbed from the charcoal by 3
washings with water-saturated n-butanol. The butanol fractions were
combined, reduced to near dryness and resuspended in a minimum amount of
water. The concentrated toxin preparation was then chromatographed
through a Bio-Gel P-2 or a Sephadex 6-15 column (1.5 x 90 cm) with water
as the solvent, at a flow rate of 0.1 ml per min. The fractions from the
column were assayed for toxicity and the toxic fractions were combined.

This was the partially purified toxin used in experiments.

11

Bioassay
Solutions were assayed for toxin using a leaf puncture method

similar to that used by French (18). Excised leaves were punctured with
a needle, and a 10 11] drop of test solution was placed inmediately on
the puncture. The petioles of the leaves were imnersed in water in
vials which were kept in humid chambers. The diameters of the resultant
lesions were measured at 24 or 48 h after treatment. All assays were
replicated on leaves of the toxin-sensitive clone 5 and the toxin-

tolerant clone 2, to determine host-specificity.

Thin Layer Chromatography

 

Both Analtech Silica Gel GF and Merck Silica Gel 60 plates were
used. Because the Merck plates had a more durable surface and also gave
better resolution, they were used in all the thin layer chromatography
(TLC). Substances on the plates were assayed either for toxin or treated
for observation. The most common method was to first Spray the plate
with a fine mist of concentrated H2304, observe the plate under longwave
UV light for fluorescent materials, then heat the plate in a 90 to 100 C
oven for 20 min and observe the charred materials. For the bioassay,

0.5 or 1.0 cm2 areas of gel were separately scraped into test tubes,
su5pended in 0.5 ml of distilled water, shaken thoroughly and the
solution assayed when the gel had settled. Typically, duplicate samples
were chromatographed on the same plate so that after assaying for toxin
the correSponding areas could be observed. All TLC plates were prewashed

with 95% ethanol.

12

High Pressure Liquid Chromatography

The machine used was a Varian LC 5060 HPLC with a Nhatman CozPell
ODS precolumn and a Varian 4.1 x 300 mm MCH-lO (C13 reverse phase, 10 pm
particles) column. A 100 pl aqueous toxin sample was injected into the
machine and eluted with gradients of increasing ethanol concentration in
water. A UV detector monitored absorption at 280 nm at the outlet of
the column. Fractions were collected, the ethanol was removed, and then

assayed for toxin after equalizing fraction volumes.

Absorbance Spectroscopy
Continuous visible-UV absorbance spectra from 800 to 200 nm were
produced with a Cary 15 recording spectrOphotometer (1 cm cell, scale 0

to 1.0 0.0.).

High Voltage Paper Electrophoresis

Toxin samples were subjected to electrophoresis on a 115 cm long
sheet of whatman #4 paper in a Savant high voltage paper electrophoresis
apparatus. The buffer used was pyridine/acetic acid/water (25:1:225) pH
6.5 (30). The paper was run in kerosene electrophoretic coolant at 7000
v and 100 ma for 1 h. The toxin was detected by cutting the paper into
sections, eluting the pieces with water, and assaying the water extract
for host-specific toxicity. A mixture of acidic, neutral, and basic
amino acids were also run as standards; they were detected by dipping the
paper in a ninhydrin solution and heating after the toxin-containing

segments had been removed.

13

Ion Exchange

 

Columns used for ion exchange chromatograhy were 5 3/4 inch glass
Pasteur pipets. A glass wool plug was used in the narrow end to support
the exchanger bed, and a serum bottle cap with a length of teflon
tubing through its center was used to seal the tap of the column. The
exchangers were slurry packed and washed with 1N HCl and 1N NaOH. A 0.2
to 0.5 ml sample of toxin was applied to the top of the column bed for

elution.

Harburg Manometry

 

Oxygen consumption was measured by a Harburg apparatus. Excised
leaves were cut into 1.0 cm disks, enclosed in cheesecloth bags, weighted
down and vacuum infiltrated in the appropriate solution for 15 min. The
treated disks were rinsed and randomized, then approximately 0.1 g (10
to 15 disks) were placed in a Harburg flask containing a paper wick and
20% KOH in the center well. After attaching the flasks to the apparatus,
they were equilibrated in the 30 C water bath, and the 02 uptake was
detected manometrically over the next few hours (44). The experiments

were carried out in darkness.

Electrolyte Leakage

Leaves used in leakage experiments were allowed to take up a 10 ul
treatment solution that was applied to the petiole end. The cut end of
the petioles were then placed in vials containing water and permitted to
transpire for 30 min. One cm disks were cut from the leaves, rinsed in
distilled water, and randomized. Approximately 0.1 g of the disks were
placed in liquid scintillation vials and suspended in 2.0 ml of distilled

water. Electrolyte leakage of the leaf disks was determined by measuring

 

14

the conductivity of the ambient solution with an Electromark conductivity

meter and a pipet-type electrode (K=1.0) (37).

RESULTS

Toxin Isolation Procedures

Early stages in toxin purification. The toxin was very soluble in
methanol, ethanol, n-butanol, acetone, and water. When partially
purified toxin or toxin in crude culture filtrates was partitioned
against butanol, more than half of the toxin was found in the butanol
fraction. Three successive partitionings extracted almost all the toxin
from aqueous solutions. Contrary to earlier reports (18,41), experiments
in this study showed that low amounts of toxic activity were extracted by
partitioning with non-polar solvents.

The toxin was a relatively stable substance. Autoclaving for 20 min
did not decrease the toxicity or affect the specificity of the toxin.
When a 50% methanol solution of toxin was left on a countertOp and
exposed to bright diffuse light for 3 months, no changes in its
host-specific activity was observed.

Toxin in aqueous solution was bound to activated charcoal; the toxin
was then desorbed from charcoal with alcohols. n-Butanol always removed
toxin bound to charcoal, but results with ethanol varied. In experiments
with 95, 50, and 10% ethanol, toxin was inconsistently desorbed; on one
occasion, all concentrations of ethanol removed much of the toxin, on
another occasion none of the solutions removed detectable toxin. The
reasons for this were not clear. Generally, the higher the concentration

of ethanol, the better the desorption of toxin from charcoal.

15

 

16

Gel filtration. Chromatography with molecular sieving gels
separated the toxin into multiple forms. Plots of toxic activity against
fraction number yielded several peaks. This was demonstrated with both
Sephadex G-15 and Bio-Gel P-Z, but the P-2 gave better resolution of the
toxins. Repeat chromatography of the collected toxic fractions gave very
similar elution patterns, indicating the multiple peaks were not due to
an inconsistent bioassay or an easily degraded molecule. The P-2 and
G-15 gels which have exclusion limits of 1,800 and 1,500 daltons,
respectively, never yielded toxin near their void volume. This suggests
the toxic substances are below 1,500 daltons in size, although
interactions leading to retardation of toxin molecules by the gel could
give a low estimate of size.

