Changes in the Concentration of an Allexivirus During the Crop Cycle

of Two Garlic Cultivars

Eva E. Cafrune, Instituto de Fitopatologa y Fisiologa Vegetal INTA, Crdoba, Argentina; Mnica Balzarini,
Facultad de Ciencias Agropecuarias, Universidad Nacional de Crdoba, Crdoba, Argentina; and Vilma C. Conci,
Instituto de Fitopatologa y Fisiologa Vegetal INTA, Crdoba, Argentina

ABSTRACT
Cafrune, E. E., Balzarini, M., and Conci, V. C. 2006. Changes in the concentration of an Allexivirus during the crop cycle of two garlic cultivars. Plant Dis. 90:1293-1296.
Garlic can be infected by a number of viruses, including allexiviruses. The coat protein sequence
of an Allexivirus was detected in Argentina and deposited in the EMBL database as Garlic miteborne filamentous virus (accession number X98991); it has high homology with Garlic virus A
(GarV-A). For reliable virus detection, plants should be sampled when virus titer is high to reduce the risk of identifying infected plants as healthy. The objective of this study was to describe
fluctuations in the concentration of this Argentine isolate of GarV-A in two garlic cultivars, Morado-INTA and Nieve-INTA, throughout the crop cycle using the double-antibody sandwich
enzyme-linked immunosorbent assay (DAS-ELISA). Over a 2-year period, for both cultivars,
virus concentration was assessed in samples from the tips section of the youngest leaves of
GarV-A-infected plants, and from basal sections of both dormant and devernalized cloves of
stored bulbs of Morado-INTA. The concentration of GarV-A varied during the crop cycle, but
peaked at the beginning and again at the end of the crop cycle. Virus concentration was slightly
higher in devernalized cloves compared with dormant cloves of Morado-INTA. No correlation
between virus concentration and mean air temperature was observed. The results of this study
recommend sampling times at the beginning of the crop cycle at 64 to 81 days after planting, and
towards the end of the crop cycle to evaluate for the presence of GarV-A by DAS-ELISA.
Additional keywords: Allium sativum, virus identification

All commercial garlic (Allium sativum

L.) cultivars can be infected by a number
of viruses, with the most prevalent viruses
transmitted by aphids (mainly Potyvirus
and Carlavirus) and mites (Allexivirus)
(4,6,2628). Although the greatest economic losses in garlic caused by viruses
have traditionally been attributed to potyviruses such as Onion yellow dwarf virus
(OYDV) (2,15), viruses that are transmitted by mites (allexiviruses) also may have
a significant economic impact on the crop
(29).
The characterization of allexiviruses of
garlic primarily has been molecular. As a
result, very little information is available
on the biological and epidemiological
aspects of these viruses. Using nucleotide
sequence analysis, different species have
been identified within the genus Allexivirus, including Garlic virus A, B, C, D,
E, and X (GarV-A, B, C, D, E, and X, respectively; 4,2123,26). The coat protein
(CP) sequence of an Allexivirus detected in
Argentina and deposited in the EMBL

Corresponding author: Vilma C. Conci

E-mail: vconci@correo.inta.gov.ar
Accepted for publication 11 May 2006.

DOI: 10.1094 / PD-90-1293

2006 The American Phytopathological Society

database as Garlic mite-borne filamentous

virus (GarMbFV; accession number
X98991) had high homology with GarV-A
(13). The CP of this virus was expressed in
Escherichia coli cells as the immunogen to
produce polyclonal antibodies, which were
used in double-antibody sandwich enzyme-linked immunosorbent assay (DASELISA) and a tissue printing assay to test
garlic plants. Thus, the virus was shown to
be present in all major garlic cultivars
grown in Argentina (13). Hereafter, we
will use the name GarV-A to refer to this
Argentine isolate. Recently, the effect on
the yield of two allexiviruses was evaluated, and GarV-A caused significant reductions in bulb weight (14 to 32%) compared
with virus-free garlic plants (3).
Fluctuations in virus concentrations are
known to occur throughout the crop cycle
(10,12,17,18,25), and plants with low virus
content may test negative for a virus if
assayed when the virus titer is lowest (7).
The most widespread technique for virus
identification is DAS-ELISA because this
process combines economical reagents and
versatility, and accommodates large-scale
testing. Unfortunately, if samples are analyzed when virus concentration is quite
low, DAS-ELISA cannot differentiate between healthy and infected plants (6,7,11).
Therefore, for accurate detection, the optimum time for sampling must be determined based on virus titer.

