Content uploaded by Osman Radwan
Author content
All content in this area was uploaded by Osman Radwan on Jul 10, 2015
Content may be subject to copyright.
Chapter
THE ROLE OF MELANIN PRODUCTION IN
GAEUMANNOMYCES GRAMINIS INFECTION
OF CEREAL PLANTS
Hanafy Fouly1, Shelby Henning2, Osman Radwan2,
Henry Wilkinson3and Bruce Martin1
1Department of Entomology, Soil and Plant Sciences, Clemson
University, SC 29506
2Department of Crop Sciences, University of Illinois, Urbana, IL
61801, and
3Department of Natural Resources and Environmental Sciences, USA
ABSTRACT
Gaeumannomyces graminis var. graminis (Ggg) is an ascomycete
that causes black sheath rot disease of rice (Oryza sativa L.) and take-all
root rot of several turfgrass species. G. g. var. graminis synthesizes
melanin and deposits it in hyphal cell walls. Our research indicates that
the nature of the association between Ggg and plant root is parasitic, but
can change to pathogenic and ultimately terminate as saprophytic.
Melanin plays several roles during fungal growth and throughout
infection and colonization of plant roots. First, hyphal morphology
(diameter, shape and melanin concentration) appears to change as the
fungus invades and colonizes the tissues of the root. Second, melanin
appears to be a determinant of fungal pathogenicity. Wild-type isolates of
Ggg were pathogenic, and colonized plants showed more severe
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
2
symptoms of infection while isolates lacking melanin were able to
ectotrophically colonize and penetrate roots as a parasite, but no
macroscopic symptoms of take-all were observed to indicate
pathogenicity.
INTRODUCTION
Gaeumannomyces graminis (Sacc.) Arx & D.L. Olivier var. graminis
(Ggg) is an ascomycete that infects roots of rice (Oryza sativa L.) and several
turfgrasses (Hawksworth, 1995; Walker, 1981). It is an aggressive pathogen of
rice causing black sheath rot disease. Gaeumannomyces graminis var.
graminis is an aggressive colonist but a somewhat weak pathogen of
turfgrasses including centipedegrass (Eremochloa ophiuroides (Munro) Hack.)
(Wilkinson, 1994), bermudagrass (Cynodon dactylon L.) (Elliot, 1991),
zoysiagrass (Zoysia japonica Steudel) (Wilkinson, 1993), and St.
Augustinegrass (Stenotaphrum secundatum (Walt.) Kuntze) (Elliot, et al.
1993). It is a primary colonist, forming a perennial association with
vegetatively cloned grasses and an annual association with rice. In general, the
pathogen acts as a primary colonist of newly formed roots and crowns. As an
aggressive colonist, it mantles the root surface with highly melanized,
ectotrophic hypha. Ectotrophic colonization is supported by endotrophic root
colonization of the epidermal and cortical tissues. Subsequent invasion and
colonization of the endodermis and stele tissues results in vascular occlusion
which compromises the host’s capacity to conduct water and store, transport,
or utilize available photosynthates (Jones & Clifford, 1978). Foliage
discoloration and root rotting are followed by plant death only when drought
and/or heat stress occur over time.. Finally, seed formation is severely limited,
or inhibited if root colonization is extensive.
Melanins are dark colored pigments produced by various organisms of all
biological kingdoms (Hill, 1992). Chemically, there are three different kinds
of melanins that are produced by living organisms (Bell and Wheeler, 1986).
Brown and black pigments manufactured from dihydroxyphenylalanine
(DOPA) are termed eumelanins. Red and yellow pigments derived from
DOPA and cysteine are known as phaeomelanins. Melanins derived from
phenols and catechols which lack nitrogen are known generically as melanin
(Bell and Wheeler, 1986). In addition to cellulolytic and pectinolytic enzymes
that aid in the infection of host cells, it is important to acknowledge the
presence of melanin in the hyphae of Gaeumannomyces. Cellular synthesis of
The Role of Melanin Production ...
3
the biopolymer melanin has been linked to the pathogenicity of fungi (Brush
and Money, 1999; Henson et al., 1999; Hill, 1992; and Hornby, 1998). For
example, melanin deficient mutants of the rice-blast fungus, Magnaporthe
grisea, have been demonstrated to be avirulent (Henson et al., 1999). Due to
the fact that Gaeumannomyces species are characteristically melanizied, the
presence of melanin in hyphae may play a role in the pathogenicity of
Gaeumannomyces graminis (Henson et. al., 1999).
Gaeumannomyces graminis melanin is formed by the 1, 8 DHN pathway
(Henson et al., 1999). Using wild-type and melanin deficient mutant isolates of
Ggg, Frederick, et al. (1999) showed melanin was deposited on Ggg hyphal
cell walls while Bell and Wheeler (1986) reported melanin was deposited as a
layer at the exterior surface of the fungal cell wall and/or as electron dense
granules distributed within the cell wall of the melanized yeast
Phaeococcomyces. The potential benefits that DHN melanins could confer to
hypha that synthesize them are considerable. Melanin protects fungal hypha
from the negative effects of UV irradiation (Bell and Wheeler, 1986),
temperature extremes (Hill, 1992), over-and under-abundance of moisture
(Hill, 1992), toxic concentrations of metal ions (Caesar-Tonthat et. al., 1998),
attack from antagonistic microbes (Henson et al., 1999), and extreme pH
conditions (Frederick et al., 1999).
While melanin has been implicated in the fungal colonization and
infection of plants, it has also been shown in work using melanin deficient
mutatnts, that that the presence of melanin in hyphae may not be required for
infection (Frederick et al., 1999; Henson et al., 1999). However, restoring
melanin production restored pathogenicity in some non-melanized, non-
pathogenic mutants that evidently depend on the presence of melanin to
penetrate host tissues. In one study, a melanin-deficient mutant of the human
pathogen, Wangiella dermatitidis was non-pathogenic (Brush and Money,
1999). However, when melanin production was restored, it was able to
penetrate and colonize animal tissues.
This research was divided into two phases. The objectives of the first
phase were to determine the role of melanin on linear growth, hyphal width
and branch formation, and to quantify melanin in wild-type isolates and
melanin-minus mutants of Ggg. The objectives of second phase were: to
observe and measure changes in the melanin content of Gaeumannomyces
graminis var. graminis (Ggg) during pathogenesis (inoculation through
colonization of the stele) to determine if melanin content had an effect on the
ability of Ggg to infect and colonize host roots and to;. A third objective of the
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
4
second phase determine the nature of the host and parasite association as it is
affected by the ability of Ggg to produce melanin.
MATERIALS AND METHODS
Fungal Isolates and Culturing
The fungal isolates of Gaeumannomyces graminis var. graminis used in
this research were designated WT1+, M1-, WT2+, and M2- (Table 1). Isolates
WT1+ and M1- were obtained from Joan M. Henson, Department of
Microbiology, Lewis Hall 109 Bozeman, MT 59717. Isolate WT1+ was
originally isolated from soybean (Glycine max L.). Isolate WT1+ is a wild-
type fungus that was used to produce the hyaline, melanin deficient mutant
M1- using nitroquinolene oxide (NQO) as the mutagenic agent (Epstein, et.al.,
1994). Isolate WT2+ is a wild-type of Ggg obtained from Monica Elliot,
University of Florida, Fort Lauderdale REC 3205 College Avenue, Ft.
Lauderdale FL 33314. Isolate M2- is a hyaline, melanin deficient mutant
produced from WT2+ by the method of Frederick, et.al. (1999). All isolates
were maintained on potato dextrose agar (PDA) (Sigma-Aldrich, St. Louis,
MO, USA) and transferred every 7 days to fresh media.
Hyphal Morphology and Vegetative Growth Rate
To determine the effect of melanin on hyphal morphology, measurements
of hyphal width (W) and distance between hyphal branches (DBB) were
recorded where isolates grew in culture. Each isolate was cultured and
evaluated on three different media: Luria-Bertani agar (LBA, 5g tryptone, 10g
NaCl, 5g yeast extract, 15g agar/1L water), vegetable juice agar (V8, 200ml
V8 juice, 1.8g CaCO3, 15g agar/1L water) and Czapek-Dox agar (CDA, 3g
NaNO3, 0.5g KCl, 0.5g MgSO4, 0.01g FeSO4, 1g K2HPO4, 30g sucrose, 15g
agar/1L water) using 3 repetitions (1 petri-plate = 1 repetition). When the
leading edge of a colony had extended to the perimeter of the petri plate, or 7
days had elapsed, measurements of hyphal width and distance between
branching were recorded. Distance between branching was defined by two
consecutive points of intersection formed between the main hypha and the
hyphal branch. Within these randomly selected areas of a culture, 10
measurements per area were recorded. Each experiment was replicated 3 times
The Role of Melanin Production ...
5
and repeated 3 times. Measurements were made using an ocular micrometer
and an Olympus BH-2 light microscope (40 x).
