Phenylalanine ammonia-lyase inhibition, autofluorescence, and localized accumulation of silicon, calcium and manganese in oat epidermis attacked by the powdery mildew fungus Blumeria

Physiological and Molecular Plant Pathology (1998) 52, 223–243Article No. : pp980148

Phenylalanine ammonia-lyase inhibition,

autofluorescence, and localized accumulation of silicon,

calcium and manganese in oat epidermis attacked by

the powdery mildew fungus Blumeria graminis (DC)

Speer.

T. L. W. C", B. J. T", M. P. R" and R. J. Z*#

" Institute of Grasslands and Environmental Research, Plas Gogerddan, Aberystwyth, Dyfed, SY23 3EB, UK#Department of Plant Pathology, 495 Borlaug Hall, University of Minnesota, St. Paul, MN 55108, USA

(Accepted for publication February 1998)

Phenolic-based autofluorogens accumulate locally in haloes and papillae formed in cerealepidermal cells responding to attack by germ tubes of Blumeria graminis (the powdery mildewfungus). The chemical elements silicon, calcium and manganese have been reportedto accumulate in spatial coincidence with autofluorogens. We tested the hypothesis thatchemical element accumulations are dependent upon the presence of autofluorogens. Light andfluorescence microscopy showed that treating seedling leaves of Selma oat (A�ena sati�a L.) withα-amino-β-phenylpropionic acid (AOPP) to inhibit phenylalanine ammonia lyase (PAL) greatlysuppressed the frequency and intensity of localized autofluorescence. Energy dispersive X-raymicroanalysis was used to identify and quantify levels of Si, Ca and Mn at 12, 16, 18, 24 and 30 hafter inoculation. Leaf specimens were cryopreserved in the frozen-hydrated state allowingelemental analysis of halo and upper papilla regions, and then freeze-dried for analysis fromdeeper within cells. Although autofluorescence was detected at 12 h, no accumulation of anyelement was detected until 18 h. Results strongly indicated that Si accumulation is a processindependent from autofluorogen accumulation. At attack sites, Mn levels were generally elevatedin association with AOPP treatment, while Ca levels deep within epidermal cells decreased.Further comparisons were made between successful and failed B. graminis penetration sites anddetailed results are presented and discussed. # 1998 Academic Press

INTRODUCTION

Conidia of Blumeria graminis (DC.) Speer (syn. Erysiphe graminis DC.) germinate on

host cereal leaves, first producing a primary germ tube (PGT) and then a second germ

tube destined to elongate and differentiate an appressorium. Haustoria are never

produced by the PGT, but it forms a short penetration peg that enters the host

epidermal cell wall and stimulates host responses. Successful infection depends upon

*To whom correspondence should be addressed.Abbreviations used in the text : AOPP, α-amino-β-phenylpropionic acid; EDX, energy dispersive X-ray

microanalysis ; PAL, phenylalanine ammonia lyase ; PGT, primary germ tube; SEM, scanning electronmicroscope.

0885-5765}98}04022321 $30.00}0 # 1998 Academic Press

224 T. L. W. Carver et al.

the appressorium from which a penetration peg passes through the host epidermal cell

wall and forms a haustorium. Defensive barriers are produced in response to attempted

penetration.

Recent reviews describe the details of germling development [6] and host response

[1, 9]. The first visible response to appressorium contact is the appearance of

autofluorescent material in the underlying host plasmalemma or cell wall [23, 30].

Within minutes an actively moving aggregation of host cell cytoplasm gathers directly

beneath the appressorium [30] and a halo region develops as the host cell wall is

modified in the area surrounding appressorial contact [2]. The cytoplasmic aggregate

is coincident with vesicle-mediated deposition of papilla materials [40]. The papilla is

a heterogeneous mass deposited between the epidermal cell plasma membrane and the

cell wall, directly beneath the site of attempted penetration. Considerable evidence

indicates that papillae provide a barrier to infection [1].

Papillae are chemically complex structures [1]. They have a callous matrix infused

with phenolic substances, proteins, hydrolytic enzymes and altered levels of specific

chemical elements including Si. Si is commonly localized in papillae and epidermal cell

wall halo regions [8, 26, 27, 28, 41, 42]. In our experience, these same host cell regions

exhibit strong localized autofluorescence (Fig. 1). This suggests that Si accumulation

depends upon the accumulation of autofluorogens. The primary objective of the

current experiments was to test this hypothesis.

Mayama & Shishyama [33] first described localized autofluorescence in barley

epidermal cells attacked by B. graminis appressoria. It is now clear that localized

autofluorescence is a normal expression of epidermal cell response to B. graminis attack

in oat, barley and wheat [2, 7, 9–14, 23, 24, 29–31, 33, 43]. Considerable physical and

biochemical evidence indicates that these autofluorogens are phenolic compounds, and

that they are probably products related to lignin biosynthesis [11, 12, 14, 43].

In a series of investigations, we examined the effects of treating cereal leaves with α-

amino-β-phenylpropionic acid (AOPP), a potent competitive inhibitor of phenyl-

alanine ammonia-lyase (PAL), which catalyses the first committed step in phenolic

compound synthesis. In oat, barley and wheat, PAL inhibition dramatically suppresses

localized autofluorescence associated with response to appressorial attack and increases

the proportion of appressoria successfully penetrating to form haustoria [7, 10–14, 43].

Kunoh and his co-workers [26–28] led early investigations of the accumulation of

inorganic chemical elements at attempted penetration sites. High levels of Si, Mn and

Ca were reported. Kunoh [25] and Zeyen et al. [41] reviewed these and other studies

and described possible roles for these inorganic ions that accumulate in the same

localized regions that exhibit autofluorescence (Fig. 1).

Mn is an essential trace element in plants. Adding Mn to barley and wheat plantings

to prevent deficiency decreased powdery mildew disease [20]. Zeyen et al. [40] pointed

out that Mn is one of the many known inducers of PAL activity in plants, and suggested

that this and its capacity to act in promoting nonenzymic polymerization of phenolic

compounds [5] may be important in resistance to attempted penetration. Conversely,

Kunoh [25] noted that Mn can promote chitin synthetase activity in fungi, and thus

it may be sequestered and used by the attacking fungus.

Ca is involved in many metabolic processes [22], and its role in resistance}susceptibility to infection is complex and not well understood. In certain situations, Ca

Oat epidermis attacked by Blumeria graminis 225

is associated with increased effectiveness of papilla-based resistance to B. graminis

[4, 16], while in other cases Ca enhances infection [19, 36, 37].

There is a consistent correlation between Si accumulation and disease resistance in

several host}pathogen systems [41]. Carver, Zeyen & Ahlstrand [8] demonstrated by

sequential use of light microscopy, scanning electron microscopy (SEM) and energy

dispersive X-ray microanalysis (EDX) [41] that the rate of Si accumulation in papillae

formed in barley may influence the outcome of attempted penetration from appressoria.

Thus, when Si accumulated to high levels relatively early in the attack, penetration

tended to fail, whereas where penetration succeeded the accumulation was delayed.

