Volume 55, Issue 3 p. 458-467
Free Access

Interactions between a crinivirus, an ipomovirus and a potyvirus in coinfected sweetpotato plants

S. B. Mukasa,

Department of Crop Science, Makerere University, PO Box 7062, Kampala, Uganda;

Department of Plant Biology and Forest Genetics, Swedish University of Agricultural Sciences (SLU), Box 7080, SE-750 07 Uppsala, Sweden

Search for more papers by this author
P. R. Rubaihayo,

Department of Crop Science, Makerere University, PO Box 7062, Kampala, Uganda;

Search for more papers by this author
J. P. T. Valkonen,

Department of Plant Biology and Forest Genetics, Swedish University of Agricultural Sciences (SLU), Box 7080, SE-750 07 Uppsala, Sweden

Department of Applied Biology, University of Helsinki, PO Box 27, FIN-00014 Finland

E-mail: jari.valkonen@helsinki.fi

Search for more papers by this author
First published: 23 February 2006
Citations: 85


Novel and severe symptoms of chlorosis, rugosity, leaf strapping and dark green islands, designated as sweetpotato severe mosaic disease (SPSMD), were caused by dual infection of Sweet potato mild mottle virus (SPMMV; Ipomovirus) and Sweet potato chlorotic stunt virus (SPCSV; Crinivirus) in three East African sweetpotato cultivars (Tanzania, Dimbuka and New Kawogo). The storage root yield was reduced by ∼80%, as compared with healthy plants under screenhouse conditions in Uganda. Plants infected with SPMMV or SPCSV alone showed nonsignificant or 50% yield reduction, respectively. SPCSV reduced resistance to SPMMV in sweetpotato, similar to the situation with resistance to Sweet potato feathery mottle virus (SPFMV; Potyvirus) that breaks down following infection with SPCSV, followed by development of sweet potato virus disease (SPVD). In single virus infections with SPMMV and SPFMV or their coinfection, cvs Tanzania and Dimbuka were initially systemically infected, displayed symptoms and contained readily detectable virus titres, but new leaves were symptomless with very low virus titres, indicating recovery from disease. In contrast, cv. New Kawogo remained symptomless and contained low SPMMV and SPFMV titres following graft inoculation. These moderate and high levels of resistance to SPMMV and SPFMV, respectively, were lost and cultivars succumbed to a severe disease following coinfection with SPCSV. The synergistic interactions increased titres of SPMMV and SPFMV RNA by ∼1000-fold as quantified by real-time PCR, whereas SPCSV titres were reduced twofold, indicating an antagonistic interaction. Coinfection with SPMMV and SPFMV caused no detectable changes in virus titres or symptom severity.


Coinfection of plants with two or several unrelated viruses often results in a more severe disease than the sum effect of infection with each of the viruses alone and is known as viral synergism (Goodman & Ross, 1974; Vance et al., 1995; Pruss et al., 1997). Synergism has also been observed between viruses and their satellite viruses or RNAs (Rodriguez-Alvarado et al., 1994; Sanger et al., 1994; Scholthof, 1999), and between a virus and a viroid (Valkonen, 1992). The mechanisms of synergism may vary. One of the viruses may aid the movement of another virus, thereby enabling it to invade tissues it could not otherwise infect, or the replication and accumulation of one of the viruses may be enhanced (Barker, 1989; Savenkov & Valkonen, 2001). It is also possible that both viruses benefit from the coinfection (Scheets, 1998; Fondong et al., 2000). Antagonistic interactions are sometimes observed, in which the unrelated viruses suppress the infection of each other (Poolpol & Inouye, 1986). However, this is different from cross-protection that takes place between closely related viruses or virus strains (Fraser, 1998).

Many examples of viral synergism include a potyvirus, the titres of which are unaffected or decline while titres of the coinfecting virus increase (Rochow & Ross, 1955; Poolpol & Inouye, 1986; Goldberg & Brakke, 1987; Pruss et al., 1997; Vance et al., 1995; Scheets, 1998). In contrast, only in a few cases has coinfection with another virus led to increased potyvirus titres (Valkonen, 1992; Scheets, 1998). One such case is the sweetpotato virus disease (SPVD) caused by a dual infection with Sweet potato feathery mottle virus (SPFMV; genus Potyvirus, family Potyviridae) and Sweet potato chlorotic stunt virus (SPCSV; genus Crinivirus, family Closteroviridae), in which the titres of SPFMV may increase by over 600-fold whereas no increase is observed in the titres of SPCSV compared with sweetpotato plants (Ipomoea batatas) infected with each virus alone (Karyeija et al., 2000a). SPVD is the most severe disease of sweetpotatoes and can reduce yields by 56–98% (Gibson et al., 1998; Karyeija et al., 1998).

