Plant Physiol. (1999) 119: 417-428
Cell-Specific Production and Antimicrobial Activity of
Naphthoquinones in Roots of Lithospermum
erythrorhizon1
Lindy A. Brigham2, *,
Paula J. Michaels, and
Hector E. Flores
Department of Plant Pathology, The Pennsylvania State University,
University Park, Pennsylvania 16802-4507
 |
ABSTRACT |
Pigmented
naphthoquinone derivatives of shikonin are produced at specific times
and in specific cells of Lithospermum erythrorhizon roots.
Normal pigment development is limited to root hairs and root border
cells in hairy roots grown on "noninducing" medium, whereas
induction of additional pigment production by abiotic (CuSO4) or biotic (fungal elicitor) factors increases the
amount of total pigment, changes the ratios of derivatives produced, and initiates production of pigment de novo in epidermal cells. When
the biological activity of these compounds was tested against soil-borne bacteria and fungi, a wide range of sensitivity was recorded. Acetyl-shikonin and
-hydroxyisovaleryl-shikonin, the two
most abundant derivatives in both Agrobacterium
rhizogenes-transformed "hairy-root" cultures and
greenhouse-grown plant roots, were the most biologically active of the
seven compounds tested. Hyphae of the pathogenic fungi
Rhizoctonia solani, Pythium
aphanidermatum, and Nectria hematococca induced
localized pigment production upon contact with the roots. Challenge by
R. solani crude elicitor increased shikonin derivative
production 30-fold. We have studied the regulation of this suite of
related, differentially produced, differentially active compounds to
understand their role(s) in plant defense at the cellular level in the
rhizosphere.
 |
INTRODUCTION |
Plants communicate with their environment (the soil, climate
conditions, neighboring plants, microorganisms, insects, etc.) by
producing a diverse array of chemicals. These secondary metabolites are
often characteristic of specific plants and plant families. Many
members of the Boraginaceae family produce naphthoquinones in their
roots. Naphthoquinones are colored substances derived from
phenylpropanoid and isoprenoid precursors (Gaisser and Heide, 1996
).
Plants of the borage family are distributed worldwide and the
naphthoquinones from many of these plants have been used in diverse
cultures as colorants for cosmetics, fabrics, and foods (Jain and
Mathur, 1965
; Ballantine, 1969
; Tabata and Fujita, 1985
), and for
medicinal applications, including antitumor, antiinflammatory, and
antimicrobial agents (Papageorgiou, 1978
; Tabata and Fujita, 1985
). The
chemicals involved in the antimicrobial activities studied to date are
all derivatives of shikonin and its enantiomer alkannin from the
European dye plant Alkanna tinctoria. Figure 1 illustrates
the chemical structure of shikonin, its derivatives, and the proposed
biosynthetic pathway. The pattern of shikonin derivatives may differ in
any one plant, root culture, or cell-suspension culture (summarized
from a literature survey; specific refs. are given in Table
I).

View larger version (28K):
[in this window]
[in a new window]
| Figure 1.
Proposed biochemical pathway for shikonin and
shikonin derivative synthesis (after Gaisser and Heide, 1996 ).
Derivatives of shikonin are formed by replacement of the R-group with
the various fatty acid chains listed.
|
|
View this table:
[in this window]
[in a new window]
|
Table I.
Shikonin derivatives detected in plants, cell
suspension, and callus cultures
X's indicate that the derivative was present.
|
|
Because of its value as a pharmaceutical agent, most research to date
has focused on understanding the production and regulation of shikonin
with the goal of increasing production of the compound, but the
biological significance of shikonin in the plant has been virtually
ignored. Because shikonin and its derivatives have biological activity
against microorganisms, it seems likely that these compounds may play a
role in plant defense in the rhizosphere. Naphthoquinones have been
shown to function in allelopathy (juglone; Binder et al., 1989
),
plant-insect interactions (plumbagin; Kubo and Klocke, 1986
), and
electron transport (phylloquinone [vitamin K1];
Hauska, 1988
). Many of the products of the phenylpropanoid and
isoprenoid pathways are produced in response to multiple stresses (for
review, see Chappell, 1995
; Dixon and Paiva,
1995
) and frequently lead to the
production of antimicrobial compounds (Bennett and Wallsgrove, 1994
;
Rhodes, 1994
; Chappell, 1995
; Herrmann, 1995
). Regulation under
numerous conditions has been studied using cell-suspension cultures (Table
II), but nothing is known about the
regulation of production of the specific derivatives of shikonin or
of the regulation in specific cell types within the root. Shikonin
accumulates in the cork layer of mature roots (Fujita and Yoshida,
1937
) and its formation in significant quantities gives the roots a
characteristic red or purple color. Production and accumulation in new
roots has not been systematically studied, nor has the biological
activity of these compounds in relation to soil-borne
microorganisms.
In view of the observations on the antimicrobial properties reported
for shikonin and its derivatives, the production of numerous derivatives of shikonin in the roots of the plant, and the existing information on regulation of the pathways in common with many other
plant defense responses, we investigated the regulation and biological
activity of shikonin and its derivatives in hairy-root cultures of
Lithospermum erythrorhizon. Because these compounds are
brightly colored, the timing and cellular specificity of their production can be followed with simple equipment such as a dissecting microscope. The suite of compounds can be studied in the context of
understanding the multiple actions of plants in fine tuning their
response to their environment. The objectives of this study were: (a)
to characterize the production of naphthoquinone pigments in L. erythrorhizon hairy-root cultures, (b) to analyze the effect(s) of
shikonin and its derivatives on soil-borne microorganisms, and (c) to
determine if the L. erythrorhizon hairy-root cultures could
respond to microorganisms by modifying the production of shikonin and
its derivatives.
 |
MATERIALS AND METHODS |
Plant Material
Seeds of Lithospermum erythrorhizon (Sieb. and Zucc.)
were obtained from the Institut fur Pflanzengenetik und
Kulturpflanzenforschung (Gatersleben, Germany), germinated, and
grown in sterile potting soil (Pro Mix BX, Premier Horticulture, Red
Hill, PA). The plants were maintained in the greenhouse at
60oC to 80oC with supplemental lighting
from 1000-W high-pressure sodium lamps for 8 h daily from October
through March and in ambient lighting the remainder of the year. Plants
were watered three times per week. Plants were fertilized twice a
week with Peters 15-16-17 Peat Lite Special (Scotts-Sierra
Horticultural Products, Marysville, OH) at 200 ppm N.