Despite this correlation, the function of Si remains a matter of speculation. For

example, in the non-host French bean}cowpea rust interaction, it was suggested that

Si may present a physical barrier to penetration and}or restrict the flow of materials

to an attacking fungus that prevents its premature senescence [17].

An important question is whether the accumulation of inorganic elements is

dependent upon the accumulation of the autofluorogenic compounds. For Si, at least,

there is reason to suppose this might be so from evidence by Heath [18], who found

insoluble silicon in necrotic cowpea cells that died as a result of infection with

incompatible cowpea rust. Further, in the French bean}cowpea rust fungus system,

Perumalla and Heath [36] found that PAL inhibition (AOPP treatment), reduced the

incidence of autofluorescence in mesophyll cell walls and this was associated with

reduced incidence of Si deposition as judged by cell wall refractivity using light

microscopy. It was argued that this is not surprising, since Si is taken up and

transported in plants in the form of monosilicic acid [see 8], and monosilicic acid can

form strong complexes with organic hydroxy compounds [21]. The formation of such

complexes was also used to explain Si accumulation in barley cells that died and

accumulated autofluorescent phenolics as a result of hypersensitivity to avirulent B.

graminis attack [24].

The present study used X-ray microanalysis to compare Si accumulation in time-

course studies where susceptible oat (A�ena sati�a L.) leaves attacked by B. graminis were

treated with AOPP to inhibit PAL and suppress localized autofluorescent host cell

responses [7]. We also studied the effects of the treatment on accumulation of Mn and

Ca. Analyses were performed using both frozen-hydrated and freeze-dried samples. In

frozen-hydrated material, electron beam penetration, which generates element-specific

X-ray emission, is restricted [3, 41]. Thus, X-ray analysis from frozen-hydrated

material reflects the composition of the epidermal wall halo regions and papillae at

appressorial attack sites. Freeze-drying allows greater beam penetration into epidermal

cells and thus provides information on elements present within attacked cells.

MATERIALS AND METHODS

Host and pathogen material

Seedling leaves of the powdery mildew susceptible spring oat cv. Selma were used

throughout. Seeds were sown in John Innes No. 3 potting compost and grown in a

spore-proof glasshouse under natural lighting conditions and a minimum temperature

of 12 °C. Seedlings were grown until the second leaf was fully expanded, then first-

formed leaves were excised for experimentation. Blumeria graminis f. sp. a�enae (Marchal)


race 5 was maintained on young plants of Selma. One day before experimentation,

Selma leaves bearing conidia were shaken to displace older spores, ensuring that freshly

formed conidia were available for inoculation.

The PAL inhibitor and its application to lea�es, inoculation and incubation procedures

The PAL inhibitor α-aminooxy-β-phenylpropionic acid (AOPP) (Cambridge Research

Biochemicals Ltd., Northwich, Cheshire, U.K.) was used as a 1±0 m aqueous solution.

At this concentration, AOPP treatment effectively inhibits localized autofluorescent

host cell responses to B. graminis in oat [7]. Sterile distilled water was the control

treatment.

Twenty-four hours before inoculation, first-formed leaves of Selma were cut through

the sheath at soil level, the second leaf was discarded, and the cut ends of the first leaves

were submerged either in AOPP solution or in water. The submerged cut ends were

trimmed to remove air embolisms from the vascular tissue. Leaves were allowed test

solution uptake for 24 h in a growth cabinet at 20°³0±5 °C, with a 12 h light}dark

cycle. Light was supplied by white fluorescent tubes supplemented with incandescent

bulbs to give 150 µmol m−# s−" photon flux density. The RH was approximately 48%

during the light and 54% during the dark periods. After 24 h, the leaves were removed

from test solutions and their adaxial surfaces were inoculated in a spore settling tower

to give approximately 20 conidia mm−#. Leaves were then replaced in the appropriate

test solution, their cut ends trimmed to remove air embolisms, and returned to the

growth cabinet for incubation.

At the end of the experimental incubation periods, 30 mm segments were cut from

the centre of the leaf blades. The acropetal 20 mm portions were prepared for light

microscopy and the 10 mm basipetal portions were used for SEM and EDX.

Tissue fixation and preparation for light and incident fluorescence microscopy

The 20 mm acropetal leaf segments were fixed and cleared as previously described [9]

to avoid displacement of fungal germlings. The unstained leaves were mounted on glass

slides without coverslips. Differential interference contrast microscopy, with a 40X ‘no-

coverslip ’ objective, allowed resolution of fungal PGTs, appressoria, haustoria and

secondary hyphae. Localized autofluorescent host responses associated with PGT and

appressorial contact sites were observed by incident fluorescence microscopy (blue

exciter filter, max. transmittance 400 nm; dichroic mirror and barrier filter

transmittance range 500–800 nm).

Tissue fixation and preparation for scanning electron microscopy and energy dispersi�e X-ray

microanalysis

The 10 mm basipetal leaf segments were mounted inoculated (adaxial) surface up on

individual copper SEM stubs using colloidal carbon. Stubs were immediately immersed

and held in liquid nitrogen to cryofix tissues until viewed and analysed as frozen-

hydrated specimens.

For observation of frozen-hydrated specimens, stubs were mounted in a pre-cooled

(®196 °C) copper stub holder and transferred immediately to the cold stage

(®160 °C) of a JEOL 840A SEM, and observed without coating. To remove ice

crystals formed on the specimen and stub holder during transfer, the cold stage was


warmed to ®70 °C until ice sublimed (1–3 min) and then re-cooled to ®160 °C before

observations were made.

The SEM was fitted with a Kevex EDX microanalysis system using a Quantum

light-element detector. To standardize X-ray collection conditions, an electron

accelerating voltage of 10 kV was selected and the electron beam current was adjusted

to a fixed value by stopping the scanning raster and positioning the beam on the copper

stub holder at 3000¬ magnification. Copper K-α X-rays were collected and the beam

current adjusted using the probe current control (beam spot size ; condenser lens) to

give an X-ray emission rate of 3000 counts s−", equivalent to a specimen current of

558 pA when measured with a carbon Faraday cup. With these instrument conditions

held constant, and with proper detector location, sufficient X-rays were produced over

100 s collection from leaf tissues to allow reliable comparisons between sites of analysis

[42, 43].

To collect X-ray data from host tissues associated with fungal appressoria, the

magnification was adjusted to 7000¬ and a raster area of 1±15¬0±8 µm was positioned

on the host cell at the tip of the first appressorial lobe (Fig. 1) and X-rays were collected

for 100 s. X-ray data were also collected from unattacked epidermal cells by

positioning the raster on the wall of a cell which was not in contact with any fungal

F. 1. Micrographs of germinated B. graminis conidia on Selma oat 24 h after inoculation,showing coincidence of localized autofluorescence and Si accumulation in epidermal sitescontacted by germ tubes. (A) A fungal germling with a primary germ tube (PGT) and anappressorial germ tube which has stimulated a papilla response in the subtending epidermal cell(Pa). Differential interference contrast microscopy. (B) The same germling as in A showinglocalized autofluorescence at the PGT and appressorium contact sites of the epidermal cell.Incident fluorescence microscopy. (C) Scanning electron micrograph of an uncoated, freeze-driedfungal germling. (D) Silicon X-ray map showing localized accumulation at both the PGT andappressorium epidermal cell contact sites of the germling shown in C.


structure. X-ray data were processed using Kevex software for automatic background

(Bremsstrahlung X-rays) subtraction for bulk specimens, and calculations of Si, Ca,

and Mn peak intensities in counts s−" was by Gaussian deconvolution.