Recent surveys around Lake Victoria in East Africa (Mukasa et al., 2003a; Ateka et al., 2004b; Tairo et al., 2004) indicate that in many sweetpotato plants, severe diseases are associated with infection with SPCSV and Sweet potato mild mottle virus (SPMMV; genus Ipomovirus, family Potyviridae; Colinet et al., 1996). Plants seropositive for SPMMV alone show mild mottling and mosaic symptoms (Hollings et al., 1976; Mukasa et al., 2003a), whereas plants seropositive for SPMMV and SPCSV display severe symptoms of leaf malformation and poor growth (Mukasa et al., 2003a). These observations suggest that SPMMV may synergize with SPCSV. However, the involvement of other, unknown viruses cannot be excluded when observing symptoms of plants infected in the field. Furthermore, considering the more widely studied potyviruses (genus Potyvirus), recent studies are not unanimous as to which of them SPCSV can synergize within sweetpotatoes. Virus isolates designated as LSU2 and LSU5 in the US (Souto et al., 2003) are phylogenetically closely related to SPV-2 from Nigeria (Ateka et al., 2004a) and all may belong to the proposed species Sweet potato virus Y (Ateka et al., 2004a), but only LSU2 and LSU5 could synergize with SPCSV (Souto et al., 2003; Ateka et al., 2004a). Similarly, the East African strains of SPFMV synergize with SPCSV (Gibson et al., 1998; Karyeija et al., 2000a, 2000b), in contrast to the American strain C of SPFMV (Souto et al., 2003). In Argentina, triple infection with SPCSV, SPFMV and Sweet potato mild speckling virus (SPMSV, genus Potyvirus) results in a severe sweetpotato chlorotic disease (Di Feo et al., 2000).

SPMMV was first described from East Africa (Kenya, Uganda and Tanzania; Hollings et al., 1976) and later detected serologically in South Africa, Egypt, Indonesia, New Zealand and Peru (Fletcher et al., 2000; Tairo et al., 2005), but there is little information about its interactions with other viruses in coinfected plants. Therefore, the recent findings of its common occurrence in sweetpotato crops in East Africa and the possible implications of synergism to virus epidemiology (Zhang et al., 2001) prompted us to advance understanding of the nature of diseases SPMMV may cause. The aim of this study was to investigate whether SPMMV interacts with SPCSV and/or SPFMV in coinfected sweetpotato plants of three East African cultivars in terms of altering virus accumulation in infected tissues and causing higher yield losses.

Materials and methods

Plant material

The East African sweetpotato cvs Tanzania and New Kawogo are ranked among the five superior cultivars grown in Uganda (Mwanga et al., 1995) and are widely grown, whereas cv. Dimbuka dominates in the central and eastern parts of Uganda (Mukasa et al., 2003a). These cultivars have several phenotypic differences indicating genetic diversity (Table 1). Tanzania and New Kawogo are moderately resistant and resistant to SPFMV, respectively (Karyeija et al., 2000b; Mwanga et al., 2002), but resistance to SPFMV in cv. Dimbuka was not known. The responses of all cultivars to SPMMV were unknown.

Table 1. Phenotypic characteristics of three sweetpotato cultivars used in this study
Cultivar descriptora Sweetpotato cultivar
Tanzania New Kawogo Dimbuka
Growth habit Semi-erect Semi-erect Spreading
Vine anthocyanin pigmentation Few purple spots Purple spots None
Leaf petiole pigmentation Light purple Purple None
Number of leaf lobes Five Five One
General leaf outline Hastate Lobed Triangular
Storage root skin colour Cream Brownish orange Cream
Storage root flesh colour Pale yellow White White
Maturity period 5–6 months 6–7 months 4–5 months

Healthy in vitro plantlets of cvs Tanzania and New Kawogo were obtained from the International Potato Centre (CIP), Lima, Peru. The vines were propagated in plastic pots containing composite soil and maintained at Uppsala Genetics Centre, Sweden, at 25–30°C in a glasshouse. Sodium halide lamps were used to extend day length to 16 h during winter.

Vigorous and healthy-looking vines of cv. Dimbuka were collected from farmers’ fields in Uganda and indexed for viruses by grafting on to plants of Ipomoea setosa, a nearly universal indicator plant for sweet potato viruses (Schaefers & Terry, 1976), which were observed for symptoms and tested by enzyme-linked immunosorbent assay (ELISA) as described below. The plants were also indexed with PCR using potyvirus-, begomovirus-, carlavirus- or crinivius-specific primers, as described previously (Mukasa et al., 2003a). After repeated tests, plants of this cultivar that did not react positively to viruses in any tests were propagated through stem cuttings and used for study.