Repeated partitioning of aqueous toxin solution with chloroform or
ethyl acetate extracted varying amounts of the toxin, depending upon the
age of the culture harvested. Toxin derived from 2-week-old cultures had
approximately half of the total toxic activity extracted with 6
partitions of chloroform, whereas the same procedure gave several-fold
less toxic activity from 7-week-old cultures. The temporal change in the
amount of lipophyllic toxins recovered was related to certain peaks of
toxic activity eluted from Bio-Gel P-2 columns. The first toxins to
emerge from the column were partially soluble in chloroform, but toxins
eluting later appeared to be completely insoluble. Gel filtration of
chloroform extracts of toxin gave peaks of activity which corresponded
with the earliest toxins to emerge from the column.

Toxins recovered from .H. mammatum cultures were, therefore, related
to the age at which the cultures were harvested. This was demonstrated

clearly with Bio-Gel P-2 columns by comparing the elution patterns of

 

17

toxin harvested at 2, 4, and 7 weeks (Figure 1). Two week cultures
yielded predominately the larger-sized toxins (largest substances are
eluted first in gel filtration) that were partially soluble in
chloroform. The toxins from 7 week cultures were less lipophyllic,
smaller molecules, and toxins harvested after the intermediate 4 week
period included those found at both 2 and 7 weeks.

Thin layer chromatography. Attempts at further purifying the toxin

 

by TLC were disappointing. Generally, amounts of the toxin easily
detected in assays overloaded the plate and produced poor resolution and
streaking. This was most likely due to the high level of impurities
found with the toxin even after passage through a Bio-Gel P-2 column.

The TLC solvent systems used by Schipper (41) gave results in my
study similar to what he reported. The Rf values for toxic activity
was the same as that reported by Schipper, but fewer areas of separated
activity were detected. Most visible spots had Rf values similar to
toxin. Ninhydrin positive or UV florescing materials were not correlated
with areas of toxic activity.

Silicic acid column chromatography. A 1.5 x 90 cm glass column was
slurry packed with 135 ml of Bio-Sil A (100 to 200 mesh, Bio-Rad Labora-
tories). A toxin sample from a Bio-Gel P-2 column was reduced to near
dryness, dissolved in 1 to 2 ml of distilled water, and applied to the
silicic acid column. Using a method modified from that of Schipper (41),
the toxin was eluted from the column with n-hexane/isopropanol/ acetone
(6:4:1), and 4.5 ml fractions were collected. After elution, most of the
yellow pigmentation found in the toxin preparation appeared to remain in
the column. A plot of toxic activity against fraction number resulted in

several peaks, representing several host-specific toxins (Figure 2). The

 

 

 

 

 

 

 

 

 

 

 

 

 

16 CT I ' l ' I ' l ' TT T l
14" 2weeks ‘
12— --
10- ..
8'— —1
6'- .1

a 4" ‘

5

g 2"!0 n Ill

g. 0- 1 4 n 1 l 1 1 l l J 1 _

a; 16— -

E 14' 4weeks ‘1

E

_ 12-— -—

.2

o 10*- ..

E

.2 a— —

D

c 61- ..

.2

g 4- a

.1

g 2- :0 -l

'5 Oh I 1 I I J I l I 1 .l 1 I -

0 I— —

< 16

{5‘ 14" 7weeks ‘

O

r- 12-— -
10- -
8— fl
6" —I
4— —
2" :0 'I
0' 1 1 I1" 4 l 4 L_1—l ""

 

 

 

 

 

1 J J
25 30 35 40 4 50 55 60 65 70 75 80
Fraction No. (2.0 ml)

Figure 1. Effect of culture age on H. mammatum toxins recovered from
culture fluids. Stationary cultures were grown on modified Fries medium
at 24 C and harvested at 2, 4, and 7 weeks. The toxins were

chromatographed with a Bio-Gel P-2 column (1.5 x 90 cm).

19

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l l l l l I I

7 - ..
E 6" d
If
.2
g 5” l a
.2
D
C
.2 4- -
m
0
_I
s 3" I d
IE
‘6
< -
c 2-
"i
O
'-

1 '- cl

0 L:v-:‘:-i----v‘- ‘i‘:‘ A- l 1L...'ool «U 1‘

50 60 70 80 90 100 110

Fraction No. (4.5 ml)

Figure 2. Chromatography of H. mammatum toxins on a silicic acid column.
Stationary cultures were grown on modified Fries medium at 24 C and
harvested at 7 weeks. The toxins were chromatographed on a 1.5 x 90 cm
column packed with Bio-Sil A (100 to 200 mesh) and eluted with n-hexane/

isopropanol/acetone (6:4:1).

20

findings were similar to those reported by Schipper who observed 7
host-specific toxins produced by H, mammatum.

High pressure liquid chromatography. The usefulness of HPLC to
further purify the toxin was evaluated. One ml of toxin solution
recovered from a P-2 column was reduced to 100 pl and injected into the
HPLC column. The toxin was eluted with an ethanol gradient: 100% water
for 1 min, followed by increasing ethanol concentration to 30% at 31 min,
and finally a steeper gradient for 30 min more reaching 100% ethanol at 61
min. The machine shut down at 39.6 min because pressure to maintain the
1.0 ml per min flow rate exceeded 350 atm. The rest of the materials were
eluted after re-programming for a 0.5 ml per min flow rate. Monitoring
the eluate with UV at 280 nm indicated approximately 70 compounds eluted
from the column, many exceeding full scale (1 AU) on the chart recorder.
An assay of the fractions collected showed at least 5 host-specific
toxins.

The toxin-containing fractions were collected, reduced to 100 pl, and
rerun by HPLC. All conditions were the same as before, except this time a
1.0 ml per min flow rate was maintained for the entire elution. The
pressure did, however, rise near 350 atm at 40 min. In both cases this
transitory rise in pressure, due to increased flow resistance, occurred
just prior to emergence of toxins from the column. Apparently, the
viscosity of materials increased with rising concentrations of ethanol,
and the pressure drapped when these materials eluted from the column.

The rechromatography resulted in a reduction in the number of UV
absorbing substances detected to 6. And as shown by a marked reduction in
degree of absorption, the amount of these 6 substances was substantially

reduced as well. Moreover, the amount of toxin recovered was essentially

21

the same as that recovered from the first HPLC run. This demonstrated
that HPLC achieved a large increase in the purity of the toxin.

Absorption spectra. A visible-UV spectrum from 800 to 200 nm was
obtained for HPLC fractions recovered after the second passage through
the machine. These were 7 sequential fractions, some which contained
detectable toxin, and others which did not. A possible relationship was
seen between toxic activity of the fractions and their absorption at 240
nm (Table 1). The lack of an exact correspondence between absorption and
toxic activity could be due to variation in the bioassay. Other
wavelengths of light did not show this correlation. The fractions
absorbed no light in the visible range.