To freeze-dry specimens previously analysed in a frozen-hydrated state, the

specimens were left on the microscope stage under vacuum, and the stage temperature

was allowed to rise to ambient (ca. 20 °C) overnight. Resultant freeze-dried specimens

were examined either the next day or were stored over anhydrous silica gel until

analysed. All freeze-dried specimens were examined uncoated, using the identical X-

ray collection conditions used for frozen-hydrated specimens.

X-ray analysis of the frozen-hydrated material caused some beam damage and loss

of mass at the appressorial contact sites under examination. Therefore, these germlings

were avoided in the subsequent analyses of the same specimens following freeze-drying.

Experimental details

Experiment 1. Fungal development, autofluorescent host cell responses to fungal germ tube contact and

EDX microanalysis of material sampled 12 and 16 h after inoculation. Experiment 1 was a

preliminary, exploratory experiment designed to determine when autofluorescence and

haustorial formation began, and to determine the earliest time of appearance of Si, Ca

and Mn at appressorial contact sites.

First-formed leaves of 12 phenotypically similar Selma plants were excised; six were

treated with 1±0 m AOPP, and six with sterile distilled water as controls. Three

replicate leaves from both treatments were selected 12 and 16 h after inoculation, and

segments from each were prepared for light microscopy and SEM}EDX microanalysis

of frozen-hydrated specimens.

On each leaf segment prepared for light microscopy, 100 germlings with a mature

appressorium were examined. Germlings were accepted for examination only if the

host cell subtending the appressorium was either adjacent to a stomatal complex or no

more than two cells distant from a stomatal complex, and then only if the host cell had

contact with only one germ tube. These selection criteria were necessary because the

oat epidermis contains several types of epidermal cell, whose responses to attack vary,

and we wished to avoid confounding effects by having only one germling attack per cell

so that an epidermal cell was only responding to that single attack and not multiple

attacks. Differential interference contrast microscopy was used to determine whether

appressoria had penetrated the host cell successfully to form a haustorium. Each

germling was also examined by incident fluorescence microscopy to determine whether

the PGT and appressorium were associated with a localized autofluorescent host cell

response. Replicate mean data for each character were calculated in percentages and

transformed to arcsines to normalize data prior to analysis of variance [7].

The intensity of localized autofluorescence associated with appressoria was classified

visually using a 0–3 scale where: 0¯no fluorescence, 1¯ faint, 2¯moderate and

3¯ intense fluorescence [7]. Fluorescence intensity data were not analysed statistically.

When SEM was used to examine the fungus at these early stages of development (12

and 16 h), it was impossible to determine whether appressoria had formed haustoria

because they would be concealed within epidermal cells and no hyphae were formed.

For each specimen taken 12 h after inoculation, X-ray data were collected from five

unattacked cells and from sites adjacent to the tips of five different appressorial germ


tubes. For specimens taken 16 h after inoculation, X-ray data were gathered from a

raster positioned at the tip of 20 appressoria on each leaf segment and from five

unattacked epidermal cells.

Experiment 2. Pathogen development, autofluorescent host cell responses to fungal germ tube contact

and EDX microanalysis of material sampled 18, 24 and 30 h after inoculation. Nine first-formed

leaves were treated with 1±0 m AOPP to inhibit PAL, and nine with water as controls.

Three replicate leaves from both treatments were selected 18, 24 and 30 h after

inoculation, and segments from each were prepared for light microscopy and

SEM}EDX microanalysis as in Experiment 1. Light and fluorescence microscope

observations were made in the same way as for Experiment 1.

When the SEM was used to examine material sampled 18 h after inoculation, it was

again impossible to determine whether appressoria had formed haustoria since no

hyphae had developed at this early stage. Therefore, at this time, X-ray data were

gathered from a raster positioned at the tip of 20 appressoria on each leaf segment.

When the SEM was used to examine 24 and 30 h samples, it was possible to

discriminate between appressoria which had penetrated host cells to form functional

haustoria (because hyphae are formed by these times), and appressoria which failed to

do so (hyphae absent). From each leaf segment cryo-fixed at 24 and 30 h, X-ray data

were collected from a raster positioned at the tip of 10 appressoria which had

successfully formed a haustorium (hyphae present) and 10 appressoria which had failed

to form a haustorium (hyphae absent). X-ray data were collected from five unattacked

epidermal cells on each leaf.

Statistical analyses of X-ray data

X-ray count data in counts s−" for each element of interest (Si, Ca, and Mn) collected at

each sample time (18, 24 and 30 h) were subjected to statistical analysis of variance,

and weighted orthogonal comparisons were made to compare data between treatments

(water control vs. AOPP treatment), between sample sites (unattacked cells vs.

appressorial lobe sites), and between sites associated with successful (hyphae present)

and unsuccessful appressoria (hyphae absent) (24 and 30 h samples only). Residuals

from analyses of raw data were inspected to determine whether the data conformed to

normal distributions and whether the variances were constant over the data range;

where they were not, the raw data were transformed to logarithms and the statistical

analysis was performed on this log data. Because of these complexities, no attempt was

made to compare data between specimens from different sample times by statistical

analysis.

RESULTS

Differential interference microscopy: fungal development

In Experiment 1, appressoria had formed a distinct lobe by 12 h after inoculation, but

no haustoria were evident at 12 or 16 h in either AOPP-treated or control leaves

showing that the outcome of penetration attempts had not been determined by 16 h.

In Experiment 2, a proportion of appressorial penetration pegs had penetrated

successfully to produce haustoria in leaves from both treatments at 18 h. However, at


this time, the haustoria were rudimentary, spherical to ellipsoidal bodies without

digitate processes. They were at an early stage of development. No hyphae were

observed at this time. The proportion of successful appressoria increased significantly

(P! 0±01) between 18 and 24 h, but there was no statistical increase between 24 and

30 h (Fig. 2). This indicated that the outcome of all penetration attempts was

determined by 24 h. Also, by 24 h all haustoria had developed some digitate processes

and hyphae were evident, showing that haustoria were functional. The presence or

absence of hyphae at 24 and 30 h was therefore used in subsequent SEM investigations

as an indicator of success or failure of attempted penetration.

30

80

018

Time (h) after inoculation

Per

cen

t ap

pres

sori

a w

ith

hau

stor

ia

60

40

20

24

SED

F. 2. Percentage (transformed to arcsines for statistical analysis) of B. graminis appressoriapenetrating to form haustoria in oat seedling leaves fixed 18, 24 and 30 h after inoculation andtreated with water (controls, *) or with 1±0 m AOPP(+) to inhibit PAL activity. Means basedon observation of 100 appressoria on each of three replicate leaves in each case. The standard errorof differences of means SED [12 d.f.] are presented for comparison of treatment means withinsample times and of sample time means within treatments.