Virus isolates and combinations

The identity of the virus isolates SPMMV-Tor (Mukasa et al., 2003b), SPFMV-Nam1 (Kreuze et al., 2000) and SPCSV-Ug (Kreuze et al., 2002), all of which originated in Uganda, was determined using reverse transcriptase polymerase chain reaction (RT-PCR) and partial sequencing. The individual viruses were maintained in plants of cv. Tanzania, each in a separate glasshouse compartment at the Uppsala Genetics Centre, Sweden. Plants infected with SPMMV-Tor, SPFMV-Nam1 or SPCSV-Ug were top-grafted with each other in order to create all possible virus combinations: SPMMV + SPCSV, SPFMV + SPCSV, SPMMV + SPFMV, and SPMMV + SPFMV + SPCSV. Each virus or the combinations were graft-transmitted to plants of the three sweetpotato cultivars and I. setosa. Five weeks after graft inoculation, serological and RT-PCR detection were carried out to confirm systemic infection. Foliar symptoms and symptom severity were recorded weekly until 10 weeks postinoculation.

Serological detection of viruses

Detection of SPMMV and SPFMV was carried out using nitrocellulose membrane ELISA (NCM-ELISA), as described previously (Gibb & Padovan, 1993). For virus quantification, double-antibody sandwich ELISA (DAS-ELISA) (SPMMV) or triple-antibody sandwich ELISA (TAS-ELISA) (SPFMV and SPCSV) were used, as described previously (Gibson et al., 1998; Karyeija et al., 2000b). Leaves were tested by taking three discs (diameter 1 cm) per leaf and grinding them in a polyvinyl bag in ELISA extraction buffer (1 g leaf per 3 mL buffer). The polyclonal antibodies (PAb) to SPMMV coat protein (CP) were kindly provided by Dr H. J. Vetten (Federal Biological Research Centre for Agriculture and Forestry, Germany). Virus-specific PAb and monocolonal antibodies to the CP of SPFMV and SPCSV were provided by CIP. Anti-mouse alkaline phosphatase-conjugated goat IgG used in TAS-ELISA was obtained from Sigma Chemical Co. Microtitre plates were purchased from Greiner Labortechnik (Germany). Absorbances (A450) were measured at 405 nm 2 h after adding the substrate (p-nitrophenyl phosphate) using a Benchmark Microplate reader (Bio-Rad Laboratories Inc.).

Detection of viral RNA by RT-PCR

Total RNA was extracted from leaf tissue (200 mg) using the TRIzol LS Reagent (Invitrogen Ltd) according to the manufacturer's instructions and quantified using a spectrophotometer. An average of 500 µg of RNA per 1 g of leaf tissue was obtained. The concentrations of RNA samples were equalized. RT-PCR was done using a RevertAid First Strand cDNA Synthesis Kit (MBI Fermentas) with the following CP gene-specific forward and reverse primers: MMA1 (5′-CCATTCAGAACAAGGAGC-3′) and MMD2 (5′-TTGAGCTCCTCTCAGACT-3′) for SPMMV, CP1 and CP2 for SPCSV (Alicai et al., 1999) and CP106 and UTR34 for SPFMV (Kreuze et al., 2000). PCR products were resolved in 1% agarose gels and visualized by ethidium bromide staining under UV light. To confirm the identity of the PCR products, they were cloned in the pCRII vector in Escherichia coli, strain 10F′ (TOPO TA Cloning Kit; Invitrogen) for sequence analysis. Plasmid DNA was purified using the QIAprep Spin Miniprep Kit (Qiagen Inc.). Sequencing was done using the DYEnamicTM ET Terminator Cycle Sequencing Kit (Amersham Pharmacia Biotech AB) and an automated ABI Prism 377 DNA sequencer (Perkin-Elmer Applied Biosystems).

NASH detection and quantification of viral RNA

Purified plasmid DNA containing the sequenced CP gene fragments were used for PCR synthesis of virus-specific probes for nucleic acid spot hybridization (NASH) as described previously (Brandsma & Miller, 1980). PCR products were purified using a QIAquick Purification Kit (Qiagen Inc.) and labelled with [32P]dCTP, using the RediprimeII Random Prime Labeling System (Amersham Pharmacia Biotech) according to the manufacturer's instructions. A dilution series was prepared from total plant RNA, starting from 2000 ng µL−1 and diluting by a factor of two down to 2 ng µL−1. One microlitre from each RNA dilution was dotted directly on to a nylon membrane and UV cross-linked, prehybridized, hybridized and washed in hybridization tubes, as described by Sambrook et al. (1989). After washing, the blots were exposed to a PhosphorImager exposure cassette (Molecular Dynamics) for 3 h. The cassette was scanned with a PhosphorImager and the results were analysed using the ImageQuant software. Relative genomic viral RNA levels were also determined by comparing the lowest dilution level at which a signal was observed.

