N-114 Interactions Between Soil Microorganisms and Weed Seeds: Novel Implications for Plant-Microbe Relationships

1
N-114 Interactions Between Soil
Microorganisms and Weed Seeds Novel Implications
for Plant-Microbe Relationships
Joanne C. Chee-Sanford1,2, Yim F. So2, Lynn M.
Connor1, Teresa J. Holman1, Robert A. Sanford3,
and Gerald K. Sims1,2 1U.S. Department of
Agriculture-Agricultural Research Service and
2Dept. of Crop Sciences, 3Dept. of Geology,
University of Illinois, Urbana, IL 61801
INTRODUCTION
RESULTS
SUMMARY AND CONCLUSIONS
Recent interest in integrated weed management
strategies that entail more efficient use of
herbicides and minimize chemical input into the
environment has shifted attention toward the
development of more effective biologically-based
methods to manage invasive plant species.
Exploiting soil-related microbial processes to
target weeds is one approach. Many annual
weed species produce high numbers of seeds, with
abundant, long-lived proportions persisting in
soil each year. Buried seed reserves, referred to
as the soil seed bank, create a long-standing
reservoir for future weed emergence. The
conditions of seed banks present opportunity for
numerous interactions to occur between native
soil microorganisms and seeds, ranging in
possible outcomes that may affect the seed,
microbe, or both. In previous studies(1-3), we
found a range of susceptibility to decay among
several weed species. Velvetleaf seeds, for
example, undergo decay readily once exposed to
soil microbial populations, particularly when
seeds provide the major carbon nutritional
resource under incubation conditions. Seeds can
also become densely colonized by microbes, yet
remain intact and viable. These early results
suggest possible growth advantages for soil
microbes in association with seeds, with the
outcome potentially detrimental to seeds, but not
always. The types and mechanisms of interactions
between microorganisms (bacteria and fungi) and
seeds continue to be the focus of our
studies. This study further investigates the
characteristics of interactions between native
soil microorganisms and different species of weed
seeds. Knowledge of the presence, identity,
distribution, and relative abundance of important
microorganisms and the activities that are
related to the fate of weed seeds in soil will
allow us to understand how such activities may be
exploited or modified for weed management use.
Further, this study addresses an area of
plant-microbe relationships that is not well
understood, and investigates fundamental aspects
of seed-microbe interactions that contribute to
soil microbial ecology and plant community
development.

• Seeds can support diverse, and often dense,
  microbial associations, where the outcome is
  often not detrimental to the seed, with the
  exception of velvetleaf. This is consistent with
  numerous studies that demonstrate seed
  persistence of many annual weeds in seed bank
  soils. Velvetleaf seeds, while susceptible to
  decay under controlled lab conditions, also
  persist in seed banks, suggesting environmental
  factors that limit occurrence of decay in nature.
  
• Total bacterial communities generally differed
  between seeds of different weed species, with
  assemblages tending to cluster with respect to
  weed species, regardless of soil origin (in this
  study). This suggests that certain interactions
  may be favored under appropriate conditions, with
  benefit for many commonly found soil bacteria,
  such as members of the Bacteroidetes and
  Proteobacteria. Benefit to the seed is currently
  unknown.
• Velvetleaf seeds, which undergo extensive decay
  in the presence of soil microorganisms, tend to
  have similar dominant bacterial associations, in
  contrast to seeds of Pennsylvania smartweed,
  woolly cupgrass, and jimsonweed, which have
  limited susceptibility to decay. The dominance
  of certain bacteria on velvetleaf seeds suggest
  these microbial species may be specifically
  involved in the process of seed decay. The C and
  N content of velvetleaf seeds, if accessed,
  suggests a reasonable source of nutrition for
  soil microorganisms.
• In contrast to fairly diverse bacteria found on
  seeds, ascomycota sequences predominated in
  fungal associations with seeds of several weed
  species. Further, regardless of soil origin,
  identical or closely related sequences comprising
  a few genera appear to specifically form
  interactions with certain weed species. These
  relationships include Chaetomium sp. and woolly
  cupgrass Cordyceps sp. and Pennsylvania
  smartweed and Cephaliophora sp. and velvetleaf.
• Water soluble exudates were recovered from seeds
  of all species tested. These exudates are likely
  to be complex mixtures, like that found with
  kochia. It is currently unknown how the
  physiological status of the seeds affect
  production of the exudates or the extent of
  exudate production under in situ seed bank
  conditions.
• Chemical exudates from kochia seed exhibit
  significant inhibition of Colletotrichum
  graminicola, a known corn pathogen of soil
  origin. These data suggests a potentially
  important role of seed exudates in plant-microbe
  relationships, and could be further hypothesized
  to have a role in regulating microbial
  associations, soil microbial communities, and
  plant development. Exudates may also provide a
  nutritional benefit for some microbes.
• Seed-microbe interactions appear highly complex,
  and may involve numerous mechanisms intrinsic to
  the growth and survival of microorganisms and
  plants (see figure below).