Figure 2 also shows that AOPP treatment was correlated with increased penetration

success ending in increased presence of haustoria at all time intervals tested. The effect

of AOPP treatment in this experiment was consistent with previous findings

[7, 10, 11, 13].

Incident fluorescence microscopy: localized autofluorescent host cell responses

In control leaves from both experiments, localized yellow-green autofluorescence was

evident in epidermal cell wall regions subtending the tips of fungal germ tubes. This

was as expected and as described previously [7, 10, 11, 13, 14]. Generally, the

autofluorescent response was more intense at appressorium contact sites than at PGT

contact sites (Fig. 1).


In both experiments, AOPP treatment reduced significantly (P! 0±001) the

percentage of PGTs and appressoria associated with localized autofluorescent responses

in host cells (Fig. 3 shows data from Experiment 2). As in previous studies, PGT

contact sites were particularly sensitive to suppression of autofluorescence by AOPP

treatment. Also, as in previous studies, treatment with AOPP increased the proportion

of germlings forming haustoria (Fig. 2).

30

100

018

Time (h) after inoculation

Per

cen

t ge

rm t

ube

s as

soci

ated

wit

h a

uto

flu

ores

cen

ce

60

40

20

24

SED PGTs

80

SED appressoria

F. 3. Percentage (transformed to arcsines for statistical analysis) of B. graminis appressoria(square symbols) and PFTs (diamond symbols) associated with localized autofluorescent responsein epidermal cells of oat seedling leaves fixed 18, 24 and 30 h after inoculation. Leaves weretreated with water (controls ; open symbols) or with 1±0 m AOPP to inhibit PAL (solid symbols).Means based on observation of 100 appressoria on each of three replicate leaves in each case.SED’s (12 d.f.) for comparison of treatment means within sample times and of sample time meanswithin treatments.

In Experiment 1 (data not shown), relatively few appressoria were associated with

localized autofluorescence at 12 h in leaves from either treatment. The proportion

increased by 16 h in both treatments, but while responses remained faint in AOPP-

treated leaves, in water controls some moderate and a few intense responses were

evident.

In Experiment 2 the frequency of autofluorescence in water controls remained

unchanged from 18 h (Fig. 3), indicating that the processes initiating autofluorogen

accumulation were underway by 18 h. The effect of AOPP treatment in suppressing

autofluorescence was evident at appressorial contact sites at all sample times. With

AOPP treatment, where autofluorescence was detected it was almost invariably faint

at 18 h and remained so at 24 and 30 h (Fig. 4). By contrast, in water controls, a high

proportion of autofluorescent responses were moderate or intense by 18 h, and in


80

00

Fluorescence intensity score

60

40

20

321 0

30 h after inoculation

321

80

00

Per

cen

tage

of

appr

esso

rial

con

tact

sit

es w

ith

loca

lize

d au

tofl

uor

esce

nce

60

40

20

321 0


321

80

00

60

40

20

321 0


321

Water (Controls) AOPP (1.0 mM)

F. 4. Intensities of localized autofluorescent responses associated with B. graminis appressoriain epidermal cells of oat seedling leaves fixed 18, 24 and 30 h after inoculation and treated withwater (controls) or with 1±0 m AOPP to inhibit PAL activity. Data based on observation of 100appressoria on each of three replicate leaves in each case. 0¯no visible fluorescence; 1, 2 and3¯ faint, moderate and intense fluorescence, respectively.


subsequent samples the majority of appressoria were associated with an intense

response (Fig. 4).

In summary, AOPP treatment suppressed the frequency and intensity of localized

autofluorescent host cell responses, and this effect was correlated with increased

successful penetration (approximately doubled, Fig. 2). Light microscopy further

revealed that the processes of attempted penetration were underway by 18 h after

inoculation, and the outcome of penetration attempts was fully determined by 24 h.

EDX microanalysis of elements present in lea�es attacked by B. graminis

Analyses from Experiment 1 showed that at 12 and at 16 h the X-ray count rates

obtained for all elements at appressorial lobe contact sites were similar to unattacked

cells which had no contact with the fungus. Therefore, for brevity, X-ray data from

Experiment 1 are omitted and only data from Experiment 2 are given for Si, Ca, and

Mn.

P < 0.05†

P < 0.01†

NS

P < 0.05

NS

300

0

Unattacked cells

Si X

-ray

cou

nts

s–1

18

200

10

24 30

NS NS

Appressorium contact sites

18 24 30

NS

Freeze-dried

Time after inoculation (h)

300

0

Si X

-ray

cou

nts

s–1

18

200

100

24 30

NS NS NS

18 24 30

NS

Frozen-hydrated

F. 5. Silicon X-ray counts gathered when an electron raster scan was focused either onunattacked oat leaf epidermal cells or adjacent to B. graminis appressoria on attacked cells in leavestreated with water (controls ; ) or with 1±0 m AOPP ( ). Data were collected at 18, 24 and 30 hafter inoculation, and are means from analyses of 20 appressorial contact sites and from fiveunattacked cells. Each sample time and treatment contained three replicate leaves. At 18 h,analyses were from randomly selected appressorial sites, but for 24 and 30 h analyses were from10 sites where appressoria had successfully penetrated (hyphae present), and from 10 sites whereappressorial penetration had failed (no hyphae). Initial analysis done on frozen-hydrated leavesfollowed by analysis in the freeze-dried state. NS above a pair of columns indicates no significantdifference, while the P-value indicates the probability of difference at the given level ofsignificance.

† Indicates data required log transformation prior to analysis.


Silicon. Figure 5 shows the average Si X-ray counts s−" from raster scans at the different

localities (appressorial contact sites and unattacked cells) and times (18, 24 and 30 h

post-inoculation).

In frozen-hydrated leaves, unattacked cells that had no contact with B. graminis

always gave less than 4 counts s−" When freeze-dried, the count rate for unattacked

cells was always less than 7 counts s−" There was no difference in the Si count rate

obtained from unattacked cells in water-treated (control) leaves and AOPP-treated

leaves, except in freeze-dried leaves fixed at 30 h after inoculation when the Si count

rate was marginally higher (P! 0±01) in water-treated (6±04 counts s−") than in

AOPP-treated leaves (2±47 counts s−").

The difference between unattacked cells and appressorial contact sites was marked

beginning at 18 h after inoculation (P! 0±001), and this difference was maintained at

subsequent sample times. No direct statistical comparison was made between

preparation procedures, but higher count rates for Si were consistently obtained in

freeze-dried, where electron beam penetration was greater, compared to frozen-

hydrated material where electron beam penetration was less [41].

Figure 5 also shows that there was very rarely any difference between water-treated

controls and AOPP-treated leaves in Si count rates at appressorial contact sites. In

frozen-hydrated material sampled at 18 h slightly higher count rates (P! 0±05) were

associated with appressorial contact sites in water controls when compared with

AOPP-treated leaves. However, this difference was not seen when the same tissue was

freeze-dried. Also, in freeze-dried material from 30 h samples, slightly higher count

rates (P! 0±05) were associated with appressorial contact sites in water controls when

compared with AOPP-treated leaves. It is questionable whether these small statistical

differences were of biological significance.

