Virus Research 174 (2013) 148–151

Contents lists available at SciVerse ScienceDirect

Virus Research
journal homepage: www.elsevier.com/locate/virusres

Short communication

Within-host competition between barley yellow dwarf-PAV and -PAS

Gerod S. Hall a,∗ , Damon P. Little b,1
a
Cornell University, Department of Ecology and Evolutionary Biology, Corson Hall, Ithaca, NY 14853, USA
b
The New York Botanical Garden, 2900 Southern Blvd., Bronx, NY 10458, USA

a r t i c l e i n f o a b s t r a c t

Article history: The PAV and PAS species of barley yellow dwarf virus (Luteoviridae) share hosts, vectors, and have sym-
Received 7 December 2012 patric distributions. To better understand how competition between species influences virus growth
Received in revised form 17 March 2013 within the host, transmission rate between hosts, and ultimately virus population structure two exper-
Accepted 21 March 2013
iments were conducted. The first experiment varied the order of PAV and PAS inoculation and the time
Available online 29 March 2013
interval between the first and second inoculation. Relative virus concentration was measured at 8, 20,
33, and 45 days after primary virus inoculation (dpi). Regardless of the order of inoculation or the length
Keywords:
of time between inoculations, PAV dominated the virus population by 33 dpi (PAV concentration ranged
Luteovirus
BYDV from 55% to 89%). The second experiment measured the rate of vector transmission from single and
Cross-protection multiple infections. From single infections, the transmission rate was 67% for PAV and 60% for PAS. PAV
Interspecific competition had significantly greater odds of transmission for all competition treatments—except if PAS was given
a 15-day head start before inoculation with PAV. In the latter treatment, PAS was transmitted with a
greater frequency than PAV, but the difference was not statistically significant. Our data show persistent
co-infection between PAV and PAS, but PAV is more likely to be transmitted from mixed infections. Thus,
within-host interactions between PAV and PAS create conditions that promote both the competitive
exclusion of PAS, as well as co-existence between species and the maintenance of genetic diversity in the
host community.
© 2013 Elsevier B.V. All rights reserved.

Competition between pathogen species infecting the same host
can influence pathogen population dynamics and the wider host
community. Within host competition may lead to a change in
the pathogen transmission rate; the evolution of more virulent
pathogen types (Raberg et al., 2006); or produce more severe symp-
toms on the host than a single pathogen infection (Pruss et al.,
1997). The PAV and PAS species of barley yellow dwarf virus (BYDV,
Luteovirus, Luteoviridae) share hosts, vectors, and geographic distri-
butions (Hall, 2006; Hall et al., 2010). The present study investigates
interactions between the two species during multiple infection and
the attendant consequences for vector transmission.
The genus Luteovirus has three species—MAV, PAV, and PAS.
MAV is efficiently transmitted by Sitobion avenae F. only, while PAV
and PAS are transmitted by Rhopalosiphum padi L. and S. avenae. PAV
and PAS are indistinguishable by ELISA with polyclonal antibodies
(Chay et al., 1996). Thus, there is a gap in our understanding of PAV
and PAS epidemics as many BYDV field surveys utilize ELISA and in
so doing conflate the prevalence of the two species. PAV and PAS
∗ Corresponding author. Tel.: +1 212 8415301.
E-mail addresses: ghall@casacolumbia.org (G.S. Hall), dlittle@nybg.org
(D.P. Little).
1
Tel.: +1 718 817 8521; fax: +1 718 817 8101.
are distinguished by a maximum of 13% nucleotide divergence in
their coat protein gene (Bencharki et al., 1999) and 22% nucleotide
divergence in their RNA-dependent RNA polymerase gene (Hall,
2006).
It is not uncommon for grasses to be infected with more than
one BYDV species. In fact, the decline in the prevalence of MAV
in New York State has been attributed to competitive interactions
between PAV and MAV (Zhang and Holt, 2001). From 1957 to 1976,
the percentage of plants infected with MAV declined from 90% to
0%, while the percentage of PAV infections rose from 3% to 98%
(Rochow, 1979). There was not a corresponding shift in the den-
sity of R. padi or S. avenae over the same time period. Experimental
studies have established that during co-infection PAV has a more
negative effect on the within host growth of MAV than vice versa
(Wen et al., 1991). Thus, the historical shift in virus dominance
may be due to a decrease in MAV transmission from mixed infec-
tions (Gray et al., 1991). Field studies conducted using nucleic acid
based methods have identified plants with mixed PAV/PAS infec-
tions (Hall et al., 2010; Mastari et al., 1998). However, no study has
yet investigated within-host interactions between PAV and PAS or
been able to relate virus population size in BYDV co-infections to
the probability of aphid transmission. To describe the outcome of
competitive interactions between virus species, this study varied
the order of inoculation with PAV and PAS and the time interval