Hairy-Root Production and Growth Conditions
Seeds of L. erythrorhizon were sterilized with 10%
commercial bleach for 20 min, rinsed in five changes of sterile water, and germinated under sterile conditions on solid Murashige and Skoog
medium (Murashige and Skoog, 1962
). Seedlings were grown in GA7
hydroponic culture vessels (Magenta Corp., Chicago, IL), and stems were
inoculated with Agrobacterium rhizogenes strain 15834 (ATCC)
to produce hairy roots. Root cultures were obtained from 16 independent
transformation events. All root culture lines appeared morphologically
indistinguishable, and experiments described in this paper were
performed on a single root line. Root cultures from the selected line
were grown in the dark at 24°C on M medium (Bécard and Fortin,
1988
) or M-9 (Fujita et al., 1981a
) medium. Roots grown on solid or
liquid M medium did not produce shikonin in sufficient quantities to be
observable on visual inspection. M-9 served as an induction medium for
shikonin because it contains 2.3 times the amount of copper sulfate as
the M medium. Liquid root cultures were grown in the dark on a rotary
shaker (model G-53, New Brunswick Scientific, Edison, NJ) at 90 rpm at
24°C. Solid root cultures were grown on medium solidified with 0.3% (w/v) Phytagel (Sigma).
Microscopy
Roots and fungal-growth plates were viewed and photographed using
a Zeiss STEMI SV8 dissecting microscope and Kodak 160T slide film. Root
hairs were viewed and photographed on a Zeiss inverted microscope and
Kodak 160T slide film.
Bacterial and Fungal Inhibition Assays
Bacterial Growth Inhibition
Bacterial cultures were grown overnight at 24°C with shaking in
Luria-Bertani broth (Luria and Burrows, 1957
) to an optical density of
0.2 at 600 nm. A 100-µL aliquot of this liquid culture was spread
onto 82-mm plates with solid Luria-Bertani medium. A filter disk
saturated with 50 µg of shikonin or shikonin derivative (dissolved in
chloroform and allowed to air dry) was placed on the bacterial lawn.
The plates were incubated in the dark at 24°C. The clear zone from
the filter disk to the bacteria was measured. Each experiment was
performed three times for each isolate tested. The following bacterial
strains were obtained from a collection of field isolates maintained in
the laboratory of Dr. Leland S. Pierson III (Department of Plant
Pathology, University of Arizona, Tucson): Bacillus subtilis
613R, Bacillus thuringiensis Gnatrol, Clavibacter
michigenensis subsp. nebraskensis CN74-1,
Agrobacterium radiobacter K84, Agrobacterium
tumefaciens C58, Burkholdaria cepacea Deny,
Escherichia coli ESS, Erwinia amylovora,
Erwinia carotovora ATCC 15713, Pseudomonas
aureofaciens 30-84, Pseudomonas fluorescens 2-79, Pseudomonas syringae B, Pseudomonas syringae pv
phaseolicola, Ralstonia solanacearum,
and Serratia marsecens. Erwinia
herbicola was isolated from pea seedling roots (L.A. Brigham,
unpublished data). A. rhizogenes ATCC 15834 was maintained
by Paula Michaels at The Pennsylvania State University (University
Park). Xanthomonas campestris pv pelargonii was
from the collection of Dr. Gary Moorman (The Pennsylvania State
University).
Fungal Growth Inhibition
Fungal isolates were maintained on V-8 medium (200 mL of
commercial vegetable juice, 2 g of CaCO3,
and 1 g of Glc per liter of water) solidified with 2% agar (Difco
Laboratories, Detroit, MI) in the dark at 24°C. Inhibition assays
were performed by dissolving shikonin in solid M-2 medium (modified
Martin's medium: 10 g of Glc, 5 g of peptone, 1 g of
KH2PO4 0.5 g of
MgSO4, 22 g of agar per liter of
double-distilled water) in 35-mm plastic Petri dishes. A 4-mm plug of
fungal hyphae was placed on one side of the plate and hyphal length
was measured at 24-h intervals. Each isolate was tested on all
concentrations in three separate replicates. Fungal isolates
Nectria hematococca 34-18 and N. hematococca T488 were from the laboratory cultures of Dr. Hans Vanetten (Department of
Plant Pathology, University of Arizona, Tucson). All of the other
fungal isolates were from Scott Rasmussen (Department of Plant
Pathology, University of Arizona).
Elicitation of Shikonin Derivatives in Liquid Root Cultures
Fungal elicitors were prepared as described previously (Flores et
al., 1988
). Mycelium from a 2-week-old fungal culture was inoculated
into a 2-L Erlenmeyer flask containing 1 L of Schenk and Hildebrandt
medium and grown in the dark on a rotary shaker (model G-53, New
Brunswick Scientific) at 90 rpm at 25°C for 3 weeks (Schenk and
Hildebrandt, 1972
). Mycelium was filtered off, resuspended in distilled
water, homogenized, centrifuged at 18,000 rpm for 30 min, autoclaved at
121°C for 30 min, and stored at
20°C. A 2.5-mL aliquot of this
filtrate was used per 50 mL of root-culture medium. Total shikonin was
quantified by treatment of derivatives with 2.5% KOH for 10 min
and spectrographic reading at 622 nm (Mizukami et al., 1977
).
Identification and Quantification of Compounds
TLC Analysis
Shikonin derivatives were extracted from the roots and liquid
media with chloroform. The chloroform layer was collected and evaporated to dryness. The residue was redissolved in chloroform and
spotted onto TLC plates (channeled Kieselgel 60CF254, Merck, Darmstadt,
Germany) using chloroform as the mobile phase. Standards were obtained
from TCI America (Portland, OR). Specific compounds were extracted from
the plates by scraping off and grinding the silica with a mortar and
pestle, and dissolving in chloroform. The identification of the
compounds was verified by comparison of retention time with known
standards by HPLC.
HPLC Analysis
HPLC was performed on a C18 60-Å 4-µm
column (3.9- × 300-mm, Nova-Pak, Waters) fitted to a HPLC device
comprising a system controller (model 600E), a Waters Intelligent
Sample Processor (model 712), and a photodiode array detector
(model 990, all Waters). The isocratic solvent system was
CH3CN:H2O:CH3COOH:Et3N
(630:370:3:3, v/v) Chromatography was performed at a flow rate of 1.2 mL min
1, a pressure of 30 kg/cm2, and a column temperature of 23°C.