Si count rates were also compared from sites adjacent to appressoria where

penetration had succeeded (hyphae present) and sites where penetration had failed

T 1Mean Si X-ray data (counts s−") from epidermal cell areas directly adjacent to B. graminis appressoria on oat

lea�es treated with water (controls) or with 1±0 mM AOPP

Specimen preparation procedure

Frozen-hydrated material Freeze-dried material

Time after Successful Failed Successful Failedinoculation Treatment penetrations penetrations penetrations penetrations

24 h Water 159 149 265 221AOPP 164 164 233 210

30 h Water 171 167 318 258AOPP 164 150 244 229

Appressoria were classified as successful (hyphae present) or failed (hyphae absent) at 24and 30 h after inoculation. X-ray analyses were first applied to frozen-hydrated specimens thatwere then freeze-dried for further analyses. Means based on 10 appressoria from each of threeleaf samples in each case. Analysis of variance showed no significant differences between countrates from successful and failed appressorium classes when comparisons were made withinpreparation procedure, treatment and sample time.


(hyphae absent) in samples from 24 h and 30 h after inoculation (Table 1). Analyses

of variances showed that there were no differences in Si count rates between successful

and failed penetration sites.

Taken together with Experiment 1, the data suggest that Si accumulated rapidly in

appressorial contact sites between 16 and 18 h, when light microscopy revealed that the

most advanced fungal germlings were penetrating epidermal cells to form haustoria.

While treatment with AOPP greatly reduced the frequency and intensity of

autofluorescence, it had almost no effect on the accumulation of Si. This was true

regardless of the success or failure of penetration.

Calcium. Figure 6 shows the average Ca X-ray counts s−" from raster scans at the

different localities (appressorial contact sites and unattacked cells) and times (18, 24

and 30 h post-inoculation).

In frozen-hydrated samples, where electron beam penetration was shallow, Ca count

P < 0.05†

P < 0.001†

40

0

Unattacked cells

Ca

X-r

ay c

oun

ts s

–1

18

30

20

10

24 30

NS

NSNS


18 24 30

P < 0.001†

Freeze-dried


40

0

Ca

X-r

ay c

oun

ts s

–1

18

30

20

10

24 30

NS P < 0.05†

NS

18 24 30

NS NS NS

Frozen-hydrated

F. 6. Calcium X-ray counts gathered when an electron raster scan was focused either onunattacked oat leaf epidermal cells or adjacent to B. graminis appressoria on attacked cells in leavestreated with water (controls ; ) or with 1±0 m AOPP ( ). Data were collected at 18, 24 and 30 hafter inoculation, and are means from analyses of 20 appressorial contact sites and from fiveunattacked cells. Each sample time and treatment contained three replicate leaves. At 18 hanalyses were randomly selected appressorial sites, but for 24 and 30 h analyses were from 10 siteswhere appressoria had successfully penetrated (hyphae present) and from 10 sites whereappressorial penetration failed (no hyphae). Initial analysis done on frozen-hydrated leavesfollowed by analysis in the freeze-dried state. NS above a pair of columns indicates no significantdifference while the P-value indicates the probability of difference at the given level ofsignificance.

† Indicates that data required log transformation prior to analysis.


rates were relatively low. In frozen-hydrated specimens there was no difference

between Ca count rates obtained from unattacked cells and from sites of appressorial

contact, or between water- and AOPP-treated leaves except for unattacked cells at

24 h after inoculation when slightly higher count rates were obtained from AOPP-

treated leaves (Fig. 6).

In freeze-dried material, where beam penetration occurred to a greater depth than

in frozen-hydrated material, substantially higher Ca count rates were obtained.

However, as in frozen-hydrated material, count rates from unattacked cells were

similar in water- and AOPP-treated leaves. By contrast, at appressorial contact sites in

freeze-dried specimens, consistently lower Ca count rates were obtained from AOPP-

treated leaves than from water controls (Fig. 6). Further, at 30 h the Ca count rate

from appressorial contact sites was lower than from unattacked cells in both the water-

and AOPP-treated leaves (P! 0±05).

Ca count rates were also compared from sites adjacent to appressoria where

penetration had succeeded (hyphae present) and sites where penetration had failed

(hyphae absent) in samples from 24 h and 30 h after inoculation (Table 2). In frozen-

hydrated material, at 24 h there was no difference between sites adjacent to successful

and failed penetration sites. However, in 30 h frozen-hydrated samples higher Ca

count rates were obtained from sites adjacent to failed than successful penetrations in

both water controls and AOPP-treated leaves. In freeze-dried material, data were

variable, and there were no significant differences between successful and failed sites of

penetration.

T 2MeanCaX-ray data (counts s−") from epidermal cell areas directly adjacent toB. graminis appressoria on oat

lea�es treated with water (controls) or with 1±0 mM AOPP




24 h Water 3±4 4±7 35±1 33±5AOPP 5±2 6±1 14±3 18±0

30 h Water 2±3 a 6±7 b 16±8 26±9AOPP 4±9 a 7±3 b 13±5 14±0

Appressoria were classified as successful (hyphae present) or failed (hyphae absent) at 24and 30 h after inoculation. X-ray analyses were first applied to frozen-hydrated specimens thatwere then freeze-dried for further analyses. Means based on 10 appressoria from each of threeleaf samples in each case. Within preparation procedure, treatment and sample time, valuesfollowed by different letters differed significantly (P! 0±01) as determined from analysis ofvariance of log transformed data; in other cases, values did not differ statistically.

Manganese. Figure 7 shows the average Mn X-ray counts s−" from raster scans at the

different localities (appressorial contact sites and unattacked cells) and times (18, 24

and 30 h post-inoculation).

Overall count rates for Mn were relatively low in unattacked cells in frozen-hydrated

and freeze-dried samples. There was no difference in Mn count rates between


P < 0.001†

P < 0.001†

P < 0.01†

15

0

Unattacked cells

Mn

X-r

ay c

oun

ts s

–1

18

10

5

24 30

NS NS NS


18 24 30

P < 0.05†

Freeze-dried


15

0

Mn

X-r

ay c

oun

ts s

–1

18

10

5

24 30

NS P < 0.05 NS

18 24 30

NSP < 0.01

Frozen-hydrated

F. 7. Manganese X-ray counts gathered when an electron raster scan was focused either onunattacked oat leaf epidermal cells or adjacent to B. graminis appressoria on attacked cells in leavestreated with water (controls ; ) or with 1±0 m AOPP) ). Data were collected at 18, 24 and 30 hafter inoculation, and are means from analyses of 20 appressorial contact sites and from fiveunattacked cells. Each sample time and treatment contained three replicate leaves. At 18 h,analyses were from randomly selected appressorial sites, but for 24 and 30 h analyses were from10 sites where appressoria had successfully penetrated (hyphae present) and from 10 sites whereappressorial penetration had failed (no hyphae). Initial analysis done on frozen-hydrated leavesfollowed by analysis in the freeze-dried state. NS above a pair of columns indicates no significantdifference, while the P-value indicates the probability of difference at the given level ofsignificance.

† Indicates data required log transformation prior to analysis.

unattacked cells from water- and AOPP-treated leaves, with the exception of frozen-

hydrated samples at 24 h when slightly greater (P! 0±05) count rates were obtained

from AOPP- than water treated leaves (Fig. 7).