Electrophoresis. After the absorption spectra were run, the toxin-
containing fractions were combined, reduced in volume and subjected to
high voltage paper electrophoresis. An assay revealed that the toxin had
migrated towards the positive electrode. All toxic acivity was found in
a 3 cm section 5 to 8 cm from the origin. The amino acid standards were
found in 3 areas: one spot 3.75 cm from the origin towards the cathode,
3 spots 31.5, 32.5, and 34.5 cm from the origin towards the cathode, and
2 spots 36.5 and 30.0 cm from the origin towards the anode. This
experiment was repeated with a less purified toxin sample, giving very
similar results.

Ion exchange, As might be expected from the negative charge seen in
electrophoresis, the toxin did not bind to the cation exchangers (Dowex
SON-X2 and Sephadex SP C—25). The toxin was eluted at the void volume of
the columns by distilled water. Pigmented impurities emerged from the
column with the toxin.

Results from anion exchangers were not as simple. No toxic activity

22

Table 1. Light absorption and relative toxicity of fractions from HPLC

of H. mammatum toxinsa

 

Optical Density

Toxin Activity

 

Fraction No. 240 nm 280 nm (lesion dia., mm)
38 0.000 0.000 0
39 0.005 0.000 0
40 0.065 0.015 7
41 0.028 0.010 0
42 0.130 0.040 14
43 0.063 0.002 12

 

aToxin was eluted from a Varian 4.1 x 300 mm MCH-lO column with a

gradient of increasing ethanol concentration in water.

23

was recovered from the Dowex 2-X8 or the Sephadex QAE A-25. However, the
yellow pigmentation could be eluted with 0.5 M NaCl. Because there was a
possibility that the anion columns were degrading the toxin molecules, a
weaker anion exchanger (Sephadex DEAE A-ZS) was tried. Elution with
approximately 3 bed volumes (10 ml) of water left both toxin and yellow
impurities on the column. The toxin was, however, readily eluted with 10
ml of 0.05 M NaCl, leaving the pigmentation at the top of the column. L 1
The colored materials could be eluted with 0.5 M NaCl. Purity of the

toxin solution was visibly improved by the Sephadex DEAE exchanger.

 

Toxin Bioassays ,.1

Leaf puncture assay. The toxin solution recovered from a P-2 column
caused necrosis on clone 5 leaves in dilutions out to 256-1, while the
undiluted toxin caused no necrosis on clone 2 leaves. Occasionally a
highly concentrated toxin preparation did cause small necrotic lesions on
the tolerant clone 2 leaves, but whether this was due to toxin or the
concentrated impurities was not know. Repeated dilution end point assays
indicated an approximate 250-fold difference in sensitivity to toxin
between clone 2 and clone 5 leaves. The leaves from clone 3 plants were
intermediate in sensitivity to toxin.

The effect of light or benzyl adenine (BA), a cytokinin, on the
assay was studied. Different concentrations of BA were mixed with small
volumes of toxin to give BA solutions of 10'3, 10'4, 10-5, 10'5, 10'7,
and 0 g per liter. Benzyl adenine had no effect on toxin-induced lesion
development in leaf tissues. Leaves pretreated by permitting them to
take up the various amounts of BA from vials and then assayed, also gave

lesions not significantly different from controls. The influence of

24

light was investigated by taking 2 identically treated assays contained in
humid chambers and placing one under a fluorescent light fixture and the
other in a dark cabinet. There appeared to be a slight increase in lesion
size with the light treatment, but the differences were not statistically
significant.
A bioassay of 2 common substances gave surprising results. Both NaCl
(1.0 M) and methanol gave a clone-selective reaction similar to that of
the toxin (Table 2). The NaCl solution and methanol produced the largest
necrotic lesions on the toxin-sensitive clone 5 leaves, and the smallest
lesions on toxin-tolerant clone 2 leaves. The clone 3 leaves were of . 1
intermediate sensitivity, much like their reaction to toxin. However, the
clonal differences were much less striking than were obtained with toxin.
Stem assay. _H. mammatum attacks the trunk and branches of suscepti-
ble trees, not their leaves. In nature the toxin would presumably affect
these woody parts during disease development. This made it desirable to
study the effects of toxin on the stems of aspens. Also, it was important
to know the relationship between leaf and stem sensitivity of a clone to
toxin because the leaf assay has been used to screen for resistance to
Hypoxylon canker (13,18). An untested assumption by these researchers was
that the reaction of a leaf to toxin reflected the reaction of the stem.
The assay was performed by moistening a small ball of cotton with 50
pl of the test solution and placing it over a 2 mm wound that exposed the
wood. The wound was made by removing a small plug of bark with a sawed-
off hypodermic needle. The cotton ball was held in place by wrapping the
stem with a strip of parafilm. Each treatment was replicated 4 times.
The stem of a tree gave a reaction similar to that of its leaves

(Table 3). One week after the experiment began clone 5 stems treated with

 

25

Table 2. Toxicity of H. mammatum toxin, NaCl, and methanol to leaves

from three clones of E. tremuloides

 

Mean lesion diameter(mm)a

 

 

 

 

Clone 2 Clone 3 Clone 5
Test Solution 24 h 48 h 24 h 48 h 24 h 48 h
1.0 M NaCl 9 b b 2.75 b 3.25
Toxin O 1.0 0 3.5 7.0 14.0
Methanol 3.25 3.5 5.0 7.0 7.5 9.0

Water 0 1.0 O 0.75 O 1.0

 

aEach value is the mean of 4 replicates from the leaf puncture assay.
Lesions were measured 24 and 48 h after application of 10 ul of test

solution.

bThere was a 3 mm water-soaked area under the test sample, but there

was no necrosis.

 

-I
I lit

I‘D t o”

l

.moe_u a umuau_PamL we: acmEHcogu scum
.epwemcma cur: mamum as» mcpqamcz >5 moo—g cw u_m; mm: couuou as» .uooz mg» ummoaxm one» acne: as N a

La>o a. mcwuopa new copu:_om ummp we p: cm zuwz couuoo eo _—mn FPuEm a mcpcmumpoe x5 umummeu mew: msmumw

 

m=_—mo o: .xcmn on»

 

 

mappmu o: .AEE om ow muv Love: ucaoz we» ucaogm mcopmmp o:

mcopmm— cascam _m>o eon—m Ass m op HV mmcpc gouFm .A.e_u.==_mv m=__uu ucaoz :wxop
m:_Feu
6

9. an wovcaoeesm Amco— ==_aH m:_Pou x: cmucsoecam m=__mu an uwuczoecsm

on oHV meowmmp cmxcam Amcop 55 m on 0v mcovmm— Am:o_ as m cu av
uwaocmnmpucpam co _m>o cmxcam cmnmsmampv=Fam meowmmp cmxcsm _e>o _ocmsumz
A.mwu as mv mappwu ucaoz A.mvu as my map—mu ease: A.m_c as my m:_—mu ease: capo:

m ago—u m mac—o N meopu cowu:_om amok
mmmconmm

 

 

 

amouwo_=emeu .m.eo mace—u omega eo msmum co Lopez new .Focmzume .cwxou ssumesue .m.mo uomemu .m m_am»

27

toxin had 25 to 30 mm black sunken lesions centered around the wound.