A520 was monitored, and peaks were compared
with known standards (TCI America).
 |
RESULTS |
Development of Naphthoquinone Pigments in Pot-Grown Plants and in
Hairy-Root Cultures
Pigment Development under "Noninducing" Conditions
L. erythrorhizon hairy-root cultures grown on solid M
medium in the dark at 24°C appeared white (Fig.
2A). Roots grew at a rate of 1 to 2 mm/d.
Under magnification pigmentation was apparent in two cell types within
these roots: the root hairs (Fig. 2, C and E) and the root border cells
(Hawes and Brigham, 1992
; Brigham et al., 1995
; Hawes et al., 1998
)
(Fig. 2, C and K). Pigment granules in the root hairs were seen at the
apex of the root hairs as they emerged (Fig. 2E). Pigmentation was not
apparent before the hair had attained a length of at least 1 mm. As the
root hair elongated, the pigment appeared to remain at the same
distance from the epidermal surface as when it was first formed. The
pigment was limited to a very specific region and appeared to be a
ring, perhaps because the vacuole occurred at this region of the root
hair. In the root-cap region, pigments were confined to the border
cells alone (Fig. 2, K and M). Removing the border cells from the cap
revealed that the pigment was formed in the border cells subsequent to
separation from the cap, because the cap cells themselves were
completely devoid of pigment. Border cells were left behind as the root
grew through the medium and, in some cases, appeared to outline the root (Fig. 2L). When border cells were visualized under light microscopy in M medium, they appeared deep purple, as opposed to the
white color of the underlying cap cells (Fig. 2M). Cells of the cap
take up the vital stain fluorescence diacetate, whereas those of the
detached border cells do not (Fig. 2N).

View larger version (87K):
[in this window]
[in a new window]
| Figure 2.
L. erythrorhizon hairy-root
cultures showing pigment formation in different cells under varying
conditions. A, Transformed roots growing on M medium. B, Transformed
roots growing on M-9 medium. Pigment diffusion from the roots into the
medium was apparent at 3 weeks. C, Root tip grown on M medium showing
normal pigment production pattern in border cells and root hairs. D,
Root tip grown on M-9 medium showing increased pigmentation in border
cells and root hairs. E, Pigment deposition patterns in M-grown roots.
Emerging root-hair tips have a cap of pigment, which remains at about
that same distance from the epidermis as the root hair grows. F, Root
hairs of M-9-grown roots showing exudation of pigment in droplets all
over the hairs. G, Lateral root formation in M-grown roots. Pigment is
apparent in the region of root emergence. H, Lateral root formation in
M-9-grown roots. Numerous roots emerge at a single point and
pigmentation is more intense than in M-grown roots. I, Pigment
deposition in a few of the cells near lateral root eruption. J, Pigment
production in all epidermal cells of M-9-grown root. K, Border cell
pigmentation on root cap of M-grown root. Pigments are confined to
border cells. L, Border cell deposition along growing root in solid M
medium. M, Light micrograph of normal (untransformed) root tip placed
in water showing dispersion of border cells from the cap. Border cells
are purple, cap is white. N, UV fluorescence of root stained with
fluorescein diacetate showing living cells of root. Most border cells
do not take up the stain.
|
|
Border cells with pigment also displayed less distinct cytoplasmic
characteristics and failed to exhibit cytoplasmic streaming. Plasmolysis was often evident and in these cases pigmentation was
confined to the region of the cytoplasm and the cell walls appeared
clear (data not shown). The pigments appeared red when roots were grown
in air, away from contact with the medium (Fig. 2K), and deep blue to
purple when the cells of the roots were in contact with the medium
(Fig. 2L). Shikonin derivatives chelate iron in the medium, turning the
compounds purple in the process (L.A. Brigham, unpublished data). As
the roots aged, pigmentation occurred at the points of lateral root
emergence (Fig. 2G). Higher magnification revealed pigment granules
that tended to accumulate in the cell walls (Fig. 2I). Shikonin
production was also seen in the oldest parts of the root. The exact
cellular location of the pigment was not obvious from visual
inspection, because the older root segments also developed a brown
coloration, probably as a result of the production of other phenolic
compounds. Comparable patterns of red pigment formation were observed
in roots grown in pots in the greenhouse. Purple pigmentation was never
observed in these roots.
Induction of Shikonin Derivatives
Roots grown on the "inducing" medium (M-9) formed pigments
sooner, in greatly increased amounts, and in more cell types than those
grown on the noninducing medium (M) described above. L. erythrorhizon hairy-root cultures grown in the dark on M-9 medium at 24°C appeared fully pigmented, and the pigments had diffused into
the medium, giving it a pinkish color (Fig. 2B). The root cap appeared
to have greatly increased pigmentation and the border cells were not as
distinct, probably because of increased mucilage production (Fig. 2D).
The root hairs appeared to contain more pigment than the noninduced
hairs, and the pigment was distributed throughout the cell (Fig. 2F).
The amount of pigment production was increased to the point that it was
exuded from the root hair and appeared in droplets on the surface of
the hair. Lateral root formation occurred at shorter intervals than in
the controls, and clusters of lateral roots were commonly observed
emerging from a single site with greatly increased pigment production
(Fig. 2H). In contrast to the noninduced state, in which these were the
only two cell types that produced pigment, in the induced state pigment
was produced in all of the epidermal cells of the root much earlier and
before other phenolic products had turned the roots brown (Fig. 2J).
The pigments appeared to accumulate in the cell wall and were seen
throughout the apoplast.
Effects of Shikonin on Bacterial and Fungal Growth
Because shikonin has been used as an antimicrobial agent in human
applications, and because shikonin and its derivatives are produced
exclusively by the root of the plant, we tested whether shikonin was
inhibitory to soil-borne microorganisms. Both bacterial and fungal
isolates were assayed.