Real-time PCR detection and quantification of viral RNA

Primers were designed to generate amplicons of 120–135 bp from the CP gene of SPMMV, SPFMV or SPCSV (Table 2). Using the ABI Prism 7000 Sequence Detection System (Applied Biosystems), the real-time quantitative RT-PCR (rt-Q-PCR) (Heid et al., 1996) was done in a total reaction volume of 25 µL that contained SYBR Green I PCR Master Mix (Applied Biosystems), 1·0 µm each of forward and reverse primers, and 20 ng of cDNA template that was synthesized using random primers. Each cDNA sample (20 ng) was loaded in triplicate on a single 96-well optical plate (Applied Biosystems). The PCR reaction was initiated with a preincubation at 50°C for 2 min and denaturation at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 s and annealing plus extension at 52°C for 1 min per cycle. Immediately after the final PCR cycle, a melting curve analysis was done to determine specificity of the reaction by incubating the reaction at 52°C for 1 min and then slowly increasing the temperature to 87°C over 20 min. The rt-Q-PCR analysis was carried out twice on all samples in two experiments.

Table 2. Primers used in real-time PCR for quantification of viral RNA
Target virus or RNA Forward and reverse primers Expected amplicon size (bp)
Primer name Primer sequence (5′- to 3′)
  • a 26S rRNA primers were designed according to the partial sequence of Ipomoea batatas (sweetpotato) 26S rRNA gene determined in this study (EmBL database accession no. AJ972410).

The Ct value, the PCR cycle number at which a threshold signal in the log phase of the amplification curve is observed, was determined for each sample (well). Quantification was done using the comparative Ct(2−ΔΔCt) method (Livak & Schmitgen, 2001), in which the change in the amount of the target viral RNA was normalized in relation to the amount of sweetpotato 26S rRNA, the endogenous control. Degenerate primers were designed according to the 26S rRNA gene sequences of other plant species available in databanks, a 700-bp partial sequence of the gene amplified, cloned, sequenced and used for design of primers for rt-Q-PCR (Table 2). Validation experiments were done according to the manufacturer's instructions (Applied Biosystems) to ensure that amplification efficiencies with the virus-specific primers were approximately equal to the efficiency of the endogenous/normalizer 26S rRNA primers. The ΔCt(T) values [Ct(target) minus Ct(normalizer)] were calculated for each sample. Subsequently, the difference in viral RNA amounts between a sample coinfected with viruses and a sample infected with a single virus was calculated as the difference of the two, and the resultant value ΔΔCt was transformed to an absolute value using the comparative expression level formula of 2−ΔΔCt (Livak & Schmitgen, 2001).

Measurement of plant growth and yield

A randomized complete block design experiment with three replicates was carried out to test the effect of different virus-infection treatments on sweetpotato growth attributes and storage root yield. A preliminary glasshouse experiment at SLU, Sweden, evaluated three cultivars at 10 weeks after planting (WAP) for plant height, number of internodes, weight of foliage and number of branches from the main vine. Subsequently, two experiments were carried out in a screenhouse at Makerere University Agricultural Research Institute, Kabanyolo, Uganda. The first included evaluation of the plant growth attributes. The second involved only cv. Tanzania (due to space constraints and the need for replicates) that were maintained up to 20 WAP. All experiments were carried out using the same, previously characterized virus isolates.

Two sweetpotato vines planted in 100 L of soil in a wooden box (60 × 60 × 30 cm) constituted an experimental unit. These vines were taken from source plants that were infected with the individual viruses or their combinations, as described before. The volume of soil corresponded to a mound to which farmers commonly plant sweetpotato in the field. Eight wooden boxes corresponding to the eight treatments constituted an experimental block. The plants in the screenhouse were sprayed twice a week with broad-spectrum insecticides (Dimethoate) and observed daily for insects to ensure freedom from virus vectors.

Foliar symptoms, symptom severity, plant height (length of the main vine), number of branches from the main vine and number of leaf internodes were recorded at 3, 5, 7 and 10 WAP for the three cultivars. Severity of symptoms was assessed by visual inspection and recorded using the scale of 1–5 (Hahn et al., 1981) where 1 corresponded to symptomless plants and 5 to stunted and severely diseased plants. At 20 WAP, plants were harvested and fresh storage root weight and weight of fresh foliage determined. Data were subjected to a one-way analysis of variance in randomized blocks (Mead et al., 1993) and analysed using SigmaStat software (SPSS Inc.). Two-way analysis of variance was also carried out to determine cultivar differences in reaction to different virus infection treatments.