0168-1702/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.virusres.2013.03.013
 G.S. Hall, D.P. Little / Virus Research 174 (2013) 148–151
between the first and second inoculation. The potential effects of
competition on virus population structure were evaluated by aphid
transmission assays of virus from doubly infected plants.
We grew individual wheat (Triticum aestium L. cv. Lud) plants
from seeds in 6 in. pots and randomly assigned each to one of seven
inoculation treatments: single inoculation with PAV, single inocu-
lation with PAS, simultaneous inoculation with PAV and PAS, PAV
challenged with PAS after a 3-day delay, PAV challenged with PAS
after a 15-day delay, PAS challenged with PAV after a 3-day delay,
and PAS challenged with PAV after a 15-day delay. PAV and PAS
isolates used in this study were collected from agricultural fields in
central New York State. Isolate PAS-129 was obtained in 1992 from
migrating alate R. padi alighting on winter wheat (Chay et al., 1996).
PAV isolate Fa2k298 was obtained in 1998 from apterous R. padi col-
lected from oat. Viruses were maintained in continuous culture in
greenhouse grown oat plants since their initial isolation. Nine virus
source plants per treatment were inoculated according to the fol-
lowing protocol. Primary inoculations were carried out when plants
were 10 days old. R. padi from laboratory maintained disease-free
colonies were fed for 48 h on detached oat leaves infected with PAV
or PAS isolates. Aphids were then transferred to individual wheat
seedlings (eight aphids for single inoculation treatments and four
aphids for multiple inoculation treatments). Three or fifteen days
after the primary inoculation plants in the multiple inoculation
treatments were challenged with a second virus species. Aphids
were allowed an inoculation access period of 5 days after which
pots were fumigated. Virus source plants were tested for the pres-
ence of virus using ELISA (Gray et al., 1991) with PAV polyclonal
antibodies (Agdia Elkhart, IN). Reverse transcription-PCR (Hall and
Little, 2007) followed by restriction enzyme analysis was used to
verify the presence of multiple infections in source plants. RT-
PCR products were digested with restriction enzymes Ase I (PAV
specific) or Nci I (PAS specific) following the supplier’s recommen-
dations (New England Biolabs, Ipswich, MA). Digestion products
were viewed after electrophoresis in a 2% agarose gel.
Plants not infected with both PAV and PAS were removed from
the study. As a result, leaves were harvested from eight source
plants in two treatments and nine source plants in all others. In
all treatments it was necessary to sample some plants more than
once: for quantitative sequencing, the youngest leaf was harvested
from three plants per treatment at 8, 20, 33, and 45 days post pri-
mary virus inoculation (dpi) and stored at −20 ◦ C. For the aphid
transmission study, the youngest leaf was harvested at 30 dpi from
two to six plants per treatment (Table 1) and used immediately for
aphid feeding.
The quantitative sequencing assay developed for this study
uses terminator sequencing to distinguish between PAV and PAS
in pooled samples. As noted above a single leaf was collected
from virus source plants and the total nucleic acid extraction was
performed on 200 mg of tissue, except for treatment “one:one”
Table 1
Virus transmission by single aphids from plants in a given experimental condition.
Primary Challenge Inoculation Number of Number of
inoculation inoculation interval (days) virus source plants
plants inoculated
PAV 6 57
PAS 6 60
PAV PAS 6 60
PAV PAS 3 4 37
PAV PAS 15 2 20
PAS PAV 3 4 39
PAS PAV 15 2 16
Note. Comparisons of PAV and PAS transmission rate within treatment were made with Fisher’s exact test.
a
Odds ratio PAV transmission = odds PAV transmission/odds of PAS transmission.
*
p < 0.001.
149
where 100 mg of leaf tissue from plants singly inoculated with
PAV and PAS were combined into one sample. The gene region
sequenced has nine nucleotide sites that differentiate PAV and
PAS. Five sites that were consistently resolved after quality trim-
ming (p = 0.10) with PHRED version 0.020425.c (Ewing et al., 1998)
were used for analysis. PHRED (with the assistance of MUSCLE
(Edgar, 2004)) as invoked by polySNP (distributed by authors:
www.nybg.org/files/scientists/dlittle/polySNP.html) was used to
calculate area of the chromatogram peak associated with the PAV
and PAS polymorphism. A standard curve for each of the five sites
was generated by mixing known quantities of in vitro transcribed