The objectives of this study were to describe fluctuations in the concentration of

GarV-A throughout the crop cycle of two
garlic cultivars for two consecutive years
in Crdoba, Argentina, and to identify the
best time for reliable analysis of this virus.
Additionally, fluctuations in concentration
in relation to air temperature at the time of
sampling and during the 5 or 10 days prior
to sampling were examined.
MATERIALS AND METHODS
Plant material and sampling procedure. Two commercial garlic cultivars
were tested, Nieve-INTA (hereafter,
Nieve), a white garlic with a crop cycle of
235 to 245 days, and Morado-INTA (hereafter, Morado), a purple Asiatic garlic with
a crop cycle of 230 days. The field trials
were carried out in the Instituto de Fitopatologa y Fisiologa Vegetal of the Instituto
Nacional de Tecnologia Agropecuaria
(IFFIVE) in Crdoba, Argentina. In each
of two consecutive years (2000 and 2001),
150 cloves of each cultivar were planted in
three furrows (three replicates of 50 cloves
per cultivar) together, with 15 cm between
cloves within the furrow. Plants were furrow irrigated every 7 to 10 days.
At the beginning of the experiment, all
plants of both cultivars were tested by
immunosorbent electron microscopy with
decoration (ISEM-D) according to the
method described by Milne and Luisoni
(19), using GarV-A antiserum from stock
at IFFIVE that was produced as described
by Helguera et al. (13). Only plants that
tested positive were selected to evaluate
fluctuation in virus concentration. In all,
92 plants of Nieve and 80 plants of Morado were assayed during the first year,
and 53 plants of each cultivar were assayed
during the second year. Each of these
plants was marked with a label. Virus concentration and percentage of plants positive by DAS-ELISA were determined at
various sampling dates throughout the crop
cycle. In addition, these were assessed
during bulb storage for Morado.
In 2000, Nieve was planted on 11 April
and harvested on 20 November. Samples
were collected at 29 (10 May), 81 (1
July), 155 (13 September), 183 (11 October), and 211 days (8 November) after
planting. In 2001, Nieve was planted on
20 April and harvested on 30 November.
Samples were collected 40 (30 May), 81
(10 July), 146 (13 September), 181 (18
Plant Disease / October 2006

1293

October), and 214 days (20 November)

after planting.
Similarly, in 2000, Morado was planted
on 6 June and harvested on 28 November.
Samples were collected 29 (5 July), 64 (9
August), 100 (14 September), 127 (11
October), and 155 days (8 November) after
planting. In addition, samples of dormant
cloves were collected 211 days after planting (3 January 2001) and devernalized
cloves were sampled 294 days after planting (27 March 2001). In 2001, cloves of
this cultivar were planted on 20 April and
harvested on 25 November. Samples were
collected at 40 (30 May), 81 (10 July), 108
(6 August), 146 (13 September), and 181
days (18 October) after planting. Samples
of dormant cloves were collected 279 days
after planting (24 January 2002) and devernalized cloves were sampled 370 days
after planting (25 April 2002). Cloves were
regarded as dormant immediately after
harvest. Then, the bulbs were marked and
stored in a dry place at room temperature
(ranging from 17 to 27C) and they were
considered as devernalized when the