Table 1. Gaeumannomyces graminis var. graminis isolates
Desig-
nation*
Plant
Source
Descri-
ption
Coloration
of Adult
Thallus
Hypho-
podia
Reference/Source
WT1+
Soybean
wild type
black
lobed,
melanized
Frederick, 1999
M1-
Soybean
NQO
mutant of
WT1+
hyaline
lobed,
melanized
Frederick, 1999
WT2+
Bermu-
dagrass
wild type
black
lobed,
melanized
M. Elliot, FL.
M2-
Bermu-
dagrass
NQO
mutant of
WT2+
hyaline
simple,
hyaline
S. Henning
* WT= wild-type; reference number; + = pigmented; and - = non-pigmentation,
NQO = 4-Nitroquinoline-1-oxide.
To determine the effect of melanin on the vegetative growth of isolates in
culture, studies comparing radial growth on agar media were conducted. Each
isolate was cultured on LBA, CDA, and V8A media. The diameter (mm) of
each colony was recorded using digital calipers. Measurements were recorded
every 24h from the time of seeding and until the leading edge of a colony had
reached the edge of the petri plate or 7 days had elapsed. Growth experiments
were repeated 3 times, with 3 replications per experiment. The data were used
to calculate mean daily growth rate (mm).
MELANIN QUANTIFICATION
Purification of Melanin from Wild-type Hyphae
Melanin concentration was estimated using Azure A as a melanin binding
agent. Melanin was produced by culturing WT1+ in LB broth (LBB, 5g
tryptone, 10g NaCl, 5g yeast extract/1L water). The LBB was seeded with 10
culture plugs (1.0 mm diameter) of WT1+ taken from leading edge of a colony
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
6
growing on LBA. The LBB cultures were incubated at room temperature (20-
23 °C) on an orbital shaker (150 rpm) for 7 days. The LBB medium was then
separated from the hyphae by gentle vacuum filtration and discarded. The
fungal mat was cut into 5 mm pieces, submersed in acetone and heated to 50
°C for 30 minutes. The acetone was then separated from the hyphae by
vacuum filtration and discarded. The fungal mass was then washed three times
by pouring 100 ml distilled water (20 °C) over the fungal tissue. The fungal
tissue was then immersed in 200 ml absolute ethanol and heated to 85 °C for 3
hours in a hot water-bath. The fungal tissue was separated and washed as
described previously. The fungal tissue was then placed into a 500 ml single
neck boiling flask equipped with a dry reflux condenser. To the fungal mass,
200 ml 38% HCl were added via the reflux condenser and heated to 85 °C for
18 hours in a fume hood. The resulting melanin granules were collected from
the resulting black suspension by ultra-centrifugation (13,200 rpm, 60 sec.),
washed 3 times with 38 ml distilled water, dried over anhydrous CaCl2 under
vacuum and stored at -80 °C. The resulting material was considered
concentrated melanin.
Absorption of Azure A by Melanin
The melanin-Azure A binding coefficient was determined from a reaction
of concentrated melanin and a stock solution of Azure A (4.75 mg Azure A/1L
0.1M HCl). The Azure A solution (4.75 µg Azure A/1ml 0.1M HCl) had an
absorbance of 0.6 O.D. at 610 nm. Serial dilutions of the reaction solution
resulted in a proportional decrease in absorbance with a lower limit of
detection estimated at 10 µg/ml. Triplicate samples of melanin (250, 500, 1000
µg) were each placed into 15 ml Corex tubes and 3 mls of Azure A stock
solution was added. The reactions were incubated for 60 minutes at 20 °C
with slight shaking (50 rpm). The melanin-Azure complex was then separated
from the Azure A solution using an ultra-centrifuge (13,200 rpm, 60 sec.). The
optical density of the Azure A remaining in solution was measured at 610 nm
and recorded. It was calculated that a 1 milligram of concentrated melanin
absorbed 873 µg of Azure A in solution and that 1 milligram of melanin would
decrease the optical density (610 nm) of the Azure A stock solution by 0.13
units.
The Role of Melanin Production ...
7
Quantification of Melanin in Mycelia
Measurement of mycelial melanin was made using a modification of
melanin quantification reported by Butler and LaChance (1986). Fungal tissue
used for melanin quantification was cultured in Erlenmeyer flasks (125 ml)
containing 60 ml LBB. Cultures were started with 3 plugs of an isolate taken
from the leading edge of a colony on LBA using a Pasteur pipette and sterile
technique as previously described. Cultures were shaken at 150 rpm and
maintained at laboratory temperature (20-22 °C) for 7, 14, 21, or 28 days. At
the end of each growth period, fungal material was collected by removing the
LBB using vacuum filtration, washed as previously described and then
lyophilized for 24 hours. Hyphal melanin was assayed by reacting triplicate
samples of lyophilized hyphae (2000 µg) and Azure A stock solution as
described above. The reactions were incubated at 20 °C for 60 minutes with
orbital shaking (50 rpm). The hyphae were then separated from the Azure A
solution using an ultra-centrifuge (13200 rpm, 60 sec.). The optical density of
the Azure A solution was measured (610 nm) and recorded. The loss in optical
density of the Azure A solution was compared with losses in optical density
resulting from concentrated melanin to determine the melanin concentration
(µg melanin/mg hyphae).
Morphological experiments were complete randomized designs with sub-
sampling and three replicates. Isolate and media type were the fixed factors for
both experiments with treatment comparisons performed using contrast
statements. Growth rate and melanin content were analyzed over time for
isolate and media combinations using linear regression. All statistics were
performed using general linear model or regression procedures of SAS
statistical software (SAS Institute Inc., Cary, NC, USA). Every experiment
was repeated at least once.
PATHOGENICITY TESTS
Inoculation of Rice Using Conetainer Assay
The fungal isolates of Gaeumannomyces graminis var. graminis WT2+
and its melanin deficient counterpart isolate WT2- were used in this study.
Inoculum was produced in Erlenmeyer flasks containing millet (Panicum
miliaceum L.) seed (50 ml) and deionized water (50 ml), autoclave-sterilized
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
8
(32 psi, 161°C) for 1 hour. The moist grain was allowed to cool for 24 hours
and then was autoclaved a second time. Upon cooling, the grain was seeded
with culture plugs of WT2= or WT2- (5 plugs ca. 5 mm square) excised from
the leading edge of a young fungal colony. The flasks were vigorously shaken
every 24 hours for the first 2 days to uniformly distribute inoculum with the
millet. After the millet appeared covered with fungal mycelium, it was
removed from the container and dried under a laminar flow hood for 48 hours.
The dried inoculum was stored at room temperature in the dark and
periodically evaluated for contamination and viability by plating 1-20 millet
kernels on the surface of PDA.
The procedure used for host inoculation was a modification of a
conetainer assay previously reported by Wilkinson et al. (1985). A cotton ball
was placed at the bottom of a small conetainer (16 x 4 cm) (Ray Leach, Inc.
Canby, OR). The conetainer was then filled to within 4 cm of the top with
double-autoclaved vermiculite. Five colonized millet seeds were placed on top
of the vermiculite layer. Upon this layer of inoculum, a 0.5 cm thick layer of
double-autoclaved vermiculite was added. Three surface-sterilized rice (Oryza
sativa ‘Cypress’) seeds (1 minute soaking in 2.5% sodium hypochlorite, rinsed
with sterile water until no smell of bleach remained) of a host species are
placed on top of the vermiculite and covered with an additional 0.5 cm thick
layer of double-autoclaved vermiculite. The filled conetainer was then placed
into a holding rack. A total of 16 conetainers were prepared for each treatment.
Eight additional conetainers were prepared lacking the pathogen and these
served as control treatments. The conetainers were placed on a mist bench (10
seconds misted water/15 minutes) until the vegetative growth of each species
was approximately 2.54 cm tall. Then, the conetainers were placed in a growth
chamber (15 or 30C, 18 hours of light/6 hours of darkness cycle). Each
conetainer was kept moist by topical applications of distilled water.
Inoculated rice plants were rated for disease severity using a modified
version of a previously reported assay (Wilkinson et al., 1985) using a
randomized block design. Each week, for a total of 4 weeks after being placed
in the growth chamber, 4 conetainers per treatment as well as 2 non-inoculated
control treatments were randomly removed from incubation, the roots washed
free of vermiculite, and the roots rated for disease severity as follows: (no
disease present, DS=0); (1-25% of roots with necrotic tissue, DS=1); (26 –
50% necrotic, DS=2); (51-75% necrotic, DS=3); (76-100% necrotic, DS=4).
All treatments were replicated 3 times as 3 independent biological replicates.
Statistical analysis of the data was performed by SAS statistical analysis
The Role of Melanin Production ...
9
software (SAS Institute Inc., Cary, NC, USA) using analysis over time and
standard deviations are given.
Inoculation of Rice Using a Petri-plate Assay
Rice was also inoculated using a Petri plate assay. Isolates used for the
plate assay were wild-type WT2+ and WT1+, and their respective melanin-
deficient counterparts M2- and M1-. Rice seeds were prepared by removing
the outer husk and surface disinfesting them in 3 percent aqueous solution of
hydrogen peroxide containing 100 ul of polyoxyethylenesorbitan (Sigma#
P139). The disinfesting solution was decanted and the seeds allowed to dry on
sterile paper towels. Three surface disinfested seeds were then placed at the
outer edges of 90 mm Petri plates containing 10 ml potato dextrose agar.