When X-ray analyses were applied to sites adjacent to appressoria, in all but one

case significantly higher Mn count rates were obtained from AOPP-treated than from

water-treated leaves. The exception was in 18 h samples from AOPP-treated, frozen-

hydrated leaves, where no difference was detected.

Mn count rates were also compared from sites adjacent to appressoria which had

succeeded in penetration (hyphae present) and sites of failed penetration (hyphae

absent) in 24 and 30 h samples (Table 3). In frozen-hydrated preparations, sites

adjacent to appressoria which had failed to penetrate had Mn count rates significantly

higher than those where penetration had succeeded. This was true in both AOPP- and

water-treated leaves. In freeze-dried preparations a similar trend was evident in most

cases, but the data were variable and no statistical differences were detected.


T 3Mean Mn X-ray data (counts s−") from epidermal cell areas directly adjacent to B. graminis appressoria on

oat lea�es treated with water (controls) or with 1±0 mM AOPP




24 h Water 0±40 a 3±63 b 1±77 3±38AOPP 1±74 a 4±51 b 3±83 6±23

30 h Water 0±04 a 4±90 b 1±29 6±17AOPP 14±60 11±90 17±30 10±19

Appressoria were classified as successful (hyphae present) or failed (hyphae absent) at 24and 30 h after inoculation. X-ray analyses were first applied to frozen-hydrated specimens thatwere then freeze-dried for further analyses. Means based on 10 appressoria from each of threeleaf samples in each case. Within preparation procedure, treatment and sample time, valuesfollowed by different letters differed significantly (P! 0±01) as determined from analysis ofvariance of log transformed data; in other cases, values did not differ statistically.

DISCUSSION

Oat PAL is efficiently inhibited by AOPP [7]. As in previous studies [7, 10, 11, 13, 14],

its application to leaves greatly reduced both the frequency and intensity of localized

autofluorescent responses in oat epidermal cell sites contacted by PGTs and appressoria

of B. graminis (Figs 1, 3, 4). The dramatically increased penetration and reduced

frequency and intensity of autofluorescence resulting from AOPP treatment (Fig. 2),

reinforced previous findings from oat [7, 10, 11, 13, 14], barley [12, 13, 31, 43] and

wheat [13]. These findings all support the view that these localized autofluorogens,

which also have ultraviolet absorption and emission characteristics of phenolic

compounds [33, 34], are indeed phenolic in nature and in the B. graminis—cereal

systems contribute to penetration resistance at the individual epidermal cell level in

both susceptible and resistant cereal hosts. No evidence, such as the direct extraction

and identification of phenolics from these microscopic, localized autofluorescent sites in

epidermal cell walls, exists.

The localized, phenolic-based autofluorescence of epidermal cells attacked by B.

graminis is spatially coincident with accumulation of Si (Fig. 1). In French bean

attacked by the cowpea rust fungus, the deposition of Si in attacked mesophyll cell

walls appeared dependent upon the presence of phenolic compounds [36]. These

observations led us to test the hypothesis that localized Si deposition is dependent upon

localized accumulation of autofluorogenic phenolics in oat epidermal cells attacked by

B. graminis. We decided to test this linkage hypothesis for Si and for two other elements,

Ca and Mn. These three chemical elements have been associated with halo and papilla

formation in cereals [4, 8, 16, 25, 26–28, 32, 42, 43].

The use of EDX [39, 40] allowed the identification and quantification of Si, Ca and

Mn in the regions of B. graminis attacked oat cells. We initially analysed frozen-hydrated


epidermal cells, and then freeze-dried the same specimens for further analysis. Since

electron beam penetration at 10 kV accelerating voltage in frozen-hydrated plant cells

is expected to be ! 3±5 µm [42], EDX analysis was confined to the epidermal cell wall

apoplast and upper portions of the subtending papillae. Freeze-drying results in far

greater electron beam penetration, " 50 µm at a 10 kV accelerating voltage. Thus,

characteristic X-rays of elements from the leaf specimens when freeze-dried came from

within the epidermal cells [42].

Despite the marked effects of AOPP on suppressing the incidence and intensity of

localized autofluorescence, AOPP treatment had little or no effect on the spatially

coincident accumulation of Si. This was true when specimens were frozen-hydrated or

freeze-dried (Fig. 5). Simply put, Si accumulation appears independent of auto-

fluorgenic phenolic accumulation, even though they are spatially coincident.

Our findings with Si contrast somewhat with the situation in French bean (Phaseolus

�ulgaris L.) attacked by the cowpea rust fungus (Uromyces �ignae), which is an

inappropriate formae specialis of rust for French bean. Here, Perumalla and Heath [36]

showed that AOPP treatment, which prevented mesophyll cell wall autofluorescence,

apparently prevented silica deposition as assessed by the refractivity of light passing

through attacked mesophyll cell wall areas. While Perumalla & Heath’s data [36]

suggested that silica deposition in French bean was linked to phenolic-based

autofluorescence in mesophyll cells of bean, they also suggested that Si accumulation

may not be solely dependent on the presence of phenolic compounds. They suggested

roles for other factors, including the presence of hydroxyproline-rich glycoproteins in

accumulating Si in attacked mesophyll cells.

We do not know what factor(s) control localized Si accumulation in oat epidermal

cells attacked by B. graminis, but Si may well complex with various organic constituents

in the halo}papilla regions. Si with its four active valences (Si%+) is almost as adept at

forming chemical bonds as is carbon. Si can be covalently combined with alginic and

pectic acids and some polysaccharides as well as with proteins [15]. Alternatively, Si

may require only minute quantities of phenolics or other organic constituents leading

to insoluble, localized deposits in epidermal cell walls and papillae [42]. If only minute

quantities of organic compounds were needed to stimulate Si accumulation, then it

would be very difficult to prove or disprove any phenolic linkage hypothesis by the

cytological means we employed.

At 24 and 30 h it is too late to determine whether Si accumulation had influenced

attempted penetration, since penetration commenced between 16 and 18 h after

inoculation. Thus, the present study did not allow us to determine whether Si

accumulation contributed to penetration resistance in oat. We were first able (by

SEM) to discriminate successful penetrations (hyphae present) from failed penetrations

(hyphae absent) at 24 h after inoculation. Our Si X-ray count data showed no

significant differences between success and failure at either 24 or 30 h post inoculation

(Table 1). With barley it has been shown that Si in successful penetrations continues

to accumulate in papillae which now become haustorial neck collars [8]. In view of

this, it is not surprising that comparisons between count rates associated with successful

and failed penetration sites showed little difference (Table 1).

In barley leaves and coleoptiles, various studies have found that Ca is present in

papillae and haloes during B. graminis attack [2, 28, 32]. The probable role of Ca in


papilla formation and effectiveness of papillae was reviewed recently [1, 25]. Ca is not

always found at elevated levels in halo}papillae regions. Zeyen et al. [42] found Ca

levels at appressorial contact sites of resistant and susceptible barley were, if anything,

slightly lower than in unattacked cells. In our current study of oat, Ca count rates at

halo}papilla regions did not increase above the level seen in unattacked cells, nor were

Ca levels reduced by AOPP treatment in unattacked cells (Fig. 6).