Symptoms caused by single-virus infections

The sweetpotato cvs Tanzania, Dimbuka and New Kawogo selected for this study show grossly different phenotypes reflecting genetic differences. They are important in sweet-potato production in Uganda and East Africa but their responses to SPMMV were unknown. SPMMV induced mild vein chlorosis and/or diffuse mottling in the leaves of cv. Tanzania (Fig. 1A) and mottling in the leaves of cv. Dimbuka, whereas SPFMV induced mottling on both cultivars. These symptoms corresponded to class 2 according to the symptom severity range (SSR) index of Hahn et al. (1981). Both viruses were readily detected in diseased leaves by NCM-ELISA. The symptoms were transient, lasting for 2–4 weeks in the young leaves of the plants growing from virus-infected, rooted cuttings, after which new leaves were both symptomless and virus-negative in NCM-ELISA. This phenotype indicated that cvs Tanzania and Dimbuka are moderately resistant to SPFMV and SPMMV, which is expressed as a recovery from disease. However, the recovered top of the shoot contained low virus titres below the detection threshold with NCM-ELISA, because in cuttings taken from the symptomless upper parts of the vines and planted in soil, symptoms re-appeared and leaves tested virus-positive.


Virus-induced symptoms in sweetpotato cv. Tanzania. (A) Vein chlorosis caused by Sweet potato mild mottle virus (SPMMV). (B–D) Chlorotic symptoms caused by coinfection with SPMMV and Sweet potato chlorotic stunt virus (SPCSV) (B); Sweet potato feathery mottle virus (SPFMV) and SPCSV (C); and triple infection with the three viruses (SPMMV, SPFMV and SPCSV) (D). (E) Dark green islands (arrows) associated with coinfection by SPMMV and SPCSV. (F–H) Increased symptom severity and stunting were observed in plants coinfected with SPMMV and SPCSV (F); SPFMV and SPCSV (G); and all three viruses (H).

No symptoms or virus was detected by NCM-ELISA (or TAS-ELISA) following graft inoculation with SPFMV or SPMMV, for up to 10 weeks postinoculation in cv. New Kawogo. However, these symptomless plants were infected with a titre of virus below the detection threshold of ELISA, as grafting a scion of I. setosa to the top of each cv. New Kawogo plant produced typical symptoms and then tested positive for the respective virus in NCM-ELISA. These results suggest that cv. New Kawogo expresses higher levels of resistance to SPMMV and SPFMV than the other two cultivars tested.

SPCSV caused similar mild, diffuse chlorotic patches or general mild chlorosis that gave a pale appearance of the leaves in all three sweetpotato cultivars at 5–7 WAP (SSR = 2). At later stages of growth, purple spots and flecks developed in the older leaves which tested positive for SPCSV in TAS-ELISA. Thus, no cultivar was resistant to SPCSV.

Symptoms caused by coinfection of viruses

Coinfection with SPMMV and SPFMV caused no alteration in the severity of symptoms in the three cultivars, compared with infection by each virus alone. In cvs Tanzania and Dimbuka, the types of symptoms in coinfected plants were initially a combination of those observed in single virus-infected plants, and recovery from disease followed as before. New Kowogo expressed similar higher levels of resistance to coinoculation of SPFMV and SPMMV as observed following graft-inoculation with each virus alone.

Regardless of the high or moderate levels of resistance expressed to SPMMV or SPFMV in the three sweetpotato cultivars, all of them developed severe symptoms following coinfection with SPCSV and SPMMV or SPFMV (SSR = 4). Coinfection of SPMMV with SPCSV (Fig. 1B and E) induced similar symptoms of distortion, strapping, rugosity, chlorosis and purpling of leaves and stunting of plants in all three cultivars. In addition, dark green islands (DGIs) (Fig. 1E) developed in the leaves. Symptoms induced by coinfection with SPFMV and SPCSV (Fig. 1C) were severe and similar in all cultivars, and resembled those caused by SPMMV + SPCSV, except that DGIs and leaf strapping were not observed. Symptoms were persistent in the coinfected plants and no recovery was observed until termination of the experiment at 20 WAP.

The triple infections (SPMMV + SPFMV + SPCSV) resulted in even more severe symptoms (Fig. 1H) (SSR = 5) than the aforementioned dual infections (cf. Fig. 1H with Fig. 1G and F). Triple infections were always associated with leaf strapping (Fig. 1D) and thinning. Cuttings taken from the triple-infected plants tended to root and establish poorly.

Plant growth and storage root yield

Cultivar × virus-infection treatment interactions were not statistically significant, indicating that the effects of viruses (in single or mixed infection) on plant growth attributes were similar for the three cultivars, regardless of some specific differences in the types of symptoms expressed. Therefore, results for the experiment involving only cv. Tanzania plants are presented (Table 3) at the time of harvesting foliage and storage of roots at 20 WAP.