The clone 3 stems treated with toxin had a 1 to 3 mm ring of necrotic
sapwood around the wound. Neither clone 5 nor clone 3 developed any
wound callus with toxin. The toxin-treated clone 2 stems reacted the
same as water controls; callus was formed around the wound and there was
no necrosis. Treatment with methanol produced sunken lesions of the same
magnitude seen in leaf bioassays. All methanol-treated wounds also
produced wound callus. At 2 weeks the water and methanol treatments, and
the toxin-treated clone 2 stems were developing more callus and healing.
At 2 weeks the toxin-treated stems of clones 5 and 3 appeared to be no
different than at 1 week.

The severity of Hypoxylon canker has been directly correlated to
moisture stress. A modest experiment was carried out to determine
whether or not moisture stress caused increased sensitivity to toxin.
Stems of four clone 2 and two clone 5 plants were treated with toxin, as
described above, but this time the plants were watered only after obvious
wilting occurred. Control plants received ample water. After 1 week the
water-stressed clone 2 and clone 5 plants had lesions identical to their
respective controls. Evidently, drought conditions had no effect on
toxin sensitivity.

Dormant potted aspen plants held outside in a cold frame, were stem
assayed. Stems were treated with toxin as described above. This was
during early December when daily temperatures were around 7 C. Six young
trees, 2 each of clone 2, clone 3, and clone 5 were tested with both
water and toxin. No lesions deveIOped on any plant and callus was
absent. Toxin and water treatments were similar on all clones. These

aspen plant had given a typical host-specific reaction to toxin the

28

previous summer. The dormant plants appeared to be insensitive to the
toxin.

Leaf1petiole assay. A more sensitive bioassay resulted when toxin
test solution was applied to the cut petiole of an excised leaf. The 10
pl sample was placed on the petiole end and was taken up by the leaf.

The treated leaves were then placed in vials contained within a humid
chamber as before. Twenty-four hours later the toxin-treated clone 5
leaves exhibited degrees of necrosis from scattered flecks to complete
necrosis of the leaf. Toxin-treated clone 2 leaves and the water
controls showed no necrosis.

To determine the speed with which toxin would be distributed in a
petiole-treated leaf, cuts were made perpendicular to the midvein 3, 1,.
1/2, and 0 h after toxin application. All sections from leaves cut 3, 1,
and 1/2 h after treatment were totally necrotic, indicating the toxin had
spread throughout the leaf by these times. However, the leaf that was
cut inmediately after exposure to toxin also developed considerable
necrosis. The actual time to complete the manipulation was 3 to 4 min,
and during this time enough toxin had moved up the leaf prior to making
the cut, to cause complete necrosis in the basal half and about 40%
necrosis of the apical half by 20 h. Apparently, transpiration moves
toxin very rapidly through aspen leaves.

Dry weights were determined for toxin present in selected fractions
from a Bio-Gel P-2 column fractions, and the toxin was used in a dilution
end point assay. Diluting the toxin in 10-fold increments, the highest
dilution that showed some toxicity to clone 5 leaves was 10'2. No
necrosis was seen in clone 2 replicates. A leaf puncture assay performed

at the same time could only detect the undiluted toxin. The dry weight

29

of the toxin preparation was 0.1 mg per ml, meaning the petiole assay
detected toxin in a 1.0 pg per ml solution. In other words, a 10 pl

impure sample of the toxin containing 10 ng of materials caused some

necrosis in sensitive leaves.

When toxin-treated leaves of clone 5 were submerged in water, they
did not develop the characteristic black necrOSis. Instead, both the
toxin-treated and control leaves had brown mottle. Submersion was
accomplished by placing tubing on the leaf in a water-filled Petri dish.
Inhibition of phenolic oxidation was assumed to be responsible (34), but
the possible involvement of carbon dioxide, ethylene, or osmotic effects
was not investigated.

In addition to being a more sensitive method, the leaf petiole
bioassay appeared to be less affected by toxic impurities in the
preparation. Ten pi samples of 0.5 M NaCl, 0.5 N HCl, and 0.5 N NaOH all
failed to produce necrosis in either clone 5 or clone 2 leaves, but did
cause considerable necrosis in the leaf puncture assay. The petiole assay
detected host-specific toxin activity (necrosis only on toxin-treated
clone 5, but not on treated clone 2 leaves) when 10 pl drops of solutions
containing toxin plus 0.5 M NaCl, 0.5 N HCl, or 0.5 N NaOH. The assay
results were the same as when only the toxin was present.

With the petiole assay, applied samples were distributed throughout
the leaf, resulting in dilution of assay materials. This distribution
was probably responsible for the lack of injury caused by contaminates
in the petiole assay. Also, all the toxin can enter the leaf, unlike
the puncture assay where most of the sample remains on the leaf surface.
Amounts of toxin 1000 times that required to cause slight necrosis in

clone 5 leaves did produce a similar degree of necrosis in clone 2 leaves.

30

Effects of Toxin on Tissues

Toxin-induced changes in respiration. Host-specific toxins produce
many changes in susceptible tissues that are caused by fungal infections;
examples are increases in respiration and in leakage of electrolytes
(36). Treatments consisted of vacuum infiltration of leaf disks with
distilled water or with toxin solution. Each treatment had two repli-
cates for each clone. Measurements of rates of 02 uptake with a Warburg
apparatus showed that toxin-treated clone 5 tissues had 37% higher
respiration rate than did control tissues (Figure 3). The toxin-treated
clone 2 tissues and water control tissues gave similar results, yielding
an 02 consumption rate near 3 pl per min. The toxin-treated clone 5 leaf
disks took up 02 at aproximately 4 pl per min. The toxin-induced
increase in respiration was evident at the first reading (60 min after
infiltration). A repeat experiment gave similar results.

Toxin-induced changes in electrolyte leakag_. Other toxins are
known to cause leakage of electrolytes from susceptible tissues, and this
has been used as a bioassay (37). Early attempts at detecting changes in
electrolyte leakage caused 9Y.fl- mammatum toxin were unsuccessful. When
clone 2 and clone 5 leaf disks were infiltrated with distilled water or
toxin solutions, as in the respiration experiments, no differences were
seen between the treatments. Vials containing the leaf disks from the
different treatments all gave conductivity measurements of 22 to 30 pmhos
at 25 h. Leaf disks were then vacuum infiltrated and held in the toxin
solutions for 1 h before they were rinsed and monitored for leakage;
there was a slight increase by 24 h in electrolyte leakage of clone 5
disks treated with toxin solution. The disks also showed some necrosis.