When Ca levels were compared at sites of successful and failed penetration, the only

statistically significant differences noted were in frozen-hydrated samples at 30 h after

inoculation, long after success or failure of penetration was determined. Perhaps, as

Zeyen et al. [42] suggested from barley studies, the low levels of Ca found in contact

sites are sufficient to support its role in papilla formation.

In freeze-dried material (where the electron beam penetrated deeply into cells) Ca

count rates from water-treated leaves were similar in attacked and unattacked cells at

18 and 24 h after inoculation. However, at 30 h they were lower in attacked cells. In

freeze-dried, AOPP-treated, leaves Ca count rates were always substantially lower

within attacked when compared with unattacked cells (Fig. 6). Although this effect

clearly relates to pathogen attack, its relevance, if any, to success or failure of

penetration is unclear.

As reported by others working with barley [25, 42], we found that Mn accumulated

in oat at appressorial contact sites. Mn had already accumulated above levels seen in

unattacked cells by 18 h when light microscopy showed that fungal penetration was

underway. Figure 7 shows that similar count rates were obtained in both frozen-

hydrated and freeze-dried samples from the same times after inoculation. This

contrasts with the situation for both Si and Ca where there were substantially higher

count rates from freeze-dried cells. This indicates that the majority of Mn was located

in the upper portions of halo}papilla regions rather than deeper within the cells.

Zeyen et al. [42] speculated that Mn may be involved in phenolic compound

synthesis related to plant defence. Mn is one of many known inducers of PAL, and it

also stimulates the production of hydrogen peroxide required for the polymerization of

monophenols during lignin formation [5]. In both AOPP- and water-treated leaves,

higher Mn count rates were generally obtained from sites adjacent to failed appressoria

(Table 3). This supports the idea that Mn accumulation plays a role in resistance.

The composition of the halo}papilla response site is clearly complex. The

physiological processes governing the accumulation of elemental and organic

components is the focus of continuing research. To understand the complicated

relationship between localized accumulation of phenolic autofluorogens and of Ca and

Mn will require further investigation. However, the current results indicate that

localized Si accumulation occurs independently of the localized accumulation of

phenolic autofluorogens. This suggests that it may prove possible to enhance the

accumulation of each of these components to increase the efficacy of host cell responses

that prevent successful pathogen attack.

A co-operative investigation of the Institute of Grassland and Environmental Research,

Plas Gogerddan, Aberystwyth, U.K.; and the Department of Plant Pathology,

University of Minnesota, St. Paul, MN, U.S.A. Research supported by MAFF (UK),

NATO Collaborative Research Grant 900441, by the Lieberman-Okinow Chair for


Disease Resistance in Cereals, and by Projects 22-70 and 22-62 of the Minnesota

Agricultural Experiment Station. Published as Paper 191220022 of the scientific

journal series of the Minnesota Agricultural Experiment Station. Mention of a

trademark name or proprietary product does not constitute a guarantee by the

University of Minnesota or by IGER.

REFERENCES

1. Aist JR, Bushnell WR. 1991. Invasion of plants by powdery mildew fungi, and cellular mechanismsof resistance. In: Cole GT, Hoch HC, eds. The Fungal Spore and Disease Initiation in Plants and Animals.New York: Plenum Press, 321–345.

2. Aist JR, Israel HW. 1986. Autofluorescent and ultraviolet absorbing components in cell walls andpapillae of barley coleoptiles and their relationship to disease resistance. Canadian Journal of Botany 64 :266–272.

3. Ahlstrand GG, Zeyen RJ, Carver TLW. 1991. Preparation of barley epidermis for X-raymicroanalysis. In: Microbeam Analysis—1990. San Francisco: San Francisco Press, 463–466.

4. Bayles CJ, Aist JR. 1987. Apparent calcium mediation of resistance of an ml-o mutant to powderymildew. Physiological and Molecular Plant Pathology 30 : 337–345.

5. Burnell JN. 1988. The biochemistry of manganese in plants. In: Graham RD, Hannam RJ, Uren NC,eds. Manganese in Soils and Plants. Dordrecht, The Netherlands : Kluwer Academic Publishers,pp 125–137.

6. Carver TLW, Ingerson SM, Thomas BJ. 1996. Influences of host surface features on developmentof Erysiphe graminis and Erysiphe pisi. In: Kerstiens G., ed. Plant Cuticles–an integrated functional approach.Oxford: Bios Scientific Publishers, 255–266.

7. Carver TLW, Robbins MP, Zeyen RJ. 1991. Effects of two PAL inhibitors on the susceptibility andlocalized autofluorescent host cell responses of oat leaves attacked by Erysiphe graminis DC. Physiologicaland Molecular Plant Pathology 39 : 269–287.

8. Carver TLW, Zeyen RJ, Ahlstrand GG. 1987. The relationship between insoluble silicon and successor failure of attempted primary penetration by powdery mildew (Erysiphe graminis) germlings onbarley. Physiological and Molecular Plant Pathology 31 : 133–148.

9. Carver TLW, Zeyen RJ, Lyngkjaer MF. 1995. Plant cell defences to powdery mildew of Gramineae.In: Walters DR, Scholes JD, Bryson RJ, Paul ND, McRoberts N, eds. Aspects of Applied Biology, Vol.42. Physiological Responses of Plants to Pathogens. Warwick: Association of Applied Biologists, 257–266.

10. Carver TLW, Robbins MP, Zeyen RJ, Dearne GA. 1992. Effects of PAL-specific inhibition onsuppression of activated defence, and on quantitative susceptibility of oats to Erysiphe graminis.Physiological and Molecular Plant Pathology 41 : 149–163.

11. Carver TLW, Zhang L, Zeyen RJ, Robbins MP. 1996. Phenolic biosynthesis inhibitors suppressadult plant resistance to Erysiphe graminis in oat at 20° and 10 °C. Physiological and Molecular PlantPathology 49 : 121–142.

12. Carver TLW, Zeyen RJ, Bushnell WR, Robbins MP. 1994. Inhibition of phenylalanine ammonialyase and cinnamyl-alcohol dehydrogenase increases quantitative susceptibility of barley to powderymildew (Erysiphe graminis DC). Physiological and Molecular Plant Pathology 44 : 261–272.

13. Carver TLW, Zeyen RJ, Robbins MP, Dearne GA. 1992. Effects of PAL inhibition on oat, barley,and wheat cell responses to appropriate and inappropriate formae specialis of Erysiphe graminis DC.Physiological and Molecular Plant Pathology 41 : 397–409.

14. Carver TLW, Zeyen RJ, Robbins MP, Vance CP, Boyles DA. 1994. Suppression of host cinnamyl-alcohol dehydrogenase and phenylalanine ammonia lyase increases oat epidermal cell susceptibilityto powdery mildew penetration. Physiological and Molecular Plant Pathology 44 : 243–259.

15. Epstein E. 1994. The anomaly of silicon in plant biology. Proceedings of the National Academy of Science USA91 : 11–17.