Table 3. Effect of Sweet potato mild mottle virus (SPMMV), Sweet potato feathery mottle virus (SPFMV), Sweet potato chlorotic stunt virus (SPCSV) and their mixed infections on growth and yield of sweetpotato cv. Tanzaniaa
Virus infection Number of branches Length of vine (cm) Weight of foliage (g) Storage root weight (g)
12·3 d 227 a 340 a 343 a
SPMMV 15·7 cd 220 a 430 a 337 a
SPFMV 20·0 bc 222 a 416 a 255 ab
SPCSV 23·3 b 232 a 421 a 192 b
SPMMV SPFMV 23·0 b 236 a 400 a 245 ab
SPMMV SPCSV 16·3 cd 148 b 164 b 72 c
SPFMV SPCSV 16·3 cd 150 b 152 b 53 c
SPMMV SPFMV SPCSV 31·3 a 85 c 103 b 11 c
LSDb 5·40 58·8 151 110
  • a Values are means of six plants measured 20 weeks after planting. Means followed by the same letter are not significantly different (P < 0·05).
  • b LSD, least significant difference (P < 0·05).

Virus-infection treatment had a significant effect on plant growth. Vine length was significantly reduced (35%) following coinfection with SPMMV or SPFMV with SPCSV (Table 3). In triple infection, vine growth was reduced by 63% compared with noninfected and single-virus-infected plants. The weight of foliage was also significantly reduced in dual and triple virus infections (Table 3). The highest number of side branches was observed in the most heavily stunted plants with triple infection (Table 3; Fig. 1H). The reduced vine length and increased branching resulted in a bushy appearance characteristic of SPVD. The numbers of internodes on the main vine were not affected (data not shown).

Infection with SPMMV, SPFMV or their coinfection did not significantly reduce the storage root yield in cv. Tanzania (P < 0·05). Coinfection of these viruses with SPCSV resulted in yield reduction by 79–86% and 62–72% compared with healthy plants and plants infected with SPCSV, respectively (Table 3). Infection with SPCSV alone caused a significant yield reduction (54%). Triple infection greatly limited storage root formation and only very small and poorly developed storage roots were produced (Table 3).

Virus accumulation

Detailed studies on virus accumulation in cv. Tanzania allowed comparison with previous studies on SPVD (Karyeija et al., 2000a). Samples generating A405 values three times higher than the mean of noninfected control plants were considered to be virus-positive.

Titres of SPMMV were determined by DAS-ELISA in leaf 1 (uppermost fully opened leaf), 3, 5, 7, 9 and 13 as descending from the shoot tip at 5 WAP in independent experiments. The leaves at the highest positions (1–5) usually tested virus-negative because of low virus titres in leaves which had recovered, but the virus was detected in lower leaves (Fig. 2) in plants infected with SPMMV alone. However, in plants coinfected with SPMMV and SPFMV, SPMMV was barely detectable in any leaves at 5 WAP (Fig. 2), indicating a faster recovery than in plants infected with SPMMV alone.


DAS-ELISA absorbance values from detection of Sweet potato mild mottle virus (SPMMV) antigen in leaves at different positions on the vines of sweetpotato cv. Tanzania infected with SPMMV alone or in combination with Sweet potato chlorotic stunt virus (SPCSV), Sweet potato feathery mottle virus (SPFMV), or both viruses 5 weeks after planting the virus-infected cuttings. Position 1 corresponds to the uppermost fully opened leaf and other positions descend on the vine accordingly. Each data point corresponds to the mean of A405 values from three leaves, each from a different plant in one representative experiment. Bars indicate standard deviation. MM, SPMMV; FM, SPFMV; CS, SPCSV.

In plants coinfected with SPMMV and SPCSV, SPMMV was readily detected by DAS-ELISA in all leaves tested, including the uppermost fully opened leaf (position 1) (Fig. 2), petioles, stem and roots (data not shown). However, following triple infection with SPMMV, SPFMV and SPCSV, the uppermost fully opened leaves (position 1, and often position 3) in the severely diseased plants remained SPMMV-negative as tested by DAS-ELISA, but the lower leaves had high titres of SPMMV (Fig. 2) and SPFMV (Fig. 3) compared with plants infected with SPMMV or SPFMV alone.


Absorbance values from detection of Sweet potato mild mottle virus (SPMMV) and Sweet potato feathery mottle virus (SPFMV) antigens by DAS-ELISA, and Sweet potato chlorotic stunt virus (SPCSV) antigen by TAS-ELISA, in leaves at the seventh position below the uppermost fully opened leaf on vines of sweetpotato cv. Tanzania 5 weeks after planting the virus-infected cuttings. All samples were tested for all viruses. Each data point corresponds to the mean of the A405 values from three leaves, each from a different plant, in two experiments. Bars indicate standard error. MM, SPMMV; FM, SPFMV; CS, SPCSV.