Bacterial Growth Inhibition
Bacterial isolates from plant roots covering a broad phylogenetic
range were tested for inhibition of growth by the filter-disk method
with 50 µg of standard shikonin per filter. The concentration was
chosen within the range shown to inhibit human pathogens (Tanaka and
Odani, 1972
). Of the 31 strains tested, 10 were inhibited to some
degree by shikonin. Table III lists a
representative sample of the strains tested. Whereas no definite
patterns emerge, it is clear that not all strains are sensitive to
shikonin and some strains are more sensitive than others. As reported
previously by Papageorgiou (1980)
, the E. coli strains were
not affected. The two pathogenic strains of Erwinia were not
sensitive, whereas the nonpathogenic E. herbicola was
inhibited. All of the Agrobacterium strains were affected,
but the nonpathogenic A. radiobacter was inhibited to a
greater degree than the pathogenic strains. All gram-positive strains
tested were inhibited. In summary, shikonin shows selective inhibition
to some bacterial strains, including important root pathogens.
View this table:
[in this window]
[in a new window]
|
Table III.
Inhibition of bacterial isolates to shikonin,
shikonin derivatives, and whole-root chloroform extractions
Data are presented as the size of the zone of inhibition. In all cases,
SD was less than 0.289.
|
|
Fungal Growth Inhibition and Morphological Changes
Hyphal growth inhibition was tested by a linear growth assay.
Twelve fungal isolates were tested and showed a wide range of growth
responses on medium containing 5, 50, 100, and 200 µg/mL shikonin
standard. Growth was measured for each isolate until the hyphae of any
treatment reached the edge of the 35-mm plate. The length of time for
the control to reach the edge of the plate varied from 1 d
(Pythium aphanidermatum) to 9 d (Colletotrichum destructivum). Figure 3 illustrates
four different growth patterns. Pythium ultimum showed
increasing sensitivity to increasing amounts of shikonin (Fig. 3A).
P. aphanidermatum was increasingly inhibited by increasing
concentrations of shikonin until the highest concentration (200 µg/mL), at which point it showed a slight increase in growth over the
100 µg/mL concentration (Student's t test, P = 0.06; Fig. 3B). Nectria hematococca 34-18 showed little
inhibition of hyphal growth even at the highest shikonin concentration
(Fig. 3C). Phytophthora parasitica grew faster than the
control at 5 and 50 µg/mL (Student's t test, P = 0.01; Fig. 3D). Comparisons of growth patterns among all of the fungi
tested were made by calculation of hyphal growth on each concentration
of shikonin as a percentage of the growth of the control (Table
IV).

View larger version (22K):
[in this window]
[in a new window]
| Figure 3.
Fungal growth on medium containing no shikonin
(control, ), 5 µg/mL shikonin ( ), 50 µg/mL shikonin ( ),
100 µg/mL shikonin (×), or 200 µg/mL shikonin ( ). A,
P. ultimum. B, P. aphanidermatum. C,
N. hematococca 34-18. D, P. parasitica.
|
|
View this table:
[in this window]
[in a new window]
|
Table IV.
Fungal growth on five different concentrations of
shikonin (in µg/mL) as a percentage of growth of the control
Data are presented as the percentage of growth compared with the
control, and represent the averages of three separate experiments, each
with three replicates. The coefficients of variation (V = [s × 100]/X; V, coefficient of
variation; s, SD; X, mean) ranged
from 2% for the controls to 16% for the treatments in which growth
was most strongly inhibited.
|
|
In addition to the variability of the effect of shikonin on hyphal
growth rates, other changes were observed. Some fungal isolates changed
the color of the medium containing shikonin from pink to varying shades
of purple and blue, with each isolate producing a slightly different
hue (data not shown). It is known that some fungi are capable of
changing the pH of their surroundings, but pH measurements were not
undertaken in this study. The hyphae of Rhizoctonia solani
formed fairly uniform sheets over the control plates (Fig.
4A). However, at increasing
concentrations of shikonin, the hyphal formations exhibited clumping
and apparent sequestration of pigment within the clumps (Fig. 4, B and
C). Aspergillus niger, which was only moderately inhibited
by shikonin, showed significant changes in both morphology and pigment
formation in its spores. Spores formed on plates with 5 µg/mL (Fig.
4D) were darker than the controls, and had a dumbbell shape similar to
the controls, but with increasing amounts of shikonin, the spores
showed less pigmentation. The spores formed on plates with 200 µg/mL
shikonin (Fig. 4E) were beige and rounded. Glomus
intraradices was not tested for growth inhibition because it was
not known at the time if the obligate mycorrhizal fungus would infect
L. erythrorhizon. However, co-cultivation of hairy roots of
carrot, G. intraradices (Chabot et al., 1992
), and L. erythrorhizon showed that this fungus also sequestered the
pigments. Red pigment could be seen moving within the hyphae and, over
several weeks, the pigments were deposited on the outside of a small
percentage (less than 5%) of the fungal spores (Fig. 4F). Thus, as in
the case of bacterial strains tested, fungal isolates exhibited
differential sensitivity to shikonin. Several known pathogenic
soil-borne fungi such as Rhizoctonia, Pythium,
and Phytophthora were significantly inhibited by shikonin. In addition, morphological changes were often observed in hyphal growth patterns and spore formation.

View larger version (94K):
[in this window]
[in a new window]
| Figure 4.
Changes in fungal morphology when grown on medium
containing shikonin. A, R. solani hyphal growth on
control plate showing an even growth pattern. B, R. solani hyphal growth on 100 µg/mL shikonin showing clumping
of hyphae. C, R. solani hyphal growth on 200 µg/mL
shikonin showing sequestration of shikonin pigment in the hyphal
clumps. D, Spores of A. niger grown on 5 µg/mL
shikonin. Spores are dark and dumbbell shaped. E, Spores of A. niger grown on 200 µg/mL shikonin. Spores are beige and
round. F, Spores of G. intraradices grown on M medium in
the presence of L. erythrorhizon hairy roots. Pigment is
not apparent in the medium by visual inspection. Pigment is transported
through the hyphae and deposited on a small fraction of the spores in
any one area.
|
|
Production of Shikonin Derivatives
The production of different shikonin derivatives in roots sold
commercially, in cell-suspension cultures, and in callus cultures has
been analyzed by several researchers in China and Japan (Table I).
Different subsets of the known shikonin derivatives are found in
different species of Lithospermum, in different sources of individual species, and in different cell lines. Not all of the known
compounds are found in a single plant or laboratory root culture.