Plates were then placed under constant fluorescent lighting until germinated
roots were approximately 6 cm long. Using sterile technique, seedling roots
were inoculated with one 2-mm2 plug cut from the leading edge of an actively
growing colony cultured on PDA. The inoculum plug was placed in the center
of the Petri pate containing germinated seeds. Inoculated plates were then
placed under fluorescent lighting and monitored to determine when the fungus
intersected a root. Plants were harvested 28 days following the initial contact
between the fungus and the root. Samples were then embedded and sectioned
for microscopicobservation.
Wax Embedment and Sectioning of Rice Roots Harvested from
Conetainer Assay
Roots harvested from the conetainer assay were submerged in formalin-
acetic acid-alcohol (FAA) solution for 48 hours to fix both host and fungus
tissues. The fixed samples were dissected. Pieces (2 cm in length) of the main
root from the area closest to the inoculum were excised, initially dehydrated in
a graded water/ethanol series, and finally dehydrated a graded ethanol/xylene
series. Dehydrated samples were prepared for sectioning by infiltrating them
with molten Paraplast (Sigma# P3558) at 60 °C over a 24-hour time period.
The infiltrated samples were placed into hand folded cube-shaped tin-foil
molds, and embedded in molten paraplast. Thin section (10 µm) were then cut
with a hand-operated rotary microtome, floated on 7% formaldehyde solution
on gelatin (Sigma# G6144) coated slides, and incubated at 30 °C for 24 h.
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
10
Paraplast was removed from sections on slides by immersing them in several
changes of xylene until no paraplast was observed when viewed at 400 X.
Sections were stained by immersing slides in hematoxylin solution (0.2%
aqueous hematoxylin (Sigma# H3136), and 0.2% potassium iodide) for 2
hours, followed by rinsing briefly under gently flowing tap-water. Sections
were further stained by placing them in 1% aqueous Fast Green (Sigma#
F758) for 30 seconds, and rinsing briefly. Hematoxylin-Fast Green stained
slides were quickly dehydrated in an ethanol/water series (70:30, 95:5, 100:0;
20-30 seconds in each solution), dipped in xylene, and allowed to air dry.
Slides were then mounted in 3 drops of permount (Electron Microscopy
Sciences, 1560 Industry Rd., Box 550 Hatfield, PA 19440), covered with 80-
mm cover slips and allowed to dry overnight previous to examination with an
Olympus BH-2 compound microscope (40-100X).
Agarose Embedment and Sectioning of Rice Roots Harvested
from Petri-plate Assay
Roots from plants harvested from the Petri plate assay were excised and
embedded in molten 3% agarose contained in 2.5 ml cryovials (Sigma#
V9380). Following agarose solidification (about 15 minutes), the embedded
roots were removed from cryovial containers and hand-sectioned underwater
using a half of a double-edged razor blade under magnification (dissecting
microscope at 40x). Sections were placed on microscope slides, observed at
100-1000X, and digital images photographically captured. Samples were
observed either stained or not. Samples were stained by placing one drop of
Azure A stain (Sigma# A918, 1g Azure A/L in 95% ethanol) on them prior to
application of a cover slip.
RESULTS
The Effect of Melanin Hyphal Width
There was no significant difference in hyphal width between Ggg wild-
type isolates when cultured on LBA or V8A media (Table 2). There were
significant differences in hyphal width between the two wild-type Ggg’s
cultured on CDA where WT1+ was wider (27%) than WT2+. Wild-type
WT1+ and its corresponding melanin-deficient mutant (M1-) displayed
The Role of Melanin Production ...
11
significant differences in hyphal width on all tested media. Wild-type WT1+
hyphae were wider than M1- on CDA (19%) and V8A (28%) media. Melanin-
deficient mutant M1- was wider (26%) compared to WT1+ on LBA media.
Wild-type WT2+ and its’ corresponding melanin-deficient mutant M2-
showed differences in hyphal width. When cultured on CDA, M2- hyphae
were wider (24%) compared to wild-type WT2+. When grown on V8A, WT2+
had wider (19%) hyphae than M2-. There were no differences in hyphal width
between WT2+ and M2- cultured on LBA medium.
Table 2. Mean Hyphal width of wild-type and melanin deficient
Gaeumannomyces graminis var. graminis isolates
Isolate Contrasts
CDA
LBA
V8A
WT1+
3.90
2.89
3.63
WT2+
2.85
3.12
3.86
**
NS
NS
WT1+
3.90
2.89
3.63
M1-
3.16
3.90
2.61
**
**
**
WT2+
2.85
3.12
3.86
M2-
3.78
3.28
3.12
**
NS
**
Mean hyphal widths (um) were calculated using the datda from three separate
experiments. Each experiment was replicated 3x and repeated 3x (n=90). * and **
represent an alpha level of 0.05 and <0.001, respectively. NS = not significant.
Ten measurements of hyphal diameter were recorded by randomly selecting 10
different hyphae from within each of those randomly selected areas of a culture.
WT = wild type and M = mutant ;CDA= Czapek-Dox Agar, LBA= Luria-Bertani
agar, and V8A= vegetable juice agar.
The Effect of Melanin on Hyphal Distance between Branches
In general, the distance between hyphal branches was longer for wild-type
isolates than their respective mutants. Distances between hyphal branches
(DBB) were significantly different between the two wild-type isolates cultured
on each of the three media tested (Table 3). On CDA, WT1+ had an 18%
longer DBB than WT2+. On LBA and V8A, respectively, WT2+ exhibited
12% and 27% longer DBB, respectively, compared to WT1+. In comparing
the wild-type with their respective mutants, WT1+ displayed longer DBB on
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
12
both CDA (22%) and LBA (26%) media compared to M1-. On V8A medium,
WT1+ and M1- showed no differences in DBB. Isolate WT2+ exhibited a
significantly longer DBB on CDA (41%), LBA (50%), and V8A (67%)
compared to M2-.
Fungal Vegetative Growth in Different Cultures
All fungal isolates were grown on LBA medium displayed differences in
hyphal growth rates (Table 4). WT2+ (15.3 mm/day) grew faster than WT1+
(12.2 mm/day) on LBA. Isolate WT2+ grew faster (15.3 mm/day) than M2-
(0.5 mm/day) on LBA medium. M1- grew faster (14.6 mm/day) than WT1+
(12.2 mm/day).
Table 3. Mean Distance between branches of wild-type and melanin
deficient Gaeumannomyces graminis var. graminis
Isolate Contrasts
CDA
LBA
V8A
WT1+
64.88
67.72
59.32
WT2+
53.29
77.11
82.08
**
*
**
WT1+
64.88
67.72
59.32
M1_
50.38
49.96
57.54
**
**
NS
WT2+
53.29
77.11
82.08
M2-
31.45
38.86
27.14
**
**
**
Mean hyphal widths (um) were calculated using the data from three separate
experiments. Each experiment was replicated 3x and repeated 3x (n=90). * and **
represent an alpha level of 0.05 and <0.001, respectively. NS = not significant.
Ten measurements of hyphal diameter were recorded by randomly selecting 10
different hyphae from within each of those randomly selected areas of a culture.
WT = wild type and M = mutant ;CDA= Czapek-Dox Agar, LBA= Luria-Bertani
agar, and V8A= vegetable juice agar.
When fungal isolates were cultured on CDA medium, there were
significant differences in growth rate between the wild-type isolates and
between each wild-type isolate and their corresponding melanin deficient
mutant (Table 5). WT1+ grew significantly faster (10.8 mm/day) compared to
WT2+ (3.6 mm/day). Wild-type isolates grew significantly slower than their
The Role of Melanin Production ...
13
corresponding melanin deficient mutants. Isolate WT1+ grew at a slower (10.8
mm/day) than M1- (15.2 mm/day) when cultured on CDA. Wild-type WT2+
grew slower (3.6 mm/day) compared to M2- (5.3 mm/day) on CDA.
Table 4. Regression equations and contrasts of analysis of growth rate of
Geaumannomyces graminis var. graminis isolates cultured on Luria-
Bertani agar
Isolate
Regression equation
WT1+
y = 12.168x -3.968
M1-
y = 14.562x -1.696
WT2+
y = 15.326x -11.88
M2-
y = 0.502x -8.26
Contrasts of Growth Rates
p-value
WT1+ vs. WT2+
< .0001
WT1+ vs. M1-
0.0025
WT2+ vs. M2-
< .0001
Regression equations were generated from regression lines fitted to growth rate
(mm/day) of wild-type (WT1+ &WT2+) and melanin deficient mutant (M1- &
M2-) Gaeumannomyces graminis var. graminis. Experiments were repeated 3
times, with 3 repetitions per experiment. Alpha level = 0.05.