Leaves from clone 2 and clone 5 plants were allowed to take up 10 pl

31

 

 

 

 

l l l
0 C5 + H20
0 05 + TOXIN
A 02 + H20 .
50° A C2 + roxm
'8 400- "‘
E
3
0)
5
_ -l
L; 300 a
3, O/‘
1 200 — w a
o
100 — -
1 I l
90 120 150

Minutes After Treatment

Figure 3. Effect of H. mammatum toxin on 02 consumption by leaves of H.
tremuloides clones 2 and 5 (C2 and CS). Leaf disks (1 cm in diameter)
were infiltrated at reduced pressure with toxin solution or water. Leaf
disks (0.1 g) were placed in Warburg flasks and the 02 uptake measured

manometrically. Each value is the mean of two replicates.

32

samples of either water or toxin solution through their cut petioles.
Next the disks were cut from the leaves and were placed in vials with
water. All treatments were replicated. The ambient solutions of
toxin-treated clone 2 and water-treated clone 2 and clone 5 preparations
increased in conductivity by approximately 0.9 pmhos per h. The ambient
solutions of toxin-treated clone 5 leaves increased in conductivity by
greater than 6.0 pmhos per h (Figure 4). Electrolyte leakage of the
toxin-treated clone 5 leaf disks, however, increased at the same rate as
the controls until 6 h after treatment, when there was an acceleration in
leakage from toxin-treated tissue. By 24 h the vials containing the
toxin-treated clone 5 leaf disks had an average conductivity of 127 pmhos
per h, whereas the controls averaged around 30 pmhos per h. Some
necrosis of toxin-treated clone 5 leaf disks was apparent by 5 h. At 24
h these leaf disks were necrotic in 40 to 100% of their area. There was
no necrosis in the control and in the toxin-treated tolerant leaves.

Isolated H, tremuloides prot0plasts. Isolated protoplasts are used

 

frequently in studies of physiological phenomena in plants. A procedure
for isolating aspen protoplasts was develOped in my study so that the
protOplasts could be used to study toxic effects. Protoplasts were
isolated from aspen leaves by enzymatic digestion. The leaves used were
approximately 1/4 the size of fully expanded leaves and had a shiny
surface not seen in the more mature tissues. Leaves cut into 2 to 3 mm
pieces were floated on enzyme solution in small Petri plates. The
digestion solution contained 4% Cellulysin, 1% Macerase, and 1% (by
volume) glusulase in 0.6 M sorbitol with 50 mM phosphate buffer, pH 7.0.
The leaf pieces were vacuum infiltrated in the solution for 15 min, then

incubated with the enzymes at 27 C for 24 h. After incubation, the

33

 

 

 

 

 

 

I I ‘1 I I l I I *1 I I
140 0 C5 + TOXIN
- 15(32 +-|120 -
A02 + TOXIN
120 F -
’8‘ ' _.
1: 100 -' ‘-
E
3 " ‘
3' 80 - ..
E a
‘6
.g 60 p -
c:
O - .-
(J
40 _. ._
20 h- ._
I I I 1 I I I I, l I 1
2 4 6 8 10 12 24

Hours After Treatment

Figure 4. Effect of H. mammatum toxin on electrolyte leakage from leaf
disks of H, tremuloides clones 2 and 5 (C2 and C5). Leaves were allowed
to take up 10 pl of toxin solution through the cut petioles. Leaf disks
(1 cm diameter) were then cut and 0.1 g of tissues were placed in vials
containing 2.0 ml of water. Conductivity of ambient solutions was

measured at intervals.

34

contents of the Petri plates were filtered though a layer of miracloth
which retained plant debris but permitted the protoplasts to pass
through. The filtrate was centrifuged at 50 xg for 4 min to pellet the
protoplasts.

The protoplast pellet was resuspended in buffered 0.6 M sorbital
containing toxin. At 24 h no differences were seen between toxin or
water treatments in clone 2 or clone 5 protoplasts. Incubation of the
toxin with the enzyme preparation did not alter the activity of the
toxin. The protoplasts were apparently unaffected by toxin. There was a
gradual death of cells over a 2 day period, as shown by staining with
Evans blue. However, the dramatic necrosis and cell death so obvious in

clone 5 leaf tissues was absent in isolated protoplasts.

Effects of Toxin on Aspen Hybrids

Leaf assay of aspen hybrids. The most convincing data showing that
toxins have significance in disease development are based on genetic
analyses of the host and pathogen (48). The following experiment with
aspen hybrids was a preliminary attempt to determine the genetic control
of sensitivity in aspen to the toxin. Hybrids of Populus grandidentata
and E, tremuloides x.H. grandidentata were tested for their reaction to
H, mammatum toxin. I used a total of 36 plants from 6 different crosses
made by Greg Reighard of the Forestry Department at Michigan State
University. The plants were 4 months old and ranged from 30 to 80 cm
tall when assayed. Leaves were tested by the leaf puncture and petiole
assay methods with 10 pl samples of toxin. Some of the hybrids from
crosses 559 x 699 and 157 x WENS were sensitive to the toxin; all the
rest were insensitive (Table 4). One plant each from crosses 565 x 1-35

and 107 x 2253 did give a slight reaction in the puncture assay at 48 h.

35

Table 4. Sensitivity of Populus hybrids to toxin from H, mammatum.

 

No. of progeny

Total no. of

 

 

 

 

 

 

 

sensitive to progeny
Cross toxina assayed
B, grandidentata x H, tremuloides
(559 x 699)b 8 11
(107 x 225B) 0c 6
B, tremuloides x H, grandidentata
(157 x WENS) 2 3
(565 x 1-35) 0C 9
g, grandidentata x g, grandidentata
(107 x 307) 0 5
(559 x 307) 0 2

 

aLeaves were tested by the leaf puncture and petiole methods with 10 pl

samples of toxin. The plants were 4 months old when assayed.

bThese are reference numbers used to identify the clones.

cOne plant did give a slight reaction in only the puncture assay at

48 h.

DISCUSSION

Isolation of H. mammatum toxins

 

Previous reports (5,18,41) were confirmed, showing that the H.
mammatum toxin was soluble in polar solvents, heat stable, and composed
of several similar host-specific toxins. Schipper (41) chromatographed
culture filtrates from H. mammatum on silicic acid columns and recovered

7 different fractions that were toxic to leaves of H. tremuloides. I

 

have obtained simiar results with a silicic acid column, indicating the
presence of several different fractions, all with the same host-specific
toxicity. Unlike work reported previously (18,41), the several different
toxic molecules were also separated by gel filtration.