viral RNA template in ratios from 1:9 to 9:1 (PAV:PAS) prior to RT-
PCR and sequencing. PAV as a proportion of the total peak area
was plotted against the frequency of template added to the RT-PCR
reaction. The equation of the best-fit line was used to calculate the
relative concentration of virus in the experimental sample when
given the peak area from the chromatogram. The complete protocol
and standard curves can be found in (Hall and Little, 2007).
The effects of multiple infection and inoculation delay on
relative virus population size were modeled with Generalized Esti-
mating Equations (GEEs). Statistical analyses were carried out in
STATA 11.1 (College Station, TX). GEE models provide robust regres-
sion estimates for repeated measures correlated within subject and
with non-normal response variables (Ballinger, 2004). The rela-
tive proportion of virus in a sample was normalized using arcsine
transformation. To test the effect of multiple infections on relative
virus population growth, we performed the contrast one:one vs.
simultaneous inoculation. To test the effect of inoculation interval
on relative virus population growth, we performed the contrasts
simultaneous inoculation vs. PAV challenged with PAS after a 3-
or 15-day delay and simultaneous inoculation vs. PAS challenged
with PAV after a 3- or 15-day delay. Treatment one:one cannot
be meaningfully compared to inoculation delay treatments due
to the confounding effects of multiple inoculation and inocula-
tion interval. Model covariates were inoculation treatment, harvest
day, and the treatment-by-harvest day interaction. GEE does not
model the between-subject variation. Hence, regression coeffi-
cients are interpreted as population means. The coefficient for
inoculation treatment measures the difference in mean propor-
tion of virus between the control and opposing treatment across
all harvest days. The coefficient for harvest day indicates how the
mean proportion of virus change over a 12-day span (the time
interval between sampling dates). A significant coefficient for the
treatment-by-harvest day interaction indicates that the effect of
inoculation treatment on the mean proportion of virus depends
upon harvest date.
Compared to singly infected hosts, simultaneously inoculated
hosts had less PAV than PAS at 8 and 20 dpi but more PAV than PAS
at 33 and 45 dpi (treatment-by-harvest day interaction: coef = 0.02,
z = 2.28, p = 0.02; Fig. 1(a)). We can conclude that the relationship
Number of % with % with Odds ratio PAV
plants infected infected infected transmissiona (95% CI:
PAV PAS lower, upper)
38 67
36 60
42 70 44 3.2 (1.4, 6.5)*
20 51 17 3.3 (2.1, 18.0)*
13 65 0
23 59 28 4.2 (1.6, 11.0)*
7 31 44 0.58 (0.14, 2.5)
150 G.S. Hall, D.P. Little / Virus Research 174 (2013) 148–151
Fig. 1. Relative population size PAV (a) or PAS (b). Treatment definitions:
one:one = leaves singly infected with PAV or PAS combined prior to nucleic acid
extraction, simultaneous = plants inoculated with PAV and PAS at the same time,
PAV 3-day head start = PAV primary virus challenged with PAS 3 days later, PAV
15-day head start = PAV primary virus challenged with PAS 15 days later, PAS 3-day
head start = PAS primary virus challenged with PAV 3 days later, PAS 15-day head
start = PAS primary virus challenged with PAV 15 days later.
between PAV and PAS is altered in mixed infections but we can-
not ascertain if absolute virus titer shifts upward, downward, or
remains the same. If PAV was given a 3-day growth advantage, it
dominated the virus population from 8 to 45 dpi (treatment-by-
harvest day interaction: coef = −0.04, z = −5.6, p < 0.001 Fig. 1(a)).
A 15-day PAV growth advantage served to increase the PAV
co-infection ratio by 9% at 20 dpi (treatment-by-harvest day inter-
action: coef = −0.03, z = −4.2, p < 0.001).
This study cannot determine the effect of competition on chal-
lenge virus population growth because for a given harvest date
the effect of treatment is confounded by the “late” inoculation of
plants with the challenge virus. For example, in treatment PAS 3-
day head start. PAV was inoculated to hosts 3 days after PAS. Thus,
at harvest date 8 dpi PAS had been allowed 8 days of replication
but PAV had been allowed only 5 days of replication. The appropri-
ate control—treatment simultaneous at 5 dpi—was not harvested.
Nonetheless, relative growth curves for the challenge virus are
presented in Fig. 1 to demonstrate the degree to which PAV is
the stronger competitor. Based on the overlapping PAV growth