sprout leaf reached 75% of the length of

the storage leaf in the clove (1).
Serological analysis. Plants were analyzed by DAS-ELISA (5,7). For each plant,
about 2 cm of the apical portion of the
youngest leaf was assayed at each sampling
date. Bulbs of Morado also were analyzed
on the last two sampling dates. One clove
from each bulb was assayed as the dormant
clove, and another clove of the same bulb as
the devernalized clove at the later sampling
period. In each case, the sample consisted of
a cube (1.0 by 0.5 by 0.5 cm) taken from the
basal portion of the clove. All samples were
ground (1:5, wt/vol) in buffer (phosphate
buffered saline: 0.02 M phosphate plus 0.15
M NaCl, pH 7.4 [PBS] + 0.05% Tween-20
+ 2% polyvinylpyrrolidone + 2% nonfat
dried milk) and tested by DAS-ELISA.
Each DAS-ELISA plate also held extracts
from five healthy plants. These plants were
obtained as virus-free garlic plants (8) and
had tested free of GarV-A by ISEM-D
(13,19). Two plants that tested positive for
GarV-A by ISEM-D were included as positive controls (8).

GarV-A antiserum (13) was used for the

DAS-ELISA. Immunoglobulin (IgG) and
the IgG conjugate with alkaline phosphatase (5). The IgG was diluted 1:1000 in
coating buffer (0.05 M sodium carbonate,
pH 9.6), and alkaline phosphataseconjugated IgG was diluted 1:1000 in conjugate buffer (PBS + 0.05% Tween-20 +
2% polyvinylpyrrolidone + 0.2% egg albumin + 2% nonfat dried milk). The substrate used was p-nitrophenyl phosphate
disodium at 0.8 mg/ml of substrate buffer.
Virus concentration. A relative virus
concentration (RC) value (7) was calculated for each sample to compare the
ELISA absorbance values (measured at
405 nm) at different sampling dates. The
RC value for a sample was calculated as
the ratio between the ELISA absorbance
value for that sample and the mean plus
two standard deviations of the ELISA absorbance of five healthy control samples
tested in the same plate. Absorbance readings were measured with a Dynatech MR
4000 spectrophotometer (Dynatech, Guernsey Channel Islands, UK). Mean values for

Fig. 1. Mean relative concentrations (RCs) of Garlic virus A (GarV-A; left Y-axis) and percentage of garlic plants that tested positive by direct doubleantibody sandwich enzyme-linked immunosorbent assay (DAS-ELISA; right Y-axis). Known GarV-A-infected plants of cvs. Nieve-INTA and Morado-INTA
were sampled in different months during the crop cycle in A, 2000 and B, 2001 and in C, 2000 and D, 2001, respectively. Dormant cloves (dc) and devernalized cloves (dvc) for C, 2001 and D, 2002 also were sampled. The threshold for positive samples was set at RC = 1.
1294

Plant Disease / Vol. 90 No. 10

the RC of virus were calculated from the

individual RC values of each replicate at
each sampling date during the crop cycle.
A threshold RC value of 1 was regarded as
positive, and was used to calculate the
percentage of plants that tested positive for
GarV-A by DAS-ELISA at each sampling
date. Statistical differences among mean
RC values from different sampling dates
were assessed by t tests according to the
Bonferroni criterion, using InfoStat (14).
Weather data. Daily mean temperatures
were recorded throughout the crop cycle
using a Metos-Dat meteorological station
(PESSL-WEIZ, Austria) located in the
field where the trials were performed.
Correlations of the mean RC values with
mean daily temperatures on the sampling
days and with mean daily temperatures
over the 5 or 10 days before each sampling
day were make for each cultivar in each
year using Pearsons correlation coefficient
(9).
RESULTS
All Nieve plant samples had RC values
ranging from 2.35 to 20.67 by DASELISA. Thus, in comparison with the
threshold RC value of 1.00, all the plants
of this cultivar tested positive for GarV-A
at each sampling date (Fig. 1A and B). In
both years, samples collected in July (81
days after planting) and November (211 to
214 days) had significantly higher mean
RC values than those collected in May (29
to 40 days), September (146 to 155 days),
and October (181 to 183 days) (P < 0.05;
Fig. 1A and B).
Morado samples had RC values ranging
from 0.47 to 24.49 by DAS-ELISA; 65 to
94% of the plants tested positive for GarVA at different sampling dates and years
(Fig. 1C and D). In 2000, the average RC
values for Morado were significantly
higher in August (64 days), October (127
days), and November (155 days) than in
July (29 days) and September (100 days)
(P < 0.05). In 2001, higher average RC
values were recorded for this cultivar in
July (81 days), September (146 days), and
October (181 days) than in May (40 days)
and August (108 days) (P < 0.05). In both
years, average RC values in devernalized
and dormant cloves were lower than those
sampled during the crop cycle. The RC
values of this cultivar (1.03 to 2.55), correlated with the variation in the percentage
of plants that tested positive by DASELISA (r = 0.70, P < 0.05; Fig. 1C and D).
At the first sampling date, when mean RC
values were low (6.09 in July 2000 and
4.03 in May 2001), the percentage of
plants that tested positive for GarV-A by
DAS-ELISA was 73 and 75% in 2000 and
2001, respectively. At the second sampling
date, the RC values increased (9.58 in
August 2000 at 64 days and 7.39 in July
2001 at 81 days) along with the percentage
of GarV-A-infected plants (88 and 87% in
2000 and 2001, respectively). At the third