Microorganisms
Seed
METHODS
Water-soluble exudates were produced by seeds of
all weed species tested (Table 1). The nature of
the exudates and their biological activity was
hypothesized, in part,to contain compounds that
are antimicrobial, more specifically, antifungal.
Of the plant/fungal combinations tested, the
water-extracted exudate from kochia demonstrated
antifungal activity against Colletotrichum
graminicola, a known plant pathogen of soil
origin.
6.
1.
MDS analysis of bacterial (16S rDNA) TRF
profiles associated with seeds of three
different weed species following exposure to four
different soils. Bacterial assemblages tend to
cluster by weed species.
Weed species used in seed-microbe association
studies Velvetleaf (Abutilon theophrasti),
giant ragweed (Ambrosia trifida), Pennsylvania
smartweed (Polygonum pennsylvanicum), jimsonweed
(Datura stramonium), woolly cupgrass (Eriochloa
villosa). Soil descriptions All four soils
used as inocula were composed of silty clay loam
(organic content 4-5) typical of the region,
from fields located in the Champaign/Urbana, IL
area. USSEU (undisturbed site, no recent history
of agriculture) ASSET, ASLOC, ASSWC (long
history of agronomic use, corn/soybean
rotation). Exposure of seeds to soil
microorganisms Seeds were embedded into a
carbon-free mineral-salts agar medium following
inoculation of the plates with a 1104 dilution
of soil. Plates were incubated for up to three
months at 25C and seeds were observed for decay.
Seeds were removed and individually assayed
using cultivation, molecular community analysis,
and microscopy. Analysis of microbial
community assemblages on seeds DNA was extracted
from individual seeds (up to 10 replicate seeds
per soil type) using a modified phenol/chloroform
extraction method. Bacterial 16S rDNA was
amplified using primers f27 (FAM labeled at the
5 end) and 1492r. The PCR products were
digested with HaeIII, HhaI, and RsaI, and
analyzed for community profiles by terminal
restriction fragment (TRF) analysis. TRF data
(HhaI digests only) were analyzed using GeneScan
software followed by cluster analysis and
multidimensional scaling (MDS) using Bionumerics
software. Similarity values were generated using
the Dice coefficient. 16S rDNA products (1400)
were cloned and digested to obtain RFLP profiles
for selection of dominant clones for sequence
analysis. To analyze fungal communities, 18S rDNA
was amplified using primers EF3 and EF4 (1540bp),
which target a high number of fungi in the four
major phyla (Zygomycota, Ascomycota,
Basidiomycota, and Chytridiomycota), as well as
Oomycota. All 18S rDNA clones obtained were
selected for sequencing. Sequencing of the
ribosomal genes were performed by the Keck Center
for Biotechnology and sequences were analyzed by
BLAST search, and assembled into phylogenetic
trees using MacVector software.
Characterization of seed exudates and testing
for antifungal activity Seed exudates were
extracted in water using 1.5 g of seed (listed in
Table 1). Extracts were filtered (0.22 mm) and
condensed to dryness by lyophilization. To test
antifungal activities, 1 mg whole extract in
water was placed on a filter disk placed on
potato dextrose agar (PDA) inoculated with a
single known fungal species (listed in Table 1).
Plates were incubated at 25C for up to one week.
Zones of inhibition around the disk positively
indicated antifungal activity. To examine the
chemical characteristics of the exudates,
extracts were analyzed using silica gel thin
layer chromatography (mobile phase chloroform
methanolformic acid water (604042)).
Following separation, silica fractions were
recovered and re-extracted with water, followed
by testing of each fraction using the antifungal
plate assay described above.
4.
Phylogenetic tree representing cloned fungal (18S
rDNA) sequences(indicated in bold type) found
associated with seeds of three weed species
following exposure to soil-derived microbial
populations. Fungal associations were dominated
by members of the Ascomycota (98-99 sequence
identities), indicating specific relationships
may occur naturally between weed seeds and soil
fungi.
Velvetleaf
Jimsonweed
Woolly cupgrass
Samples of concentrated extracts from 1.5 g seeds
Kochia
Common ragweed
Giant ragweed
Velvetleaf
Zone of inhibition around kochia exract (on
filter disk)
Cumulative TRFs for all soil inocula with each
weed species.