However, AS and HIVS were found in all samples in these studies. We
wanted to determine if changes occurred in the production of
derivatives under different conditions, including fungal challenge. All
experiments were done on root cultures that originated from a single
transformation event and were therefore genetically identical. Hairy-root cultures have been used as a tissue-culture tool for studies
of secondary metabolite production in roots and have been found by
numerous studies to "remain more or less stable in their growth
and/or secondary metabolite productivity characteristics for a number
of years" (Hamill and Lidgett, 1997
). Roots were grown in liquid
cultures of M or M-9 medium and challenged with R. solani
elicitor preparations. Roots grown in M medium remained white as long
as the medium was changed frequently. Roots grown in unchanged
(depleted) medium eventually accumulated shikonin derivatives. As shown
in Table V, roots grown in M-9 medium
formed shikonin derivatives earlier and in greater quantities than
those grown in M medium, and the ratios of HIVS to AS varied. HIVS and AS were detected in the greatest abundance in all of the roots sampled,
although the relative proportions varied by treatment. IVS and MBS were
not readily distinguishable using HPLC under the conditions we used,
but one or both were detected in most cultures in the 10% to 15%
range. Challenge with R. solani crude elicitor caused a
reversal of the ratios of HIVS and AS compared with treatments on M
medium. Shikonin derivatives from plants grown in pots in the
greenhouse were analyzed and found to contain equal amounts (38% each)
of HIVS and AS, 14% of IVS/MBS, and a small amount (2%) of DS.
Shikonin was not detected in any of the samples analyzed.
Effect of Shikonin Derivatives on Bacterial Growth
Because shikonin itself was shown to have an effect on specific
bacterial and fungal soil isolates, further tests were done to
determine their sensitivity to the shikonin derivatives and to the
combinations of derivatives produced by the roots. As has been reported
previously (Tanaka and Odani, 1972
; Tabata et al., 1975
), E. coli was not sensitive to shikonin or any of its derivatives. All
organisms that were insensitive to shikonin were also insensitive to
all of the derivatives tested, including the crude whole-root chloroform extract (Table III). However, the two Erwinia
species that were insensitive to shikonin were insensitive to the
derivatives, but were inhibited by an unknown component of the induced
chloroform extract. In most cases, induced extract (M-9) produced the
highest inhibition, followed by noninduced extracts. AS, HIVS, and
shikonin were often as inhibitory as the whole-extract preparations and were the most-active single compounds. IBS and MBS were found the least
inhibitory in most cases. As in the case of the fungal studies on
shikonin, color changes in the bacterial plates were varied (data not
shown), but pH was not determined. In summary, bacterial isolates
showed a wide range of responses to shikonin derivatives singly and in
combination. Most isolates were inhibited to the greatest degree by the
two derivatives produced in highest abundance by the root cultures.
In Situ Fungal Inhibition and Plant Response to Fungi
To analyze the dynamics of shikonin production in root cultures on
fungi, we placed plugs of fungal hyphae on 4-week-old L. erythrorhizon root cultures on solid M and M-9 medium. One-half of
them were grown in the light (to inhibit shikonin derivative production) and one-half were grown in the dark. Four isolates were
chosen based on growth rates and response to the shikonin concentration
assays (Table VI). The amount of shikonin
derivatives the fungi were exposed to on the plates was not quantified.
Pigment was apparent upon visual inspection in the roots grown on M-9 at this point. P. aphanidermatum was sensitive to both the
existing shikonin derivatives on the M-9 plates and to increased
shikonin derivative produced in the dark on both the M and M-9 plates. On the light-grown plates, shikonin production was inhibited and hyphal
growth was twice that seen on the dark-grown M plates (Table VI; Fig.
5A). R. solani, which is very
sensitive to high concentrations of shikonin (Table VI), was inhibited
on the M-9 plates, but not significantly on the M plates, even though
the fungus did induce pigment formation in these roots. N. hematococca T488, which is sensitive to shikonin at relatively low
concentrations, showed marked inhibition on dark-grown plates whether
preinduced or not. A. niger, the least-sensitive and
slowest-growing isolate, showed no significant difference between
light- and dark-grown plates, preinduced or not. This was the only one
of the four isolates tested that did not induce localized pigment
production in the roots (Table VI). Figure 5B shows an example of a
region of hyphal contact with the root. The cells in the region of
contact show considerable pigmentation. Areas normally pigmented, such
as sites of lateral root emergence, had fewer hyphae than regions in
between (Fig. 5C). However, all regions where hyphae made contact
eventually produced pigment. Whereas most of the roots on the plate
were eventually overcome by hyphal growth, the root tips were always devoid of hyphae. These tips, like those grown on M-9 medium, were
moist and very red, showing increased pigment production (Fig. 5D).
Whereas root epidermal cells in direct contact with hyphae showed
increased pigment production compared with adjacent noncontact cells,
the root caps showed increased production without direct hyphal
contact.
View this table:
[in this window]
[in a new window]
|
Table VI.
Fungal growth on hairy-root cultures
Data are from a single experiment with two replicates. The whole
experiment was performed two times with similar results.
|
|

View larger version (105K):
[in this window]
[in a new window]
| Figure 5.
In situ growth inhibition and stimulation of
pigment production of R. solani in the presence of
L. erythrorhizon hairy roots. A, Hairy roots grown in
the light on M medium (left) and grown in the dark on M medium (right).
The fungal plug is seen on the left in each plate. B, Pigment formation
in root cells at the points of contact with hyphae. C, Accumulation of
shikonin at lateral root eruptions on roots grown in M-9 medium showing
inhibition of fungal growth. D, Root tip with greatly increased pigment
production is free of hyphae, whereas the surrounding root surfaces are
completely overcome.
|
|
Pigment production in root cultures was also induced with crude
preparations of R. solani elicitor. Liquid cultures of
L. erythrorhizon hairy roots were grown for 1 month in
liquid M medium. One set of flasks received fresh M medium, and the
other set received M-9 medium. Root cultures were allowed to grow for
1 d, then 2.5 mL (per 50 mL of medium) of fungal elicitor was
added to one-half of the M flasks and one-half of the M-9 flasks.