Table 5. Regression equations and contrasts of analysis of growth rate of
Geaumannomyces graminis var. graminis isolates cultured on Czapek-Dox
agar
Isolate
Regression equation
WT1+
y = 10.774x -3.447
M1-
y = 15.169x -7.016
WT2+
y = 3.61x + 0.103
M2-
y = 5.292x +3.192
Contrasts of growth rate
p-value
WT1+ vs. WT2+
< .0001
WT1+ vs. M1-
0.0333
WT2+ vs. M2-
< .0001
Regression equations were generated from regression lines fitted to growth rate
(mm/day) of wild-type (WT1+ &WT2+) and melanin deficient mutant (M1- &
M2-) Gaeumannomyces graminis var. graminis. Experiments were repeated 3
times, with 3 repetitions per experiment. Alpha level = 0.05.
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
14
Quantification of Melanin in Gaeumannomyces Graminis var.
graminis Hypha
Melanin concentration was calculated using changes in the optical density
at 610 nm of Azure A in solution. Extracted, purified, and concentrated Ggg
melanin had a binding coefficient of 873 µg Azure A/mg melanin. Both wild-
type isolates of Ggg showed no significant difference in melanin concentration
throughout the experiment (Figure 1). Wild type isolates and their
corresponding melanin deficient mutants were not significantly different in
melanin production at the first sampling(7days of growth). Melanin
concentration appeared to reach a maximum by day 14 for the wild-type
isolates (Figures 2 & 3). After 14 days of fungal growth, each wild-type had a
mean of 233 µg melanin/mg hyphae. Melanin deficient mutants M1- and M2-
displayed means of 65 and 33 µg melanin/mg hyphae, respectively (Figures 2
& 3).
Culture age (days)
714 21 28
Melanin concentration (ug melanin/mg hyphae)
0
100
200
300
400
500
Figure 1. Melanin concentration (ug melanin/mg hyphae) of wild-type WT1+ (●) and
WT2+ (○) Gaeumannomyces graminis var. graminis cultured in Luria-Bertani broth.
Each point represents the mean of 3 repeated trials with 3 replications per treatment.
Where standard error bars cross, there is no statistically significant difference in the
data.
The Role of Melanin Production ...
15
Culture age (days)
714 21 28
Melanin concentration (ug melanin/mg hyphae)
0
100
200
300
400
500
Figure 2. Melanin concentration (ug melanin/mg hyphae) of wild-type WT1+ (●) and
melanin-deficient M1- (○) Gaeumannomyces graminis var. graminis cultured in Luria-
Bertani broth. Each point represents the mean of 3 repeated trials with 3 replications
per treatment. Where standard error bars cross, there is no statistically significant
difference in the data.
Culture age (Days)
714 21 28
Melanin concentration (ug melanin/mg hypae)
0
100
200
300
400
500
Figure 3. Melanin concentration (ug melanin/mg hyphae) of wild-type WT1+ (●) and
melanin-deficient M2- (○) Gaeumannomyces graminis var. graminis cultured in Luria-
Bertani broth. Each point represents the mean of 3 repeated trials with 3 replications
per treatment. Where standard error bars cross, there is no statistically significant
difference in the data.
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
16
Disease Severity and Ectotrophic Colonization of Rice in
Conetainer Assay
Rice plants generated by conetainer assay were examined macroscopically
after 28 days of incubation (n=24). Plants inoculated with wild-type WT2+
were all severely diseased, with all rated a 4 for mean disease severity (MDS),
..., and those inoculated with melanin-deficient M2- were free of disease, with
all rated a 0 (MDS) for each experiment (n=24). Using the conetainer assay,
wild-type WT2+ displayed extensive ectotrophic colonization (Figure 4) of the
root epidermis at the time the roots were prepared for histopathological
observation (ca 28 days incubation)(Figure 4). The roots of plants inoculated
with the wild-type isolate were a uniform black color. Ectotrophic large
diameter runner hyphae (5um) were darkly pigmented, and branched
extensively forming a mantle of mycelia on the root epidermis. Plants
inoculated with Ggg isolate M2- (melanin-deficient) displayed no evidence of
lesions, or other symptoms commonly observed for root ectotrophic
colonization after 28 days of incubation (Figure 4). Plants inoculated with M2-
were indistinguishable from controls at 28 days of age. At the time they were
prepared for sectioning, roots of M2- inoculated rice plants exhibited no
observed mycelium on the outer surface of the roots, and the roots displayed
no symptoms. Uninoculated rice plants displayed no symptoms of disease or
discoloration (Figure 4). Roots were uniform in their appearance among
treatments.
Histopathological Observations of Rice Inoculated in Conetainer
Assay
Unstained roots inoculated with the wild-type WT2+ isolate exhibited
darkly pigmented runner hyphae on their epidermal surfaces. Darkly
pigmented hyphopodia were also detected on roots inoculated with WT2+.
Unstained root sections inoculated with WT2+ did not show infection hyphae
in the epidermis, cortex, or stele. Mycelia of the melanin-deficient isolate M2-
were not detected in any tissues of inoculated, unstained, sectioned, root
material (n=24). Unstained sections of controls did not show the presence of
fungi on or in any root tissues. Colonizing mycelium in inoculated root
samples were elucidated by staining with Hematoxylin/Fast Green. Stained
hypha were readily identified both ecto- and endotrophically. Dark runner
hyphae were easily detected without staining, and their appearance was
The Role of Melanin Production ...
17
enhanced by staining with hematoxylin (Figure 5). Hematoxylin stained runner
hyphae were a rich, dark brown color. Runner hyphae were also observed to
develop infection pegs on the host surface, and these exhibited the same
staining reaction (dark brown) as runner hyphae (Figure 5). Infection hyphae
of the wild-type fungus did not appear melanized, and were visible in the
epidermal, cortical, and vascular tissue of infected plants only after treatment
with fast green (Figure 5; n=24). Infection hyphae stained by fast green were
green/blue in color, and colonized root tissues in an intracellular manner
(Figure 5).
Figure 4. Rice inoculated with Gaeumannomyces graminis var. graminis using a
container system and incubated at 15C. Uninoculated controls (A.); wild-type
inoculated WT2+ (B.); melanin-deficient M2- inoculated rice (C.).
A.
B.
C.
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
18
A
B
Figure 5. Rice roots after inoculation with Gaeumannomyces graminis var. graminis
wild-type isolate WT2+. Longitudinal section stained with hematoxylin and fast green
(A.) shows infection pegs (IP); runner hyphae (RH); and infection hyphae (IH).
Longitudinal section stained with hematoxylin and fast green (B.) shows infection pegs
(IP); runner hyphae (RH); infection hyphae (IH), and xylem (X).
The Role of Melanin Production ...
19
The fungus appeared hyaline as it traversed the endodermis and entered
the stele. A color change indicating re-melanization of hyphae at the
endodermal tissue layer was not detected in roots inoculated with the wild-
type fungus after 28 days of incubation. Hematoxylin did not appear to cause a
staining reaction in infection hyphae as they traversed the endodermis and
entered the stele. Identification of infection hyphae was easiest in
longitudinally sectioned samples. After treatment with fast green, sectioned
samples of rice roots inoculated with M2- did not exhibit stained hyphae inter-
or intracellularly colonizing the epidermis, epidermal, cortical, or vascular
tissues. Sections produced from M2- inoculated roots (n=24) and un-
inoculated controls (n=4) were microscopically indistinguishable from each
other.
Figure 6. Rice inoculated with Gaeumannomyces graminis var. graminis: un-
inoculated (A); wild-type WT2+ (B); melanin-deficient M2- (C). Red circle denotes
border of fungal colony. Germinated seeds are labeled as “S”.
Disease Severity and Ectotrophic Colonization of Rice in Petri-
plate Assay
Rice plants inoculated with Ggg and incubated in the Petri plate assay
were extensively colonized by ectotrophic hyphae after 28 days (Figure 6). As
cultures grew across the Petri-plate, they grew over and obscured the rice roots
on the plate with a mycelial mat. Inoculated plants removed from the plates
were entirely mantled by the fungi and their roots were often embedded in the
agar. The hyphae formed a much denser mantle around the roots compared to
the colonization observed on roots cultured in the conetainer assay. Wild-type
isolates WT1+ and WT2+ produced ectotrophic mycelium that was melanized
and mantled the roots. Wild-type isolates (WT1+ & WT2+) caused
macroscopic lesions on black roots in inoculated plants that coalesced (n=18
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
20
each). Melanin-deficient mutant isolates M1- and M2- were also observed
colonizing the exterior root surface and forming a mantle. These hyphae were
hyaline. The densities of the mycelial mantle of wild-type and melanin-
deficient isolates were judged to be equivalent. Melanin-deficient isolate M1-
inoculated roots showed reddish colored lesions While M2- displayed no color
change or other symptoms (n=18).
Histopathological Observation of Rice inoculated in Petri-plate
Assay
Unstained rice roots inoculated with Ggg isolate WT1+ showed uniformly
light brown hyphae in the epidermal, cortical, and vascular tissues by 28 days
of incubation. Melanized runner hyphae were not observed to be produced by
WT1+ on the surface of inoculated roots cultured in the Petri plate assay.