2.
Phylogenetic distribution ( no. of clones per
soil type) of the cloned dominant (by TRF
abundance)16S rDNA sequences on decayed
velvetleaf seeds. Populations were dominated by
members of the phyla Proteobacteria and
Bacteroidetes.

Bacteroidetes
C. graminicola whole extract (1 mg) from
morning glory seeds, displaying no fungal growth
inhibition.
C. graminicola whole extract (1mg) from
kochia seeds, displaying significant fungal
growth Inhibition.
Soil inocula
a-Proteobacteria
8
7
b-Proteobacteria
6
5
4
3
g-Proteobacteria
vvl velvetleaf wc woolly cupgrass ps
Pennsylvania smartweed
2
C. graminicola inhibition from Fraction 7
d-Proteobacteria
Firmicutes
1
Actinobacteria
Sample loaded
STUDY OBJECTIVES
Silica gel thin layer chromatography of kochia
extract (color-enhance to show compound
separation) (chloroformMeOHformic acidwater,
604042). Fractions (1-8) recovered for
antifungal assays. Fraction 7 (orange) was
positive for inhibition of C. graminicola.
Distribution of Bacterial Phyla on Velvetleaf
Seeds

• To characterize microbial communities associated
  with seeds of different weed species, and to
  determine if these associations are specific.
• To identify seed-associated characteristics that
  may influence the growth or survival of soil
  microorganisms.

MDS analysis of dominant bacterial TRF
communities (using peaks representing gt8 of
total peak intensity per profile) associated with
seeds exposed to ASLOC soil. Velvetleaf seeds
are highly susceptible (99, bottom inset-VVL) to
microbial-mediated decay, compared to
Pennsylvania smartweed (10, top inset- PS),
woolly cupgrass (1), and jimsonweed (4).
Communities of the more dominant bacteria
associated with decayed velvetleaf seeds tended
to cluster, compared to those associated with
non-decayed seeds of other weed species.
3.
Carbon and nitrogen contents of seed components
of velvetleaf and giant ragweed. Seeds are
typically comprised of proteins, starches,
carbohydrates, oils, and complex polymers
potentially rich sources of nutrition for support
of microbial growth.
5.
Seed colonization and decay. Microbial
associations can be dense and varied, with
evidence of preference for seed components
subjected to microbial attack.
Protection
Woolly cupgrass
Pennsylvania smartweed
Jimsonweed
Velvetleaf
(by wt.)
Velvetleaf
Seed surface colonization (viewed by electron
microscopy)
Conceptualized view of seed bank interactions.
Intact seed
Decayed seed
Densely colonized seed aggregate
PS
Involucre
Fungal biomass
colonized seeds
References and Acknowledgments

• 1.Chee-Sanford, J.C., M.M. Williams II, A.S.
  Davis, and G.K. Sims. 2005. Do microbes
  influence seed bank dynamics? Weed Science.
  Submitted.
• 2.Chee-Sanford, J.C. L.M. Connor, T.J. Holman,
  E.Tang, Y. So, R. Marvelli, and M.M. Williams II.
  2004. Relationships between bacterial
  populations and weed seeds implications for soil
  seed bank microbial ecology. In Abstr. 104th
  General Meeting, ASM. N-013.
• 3.Chee-Sanford, J.C., L.M. Connor, and T.J.
  Holman. 2003. Characterization of weed seed
  decay activities mediated by natural soil
  microbial populations. In Abstr. 103rd General
  Meeting, ASM. N-024.
• This work was funded in part by a UI Special
  Undergraduate ResearchExperience (SURE) grant to
  Y. So.

Cavity
Giant ragweed
Embryo
Bifurcated seed embryo decayed, involucre
primarily intact
Cavity in involucre produced by microbial
activity
Seed surface colonization (viewed by electron
microscopy)
Intact seed
VVL