curves for treatments simultaneous and PAS 3- or 15-day head start
(Fig. 1(a)), it appear that the PAS growth advantage had very little
effect on relative PAV population growth.
There was no difference in the PAS co-infection ratio between
treatments simultaneous and PAS 3-day head start (coef = −0.09,
z = −0.20, p = 0.84; Fig. 1(b)). A 15-day growth advantage for PAS
attenuated the negative effect of simultaneous inoculation on the
PAS co-infection ratio; however, the difference in co-infection ratio
decreased over time for the two treatments (treatment-by-harvest
day interaction: coef = −0.02, z = −2.2, p = 0.03; Fig. 1(b)). When PAS
was the challenge virus its relative growth was well below that
of treatment simultaneous (Fig. 1(b)). A decrease in the PAS inoc-
ulation interval from 15 to 3 days lead to increases in the PAS
co-infection ratio at 8 (5%) and 20 (7%) dpi (Fig. 1(b)).
There were two sample points where virus concentration var-
ied widely among treatment replicates. At 8 dpi in treatment PAS
3-day head start, relative PAS co-infection rate in replicate plants
was 86%, 76%, and 18%. In the same treatment sampled at 20 dpi
(but different plants than those sampled at the earlier date due
to destructive sampling) the PAS co-infection rate among repli-
cates was 89%, 61%, 38%, and 30%. This suggests that there is
variation in the ability of PAS to cross-protect (the ability of an
established pathogen to restrict infection of the host by another,
usually related, pathogen) hosts against PAV—particularly when
there is a short period of time between primary and secondary inoc-
ulation. Longer inoculation intervals enhanced the persistence of
cross-protection between genetic variants of Tobacco mosaic virus
(Tobamovirus; Tenllado et al., 1997), genetic variants of Satellite
tobacco mosaic virus (Tobamovirus; Kurath and Dodds, 1994), and
Beet soilborne mosaic virus and Beet necrotic yellow nein virus
(Furovirus; Mahmood and Rush, 1999). It is also possible that the
strength of cross-protection is related to the dose of the primary
virus or rate of phloem movement from the site of inoculation.
Neither of these factors was controlled among replicates as aphids
carrying an unknown number of virus particles were allowed to
feed on whole plants, not single leaves.
Inoculation order and the time interval between primary and
challenge inoculation influenced the relative rate of virus popu-
lation growth but not the competitive hierarchy between species.
By 45 dpi PAV was 70% of the virus population in most multiple
inoculation treatments. The only exception was treatment PAS 15-
day head start where PAV represented 59% of the virus population
at 45 dpi. Although PAS was the weaker competitor, it was not
prohibited from establishing an infection in multiple inoculation
treatments. Co-infection may serve as a mechanism for the main-
tenance of virus genetic diversity in the larger host community.
To assess the impact of within host competition on aphid trans-
mission, at 30 dpi leaves were collected from two to six virus source
plants per inoculation treatment (Table 1). Ten non-viruliferous
aphids were fed on each detached leaf for 48 h then individually
transferred to a single, healthy oat seedling. In some cases all 10
aphids did not survive the feeding the period and were not available
for transfer (Table 1). Aphids were allowed an inoculation access
period of 5 days after which pots were fumigated. At 14 dpi, virus
infected plants were identified by ELISA. Potentially doubly infected
plants (i.e. they received aphids fed on doubly infected leaf tis-
sue) were assayed for virus content by restriction enzyme analysis.
Fisher’s exact test was used to evaluate if for each inoculation treat-
ment, there were differences in transmission rate between PAV and
PAS.
The mean transmission efficiency (±standard deviation) of R.
padi from singly infected plants was 67% (±21%) for PAV and 60%
(±28%) for PAS. Aphids recovered PAV more frequently than PAS
from treatment simultaneously (OR = 3.2), PAV 3-day head start
(OR = 3.3), PAV 15-day head start (no PAS transmission), and PAS 3-
day head start (OR = 4.2). Treatment PAS 15-day head start was the
only case where there was no significant difference in transmission
rate between PAV and PAS (OR = 0.58).
Our results suggest that in a field setting, aphids will be less
likely to recover PAS from multiple infections. The two virus iso-
lates used in this study cannot represent the full diversity of each
species and further study is necessary to determine whether the