Some of the toxins were partially soluble in non-polar solvents,
contrary to earlier reports (18,41). Repeated partitioning of aqueous
toxin with ethyl acetate or chloroform extracted most of the toxins that
were eluted first from Bio-Gel P-2 columns. The smaller forms of the
toxin which emerged later from the column, appeared to be completely
insoluble in non-polar solvents. Also, yields of the different toxins in
liquid cultures varied with culture age. The larger toxins, partially-
soluble in non-polar solvents, predominated in cultures at 2 weeks, but
only small amounts were detected at 7 weeks. Differences in fungal
isolates and culture media used by different workers could influence the
nature of toxins produced by H, mammatum.

Several problems with thin layer chromatography of partially

36

37

purified toxin were encountered. Impurities in the toxin preparation
made it easy to overload the plates and give poor resolution or
streaking. When smaller amounts of the toxin-containing preparations
were used, the toxic spots on the plates were difficult to detect in the
bioassay. My results with the TLC solvent systems used by Schipper (41)
gave Rf values for spots with toxic activity which were similar to his
reported values, but I observed only 3 distinct areas with selective
toxicity. Good separations could possibly spread the various toxins,
resulting in a reduction of activity per area beyond the dilution end
point of the assay. Contrary to the report of French (18), there was no
consistent correlation between UV-fluorescence and selective toxicity.
The toxin also appeared to be ninhydrin negative.

All forms of the toxin appeared to be acidic molecules. At physio-
logical pH levels the toxins had a negative charge that caused movement
in an electrical field and binding to anion exchangers. Neutral or basic
forms of the toxin were not seen. Toxin migration in electrophoresis was
less than that of negatively-charged aspartic and glutamic amino acids.
This may have resulted from the friction of chromatographing a large
molecule, or from the presence of only a partial net negative charge on
the toxins. Strong anion exchangers apparently degraded the toxin
molecules; even elution with 0.5 M NaCl did not recover toxic activity.
The toxins bound to the weak exchanger Sephadex DEAE, and toxic activity
was eluted with 0.05 M NaCl, but not with water.

Correlative evidence indicated the toxins absorb UV light at 240 nm.
The absorption spectra of highly purified fractions from HPLC showed
increases in absorption at 240 nm directly related to the toxic activity

of the fractions. This absorption was observed in all the different

38

forms of the toxin separated by HPLC.

Adsoprtion and subsequent desoprtion of toxin with activated
charcoal was useful in the early stages of purification, but the largest
increase in purity was obtained with HPLC. A greater than 10-fold
reduction in contaminates was achieved without a reduction in toxic
activity. Chromatography of the toxin with Sephadex DEAE allowed elution
of toxin from the column without removing the visible materials from the
column. When the toxin recovered from gel filtration was subjected to
ion exchange chromatography on DEAE and then chromatographed by HPLC
through a C13 reverse phase column, the purity of the toxin preparation

was substantially improved.

Toxin Bioassays

The leaf puncture assay method showed that leaves from H, tremuloides
clone 2 tolerated at least 250-fold larger amounts of toxin than did
leaves of clone 5. Treatment with light, darkness, or various
concentrations of benzyl adenine did not appear to affect the results of
the puncture assay; neither the severity of leaf necrosis nor the clonal
specificity was changed. Impurities, however, did affect the bioassay
through their toxic effects on leaves. Moreover, clones 2 and 5
differed in sensitivity to 1.0 M NaCl and methanol. The specificity of
host-dependent lesion formation in leaves caused by the NaCl solution or
methanol was similar to that for the toxin; however, the differences
between clone 2 and 5 were not as great with NaCl and methanol as with
toxin. The significance of this host-selective reaction to NaCl and
methanol in the leaf puncture assay is not clear. A similar phenomenon

was observed with the stem assay, but not with the petiole assay.

39

Stem assay results for clones 2, 3, and 5 showed that the toxin had
the same selective effects on stems as on leaves. Establishment of this
relationship is important because it is usually much easier to use leaves
than stems to study toxic action and screen for resistance. Toxin
inhibited the formation of wound callus in stems, as reported by other
investigators (24,41). The lack of callus in toxin-treated wounded stems
probably is not caused by simple injury of the bark. In contrast to the
results of toxin-treatment, stem necrosis caused by methanol eventually
callused and healed. Evidently, the toxin has a prolonged effect on
wound callus formation in toxin-sensitive stems.

No lesions developed when toxin was applied to stems of dormant
plants (clones 2 and 5). This correlates with resistance of dormant
stems to infection by H. mammatum (7). The reason for this apparent
insensitivity of dormant plants to toxin and to infection is not obvious.
Temperatures during the toxin assay were above freezing. Stems of
growing clone 5 plants in the greenhouse at the same time developed
normal necrotic lesions after toxin was applied. Perhaps a physiological
process necessary for susceptibility to the toxin is inactive in dormant
tissues.

French (18) observed that plants of clone 2 and 5 were equally
susceptible to H, mammatum when inoculated artificially. However, these
clones differ greatly in their sensitivity to the toxin. Therefore, if
the toxin is necessary for H. mammatum infection of aspen as hypothesized
by Schipper (40), an effect of toxin other than necrosis must be impor-
tant to the infection process. Results of toxin treatment of stems of
drought—stressed plants is consistent with this. Although nany workers

have reported that susceptibility of aspen to H. mammatum infection

40

increases directly with moisture stress, the plants that were under
considerable moisture stress did not increase in sensitivity to toxin.
The necrotic effect of the toxin alone could not be reSponsible for the
increased susceptibility to infection. Bagga and Smalley (5) reported
that inoculations with ascospores of H. mammatum were not successful
unless toxin was added with the inoculum. An interesting experiment
would be to inoculate the stems of toxin-tolerant clone 2 aspens with
ascospores and toxin, and see if a successful infection could develop
without the initial necrotic lesion. Possibly, the role of the H.
mammatum metabolite in the infection process is to induce compatibility
in susceptible hosts, similar to that claimed by Oku et al. (31) for the
non-toxic determinant of pathogenicity produced by Mycosphaerella
pinodes.

The leaf petiole assay was more sensitive to toxin, yet was less
affected by impurities than was the leaf puncture assay. Uptake of the
entire sample into the leaf and dilution of the solutes throughout the
lamina were probably why the petiole assay was more sensitive and
selective. Application of a test solution to the petiole was, however,
more time consuming and required more leaves than did the puncture assay,
in which 4 samples can be tested on one leaf. The leaf puncture assay is
preferred with partially purified toxin and for experiments requiring
assays of many samples. The leaf petiole method should be used when
assaying solutions with high levels of impurities or with very low levels
of toxic activity. The petiole application method is more precise than
Schipper's leaf assay method because the petiole method treats each leaf

with a measured amount of toxin solution.