sampling, the RC values fell (to 4.92 in

September 2000 at 100 days and to 5.07 in
August 2001 at 108 days) as did the percentage of GarV-A-infected plants (68 and
71% in 2000 and 2001, respectively).
Thereafter, the mean RC values increased
(to 8.22 in October 2000 at 127 days and
to 7.88 in September 2001 at 146 days), as
did the percentage of plants infected with
GarV-A (76 and 85% in 2000 and 2001,
respectively). Results were similar for the
last sampling date of the crop cycle. In
harvested bulbs, for dormant cloves the
mean RC values in year 2000 and 2001
were 2.52 and 2.73, respectively, and the
percentage of cloves that tested positive for
GarV-A were 81 and 69%, respectively, for
the two years. For devernalized cloves, the
RC values were 3.98 and 3.18 and the
percentage of infected plants was 94 and
91% in 2000 and 2001, respectively.
Mean RC values and mean daily temperatures were not significantly correlated
for either cultivar, regardless of whether
temperature was recorded for the day of
sampling or for the 5 or 10 days preceding
each sampling period (P > 0.05) (Fig. 2).
DISCUSSION
The concentration of GarV-A in garlic
plants varied through the crop cycle in the
2 years of the experiment for both cultivars. The changes in relative virus concentration through the season were consistent
for Nieve. Typically, garlic populations
appear to be infected by viruses at different
concentrations, and the virus concentration
within a plant varies through the crop cycle
(7). Despite the fluctuations in RC values
for GarV-A during the crop cycle, all
plants of cv. Nieve tested positive for
GarV-A by DAS-ELISA at all sampling

dates, because all plants had high virus

concentration.
In Morado, the specific sampling dates
differed between the 2 years; therefore, it
is difficult to compare results directly between years. However, it still is important
to note the similar patterns of RC fluctuations for both years. The peaks in concentration of GarV-A observed occurred earlier in 2001 than 2000. This difference
probably was due to the garlic being
planted 47 days later in 2000 (6 June) than
in 2001 (20 April) (16,20,24). The fluctuations in GarV-A concentration followed a
pattern similar to changes in the percentage of plants that tested positive for GarVA. Although all plants used in the assays
initially tested positive for GarV-A, some
of them were negative by DAS-ELISA. In
harvested garlic bulbs, an increase in mean
RC values was accompanied with an increase in the percentage of GarV-Ainfected plants. Even though the concentration of GarV-A was lower for the assays
of bulbs in storage (RC = 2.52 to 3.98)
than those of leaves sampled from the field
(RC = 3.95 to 8.42), the percentages of
plants that tested positive for GarV-A were
as high as the leaves sampled from the
field. Thus, GarV-A can be detected from
bulbs, and sampling devernalized cloves is
the best choice for testing incidence of
infected bulbs. On devernalized cloves also
were a good option to detect Leek yellow
stripe virus (LYSV) in four garlic cultivars
tested by DAS-ELISA (7). This is important for growers, so that an infected bulbs
will not be planted.
Developing a general rule for choosing a
date to test garlic plants for GarV-A may
be difficult, in spite of similarity in the
concentration patterns of GarV-A detected