Hyphae inside of plant roots appeared the same diameter (6 µm) in each tissue.
A change in hyphal coloration (darkening) indicating re-melanization of
WT1+ hypha when entering the stele was not detected in any unstained
sections (n=18). Unstained sections of WT2+ infected plant material
inoculated with wild-type isolate WT2+ showed that it had colonized all
tissues of the root at the time of sectioning (Figure 7; n=18). Melanized runner
hyphae were not observed to be produced by WT2+ on the surface of
inoculated roots cultured in the Petri plate assay. In unstained sections, WT2+
hypha were a uniform brown color in each tissue (epidermis, cortex and stele)
of an infected root (n=18). These hyphae appeared to be the same diameter (6
µm) in each tissue. Sections of rice roots indicated that WT2+ hyphae did not
change hyphal coloration throughout the process of pathogenesis. In addition,
a change in hyphal coloration (darkening) indicating re-melanization of WT2+
hypha when entering the stele was not detected in sections of unstained
inoculated roots (n=18). Unstained root sections of melanin-deficient mutant
M1- inoculated plants did not indicate the presence of hyphae in any of the
roots tissues. Unstained root sections of melanin-deficient mutant M1-
inoculated plants showed lignitubers being produced by the host in epidermal
and cortical tissues at 28D of inoculation. Unstained root sections of melanin-
deficient mutant M2- inoculated plants did not indicate the presence of hyphae
in any of the roots tissues.
Staining with Azure A allowed the fungus to be readily observed in
sections produced from plants in Petri-plate assay. Azure A stained the
anticlinal plant cell walls dark purple while fungal hyphae were stained with
The Role of Melanin Production ...
21
light purple color (Figure 8). Stained roots inoculated with Ggg isolate WT1+
showed light purple hyphae in the epidermal, cortical, and vascular tissues by
28 days of incubation (Figure 8). Hyphae inside plant roots appeared the same
diameter (6 µm) in each tissue. A change in hyphal coloration (darkening)
indicating re-melanization of WT1+ hypha when entering the stele was not
detected in any stained sections. Azure A stained sections of WT 1+
inoculated plants showed colonizing mycelia infecting the root in an
intracellular manner. Stained roots inoculated with Ggg isolate WT2+ showed
light purple hyphae in the epidermal, cortical, and vascular tissues by 28 days
of incubation. Hyphae inside of plant roots appeared the same diameter (6 µm)
in each tissue. A change in hyphal coloration (darkening) indicating re-
melanization of WT2+ hypha when entering the stele was not detected in any
stained sections.
Figure 7. Rice inoculated with wild-type Gaeumannomyces graminis var. graminis
isolate WT2+ and cultured in the Petri-plate assay. The transverse sections of rice roots
(A. & B.) were unstained. Root morphology is labeled as follows: Plant cell walls
(PCW), epidermis (E); cortex (C); stele (S); vascular bundles (V). Fungal hyphae are
labeled as “H”.
Azure A stained sections of WT 2+ inoculated plants showed colonizing
mycelia infecting the root in an intracellular manner. Hyphae appeared to be
the same diameter (6 µm) in each tissue. The melanin deficient isolate M1-
could infect and colonize the epidermal and cortical cells of the root by day 28
(Figure 9). The hyphae produced by M1- during infection were hyaline in
color and could not be detected without staining. Azure A stained the
anticlinal plant cell walls dark purple, and fungal hyphae were stained a light
purple color. Infective hyphae of M1- were produced intracellularly in the host
root. At 28 days of incubation, the melanin-deficient isolate M1- was able to
infect epidermal and cortical cells, but was stopped prior to entering the stele
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
22
by l]lignified host tissue (lignitubers) around its hyphae (Figure 9; n=18). The
hyphae produced by M2- during infection were hyaline in color and could not
be detected without staining. Azure A stained the anticlinal plant cell walls
dark purple, and fungal hyphae were stained a light purple color. The melanin-
deficient isolate M2- was rarely able to infect single epidermal cells
intercellularly (Figure 10; n=1/18).
Figure 8. Rice inoculated with wild-type Gaeumannomyces graminis var. graminis
isolate WT1+ and cultured in the Petri-plate assay. The transverse sections of rice roots
(A. & B.) were stained with Azure A. Plant cell walls (PCW) are dark purple, fungal
hyphae (H) are light purple. Root morphology is labeled as follows: epidermis (E);
cortex (C); stele (S); vascular bundles (V).
DISCUSSION
Nature of Mutations Used (Pigmentation and Morphology)
Both melanin deficient mutants (M1- and M2-) were produced from their
corresponding wild-type parents using 4-nitroquinolene-1-oxide (NQO) as the
mutagenic agent. Nitroquinolene oxide is an electrophile and a powerful
carcinogen and mutagen (Sugimura, 1981). It mimics the mutagenic action of
ultraviolet light and forms charge-transfer complexes with 5'-
deoxyribonucleotides (Winkle & Tinoco, 1979). Nitroquinolene oxide (NQO)
forms DNA adducts and can cause a wide range of DNA “lesions” including
single-strand breaks, pyrimidine-dimer formation, abasic sites, and oxidized
bases. In bacteria and yeast NQO has been shown to be a base substitution
mutagen acting at guanine residues, inducing mainly guanine to adenine
transitions (Fronza et al. 1992). The genetic basis for the pigmentation
(melanin) changes in both M1- and M2- have not been determined. Due to the
nature of NQO chemical mutagenesis, mutations in addition to conferring
The Role of Melanin Production ...
23
changes in pigmentation could have occurred, but have not been identified or
characterized. For instance, the melanin mutant produced by Frederick et al.
(1999) also exhibited hyphopodia and other morphological variation different
from those of their parent cultures. Epstein et al. (1994) reported mutants that
differed not only in hyphopodia, but pigmentation, compared to the parent
culture. While the pigmentation mutations for these isolates have been
characterized, it is unknown whether they possess additional DNA mutations.
Epstein’s isolates were selected from either Benomyl (2 of 1000
transformants) or Phleomycin (1 of 42 transformants) resistant transformants.
The hyphopodial mutation was an artifact of the transformation process, not
the main goal. Analysis of the transformants indicated that there was a single
insertion in each case, but the exact location of the insertion in the
transformants was not described. Bal et al. (1977) reported that NQO is a
“good” mutagen for Aspergillus nidulans (Eidam) Winters because it induces
mutations at a high frequency (0.5% of treated cells) and generates a broad
spectrum of morphological and physiological changes.
Figure 9. Rice inoculated with melanin-deficient Gaeumannomyces graminis var.
graminis isolate M1- and cultured in the Petri-plate assay. The transverse sections of
rice roots (A., B., C., & D.) were stained with Azure A. Plant cell walls (PCW) are
dark purple, fungal hyphae (H) are light purple. Root morphology is labeled as
follows: lignituber (L); epidermis (E); cortex (C); stele (S); vascular bundles (V).
Fungal hyphae are labeled as “H”.
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
24
Figure 10. Rice inoculated with melanin-deficient Gaeumannomyces graminis var.
graminis isolate M2- and cultured in the Petri-plate assay. The transverse sections of
rice roots (A. & B.) were stained with Azure A. Plant cell walls (PCW) are dark
purple. Root morphology is labeled as follows: epidermis (E); cortex (C); stele (S);
vascular bundles (V). Fungal hyphae are labeled as “H”.
Our mutant cultures were observed to be stable for melanin content for 36
months. M1- is similar to the hyphopodial mutant reported by Epstein et al.
(1994). M1- (generated by Frederick et al. 1999) exhibited an increased
frequency in production of lobed, melanized hyphopodia on the bottom of the
polystyrene culture dish when cultivated on solid media as compared to its
WT1+ parent culture. M2- is hyaline, and did not produce hyphopodia when
cultivated on solid media, while its corresponding parent culture WT2+ is
heavily pigmented and produces lobed hyphopodia under these conditions.
M2- formed only simple hyaline hypal tips: M1- had hyaline mycelia and
lobed, pigmented hyphopodia. Further, it appears that melanin production
could be genetically segregated within a thallus of G. graminis. This
segregation appears to be non-temporal as M2- demonstrated stable hyaline
mycelia and hyphopodia for at least 36 months. Such genetic segregation
could be based in the poly-nucleic nature of ascomycetes or in the N + N
status of the thallus if melanin production is a dominant trait. Finally, it is very
interesting that WT2+, a true Ggg, when transformed to M2- appears
morphologically like Gaeumannomyces graminis var. avenae (Gga) or
Gaeumannomyces graminis var. tritici (Ggt). This raises the question about the
relationship between Gga, Ggt and Ggg. Fouly et al. (1997) showed genetic
dissimilarities between these groups, but there was also a great deal of genetic
similarity. Are these isolates really all Ggg sub-species, with Gga and Ggt
lacking some of the functional genes of Ggg?