Pigment released into the media was evident visually at the end of
2 weeks. The media were collected at 3 weeks and the shikonin
derivatives were extracted in chloroform and analyzed by HPLC. The
roots treated with a combination of both M-9 and fungal elicitor showed
the strongest response in total pigment produced (Table V). The roots treated with fungal elicitor in the inducing medium produced 3 times as
much shikonin as those with fungal elicitor in M medium and almost 30 times as much as the noninduced roots. These results are consistent
with the model that shikonin derivatives in the roots function as both
preformed and inducible antimicrobial compounds.
 |
DISCUSSION |
The transformed roots of L. erythrorhizon appear to be
a useful model in which to study the production and potential
function(s) of a suite of antimicrobial compounds in plant roots
(Rhodes et al., 1997
). The derivatives of shikonin are produced in a
highly regulated and visible manner in specific cell types of the
growing root tip. The compounds have varying degrees of biological
activity against a wide range of soil-borne microorganisms. Production of these compounds in changing ratios within the root varies with conditions of stress, including challenge by microorganisms.
The cellular pattern of shikonin derivative production depends on the
developmental stage of the root and on environmental signals. In
relatively "unstressed" root cultures pigment production is
normally evident only in two cell types, the root hairs and border
cells. The identification of pigments in these cells often requires
magnification. As the radicle of L. erythrorhizon emerges from the seed, the root border cells are already pigmented, and as the
root elongates, they remain at intervals along the root as a potential
chemical barrier to microbial infection (Hawes and Brigham, 1992
; Hawes
et al., 1998
). Even under unstressed conditions, the border cells in
the root cultures are intensely pigmented, showing granules distributed
throughout the cell (Tsukada and Tabata, 1984
). The underlying cap
cells are devoid of pigment. From observation in this study, it appears
to be a rapid development, because no intermediate states of pigment
production have been observed in border cells. The dynamics of the
development of these compounds in border cells remains to be studied.
The pigment in the root hairs appears to be localized to a region of
the hair between the epidermal surface and the midpoint of the hair.
The exact cellular location and the mechanism are unknown. The uniform placement of pigments in the root hairs at some distance from the root
surface may also serve as a chemical barrier to microbial invasion.
Preformed inhibitors are often distributed in plants in a
tissue-specific manner, often in the outer layers of cells of plant
organs (Osbourn, 1996
). But, unlike many of the preformed inhibitors
studied to date, the shikonin derivatives, in addition to being formed
as part of normal development, are highly inducible.
In the unstressed roots described above, the pigments appeared to
remain within the cells in which they were produced. In contrast, when
the roots were specifically stressed by the addition of copper sulfate,
the root hairs made significantly more pigments and the epidermal cells
produced amounts sufficient to turn the roots "red." In both these
of cell types, the material is clearly extracellular. In root hairs,
droplets formed on the surface; in epidermal cells, pigment accumulated
in the apoplast. The mechanism by which this is accomplished is
unknown. Tsukada and Tabata (1984)
have described the formation of
pigment within the cells of induced callus cultures as occurring on the
rough ER in spherical bodies that break away and migrate to the plasma
membrane, presumably releasing their contents. The droplets were found
to be composed of 27.2% shikonin, 21.5% proteins, 28.6% lipids, and
22.7% "other" (Tsukada and Tabata, 1984
). The caps of induced
roots growing aerially in our study appeared moist, possibly because of
the active export of these droplets. The amount of the compounds
produced in each cell does not appear to be the driving mechanism,
because, as we have observed, border cells from noninduced caps are
almost completely filled with pigment that is not exported. Yazaki and Matsuoka (1997)
have identified a clone from an induced cDNA library that may begin to shed light on this aspect of the regulation.
The ratio of compounds produced in different induced states differs
from that in the noninduced state, suggesting that the plant can sense
and respond to different environmental and physiological stimuli
(Bennett and Wallsgrove, 1994
). It has been reported that accumulation
of glucosinolates in oilseed rape varies by challenge organism (insect,
fungus, mechanical damage) and tissue type (for review, see Bennett and
Wallsgrove, 1994
). We found that challenge with a crude elicitor from
R. solani changes the ratio of HIVS to AS produced by the
roots irrespective of the media on which the roots were grown. The
production of a set of very similar compounds raises the question of
redundancy and whether the duplication is part of the survival strategy
to preclude development of resistance by microorganisms. Of the seven
compounds identified from L. erythrorhizon roots from potted
plants, cell suspension, callus, and root cultures, we found that only
three of the compounds were consistently produced in the root cultures
and potted plants under the conditions in which they were grown in our
laboratory. AS and HIVS, the most biologically active compounds, were
the most abundant and were produced under all of the conditions,
although in different amounts in relation to each other. The whole-root
component mix almost always had the greatest activity against the
bacteria tested, which may be attributable to the components acting
synergistically, or to additional components soluble in the
chloroform-extraction process. It is not known if the mechanisms of
action in microorganisms are the same or different for the various
compounds.
Microorganisms also influence the chemistry of the products they
encounter. Certain fungi are known to change the pH of their surroundings (Bago et al., 1996
). For example, Alternaria
solani blocks the effects of saponin-based defense chemicals by
lowering the pH of the infection site (for review, see Jackson and
Taylor, 1996
). Shikonin can function as a pH indicator in buffered
aqueous solutions (Windholz, 1983). In the hyphal inhibition
studies, in which shikonin was dissolved in the medium, there were
almost as many color changes as there were fungi studied, indicating that each fungus changed the pH to a different degree. It is not known
if pH influences the biological activity of shikonin for a specific
microorganism, but alkaline conditions convert all shikonin derivatives
to shikonin, which was the third most toxic of the derivatives tested
in this study. pH also influences enzymatic optima, membrane
permeability, and a host of other conditions that could ultimately
determine the toxicity or safety of the environment for the fungus
(Marschner, 1995
). The composition of microbial populations in the
rhizosphere of L. erythrorhizon is unknown, and further work
is required to determine the nature of the rhizosphere communities in
the various L. erythrorhizon populations.
The results from this study are consistent with the possibility that
shikonin derivatives function as both preformed and inducible microbial
inhibitors regulated in a cell-specific manner to maximize the effect
of highly toxic substances with minimal expense and harm to the plant.