competitive dominance of PAV is generally applicable. If this were
the case, one would expect a greater prevalence of PAV and, per-
haps, a gradual loss of PAS in field populations (Zhang and Holt,
2001). A one-year survey of PAV and PAS isolates in winter wheat
fields in central New York State reported a total BYD disease
 G.S. Hall, D.P. Little / Virus Research 174 (2013) 148–151
prevalence of 15%, with at least 60% of infected plants carrying
PAV, 18% carrying PAS, and 5% carrying both PAV and PAS (Hall
et al., 2010). Continued surveillance of PAV and PAS is necessary to
determine if the observed pattern represents an intermediate stage
in the competitive exclusion of PAS—as has been hypothesized for
MAV—or independent epidemics that proceed at different rates due
to intrinsic differences in transmissibility between virus species.
Acknowledgments
We thank Alison Power, Micheal Milgroom, and John Losey for
reviewing previous drafts of this article. Research support was
provided by the National Institutes of Health Ruth L. Kirschstein
individual research fellowship (to G.S.H.). Laboratory work was per-
formed in the Evolutionary Genetics Core Facility in the Department
of Ecology and Evolutionary Biology at Cornell University.
References
Ballinger, G.A., 2004. Using generalized estimating equations for longitudinal data
analysis. Organizational Research Methods 7 (2), 127–150.
Bencharki, B., Mutterer, J., El Yamani, M., Ziegler-Graff, V., Zaoui, D., Jonard, G., 1999.
Severity of infection of Moroccan barley yellow dwarf virus PAV isolates corre-
lates with variability in their coat protein sequences. Annals of Applied Biology
134 (1), 89–99.
Chay, C.A., Smith, D.M., Vaughan, R., Gray, S.M., 1996. Diversity among isolates
within the PAV serotype of barley yellow dwarf virus. Phytopathology 86 (4),
370–377.
Edgar, R.C., 2004. MUSCLE: multiple sequence alignment with high accuracy and
high throughput. Nucleic Acids Research 32 (5), 1792–1797.
Ewing, B., Hillier, L., Wendl, M.C., Green, P., 1998. Base-calling of automated
sequencer traces using phred. I. Accuracy assessment. Genome Research 8 (3),
175–185.
151
Gray, S.M., Power, A.G., Smith, D.M., Seaman, A.J., Altman, N.S., 1991. Aphid transmis-
sion of barley yellow dwarf virus-acuisition access period and virus concentraion
requirements. Phytopathology 81 (5), 539–545.
Hall, G., 2006. Selective constraint and genetic differentiation in geographically
distant barley yellow dwarf virus populations. Journal of General Virology 87,
3067–3075.
Hall, G.S., Little, D.P., 2007. Relative quantitation of virus population size in mixed
genotype infections using sequencing chromatograms. Journal of Virological
Methods 146 (1–2), 22–28.
Hall, G.S., Peters, J.S., Little, D.P., Power, A.G., 2010. Plant community diversity influ-
ences vector behaviour and barley yellow dwarf virus population structure. Plant
Pathology 59 (6), 1152–1158.
Kurath, G., Dodds, J.A., 1994. Satellite tobacco mosaic virus sequence variants
with only five nucleotide differences can interfere with each other in a cross
protection-like phenomenon in plants. Virology 202, 1065–1069.
Mahmood, T., Rush, C.M., 1999. Evidence of cross-protection between beet soilborne
mosaic virus and beet necrotic yellow vein virus in sugar beet. Plant Disease 83
(6), 521–526.
Mastari, J., Lapierre, H., Dessens, J.T., 1998. Asymmetrical distribution of barley yel-
low dwarf virus PAV variants between host plant species. Phytopathology 88
(8), 818–821.
Pruss, G., Ge, X., Shi, X.M., Carrington, J.C., Vance, V.B., 1997. Plant viral syner-
gism: the potyviral genome encodes a broad-range pathogenicity enhancer
that transactivates replication of heterologous viruses. Plant Cell 9 (6),
859–868.
Raberg, L., de Roode, J.C., Bell, A.S., Stamou, P., Gray, D., Read, A.F., 2006. The role of
immune-mediated apparent competition in genetically diverse malaria infec-
tions. American Naturalist 168 (1), 41–53.
Rochow, W.F., 1979. Field variants of barley yellow dwarf virus-detection and fluc-
tuation during 2 years. Phytopathology 69 (6), 655–660.
Tenllado, F., GarciaLuque, I., Serra, M.T., DiazRuiz, J.R., 1997. Pepper resistance-
breaking tobamoviruses: can they co-exist in single pepper plants? European
Journal of Plant Pathology 103 (3), 235–243.
Wen, F., Lister, R.M., Fattouh, F.A., 1991. Cross-protection among strains of barley
yellow dwarf virus. Journal of General Virology 72, 791–799.
Zhang, X.S., Holt, J., 2001. Mathematical models of cross protection in
the epidemiology of plant-virus diseases. Phytopathology 91 (10),
924–934.