Isolate WT2+ is a wild-type isolate of Ggg from bermudagrass (Table 1)
and a pathogen of this host. WT1+ and WT2+ both produced about the same
The Role of Melanin Production ...
25
amount of melanin in their respective hyphae. Therefore differences in their
behavior as reported here (growth, hyphal width and DDB), would not be
expected to be assigned to melanin content.
Effect of Melanin on Fungal Hyphal Morphology and Vegetative
Growth Rate
Measurements of hyphal width were used as a means to evaluate the effect
of melanin content on hyphal morphology. The basis for using hyphal width is
that melanin, a wall component, is suspected of imparting a more rigid wall
structure, which could allow for higher turgor pressure within a hypha.
Further, it has been reported (Skou, 1981) that root-infecting hypha are
melanin-less and smaller in diameter than melanized, ectotrophic hypha. This
suggests that Gaeumannomyces species have melanin regulatory mechanisms
that are environmentally sensitive. While infectious hypha appear to be devoid
of melanin, it is unclear if they still are producing low levels of this pigment.
In general, the WT isolates developed hypha with similar widths, except
when grown on CDA. The basis of this difference could reflect the
heterogenous nature of the Ggg isolates as reported by Fouly and Wilkinson
(2000). More interestingly, there were significant differences between the WT
isolates and their respective melanin-deficient mutants in terms of hyphal
width. For both isolate couplets, the WT generally displayed larger diameter
hypha than the corresponding mutant. However, there were some
inconsistencies in this pattern when considering behavior in different media.
However the general trend toward larger hypha with melanin suggests that
melanin may in fact allow hypha to grow larger while fungal walls deficient in
melanin will support smaller diameter hypha. The smaller diameter of hypha
for melanin-less isolates reported here supports observations that melanin-less
infecting hypha are also smaller diameter. Epstein et al. (1994) reported that
wild-type and corresponding single gene insertion melanin/hyphopodial
mutants did not show a differences in hyphal width when cultured on dilute
V8A. A reason for this difference compared to our measurements could have
resulted from different experimental growth conditions. Epstein cultured
isolates under thin layers of diluted V8A and measured hyphal widths after the
agar layer was removed from the hypha. Both the dilution of the V8A and the
sub-agar culturing could affect the osmotic and hydration conditions of the
fungal environment resulting in moisture limitations. If melanin functions to
control osmotic potential and/or allows for greater endogenous turgor pressure,
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
26
then in a dilute osmotic medium, the loss of melanin would not be expected to
correspond to a reduction in hyphal width. The comparison of our results with
those of Epstein et al. (1994) are further limited by the fact that the width of
wild-type and mutants reported were 6.4 +/- 0.05 µm whereas we reported that
the wild-types here were 3.9 and 3.6 µm on V8A and the mutants were 3.6 and
3.1 µm, respectively.
The primary site of nutrient uptake for a fungal colony is at the thin-
walled hyphal tip (Sietsma et al. 1995). This is also the only site where the
fungal mycelium is actively elongating (growing) using a complicated physio-
chemical process that involves hydrostatic pressure generated and controlled
by osmotic regulation. The number of hyphal tips for a given thallus is
determined by the frequency that hyphal branches are formed.
Gaeumannomyces is characterized as having both septa, and branches. Septa
are generally intercalary to the branches. In general, fungi form more branches
when exposed to optimal growth conditions. The frequency of branching or
distance between branches (DBB) is, in part, dependent on the physical-
chemical nature of the medium, and the genetic-based ability to exploit it for
growth and development (Rayner et al. 1994). As DBB increases, there are
fewer hyphal tips being produced per thallus. Our results showed that WT2+
and WT1+ did not form branches at the same rate and that each formed
branches at variable rates depending on the growth medium. To determine the
range of DBB among Ggg isolates, a large population of isolates would be
required for comparative purposes. However, the DBB data for WT and their
corresponding mutants did show that the loss of melanin resulted in
significantly shorter DBB compare to WT isolates. Further, this behavioral
pattern was only slightly affected by WT1+ in V8A medium. Epstein et al.
(1994) reported that wild-type Ggg and corresponding single insertion mutants
also displayed no differences in hyphal DBB when cultured on dilute V8A.
Their work focused on the differentiation and pigmentation of hyphal tips to
form hyphopodia. Their primary objective focused on the frequency, shape,
color and stimuli of hyphopodia. However their mutants all produced melanin,
although in different degrees. For example, isolate JH849 produced melanized
hyphopodia, but not in “sufficient quantity.” Upon further examination of their
work, JH849, did produce a reduced, but unquantified amount of melanin, and
was the only mutant reported to branch about half as often (DBB = 158 µm)
compared to wild-type or two other mutants of Ggg (DBB = 97, 96, and 71
µm, respectively). Mutant isolate JH2982, produced as much or more melanin
than the wild-type (JH2033) and exhibited the shortest DBB (71 µm). While
not statistically tested, it would appear from Epstein’s work, that both hyphal
The Role of Melanin Production ...
27
width and DBB were affected similarly by reductions in melanin content of the
hypha compared to the M1- and M2- mutants used in this study.
The three media (CDA, V8 and LBA) that were used in this research were
also used by other researchers that investigated melanin and its role in Ggg
morphology (Epstien et al. 1994; Frederick et al. 1999; Money et al. 1998).
These media are among the most commonly used for fungal cultivation and in
particular for culturing Gaeumannomyces. Each isolate was cultured and
evaluated for hyphal width and distance between branching when grown on
LBA and CDA. In general, as medium type became more defined, wild-type
isolates produced narrow hyphae and with a shorter distance between
branches. There was no single factor that could be attributed to the effect of
medium type, melanin, and growth rate, though the data indicate that in
general the more defined the medium, the slower the growth rate of isolates.
The reported effect of melanin on the vegetative growth rate of
Gaeumannomyces is variable. High levels of DNH-melanin have been
implicated in limiting the uptake of nutrients by mycelium (Frederick et al.
1999; Henson et al. 1999). A constitutive melanin producing Ggg mutant
exhibited slower radial growth in culture than a melanin minus mutant and the
corresponding wild-type (Frederick et al. 1999). In addition, the melanin
deficient mutant exhibited a faster growth rate in culture as compared to the
wild-type isolate of Ggg. Epstein et al. (1994) showed that wild-type and
melanin/hyphopodial mutant Ggg’s, when cultured on dilute V8A on glass
slides did not have different growth rates. However, the culture conditions (see
above) and the incompleteness of melanin disruption preclude this work from
being considered definitive in terms of the impact of melanin on vegetative
growth rate.
In data presented here, melanin deficient M1- was the fastest growing
isolate, growing faster than both wild-type isolates. However, the melanin
deficient isolate, M2-, was the slowest growing isolate under most tested
conditions. This phenomenon could be due to the presence of melanin as
evidenced in the pigmented hyphopodia of this isolate. Still, the effect of
melanin on growth rate based on the data collected is not definitive. Wild-type
and melanin deficient isolates displayed variable responses when tested on
different media. The pair consisting of WT1+ and M1- generally showed the
melanin deficient M1- growing faster on different media compared to WT1+.
The pair consisting of WT2+ and M2- showed the melanin deficient strain
growing slower on all but the CDA medium. In the case of CDA medium, the
mutant strain grows slightly faster than its melanin producing parent. The
complexity of the media combined with the likely multiple mutations in both
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
28
M1- and M2- preclude assigning any role of melanin in the determination of
growth rate. Single insertion mutants with disrupted genes involved in the
DHN-melanin synthesis pathway along with testing isolates for growth rate
using defined media could allow for a determination of the role of melanin in
vegetative growth rate of Ggg.
Melanin Quantification
The melanin content of the wild-type and mutants reported here was
measured using an indirect method (Butler & LaChance, 1986). The binding
of solubilized Azure A dye to hyphal melanin was used to estimate the
concentration of melanin in hyphae. Wild-type isolates were not significantly
different from each other in hyphal melanin content. The mutants M1- and
M2- both showed very little melanin per unit mass of mycelium and the
melanin content of WT isolates were considerably higher than the mutants.
Melanin concentration reached a constant value after 14 days in culture. Wild-
type WT1+ (JH2033) was also analyzed for melanin by Frederick, et al. (1999)
who reported a concentration of 155 µg melanin/mg hyphae. The melanin
concentration of the wild-type Ggg isolates reported here were 250 and 115 µg
melanin/mg hyphae, respectively, thereby supporting the use of the Azure A
method for mycelial melanin determination. While the mutants M1- and M2-
reported here produced significantly less melanin compare to their respective
parent cultures, they did produce an average of 65 and 30 µg melanin/mg
mycelium respectively at 14-28 days of age according to the Azure A melanin
assay. Bell and Wheeler (1986) and Frederick et al. (1999) also reported that
melanin-deficient mutants of Ggg showed a similar degree of Azure A binding
when compared to M1- and M2- reported here. This can perhaps be explained
by cell wall components other than melanin absorbing some of the Azure A,
but not an amount comparable to the binding coefficient of melanin.