The complexity of the rhizosphere is often decried as an almost
insurmountable challenge by plant pathologists and physiologists, and
thus research has concentrated on the effects of single compounds
produced in a plant on specific microorganisms. The system described
here may allow us to move toward understanding the complex
relationships between multiple compounds influencing multiple organisms
in the rhizosphere. Because the shikonin derivatives are produced from
two pathways under intense investigation because of their contribution
to antimicrobial products, this system may add to our understanding of
how the plant regulates these pathways in subtle ways to influence
microbial population dynamics in the rhizosphere. Studies are under way
to examine the molecular mechanisms of cell-specific expression,
regulation of compound ratios, and microbial sensitivity/tolerance
mechanisms.
 |
FOOTNOTES |
1
This research was supported by Department of
Energy/National Science Foundation/U.S. Department of Agriculture
Collaborative Research in Plant Biology award no. BIR-9220330,
"Interdisciplinary Research Training Program in Advanced Root
Biology."
2
Present address: Department of Plant Pathology,
The University of Arizona, Tucson, AZ 85721-0036.
*
Corresponding author; e-mail lbrigham{at}ag.arizona.edu; fax
1-520-621-9290.
Received May 26, 1998;
accepted October 22, 1998.
 |
ABBREVIATIONS |
Abbreviations:
AS, acetyl-shikonin.
DMAS,
,
-dimethylacryl-shikonin.
DS, deoxy-shikonin.
HIVS,
-hydroxyisovaleryl-shikonin.
IBS, isobutyl-shikonin.
IVS, isovaleryl-shikonin.
MBS,
-methyl-n-butyl-shikonin.
 |
ACKNOWLEDGMENTS |
We thank Drs. Leland S. Pierson and Derek Wood (University of
Arizona, Tucson) for providing the bacterial strains used in these
experiments, Dr. Hans Vanetten and Scott Rasmussen (University of
Arizona) for providing most of the fungal isolates, Dr. Martha Hawes
for providing laboratory space for the experiments performed at the
University of Arizona, Tom Orum (University of Arizona) for providing
assistance in the statistical analyses, and Anthony Omeis (Biology
Greenhouse at The Pennsylvania State University) for propagating the
L. erythrorhizon plants and caring for them in the
greenhouse. The carrot hairy-root cultures were a gift from Dr. Roger
Koide of The Pennsylvania State University. Dr. Gary Moorman graciously
provided a critical reading of the manuscript, and Dr. Kazufumi Yazaki
(Kyoto University, Japan) provided invaluable background information,
discussion, and advice for this research.
 |
LITERATURE CITED |
Bago B,
Vierheilig H,
Piché Y,
Azcón-Aguilar C
(1996)
Nitrate depletion and pH changes induced by the extraradical mycelium of the arbuscular mycorrhizal fungus Glomus intraradices grown in monoxenic culture.
New Phytol
133:
273-280
Ballantine JA
(1969)
The isolation of two esters of the naphthoquinone alcohol, shikonin, from the shrub Jatropha glandulifera.
Phytochemistry
8:
1587-1590
[CrossRef]
Bécard G,
Fortin JA
(1988)
Early events of vesicular-arbuscular mycorrhiza formation on Ri T-DNA transformed roots.
New Phytol
108:
211-218
[CrossRef]
Bennett RN,
Wallsgrove RM
(1994)
Secondary metabolites in plant defense mechanisms.
New Phytol
127:
617-633
[CrossRef]
Binder RG,
Benson ME,
Flath RA
(1989)
Eight 1,4-naphthoquinones from Juglans.
Phytochemistry
28:
2799-2801
[CrossRef][ISI]
Brigham LA,
Woo HH,
Hawes MC
(1995)
Root border cells as tools in plant cell biology studies.
Methods Cell Biol
49:
377-387
[Medline]
Chabot S,
Bécard G,
Piché Y
(1992)
Life cycle of Glomus intraradix in root organ culture.
Mycologia
84:
315-321
Chappell J
(1995)
The biochemistry and molecular biology of isoprenoid metabolism.
Plant Physiol
107:
1-6
[ISI][Medline]
Dixon RA,
Paiva NL
(1995)
Stress-induced phenylpropanoid metabolism.
Plant Cell
7:
1085-1097
[CrossRef][ISI][Medline]
Flores H,
Pickard J,
Hoy M
(1988)
Production of polyacetylenes and thiophenes in heterotrophic and photosynthetic root cultures of Asteraceae.
In
J Lamb,
H Breteler,
T Arnason,
L Hansen,
eds, Chemistry and Biology of Naturally-Occurring Acetylenes and Related Compounds (NOARC).
Elsevier, Amsterdam, pp 233-254
Fujita Y,
Hara Y,
Ogino T,
Suga C
(1981a)
Production of shikonin derivatives by cell suspension cultures of Lithospermum erythrorhizon. I. Effects of nitrogen sources on the production of shikonin derivatives.
Plant Cell Rep
1:
59-60
Fujita Y,
Hara Y,
Suga C,
Morimoto T
(1981b)
Production of shikonin derivatives by cell suspension cultures of Lithospermum erythrorhizon. II. A new medium for the production of shikonin derivatives.
Plant Cell Rep
1:
61-63
Fujita N,
Maeda Y,
Suga C,
Morimoto T
(1983)
Production of shikonin derivatives by cell suspension cultures of Lithospermum erythrorhizon. III. Comparison of shikonin derivatives of cultured cells and Ko-shikon.
Plant Cell Rep
2:
192-193
[CrossRef]
Fujita N,
Yoshida Y
(1937)
Ueber die Anatomie von Radix Lithospermi.
Yakugaku Zasshi
57:
368-391
Fukui H,
Tani M,
Tabata M
(1990)
Induction of shikonin biosynthesis by endogenous polysaccharides in Lithospermum erythrorhizon cell suspension cultures.
Plant Cell Rep
9:
73-76
Fukui H,
Yoshikawa N,
Tabata M
(1983)
Induction of shikonin formation by agar in Lithospermum cell suspension cultures.
Phytochemistry
22:
2451-2453
[CrossRef]
Gaisser S,
Heide L
(1996)
Inhibition and regulation of shikonin biosynthesis in suspension cultures of Lithospermum.
Phytochemistry
41:
1065-1072
[CrossRef]
Hamill JD,
Lidgett AJ
(1997)
Hairy root cultures: opportunities and key protocols for studies in metabolic engineering.
In
PM Doran,
eds, Hairy Roots: Culture and Applications.
Harwood Academic, Sydney, Australia, pp 1-29
Hauska G
(1988)
Phylloquinone in photosystem I.