41

Effects of Toxin on Aspen

 

Physiological changes induced 9Y.fl- mammatum toxin were similar to
the changes that occur with many plant infections (36). The toxin caused
increases in respiration, electrolyte leakage, and necrosis in sensitive
tissues of H, tremuloides. The increases in respiration were detected
relatively early; within 60 min the toxin-treated leaf disks of the
susceptible clone 5 had a respiration rate higher than controls.
Electrolyte leakage, however, was not detected until 5 h later. Necrosis
was first observed in leaf disks just prior to detection of increased
leakage, and the development of the necrosis was roughly proportional to
increases in electrolytes found in ambient solutions. Apparently, the
change in plasmalemma permeability of cells in toxin-treated clone 5 leaf
disks is a secondary effect of the toxin, resulting from cell death.
Earlier changes in permeability cannot be ruled out, because leaf
characteristics which prevented good vacuum infiltration of leaf disks
could possibly retard electrolyte leakage as well. The quick increase in
02 consumption of sensitive leaf disks indicates this effect was related
more closely to the primary site of action of the toxin.

Protoplasts were successfully isolated from young aspen leaves.

The isolated protoplasts were to be used to study the action of the
toxin, but protoplasts isolated from tissues of clone 5 were not visibly
affected by the toxin. The cause of this apparent loss of sensitivity
to toxin by protoplasts when isolated from tissues is not known. The
plasmalemma surface left by enzymatic digestion was probably changed.
The distribution and configuration of functional molecules left on the
membrane surface are as likely to be an expression of the digestion

procedure as they are to reflect any inherent properties of the

42

plasmalemma (22). Perhaps a toxin receptor was modified or removed from
the membrane surface during the enzymatic digestion which could render
the protoplasts insensitive to toxin.

There was much variation in toxin sensitivity of aspen hybrids, and
also between members of a toxin-sensitive hybrid. Unfortunately, the

toxin sensitivity of the parents was not known. The E, grandidentata

 

plants probably were all highly tolerant (18,38), but the sensitivity of
H, tremuloides parents could have varied (18). Hybrids between H.
grandidentata clones were all highly tolerant to the toxin. Progeny from
g, tremuloides x E, grandidentata crosses contained toxin sensitive
plants in 2 crosses and no sensitive plants in the other 2 crosses.
Because toxin-sensitive hybrids came from both male and female H.
tremuloides parents, sensitivity to toxin does not appear to be
cytoplasmically inherited. Genetic control of aspen reaction to the
toxin is likely to be complex, similar to genetic control of resistance
to Hypoxylon canker (27,45,46).

Evidence presented in my study and elsewhere support the hypothesis
that the host specific metabolites from H. mammatum are required by the
fungus for canker development in Populus species. Several lines of
evidence indicate that the toxins are a significant factor in disease
development. .2, tremuloides, the only species highly sensitive to the
toxins, is also the species most susceptible to H, mammatum infection;
all other species tested are highly tolerant of the toxins and also very
resistant or immune to infection (7,18,38). Inoculations of wounded
quaking aspen stems with H. mammatum ascospores result in canker
development only when combined with topical applications ole. mammatum

toxins (5). Similarly, when exogenous toxin is applied to mycelial

43

inoculations of an avirulent isolate of H. mammatum, successful infection
occurs; no infection occurs with the avirulent or the normally virulent
isolates when the toxins are removed (40). Isolates of H. mammatum that
produce the most toxin in culture are also the most virulent when
artificially inoculated on quaking aspen stems (18). Results from my
study show that dorman aspen stems are apparently insensitive to the
toxins; because dormant stems are also resistant to infection by H.
mammatum (7), this data is consistent with the hypothesis that the toxins
are necessary for canker develOpment. Also, the extreme biological
activity of the H. mammatum host-specific toxins, with only nanogram
amounts necessary to produce effects in sensitive tissues, argues against
just a fortuitous correlation of the toxins with Hypoxylon canker.

There are, however, observations that raise some doubts about the
role of the toxins. The tolerance of differentH, tremuloides clones to
the necrotic effect of the toxins is not correlated with their resistance
to infection by H, mammatum (18). Also, there is the question of how the
toxins act in natural infections;.H. mammatum ascospores do not contain
toxin (41), and no vector of mycelium has been identified.

I suggest that the primary effect of the toxins is to induce
compatibility in susceptible hosts; the necrotic effect of the toxins is
secondary. The action of the toxins is to permit canker development to
occur (5,40). Perhaps this is by reducing the ability of a tree to
recognize the pathogen, or by elicitation of a host response necessary
for canker development. Examination of the toxic effects of the fungal
substances could be misleading when used as a method to screen for
resistance to Hypoxylon canker.

Genetic studies on the control of pathogenicity, virulence, and

44

toxin production in H, mammatum, and genetic studies of canker resistance
and toxin sensitivity in POpulus species, would greatly improve the under-
standing of the role of the toxins in disease development. Only a genetic
analysis of canker resistance has been reported (46). Obviously more
information is needed on genetic control of these factors. Purification
of the H. mammatum toxins to homogeneity and their characterization also
is needed for thorough investigations of their action and role in disease.
The relationship between Hypoxylon canker and insects of aspen also

deserves more study.

BIBLIOGRAPHY

1.

2.

3.

6.

9.

10.

11.

12.

BIBLIOGRAPHY

Anderson, 0. L., and D. W. French. 1972. Germination of ascospores
of Hypoxylon mammatum in living aSpen. Can. J. Bot. 50:1973-1974.

Anderson, N. A., M. E. Ostry, and G. W. Anderson. 1979. Insect

wounds as infection sites for Hypoxylon mammatum on trembling aSpen.
Phytopathology 69:476-479.

Anderson, R. L., P. N. Joranson, and D. W. Einspahr. 1960.
Hypoxylon canker on European aspen. Plant Disease Reptr. 44:132.

Bagga, D. K. 1968. Physiology of Hypoxylon pruinatum and its
pathogenesis on quaking aspen. Ph.D. Thesis, University of
Wisconsin, Madison. 165 p.

Bagga, D. K., and E. B. Smalley. 1974. The development of Hypoxylon
canker of Populus tremuloides: Role of ascospores, conidia, and
toxins. Phytopathology 64:654-658. .

Bagga, D. K., and E. B. Smalley. 1974. The development of Hypoxylon
canker of Populus tremuloides: Role of interacting environmental
factors. Phytopathology 643658-662.

Berbee, J. G., and J. 0. Rogers. 1964. Life cycle and host range of

Hypoxylon pruinatum and its pathogenesis on poplars. Phytopathology
54:257-261.

Bier, J. E. 1939. Hypoxylon canker of maple. Forest Chron.
15:122-125.

Bier, J. E. 1940. Studies in forest pathology. III. Hypoxylon
canker of poplar. Can. Dep. Agr. Tech. Bull. No. 27. 40 p.

Bier, J. E. 1961. The relation of bark moisture to the development
of canker disease caused by native, facultative parasites.

VI. Pathogenicity studies of H ox lon pruinatum and Se toria musiva
on species of Acer, Populus an Sa ix. Can. J. Bot. 39:1555-1 61.