16. Gold RE, Aist JR, Hazen BE, Stolzenburg MR. 1986. Effects of calcium nitrate and chlortetracyclineon papilla formation, ml-o resistance and susceptibility of barley to powdery mildew. Physiological andMolecular Plant Pathology 29 : 115–129.

17. Heath MC, Stumpf MA. 1986. Ultrastructural observations of penetration sites of the cowpea rustfungus in untreated and silicon-depleted French bean cells. Physiological and Molecular Plant Pathology29 : 27–39.

18. Heath MC. 1981. Insoluble silicon in necrotic cowpea cells following infection with an incompatibleisolate of the cowpea rust fungus. Physiological and Molecular Plant Pathology 19 : 273–276.

19. Hirata K. 1971. Calcium in relation to the susceptibility of primary barley leaves to powdery mildew.


In: Ouchi S, ed. Morphological and Biochemical E�ents in Plant–Parasite Interactions. PhytopathologicalSociety of Japan, Tokyo, 207–228.

20. Huber DM, Wilhelm NS. 1988. The role of manganese in resistance to plant diseases. In: Graham RD,Hannam RJ, Uren NC, (eds.). Manganese in Soils and Plants. Dordrecht, The Netherlands : KluwerAcademic Publishers, 155–173.

21. Ingri N. 1978. Aqueous silicic acid, silicates and silicate complexes. In: Bendz G, Lindquist I, eds.Biochemistry of Silicon and Related Problems. New York: Plenum, 3–50.

22. Kauss H. 1987. Some aspects of calcium-dependent regulation in plant metabolism. Annual Re�iew ofPlant Physiology 38 : 47–72.

23. Kidger AL, Carver TLW. 1981. Autofluorescence in oats infected by powdery mildew. Transactions ofthe British Mycological Society 76 : 405–409.

24. Koga H, Zeyen RJ, Bushnell WR, Ahlstrand GG. 1988. Hypersensitive cell death, autofluorescence,and insoluble silicon accumulation in barley leaf epidermal cells under attack by Erysiphe graminis f.sp.hordei. Physiological and Molecular Plant Pathology 32 : 395–409.

25. Kunoh H. 1990. Ultrastructure and mobilization of ions near infection sites. Annual Re�iew ofPhytopathology 28 : 93–111.

26. Kunoh H, Ishizaki H. 1976. Accumulation of chemical elements around the penetration sites ofErysiphe graminis hordei on barley leaf epidermis II. Level of silicon in papilla around the haustorialneck. Annals of the Phytopathological Society of Japan 42 : 30–34.

27. Kunoh H, Ishizaki H. 1976. Accumulation of chemical elements around the penetration sites ofErysiphe graminis hordei on barley leaf epidermis (III). Micromanipulation and X-ray microanalysis ofsilicon. Physiological Plant Pathology 8 : 91–96.

28. Kunoh H, Ishizaki H, Kondo F. 1975. Composition analysis of ‘halo ’ area of barley leaf epidermisinduced by powdery mildew infection. Annals of the Phytopathological Society of Japan 41 : 33–39.

29. Kunoh H, Kuno K, Ishizaki H. 1985. Cytological studies of the early stages of powdery mildew inbarley and wheat. XI. Auto-fluorescence and haloes at penetration sites of appressoria of Erysiphegraminis and Erysiphe pisi on barley coleoptiles. Canadian Journal of Botany 63 : 1535–1539.

30. Kunoh H, Yamamori K, Ishizaki H. 1982. Cytological studies of early stages of powdery mildew inbarley and wheat. VIII. Auto-fluorescence at penetration sites of Erysiphe graminis hordei on livingbarley coleoptiles. Physiological Plant Pathology 21 : 373–379.

31. Lyngkjaer MF, Carver TLW, Zeyen RJ. 1996. Suppression of mlo5 powdery mildew resistance inbarley. Physiological and Molecular Plant Pathology 50 : 17–36.

32. Marshall MR, Smart MG, Aist JR, Israel HW. 1985. Chlortetracycline and barley papillaformation: localization of calcium and alteration of the response induced by Erysiphe graminis. CanadianJournal of Botany 63 : 876–880.

33. Mayama S, Shishiyama J. 1978. Localized accumulation of fluorescent and U.V.-absorbingcompounds at penetration sites in barley leaves infected with Erysiphe graminis hordei. Physiological PlantPathology 13 : 347–354.

34. Nicholson RL, Hammerschmidt R. 1992. Phenolic compounds and their role in disease resistance.Annual Re�iew of Phytopathology 30 : 369–389.

35. Perumalla CJ, Heath MC. 1991. The effect of inhibitors of various cellular processes on the wallmodifications induced in bean leaves by the cowpea rust fungus. Physiological and Molecular PlantPathology 38 : 293–300.

36. Takamatsu S, Ishizaki H, Kunoh H. 1978. Cytological studies of early stages of powdery mildew inbarley and wheat. (V). Effects of calcium on the infection of coleoptiles of barley by Erysiphe graminishordei. Canadian Journal of Botany 56 : 2544–2549.

37. Takamatsu S, Ishizaki H, Kunoh H. 1979. Cytological studies of early stages of powdery mildew inbarley and wheat. (VI). Antagonistic effects of calcium and lithium on the infection of coleoptiles ofbarley by Erysiphe graminis hordei. Canadian Journal of Botany 57 : 408–412.

38. Zeyen RJ. 1982. Applications of in situ microanalysis in understanding disease : X-ray microanalysis.Annual Re�iew of Phytopathology 20 : 119–142.

39. Zeyen RJ. 1991. Analytical electron microscopy in plant pathology: X-ray microanalysis and energy lossspectrometry. In: Mendgen K, Lesemann DE, eds. Electron Microscopy of Plant Pathogens. Berlin:Springer-Verlag Press, 59–71.

40. Zeyen RJ, Bushnell WR. 1979. Papilla response of barley epidermal cells caused by Erysiphe graminis :rate and method of deposition determined by microcinematography and transmission electronmicroscopy. Canadian Journal of Botany 57 : 898–913.

41. Zeyen RJ, Ahlstrand GG, Carver TLW. 1993. X-ray microanalysis of frozen-hydrated, freeze-dried,and critical point dried leaf specimens: determination of soluble and insoluble chemical elements atErysiphe graminis epidermal cell papilla sites in barley isolines containing Ml-o and ml-o alleles. CanadianJournal of Botany 71 : 284–296.


42. Zeyen RJ, Carver TLW, Ahlstrand GG. 1983. Relating cytoplasmic detail of powdery mildewinfection to presence of insoluble silicon by sequential use of light microscopy, SEM, and X-raymicroanalysis. Physiological Plant Pathology 22 : 101–108.

43. Zeyen RJ, Bushnell WR, Carver TLW, Robbins MP, Clark TA, Boyles DA, Vance CP. 1995.Inhibiting phenylalanine ammonia lyase and cinnamyl-alcohol dehydrogenase suppresses Mla1 (HR)but not mlo5 (non-HR) barley powdery mildew resistance. Physiological and Molecular Plant Pathology47 : 119–140.

Phenylalanine ammonia-lyase inhibition, autofluorescence, and localized accumulation of silicon, calcium and manganese in oat epidermis attacked by the powdery mildew fungus Blumeria

Documents