Trends Biochem Sci
13:
415-416
[Medline]
Hawes MC,
Brigham LA
(1992)
Impact of root border cells on microbial populations in the rhizosphere.
Adv Plant Pathol
8:
119-148
Hawes MC,
Brigham LA,
Wen F,
Woo HH,
Zhu Y
(1998)
Function of root border cells in plant health: pioneers in the rhizosphere.
Annu Rev Phytopathol
36:
311-327
[CrossRef][ISI][Medline]
Herrmann KM
(1995)
The shikimate pathway as an entry to aromatic secondary metabolism.
Plant Physiol
107:
7-12
[Medline]
Jackson AO,
Taylor CB
(1996)
Plant-microbe interactions: life and death at the interface.
Plant Cell
8:
1651-1668
[CrossRef][ISI][Medline]
Jain AC,
Mathur SK
(1965)
Bull Natl Inst Sci India
28:
52-56
Kim DJ,
Chang HN
(1990)
Increased shikonin production in Lithospermum erythrorhizon suspension cultures with in situ extraction and fungal cell treatment (elicitor).
Biotech Lett
12:
443-446
Kubo I,
Klocke J
(1986)
Insect ecdysis inhibitors.
In
MB Green,
PA Hedin,
eds, Natural Resistance of Plants to Pests.
American Chemical Society, Washington, DC, pp 206-219
Kyogoku K,
Terayama H,
Tachi Y,
Suzuki T,
Komatsu M
(1973)
Constituents of shikon. II. Comparison of contents, constituents, and antibacterial effect of fat soluble fraction between nanshikon and koshikon.
Shoyakugaku Zasshi
27:
31-36
Luria SE,
Burrows JW
(1957)
Hybridization between Escherichia coli and Shigella.
J Bacteriol
74:
461-476
[Free Full Text]
Marschner H (1995) Mineral Nutrition in Higher Plants, Ed 2. Academic Press, London, pp 537-595
Mizukami H,
Konoshima M,
Tabata M
(1977)
Effect of nutrition factors on shikonin derivative formation in Lithospermum callus cultures.
Phytochemistry
16:
1183-1186
[CrossRef]
Mizukami H,
Konoshima M,
Tabata M
(1978)
Variation in pigment production in Lithospermum erythrorhizon callus cultures.
Phytochemistry
17:
95-97
[CrossRef]
Morimoto I,
Hirata Y
(1966)
New naphthoquinone derivatives from Lithospermum erythrorhizon.
Tetrahedron Lett
31:
3677-3680
Morimoto I,
Kishi T,
Ikegami S,
Hirata Y
(1965)
Naphthoquinone derivatives from Lithospermum erythrorhizon Siebold et Zuccarini.
Tetrahedron Lett
52:
4737-4739
[CrossRef]
Murashige T,
Skoog F
(1962)
A revised medium for rapid growth and bioassays with tissue cultures.
Physiol Plant
15:
473-497
[CrossRef]
Osbourn AE
(1996)
Preformed anti-microbial compounds and plant defense against fungal attack.
Plant Cell
8:
1821-1831
[CrossRef][ISI][Medline]
Papageorgiou V
(1978)
Wound healing properties of naphthoquinone pigments from Alkanna tinctoria.
Experientia
34:
1499-1501
[Medline]
Papageorgiou V
(1980)
Naturally occurring isohexenylnaphthazarin pigments: a new class of drugs.
Planta Med
38:
193-203
[Medline]
Rhodes MJC
(1994)
Physiological roles for secondary metabolites in plants: some progress, many outstanding problems.
Plant Mol Biol
24:
1-20
[Medline]
Rhodes MJC,
Parr AJ,
Walton NJ
(1997)
Studies of secondary metabolic pathways in transformed roots.
In
PM Doran,
eds, Hairy Roots: Culture and Applications.
Harwood Academic, Sydney, Australia, pp 31-41
Schenk RU,
Hildebrandt AC
(1972)
Medium and techniques for induction and growth of monocotyledonous and dicotyledonous plant cell cultures.
Can J Bot
50:
199-204
[ISI]
Tabata M,
Fujita Y
(1985)
Production of shikonin by plant cell cultures.
In
M Zaitlin,
P Day,
A Hollaender,
eds, Biotechnology in Plant Science.
Academic Press, San Diego, CA, pp 207-218
Tabata M,
Mizukami H,
Haoe S,
Konoshima M
(1975)
Anti-microbial activity of Lithospermum erythrorhizon callus cultures.
Yakugaku Zasshi
95:
1376-1379
[Medline]
Tabata M,
Mizukami H,
Hiraoka N,
Konoshima M
(1974)
Pigment formation in callus cultures of Lithospermum erythrorhizon.
Phytochemistry
13:
927-932
[CrossRef]
Tanaka Y,
Odani T
(1972)
Pharmacodynamic study on "shiunko." I. Antibacterial effect of "shiunko."
Yakugaku Zasshi
95:
525-530
Tani M,
Fukui H,
Shimomura M,
Tabata M
(1992)
Structure of endogenous oligogalacturonides inducing shikonin biosynthesis in Lithospermum cell cultures.
Phytochemistry
31:
2719-2723
[CrossRef]
Tani M,
Takeda K,
Yazaki K,
Tabata M
(1993)
Effects of oligogalacturonides on biosynthesis of shikonin in Lithospermum cell cultures.
Phytochemistry
34:
1285-1290
[CrossRef]
Tsukada M,
Fukui H,
Habara C,
Tabata M
(1983)
Comparative studies on naphthoquinone derivatives in various crude drugs of "zicao" (shikon).
Shoyakugaku Zasshi
37:
299-305
Tsukada M,
Tabata M
(1984)
Intracellular localization and secretion of naphthoquinone pigments in cell cultures of Lithospermum erythrorhizon.
Planta Med
51:
338-341
Windholz M, ed (1983) The Merck Index: An Encyclopedia of
Chemicals, Drugs, and Biologicals, Ed 9. Merck & Co., Rahway, NJ
Yazaki K,
Matsuoka H
(1997)
cDNA cloning and functional analysis of LEDI-2, a gene preferentially expressed in the dark in Lithospermum erythrorhizon cell suspension cultures (abstract no. 1348).
Plant Physiol
114:
S-261