Hyphopodia were not produced by liquid-grown isolates, and this precludes
the affect the melanin status of these appendages could produce in the assay.
Furthermore, when living cultures of M1- and M2- were stained with Azure A
and examined at 400X, both the cell walls and the cytoplasm absorbed some
Azure A. We also immersed Pythium aphanidermatum (Edson) Fitzp. cultures
in Azure A solution, and it absorbed stain as well. The reason for this may be
due to the cell walls of Pythium being composed primarily of beta-glucans
and cellulose, not chitin (a polymer of N–acetylglucosamine), as in
The Role of Melanin Production ...
29
filamentous fungi. These compounds may absorb Azure A to a greater
extent than chitin. The melanized yeast, Phaeococcomyces, was reported to
bind Azure A to melanin located in its cell walls (Butler & LaChance, 1986).
Melanin-deficient mutants of this yeast did not take up stain at the cell wall,
though dead or impaired cells showed staining.
The issues dealing with melanin as a determinant of Ggg morphology and
vegetative growth reported here give a strong indication that melanin is
important to the basic growth and development of the fungus. To further test
this using a more definitive approach, single DNA insertion into one or more
of the enzymes of the DHN melanin pathway should be used. Using these
defined mutants would also allow non-specific binding of Azure A to be
assigned to mycelial components other than melanin.
The Effect of Melanin on the Histopathology of
Gaeumannomyces Graminis
The location and quantification (+/-) of hyphal melanin in sectioned
tissues during pathogenesis was difficult to determine. Melanized runner
hyphae were readily seen on rice plants inoculated with a wild-type isolatein
the conetainer assay. In the conetainer assay, the wild-type exhibited no
discernable melanization of hyphae when it had gained entry to the plant. Re-
melanization of hyphae prior to infection of the stele, as seen by Wilkinson
(personal communication) in regard to Ggt was not observed. Yet, it is
common knowledge that Gaeumannomyces form melanized hyphae in necrotic
plants at the later stages of pathogenesis. One aspect not included in these
studies was induction ofphysical/physiological stress applied to the host. It is
possible that had heat or drought stress been applied to the colonized host, the
hyphae might have reacted differently.
Staining of fungal elements (macro and micro-hyphae, infection pegs,
hyphopodia) was investigated though several methods in order to determine if
stains would enhance observation of fungal melanin. Some staining procedures
were not useful due to their interference with the visual detection of melanin.
Melanin specific stains such as Masson-Fontana and Schmorl’s used to stain
these samples precluded enhanced observations of melanin deposition on
hyphae because they stain all fungal tissues black and therefore obscure
melanin. This was also reported by Masatomo et al. (1998) and Gupta et al.
(1985). Periodic acid-Shiff staining was attempted to contrast melanized and
non-melanized hyphae, but was discarded as it stained all tissues bright
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
30
crimson and did not enhance the visualization of hyphal melanin. Azure A, a
melanin specific stain, was also found not to be useful in selectively staining
for melanin. Azure A stained plant tissues as well as fungal elements and
proved valuable for enhancing observation of the fungus in rice roots
inoculated in Petri-plate assay. This is interesting because infection hyphae are
hyaline in appearance. It was recorded (see Chapter 1) that melanin-deficient
strains of Ggg can show a degree of Azure A binding. The staining of hyaline
infection hyphae by Azure A appears to support this finding. Fluorescence
microscopy was also investigated. Melanin will fluoresce if oxidized (Kyatz et
al., 2001). It was postulated that by inducing fluorescence, melanin could be
pinpointed as to where it was being deposited by the fungus. Unfortunately,
oxidizing melanin with hydrogen peroxide to enhance visualization was not
useful because plant tissues fluoresced brightly as well, and melanin in hyphae
could not be discerned.
It has been stated that there are three phases for fungal nutrition during
plant infection (Solomon et al, 2003). These phases are germination,
proliferation, and sporulation. These phases can be compared to the seven
stages of pathogenesis and are useful as a context for studying the production
of melanin by Ggg in situ. The first, or germination, phase can be compared to
melanized Ggg hyphae growing towards a host from the tissues they have
overwintered upon (dissemination, stage 2). During this time, external nutrient
sources are likely to be in short supply, and the fungus is at a disadvantage to
compete with host defenses and competitive microorganisms. The production
of melanin at this time may assist the pathogen in survival during inoculation
(stage 3) and pre-penetration (stage 4), penetration (stage 5), and infection,
(stage 6). The second phase, proliferation, takes place after the fungus has
gained entry to the host. At this stage, nutrients are not limiting, and
competing microorganisms will be reduced, if not eliminated, compared to
stage one. After this stage, compatibility (stage 7) is determined, and melanin
may not be necessary for the advance of the fungus throughout the rest of the
host. At this time, the production of melanin may either hinder the progress of
disease or be unnecessary for pathogenesis. This may be why Ggg is at first
hyaline when it invades the cortex and stele of infected plants. The production
of spores (sporulation, stage 3) by Ggg is not thought to be a major
determinant in the spread of Gaeumannomyces (Skou, 1981). Instead, the re-
melanization of hyphae in host tissues post-mortem is most likely most
comparable to this stage of pathogenic nutritional requirements. At this point,
the protective role melanin biosynthesis serves may be necessary for the
pathogen in order for it to complete the disease cycle. In this case, melanin
The Role of Melanin Production ...
31
will protect the fungus as it competes with other organisms in the soil for
nutrient sources, and aid in the overwintering process, thus completing the
disease cycle. This is why it remains to be determined when Ggg resumes
melanin production between colonization of the stele and necrosis.
Effect of Melanin on Fungal Pathogenicity
Pathogenicity of Ggg isolates was affected by their ability to produce
melanin. It has been shown that melanin is necessary for Ggt to cause disease
(Kelly, 1997). Gaeumannomyces graminis var. tritici may have to re-melanize
upon encounter of the stele in order to penetrate vascular tissue (disease),
while Ggg may not. Frederick et al. (1999) showed that melanin was not
necessary for Ggg to cause disease in rice, but did not examine infected plants
histopathologically to determine melanin status of infective hyphae, or where
they were produced. Infected plants may have been colonized by the fungi
they were inoculated with, and not actually diseased. Here, melanized wild
type isolates were pathogenic, and isolates lacking in melanin were reduced in
their ability to infect and colonize the host. Melanin can bind to and inactivate
chemical agents (Hill, 1992). The inability of melanin deficient strains to
colonize all tissues of host roots (parasitic vs. pathogenic) may be a result of
failure to resist host-defense chemical mechanisms. This phenomenon has
been described in rice (Datta et al., 2001; Nishizawa et al., 1999). Also, since
melanin has an effect on hyphal turgor pressure (Brush and Money, 1999), it
may be that non-melanized isolates could not produce the necessary
mechanical force to penetrate through the lignified host response (lignitubers)
in the epidermal and cortical layers. In contrast to this, melanized isolates were
able to infect all tissues of rice roots by 28 days. Genetic analysis of the
mutations in the mutant isolates would help to characterize the mutation of the
mutant Ggg strains and perhaps their influence on the infection process.
It is important to point out that there were differences in the way the
fungus behaved between the conetainer and plate assay. In conetainer assay,
the wild-type produced runner hyphae, infection pegs, and non-melanized
infection hyphae. This differentiation of hyphae is consistent with those found
in “naturally” infected plants in the field. In Petri plate assay, only one type of
hyphae was seen. These hyphae were distinguishable from those that
proliferated on the culture medium only by the fact that they were produced in
the roots of plants. Some researchers (Frederick et al., 1999) have stated that
since melanin serves in a protective role for hyphae, unmelanized strains
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
32
would be at a disadvantage in a natural setting. This may be true since M2- did
not infect in conetainer assay, where the environment is not as conducive to
the survival of the fungus as it is in culture (Petri-plate assay). While melanin
deficient (M1- & M2-) Ggg strains were able to parasitize roots in plate assay,
designation of these fungi as parasites may be correct only in an environment
where the host is at an extreme disadvantage to repel the fungus. In a more
natural setting, it appears that melanin may be required not only as a
determinant of pathogenism, but for survival outside of culture.
This research indicates that the common rating system of virulence (i.e.,
darkening of roots) may not be adequate for assessing whether a particular
fungus is a parasite or a pathogen. Rating roots for infection by Ggg based
solely on epidermal darkening may not illustrate the actual relationship
between host and fungus. When acting as parasites, melanin-deficient strains
did not induce characteristic blackening of host roots. This interaction of
host/pathogen would not be properly characterized by using this traditional
scale.
It is undisputed that the fungus will at some point re-melanize in host
tissues (Skou, 1981.). This has been previously documented in the fact that
wild type fungi will often develop “cessation structures” if they are unable to
advance past a certain point in host roots. This is often seen in another
melanized monocot root pathogen, Magnaporthe poae. These structures are
also produced, seemingly at random, in compromised host roots by Ggg
(Skou, 1981). While their function remains unknown, it is postulated that they
may function as survival structures (overwintering) for the fungus in temperate
zones (Skou, 1981). It may be that the time-frame involved in both assays
reported here are too short for this phenomenon to be evidenced and that re-
melanization takes a longer amount of time than we used. Frederick et al.