Fig. 2. Mean relative Garlic virus A (GarV-A) concentration and mean daily temperature over the 5
days before each sampling period for garlic cvs. A and B, Nieve and C and D, Morado in A and C,
2000 and B and D, 2001.
Plant Disease / October 2006

1295

throughout the season. LYSV concentration in the garlic cvs. Blanco Mendoza and
Norteo peaked 60 days after planting,
which coincided with the second sampling
date for GarV-A in the present study (64 to
81 days after planting; 7).
The correlation between GarV-A concentration and air temperature also was
evaluated. In the present study, mean daily
temperature and GarV-A concentration
during the crop cycle were not significantly correlated.
Based on our results, optimum sampling
times for GarV-A testing appear to be at
about 64 to 81 days after planting or toward the end of the crop cycle. On the
bulbs, the best sampling time is in devernalized cloves.
ACKNOWLEDGMENTS
We thank the National Agricultural Technology
Institute (INTA), the National Council for Scientific and Technical Investigations (CONICET),
Crdoba Science Agency, and the Argentine National Secretariat for Science and Technology
(FONCYT and CABBIO) projects for providing the
funds for this study.
LITERATURE CITED
1. Burba, J. L., Mller, J. J., and Casali, V. W. D.
1983. Relaciones entre el ndice visual de superacin de dormicin (IVD) en ajo (Allium
sativum L.) con el tamao y posicin de bulbillos. Rev. Cs. Agropec. IV:99-102.
2. Canavelli, A., Nome, S. F., and Conci, V. C.
1998. Efecto de distintos virus en la produccin de ajo (Allium sativum) Rosado Paraguayo. Fitopatol. Bras. 23:354-358.
3. Cafrune, E. E., Perotto, M. C., and Conci, V.
C. Effect of two Allexivirus isolates on garlic
yield. Plant Dis. 90:898-904.
4. Chen, J., Chen, J., and Adams, M. J. 2001.
Molecular characterization of a complex mixture of viruses in garlic with mosaic symptoms
in China. Arch. Virol. 146:1841-1853.
5. Clark, M. F., and Adams, A. N. 1977. Characteristics of the microplate method of enzyme linked immunosorbent assay (ELISA)
for the detection of plant viruses. J. Gen.

1296

Plant Disease / Vol. 90 No. 10

Virol. 34:475-482.
6. Conci, V. C. 1997. Virus y fitoplasmas de ajo.
Pages 267-291 in: 50 Temas Sobre Produccin
de Ajo, Vol. 3 J. L. Burba, ed. Estacin Experimental Agropecuaria-Instituto Nacional de
Tecnologa Agropecuaria La Consulta, Mendoza, Argentina.
7. Conci, V. C., Lunello, P., Buraschi, D., Italia,
R. R., and Nome, S. F. 2002. Variations of
Leek yellow stripe virus concentration in garlic
and incidence in Argentina. Plant Dis.
86:1085-1088.
8. Conci, V. C., and Nome, S. F. 1991. Virus free
garlic (Allium sativum L.) plants obtained by
thermotherapy and meristem tip culture. J.
Phytopathol. 132:186-192.
9. Conover, W. J. 1998. Some methods based on
ranks. Pages 269-467 in: Practical Nonparametric Statistics, 3rd ed. John Wiley & Sons,
New York.
10. Dal-Zotto, A., Nome, S. F., Di Rienzo, J. A.,
and Docampo, D. M. 1999. Fluctuations of
Prunus necrotic ringspot virus (PNRSV) at
various phenological stages in peach cultivars.
Plant Dis. 83:1055-1057.
11. Dovas, C. I., Hatziloukas, E., Salomon, R.,
Barg, E., Shiboleth, Y., and Katis, N. I. 2001.
Comparison of methods for virus detection in
Allium spp. J. Phytopathol. 149:731-737.
12. Dovas, C. L., Mamolos, A. P., and Katis, N. I.
2002. Fluctuations in concentration of two potyviruses in garlic during the growing period
and sampling conditions for reliable detection
by ELISA. Ann. Appl. Biol. 140:21-28.
13. Helguera, M., Bravo-Almonacid, F., Kobayashi, K., Rabinowicz, P. D., Conci V., and
Mentaberry, A. 1997. Immunological detection
of a GarV-type virus in Argentine garlic cultivars. Plant Dis. 81:1005-1010.
14. InfoStat, Version 1.5. 2003. Manual Del
Usuario. Argentina E. B., editor. Primera
Edicin, Grupo InfoStat, Facultad de Ciencias
Agropecuarias, Universidad Nacional de Crdoba, Argentina.
15. Lot, H., Chovelon, V., Souche, S., and Delecolle, B. 1998. Effects of Onion yellow dwarf
and Leek yellow stripe viruses on symptomatology and yield loss of three French garlic
cultivars. Plant Dis. 82:1381-1385.
16. Mann, L. K. 1952. Anatomy of the garlic bulb
and factors affecting bulb development. Hilgardia 21:195-251.
17. March, G. J., Balzarini, M., Ornaghi, J. A.,