Bier, J. E., and M. H. Rowat. 1963. Further effects of bark
saprophytes on Hypoxylon canker. Forest Sci. 9:263-269.

Bruck, R. I., and P. D. Manion. 1980. Interacting environmental

factors associated with the incidence of Hypoxylon canker on
trembling aspen. Can. J. For. Res. 10:17-24.

45

13.

14.
15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

46

Bruck, R. I., and P. D. Manion. 1980. Mammatoxin assay for genetic
and environmental predisposition of aspen to cankering by HypoHylon

mammatum. Plant Disase 64:306-308.

Cooke, M. C. 1883. Hypoxylon and its allies. Grev. 11:121-140.

Flores, G., and M. Hubbes. 1979. Phytoalexin production by aspen
(Populus tremuloides Michx.) in response to infection by Hypoxylon

 

mammatum (Wahl.)Mill and Alternaria spp. Eur. J. For. Path.
97280—288

Flores, G., and M. Hubbes. 1980. The nature and role of phyto-
alexins produced by aspen (Papulus tremuloides Michx.). Eur. J.
For. Path. 10:95-103.

French, 0. W., and N. Oshima. 1959. Host bark characteristics
and infection by Hypoxylon pruinatum (Klot.) Cke. Forest Sci.
5:255-258.

French, J. R. 1976. Screening aspen for resistance to Hypoxylon
canker. Ph.D. Thesis, Michigan State University. 66 p.

Graham, 5. A., and R. P. Harrison. 1954. Insect attacks and
Hypoxylon infections in aSpens. J. Forest. 52:741-743.

Gruenhagen, R. H. 1945. Hypoxylon pruinatum and its pathogenesis
on poplar. PhytOpathology 35:72-89.

Hepting, G. H. 1971. Diseases of forest and shade trees of the
United States. Agriculture Handbook No. 386.

Heslop-Harrison, J. 1978. Cellular recognition systems in plants.
University Park Press, Baltimore. 60 p.

Hubbes, M. 1962. Inhibition of Hypoxylon pruinatum by pyrocatechol
isolated from bark of aspen. Science 136:156.

Hubbes, M. 1964. The invasion site of Hypogylon pruinatum in
Populus tremuloides. Phytopathology 54:896.

Hubbes, M. 1964. New facts on host-parasite relationships in the
Hypoxylon canker of aspen. Can. J. Bot. 42:1489-1494.

Klotsch, J. F. 1833. Sphaeria pruinata. 1H_Fungi exotici e
collectionibus britannorum. Linn 8:489.

Manion, P. 0., and F. A. Valentine. 1974. Quantitative inheritance
of tolerance to Hypoxylon mammatum in aspens. Amer.
Phytopathological Soc. Proc. 1:60-61.

Marty, R. 1972. The economic impact of Hypoxylon canker on the
Lake States resource. I! Aspen: symposium proceedings. USDA Forest
Service Gen. Tech. Report NC-l. pp. 21-26.

29.

30.

31.

32.
33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

47

Miller, J. H. 1961. A monograph of the world species of Hypoxylon.
University of Georgia Press, Athens. 158 p.

Offord, R. E. 1966. Electrophoretic mobilities of peptides on
paper and their use in determination of amide groups. Nature
211:591-593.

Oku, H., T. Shiraishi, S. Ouchi, and M. Ishiura. 1980. A new

determinate of pathogenicity in plant disease. Naturwissenschaften

Povah, A. 1924. Hypoxylon poplar canker. Phytopathology 14:140.

Pringle, R. B., and R. P. Scheffer. 1963. Purification of the
selective toxin of Periconia circinata. Phytopathology 53:785-787.

Rhodes, J. M., and L. S. C. Wooltorton. 1978. The biosynthesis of
phenolic compounds in wounded plant storage tissues. .12 Biochemis-
try of wounded plant tissues, Gunter Kahl, ed. Walter de Gruyter
and Co., New York. pp. 243-286.

Rogers, J. D. 1963. Hypoxylon canker of aspen. 1. Screening for
disease resistance. II. Developmental morphology and cytology of
Hypoxylon pruinatum. Ph.D. Thesis, University of Wisconsin,
Madison. 140 p.

Scheffer, R. P. 1976. Host-specific toxins in relation to patho-
genesis and disease resistance. IH_Physiological plant pathology,
Heitefuss and Williams, eds. Springer-Verlag, New York. pp.
247-269.

Scheffer, R. P., and R. S. Livingston. 1980. Sensitivity of sugar-
cane clones to toxin from Helminthosporium sacchari as determined by
electrolyte leakage. Phytopathology70:400-404.

Schipper, A. L., Jr. 1973. Mammatoxin, a Hypoxylon mammatum
metabolite responsible for canker development in aspen. Second Int.
Cong. of Plant Pathology, Abstracts of Papers (Abstract No. 0677).
Amer. Phyt0pathological Soc., Inc.

Schipper, A. L., Jr. 1975. Changes in dehydrogenase and peroxidase
activities of aspen infected with Hypoxylon mammatum.
Phytopathology 65:440-445.

 

Schipper, A. L., Jr. 1975. Hypoxylon pathotoxin necessary to the
infection of aspen by Hypoxylon mammatum. Amer. Phytopathological
Soc. Proc. 2:46-47.

Schipper, A. L., Jr. 1978. A Hypoxylon mammatum pathotoxin
responsible for canker formation in quaking aSpen. Phytopathology
68:866-872.

Shea, K. R. 1963. Induction of Hypoxylon cankers in aspen by
artificial inoculations. Forest Sci. 9:2-7.

43.

44.

45.

46.

47.

48.

48

Shea, K. R., and J. E. Kuntz. 1956. Effective inoculation of aSpen
Hypoxylon pruinatum. Phytopathology 46:26.

Umbreit, W. W., R. H. Burris, and J. F. Stauffer. 1957. Manometric
techniques. Burgess Publishing Co., Minneapolis. 338 p.

Valentine, F. A., and P. D. Manion. 1972. Genetic control of
susceptibility to Hypoxylon mammatum in native aspens. .lfl 19th
Northeast For. Tree Improv. Conf. Proc. pp. 97-108.

Valentine, F. A., P. D. Manion, and K. E. Moore. 1976. Genetic
control of resistance to Hypoxylon infection and canker
developmentin Populus tremuloides. In Proc. 12th Lake States
For. Tree Improv. Conf., U. S. 0ept.'Tbr. Serv. Gen. Tech. Rep.
NC-26. pp. 132-146.

Wood, F. A., and D. W. French. 1969. Ejection of ascospores by

Hypoxylon pruinatum during the winter in Minnesota. Phytopathology
52:33.

Yoder, 0. C. 1980. Toxins in pathogenesis. Annu. Rev.
Phytopathology 18:103-129.

 

MICHIGAN STRTE UNIV. LIBRRRIES
llllllllllllllllillHillHill"lllllllllllllllllllllllll
31293105444206