Amounts of viral RNA were estimated in leaves at position 5 using both NASH and rt-Q-PCR at 5 WAP. Dotting of 2000 ng total plant RNA on the membrane resulted in a positive signal for SPMMV in NASH, whereas no signal was obtained using 1000 ng of the RNA (Fig. 4). However, a signal for SPMMV was observed with as little as 62·5 ng (but not 31·25 ng) of total RNA from plants coinfected with SPMMV and SPCSV and from the triple-virus-infected plants (Fig. 4). These data indicated an increase by 32-fold in the titres of SPMMV due to coinfection with SPCSV. For SPFMV, no signals were observed using 2000 ng total RNA from plants systemically infected with SPFMV alone, whereas 31·25 ng total RNA resulted in a signal in plants coinfected with SPFMV and SPCSV (Fig. 4). These data suggested a 64-fold increase in SPFMV titres following coinfection with SPCSV. In contrast, the titres of SPCSV decreased by two- to fourfold in coinfection with SPMMV, SPFMV or both viruses (Fig. 4). Results from rt-Q-PCR were consistent with those of NASH but the estimated differences in viral RNA amounts were larger, probably due to the higher detection sensitivity of rt-Q-PCR. The amounts of SPMMV and SPFMV RNA were ∼1000- and 1100-fold higher, respectively, in double infections with SPCSV than in single infections according to rt-Q-PCR (Table 4). In contrast, the amounts of SPCSV RNA were two- to threefold lower in mixed infections than in plants infected with SPCSV alone (Table 4). There was no detectable amplification from healthy plants using any virus-specific primers.


Detection of viral RNA by nucleic acid spot hybridization (NASH) in sweetpotato cv. Tanzania infected with Sweet potato mild mottle virus (SPMMV, MM), Sweet potato chlorotic stunt virus (SPCSV, CS), or coinfected with SPMMV and SPCSV (MM + CS), or Sweet potato feathery mottle virus (SPFMV, FM) and SPCSV (FM + CS), or with all three viruses (MM + FM + CS). Total RNA was extracted from the seventh leaf below the uppermost fully opened leaf at 5 weeks after planting the virus-infected cuttings. Blots CS, MM and FM were hybridized with a virus-specific DNA probe for SPCSV, SPMMV and SPFMV, respectively.

Table 4. Estimation of viral RNA amounts in different virus infection treatments of sweetpotato cv. Tanzania using real-time quantitative PCR (rt-Q-PCR)
Virus infectiona Target virusa ΔCt(T)b 2−(ΔΔCt)c
MM MM 14·7d 1·0
FM FM 14·5 1·0
CS CS 4·0 1·0
CS + MM CS 5·0 0·5
CS + MM MM 4·8 975·5
CS + FM CS 4·6 0·7
CS + FM FM 4·3 1136·2
MM + FM MM 14·2 1·5
MM + FM FM 14·6 1·0
CS + MM + FM CS 5·4 0·4
CS + MM + FM MM 6·2 360·8
CS + MM + FM FM 5·4 546·8
  • a Virus-infection treatments and viruses quantified: MM, Sweet potato mild mottle virus; FM, Sweetpotato feathery mottle virus; CS, Sweet potato chlorotic stunt virus.
  • b Viral RNA amounts relative to the endogenous control. ΔCt(T) = Ct(T) – Ct(N), where Ct(T) is the Ct value for the target virus and Ct(N) value is the average value for sweetpotato 26S rRNA that was used as the normalizer (endogenous control).
  • c Relative difference (fold) in the viral RNA amounts between single-virus-infected plants as compared with plants infected with different virus combinations (Livak & Schmitgen, 2001). ΔΔCt = ΔCt(T) –ΔCt(Rs), where ΔCt(Rs) is the ΔCt value for the plant infected with a single virus. For ΔCt(T), see above.
  • d Values are means of two independent analyses on the leaf samples from one representative experiment (each RNA sample was analysed as triplicate).


The results of this study show that coinfection of SPMMV with SPCSV results in greatly enhanced disease severity compared with infection with either virus alone, thus revealing a novel ipomovirus–crinivirus synergism. While infection with SPMMV alone caused no loss in the yield of storage roots, coinfection with SPCSV caused up to 80% reduction compared with the healthy plants. Yield reduction caused by SPCSV alone reached 50%. Thus, SPMMV can contribute to greater yield losses when co-occurring with SPCSV in sweetpotato, which had not been previously studied. The yield losses resulting from coinfection of SPFMV with SPCSV in the field have been documented in many studies (Sheffield, 1957; Schaefers & Terry, 1976; Ngeve & Bouwkamp, 1991; Milgram et al., 1996; Gibson et al., 1998; Gutierrez et al., 2003). However, the poorest yield was obtained following infection with all three viruses. Taken together, these results and the previous studies show a relationship between viral synergism and reduced yields. Comparisons of growth and yield reduction caused by single- and mixed-virus infections, including SPFMV and SPCSV, may be more accurately documented here than in studies that were carried out in the field, where the spread of viruses is difficult to control. These experiments were done in an insect-proof screenhouse under a careful pest management scheme in a geographic area typical for sweetpotato production. Although storage root initiation and growth may be different under screenhouse conditions and in the field, it is anticipated that the results of this study provide a fair estimate of the potential negative impact which SPMMV, SPFMV or SPCSV and their mixed infections can cause on sweetpotato yields, especially when virus-infected vines are used as planting material.