18.
19.

20.

21.

22.

23.

24.

25.

26.

27.

28.
29.

Beviacqua, J. E., and Marinelli, A. 1995. Predictive model for Mal de Rio Cuarto disease
intensity. Plant Dis. 79:1051-1053.
Matthews, R. E. F. 1991. Plant Virology, 3rd
ed. Academic Press, London.
Milne, R. G., and Luisoni, E. 1977. Rapid
immune electron microscopy of virus preparations. Pages 265-281 in: Methods in Virology,
Vol. 6. K. Maramorosch and H. Koprowski,
eds. Academic Press, New York.
Portela, J. A. 1996. El ambiente como regulador del desarrollo de las plantas. Su efecto
sobre el ajo (Allium sativum L.). Av. Hortic.
1:19-40.
Song, S. I., Song, J. T., Chang, M. U., Lee, J.
S., and Choi, Y. D. 1997. Identification of one
of the major viruses infecting garlic plants,
Garlic virus X. Mol. Cell. 7:705-709.
Song, S. I., Song, J. T., Kim, C. H., Lee, J. S.,
and Choi, Y. D. 1998. Molecular characterization of the Garlic virus X genome. J. Gen. Virol. 79:155-159.
Sumi, S., Tsuneyoshi, T., and Furutani, H.
1993. Novel rod-shaped viruses isolated from
garlic, Allium sativum, possessing a unique
genome organization. J. Gen. Virol. 74:18791885.
Takagi, H. 1990. Garlic Allium sativum L.
Pages 109-146 in: Onions and Allied Crops,
Vol. 3. H. D. Rabinowitch and J. L. Brewster,
eds. CRC Press, Boca Raton, FL.
Torrance, L., and Dolby, C. A. 1984. Sampling
conditions for reliable routine detection by enzyme-linked immunosorbent assay of three
ilarviruses in fruit trees. Ann. Appl. Biol.
104:267-276.
Tsuneyoshi, T., and Sumi, S. 1996. Differentiation among garlic viruses in mixed infections
based on RT-PCR procedures and direct tissue
blotting
immunoassays.
Phytopathology
86:253-259.
Van Dijk, P. 1991. Mite-borne virus isolates
of cultivated Allium species and their classification into two new rymoviruses in the family Potyviridae. Neth. J. Plant Pathol. 97:381399.
Van Dijk, P. 1993. Survey and characterization
of potyviruses and their strains of Allium species. Neth. J. Plant Pathol. 99. Suppl. 2:1-48.
Van Dijk, P., and Van der Vlugt, R. A. 1994.
New mite-borne virus isolates from rakkyo,
shallot and wild leek species. Eur. J. Plant
Pathol.100:269-277.