(1999) reported that a non-melanized mutant derived from WT1+ (JH4300)
was pathogenic 28 days post-inoculation. Unfortunately, this isolate is no
longer living and thus is precluded from our studies. It should be noted that
their conditions for host/pathogen interaction was rated by root dry weights, at
a time of incubation at 25C for 28D. Melanized wild-type isolates were
capable, under our conditions, of colonizing the stele (i.e., causing disease) by
28 days and melanin-deficient isolates were not. It may be that after six
months the non-melanized isolates might be able to compromise host defenses,
but the apparent advantage that melanin confers to the fungi that synthesize it
are clear: melanin assists in the ability of Ggg to cause disease.
The objective of this research was to determine if melanin plays a role in
the pathogenicity of Ggg. The evidence presented here indicates that the
The Role of Melanin Production ...
33
ability of wild-type isolates to produce melanin has a marked effect on their
ability to produce disease symptoms on rice roots. This effect on disease
severity caused by Ggg is uniform over the plants evaluated. Melanin
producing wild-types caused extensive root rotting in most host roots, while
melanin-deficient strains did not. Melanin-deficient strain WT1- caused slight
discoloration of inoculated oats, wheat and rice roots (data not presented). This
differed from the coloration caused by wild-type isolates. Wild-type Ggg
caused characteristic blackening of host roots. In some cases, wild-types were
able to completely kill and rot host roots into an unrecognizable state. In
contrast, melanin-deficient strain WT1- induced a light brown, rather than
black color to inoculated roots, a phenomenon especially noticed in rice. These
lesions macroscopically appeared to be limited in their deleterious effect on
the roots (Wilkinson, personal communication).
The effect of melanin on the pathogenicity of Gaeumannomyces graminis
var. graminis based solely on symptoms of host roots as indicated by this
research is that melanin is required by Ggg to induce disease symptoms in host
roots. This research shows a dependence on melanin by Ggg in order to cause
characteristic blackened and rotted roots in host plants. In conetainer assay
conducted with melanized wild-type fungi and their melanin-deficient
counterparts, there was a significant difference in the ability of wild-type and
melanin-mutant isolates to colonize rice plants.
REFERENCES
Bal, J., Kajtaniak, E.M., & Pieniazek, N.J. 1977. 4-Nitroquinoline-1-oxide: a
good mutagen for Aspergillus nidulans. Mutat Res Fund Mol Mech
Mutagen. 56: 153-156.
Bell, A. A., and Wheeler, M.H. 1986. Biosynthesis and Functions of Fungal
Melanins. Ann. Rev. Phytopathology. 24: 411-51.
Brush, L, and N.P. Money. 1999. Invasive hyphal growth in Wangiella
dermatitidis is induced by stab inoculation and shows dependence upon
melanin biosynthesis. Fungal Genetics and Biology 20:190-200.
Butler, M.J., and Lachance, M.A. 1986. Quantitative binding of azure A to
melanin of the black yeast Phaeococcomyces. Exp. Mycol. 10: 166-170.
Caesar-Tonthat, T., Kloeke, F.V.O., Geesey, G.G. and Henson, J.M. 1998.
Melanin production by a filamentous soil fungus in response to copper
H. Fouly, Sh. Henning, O. Radwan, H. Wilkinson et al.
34
and localization of copper sulfide by sulfide-silver staining. Appl. Environ.
Microbiol. 61: 1968-1975.
Datta, K., J.M. Tu, N. Oliva, I. Ona, R. Velazhahan, T.W. Mew, S.
Muthukrishnan, and S.K. Datta. 2001. Enhanced resistance to sheath
blight by constitutive expression of infection-related rice chitinase in
transgenic elite indica rice cultivars. Plant Science v. 160, p. 405-414.
Elliot, M.L. 1991. Determination of an etiological agent of Bermudagrass
decline. Phytopathology. 83: 414-418.
Elliot, M.L., Hagan, A.K., and Mullen, J.M.. 1993. Association of
Gaeumannomyces graminis var. graminis with St. Augustine grass root
rot disease. Plant Dis. 77: 206-209.
Epstein, L., Kaur, S., Goins, T., Kwon, Y.H., and Henson, J.M. 1994.
Production of hyphopodia by wild-type and three transformants of
Gaeumannomyces graminis var. graminis. Mycologia. 86: 72-81.
Fouly, H. M., Wilkinson, H. T. & Chen, W. 1997. Restriction analysis of
internal transcribed spacers and the small subunit gene of ribosomal DNA
among four Gaeumannomyces species. Mycologia. 89: 590-597.
Fouly, H.M., & Wilkinson, H.T. 2000. Detection of Gaeumannomyces
graminis varieties using polymerase chain reaction with variety-specific
primers. Plant Dis. 84, 947-951.
Frederick, B., T. Caesar-TonThat, M.H. Wheeler, K.B. Sheehan, W.A. Edens,
and J.M Henson. 1999. Isolation and characterization of
Gaeumannomyces graminis var. graminis melanin mutants. Mycological
Research 103: 99-110.
Fronza, G., Campomenosi, P., Iannone, R., & Abbondandolo, A. 1992. The 4-
nitroquinoline 1-oxide mutational spectrum in single stranded DNA is
characterized by guanine to pyrimidine transversions. Nucleic acids res.
20, 1283-1287.
Hawksworth, D.L., Kirk, P.M., Sutton, B.C., & Pegler, D.N. (Editors).1995.
Ainsworth and Bisby's Dictionary of the Fungi. CAB International,
Wallingford.
Henson, J., Butler, M.J., & Day A.W. 1999. The dark side of the mycelium:
melanins. Annu. Rev. Phytopathology. 37: 47-71.
Hill, H.Z. 1992. The Function of melanin or six blind people examine an
elephant. BioEssays. 14: 49-56.
Hornby, D. 1998. Take-all Disease of Cereals, a Regional Perspective. CAB
International, New York. 384 pp.
Jones, D.G. & Clifford, B.C. 1978. Cereal Diseases: Their Pathology and
Control. John Wiley & Sons, New York.
The Role of Melanin Production ...
35
Kelly, C. 1997. Genetics of pigmentation and pathogenicity in G. graminis.
PhD thesis, University of Birmingham, UK.
Money, N. P., Caesar-Ton That, T., Frederick, B., & Henson, J. M. 1998.
Melanin synthesis is associated with changes in hyphopodial permeability,
wall rigidity, and turgor pressure in Gaeumannomyces graminis var.
graminis. Fungal Genet. Biol. 24: 240-251.
Nishizawa, Y., Z. Nishio, K. Nakazono, M. Soma, E. Nakajima, M. Ugaki,
and T. Hibi. 1999. Enhanced resistance to blast (Magnaporthe grisea) in
transgenic Japonica rice by constitutive expression of rice chitinase.
Theoretical Applied Genetics v. 99 p. 383-390.
Rayner, A.D.M., Griffith, G.S., & Ainsworth, A.M. 1994. Mycelial
Interconnectedness. in The Growing Fungus. Edited by N.A.R. Gow
&.G.M. Gadd Chapman & Hall, London. pp. 21-37.
Sietsma, J.H., Wosten, H.A.B., & Wessels, J.G.H. 1995. Cell wall growth and
protein secretion in fungi. Can. J. Bot., 73: 388-395.
Skou, J.P. 1981. Morphology and Cytology of the Infection Process. in Biology
and Control of Take-all. Edited by M.J.C. Asher & P.J. Shipton.
Academic Press, London. pp. 175-196.
Solomon , P. S., Tan, K. C. & Oliver, R. P. 2003. The nutrient supply of
phytopathogenic fungi; a fertile field for study. Mol Plant Pathol 4: 203–
210.
Sugimura, T. 1981. Carcinogenesis: A Comprehensive Survey. Vol. 6. Raven,
New York.
Walker, J. 1981. Taxonomy of take-all fungi and related genera and species. in
Biology and Control of Take-all. Edited by M.J.C. Asher & P.J. Shipton.
Academic Press, London. pp. 15-74.
Wilkinson, H.T. 1994. Root rot of centipedegrass (Eremochloa ophinroides
(Munro) Hack.) caused by Gaeumannomyces graminis ((Sacc.) Arx &
Olivier) var. graminis. Plant Dis. 78: 1220.
Wilkinson, H.T., & Kane, R.T. 1993. Gaeumannomyces graminis infecting
zoysiagrass in Illinois. Plant Dis. 77: 100.
Wilkinson, H.T., J.R. Alldredge, and R.J. Cook. 1985. Estimated distance for
infection of wheat roots by Gaeumannomyces graminis var. tritici in soils
suppressive and conducive to take-all. Phytopathology 75: 557-559.
Winkle, S. A. & Tinoco, I., Jr. 1979. Interactions of 4-nitroquinoline 1-oxide
with four deoxyribonucleotides. Biochemistry 18: 3833-3839.