The level of symptom severity was similar, regardless of whether SPMMV or SPFMV coinfected sweetpotato plants with SPCSV, but some differences in the types of symptoms were observed. SPVD is the name used to describe the severe disease caused by coinfection with SPFMV and SPCSV in sweetpotato plants (Schaefers & Terry, 1976; Milgram et al., 1996; Aritua et al., 1998; Gibson et al., 1998). Similar diseases are caused also by other virus combinations, including coinfection with Cucumber mosaic virus (CMV; genus Cucumovirus, family Bromoviridae) and SPCSV in Israel (Cohen & Loebenstein, 1991) and with SPCSV, SPFMV and SPMSV, a syndrome designated as sweetpotato chlorotic dwarf disease in Argentina (Di Feo et al., 2000). Because novel disease symptoms, including rugosity, leaf strapping and DGI, were observed following dual infection with SPMMV and SPCSV in the three important East African sweetpotato cultivars tested, and these symptoms were distinguishable from those of SPVD, it is proposed that the name ‘sweetpotato severe mosaic disease’ (SPSMD) is adopted for the new disease described in this study.

Resistance to SPMMV has not been studied in the cultivars included in this study, whereas responses to SPFMV and SPCSV are known in cvs Tanzania and New Kawogo (Karyeija et al., 2000a, 2000b). Cultivars Dimbuka and Tanzania expressed similar, moderate levels of resistance to SPMMV, SPFMV and their mixed infection and recovered from the disease. New Kawogo showed higher levels of resistance to both viruses and remained symptomless with only very low virus titres following graft inoculation. However, regardless of the different levels of resistance to SPMMV and SPFMV in these cultivars, resistance broke down following infection with SPCSV, and the resulting titres of SPMMV and SPFMV were greatly elevated. These data indicate synergistic interactions between SPCSV and SPMMV or SPFMV. In contrast, no interactions (alteration of virus titres or symptom severity) were observed between SPFMV and SPMMV. The titres of SPCSV slightly declined in coinfection with the other viruses. All cultivars were susceptible to SPCSV, and so far a search for resistance to SPCSV and SPVD in sweetpotato germplasm by breeders has identified only one source with modest levels of resistance (Mwanga et al., 2002). The mechanism by which SPCSV suppresses virus resistance in sweetpotato is unknown, but it may be associated with the unusual mechanism by which SPCSV suppresses RNA silencing (Kreuze et al., 2005), the fundamental antiviral defence system in plants and other multicellular organisms (Ahlquist, 2002). The recovery of cvs Tanzania and Dimbuka from SPMMV or SPFMV infection and production of the new symptomless leaves with very low virus titres are consistent with infection-induced resistance based on RNA silencing (Ratcliff et al., 1997; Yelina et al., 2002). Consequently, suppression of resistance may be associated with suppression of RNA silencing by SPCSV. Furthermore, viral proteins that suppress RNA silencing are known effectors in viral synergism (Pruss et al., 1997; Shi et al., 1997; Moissiard & Voinnet, 2004).

These data highlight the ability of SPCSV to synergize with different unrelated viruses and that this is a serious constraint to development of durable virus resistance in sweetpotato. Furthermore, current knowledge is insufficient to predict with which viruses SPCSV can exhibit synergism. SPCSV has two strains, the East African strain (SPCSV-EA) included in this study and the West African strain (SPCSV-WA); these are readily distinguishable at serological and nucleotide sequence levels (Tairo et al., 2005). A recent study on SPFMV showed that isolate 95-6 of the RC strain, in contrast to isolate 95-2 of the C strain, synergized with an isolate of SPCSV-WA (Souto et al., 2003). Similarly, isolates of the proposed potyvirus species Sweet potato virus Y (Ateka et al., 2004a) may differ in their synergistic interactions with SPCSV-WA because isolates LSU2 and LSU5 from the United States showed synergism (Souto et al., 2003), in contrast to isolate SPV2 from Nigeria (Ateka et al., 2004a). Finally, different studies have used different sweetpotato cultivars, introducing putative host effects to further complicate the picture. A more systematic study is required to understand the factors that determine synergistic interactions between SPCSV and other viruses.


We are grateful to Dr H. J. Vetten at BBA, Germany, and to Dr L. F. Salazar and S. Fuentes at CIP, Peru, for providing antibodies to sweetpotato viruses, and to Dr J. F. Kreuze, CIP, for plant materials. This study (project no. 771502) is part of the East African Regional Network for Biotechnology, Biosafety and Biotechnology Policy Development programme (BIO-EARN), which is funded by Sida/SAREC through the Stockholm Environment Institute (SEI), Sweden.