Factors influencing reversion from virus infection in sweetpotato

Viruses limit sweetpotato (Ipomoea batatas) production worldwide. Many sweetpotato landraces in East Africa are, however, largely virus-free. Moreover, some plants infected by the prevalent Sweet potato feathery mottle virus (SPFMV) may be able to revert to virus-free status. In this study, we analysed reversion from SPFMV, Sweet potato virus C, Sweet potato mild mottle virus, Sweet potato chlorotic stunt virus (SPCSV) and Sweet potato leaf curl Uganda virus using the indicator plant I. setosa and PCR/reverse-transcriptase PCR. We also investigated environmental factors (temperature and soil nutrients) that may influence reversion from virus infection. We tested reversion in the East African cultivars New Kawogo, NASPOT 1 and NASPOT 11, and the United States cultivars Resisto and Beauregard. Reverted plants were asymptomatic and virus was undetectable in assayed parts of the plant. After graft inoculation, only the East African cultivars mostly reverted at a high rate and from most viruses though cultivar Beauregard fully reverted following sap inoculation with Sweet potato virus C. None of the tested cultivars fully reverted from single or double infections involving SPCSV, and reversion was only observed in co-infections involving potyviruses. Root sprouts derived from SPFMV-reverted plants were also virus free. Reversion generally increased with increasing temperature and by improved soil nutrition. Overall, these results indicate variation in reversion by cultivar and that the natural ability of sweetpotato plants to revert from viruses is malleable, which has implications for both breeding and virus control.


| INTRODUCTION
Sweetpotato (Ipomoea batatas (L.) Lam.) is an important root crop, with world production of more than 100 million tonnes annually. In East Africa, sweetpotato is second only to cassava in importance as a starch staple (Food and Agriculture Organization, 2016).
Sweetpotato is both a staple and a food security crop in most of sub-Saharan Africa and is grown mainly by smallholder farmers (Andrade et al., 2009;Bashasha, Mwanga, Ocitti p'Obwoya, & Ewell, 1995).
East African sweetpotato varieties that had been infected with SPFMV. These reverted plants were asymptomatic, and SPFMV was undetectable by the most sensitive test-grafting to Ipomoea setosa (Gibson et al., 2014;Moyer, Jackson, & Frison, 1989)-in tested parts of the plant. It is postulated that reversion may explain why SPFMV, which is widely distributed, does not completely devastate sweetpotato crops. Rates of reversion from other, rarer viral infections may be even greater. Reversion may be considered an extreme form of the well-known phenomenon of recovery from viral infection (Gibson & Kreuze, 2015).
In this study, we investigated the incidence of viral reversion in response to temperature and soil nutrient enhancements. We used sweetpotato cultivars from East Africa and the United States, after infection or co-infection with viruses of different genera and families.

| Virus cultures
The East African sweetpotato viruses used were obtained from the National Crops Resources Research Institute in Namulonge, Uganda and Makerere University Agricultural Research Institute in Kabanyolo (MUARIK), central Uganda. MUARIK is 19 km north of Kampala, latitude 0 27 0 60 00 N, longitude 32 36 0 24 00 E, at an altitude of 1,250-1,320 m above sea level (Yost & Eswaran, 1990). MUARIK receives annual rainfall of about 1,300 mm.

| Nucleic acid extraction and virus identification
For SPLCUV, total nucleic acid extraction was performed from leaves using the CTAB method (Maruthi, Colvin, Seal, Gibson, & Cooper, 2002). For RNA viruses, RNA was extracted from leaves and storage roots using the TRI Reagent protocol following the supplier's manual (Bio Labs, Jerusalem, Israel). Nucleic acids were quantified using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific; Bargal Analytical Instruments, Airport City, Israel) and evaluated on a 1.5% agarose gel. Complementary DNA (cDNA) was synthesised using Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Scientific, Tamar, Israel) following the manufacturer's manual.
Plants in all experiments in Uganda were confirmed infected 1-2 weeks post-inoculation (wpi) using I. setosa as an indicator plant and/or by PCR (SPLCUV) or RT-PCR (SPMMV, SPFMV, SPVC, SPCSV); in Israel, infection was confirmed by RT-PCR. For amplification we used a 25-μl PCR master mix of 8.5 μl of water, 12.5 μl of PCR mix (HyLabs Ready Mix [×2], HyLabs, Rehovot, Israel), 1 μl of each primer (10 pmol) and 2 μl of DNA/cDNA (500-1,000 ng/μl). The PCR products were assessed by electrophoresis in a 1.5% agarose gel in 0.5% Tris-acetate buffer, stained with ethidium bromide and viewed under UV light and documented by an OmniDoc gel documentation system (Clever Scientific, Image Care, Uganda).

| Reversion from single viral infections
At MUARIK, healthy vine cuttings of 200-250 mm with 3-4 nodes were established in pots with about 1 kg of soil mixture (3:1:1 ratio of black soil: lake sand: cow manure). Plants were grown for 2 weeks in a screenhouse and watered daily.
Plants were side-graft-inoculated with SPLCUV, SPFMV, SPMMV or SPCSV individually using scions of infected I. setosa (each scion was~25 mm). The scions remained on the grafted plants for the experimental period. One mock-inoculated plant per cultivar per virus was included as a control. Plants were tested using I. setosa and PCR/RT-PCR at 1 wpi to confirm virus infection. Ten successfully inoculated plants of each cultivar were then evaluated for reversion.
Evaluation was done through cutting shoot tips (~50 mm) every 2 weeks for 10 weeks to test for virus infection by side-grafting to I. setosa and then testing for the virus presence in I. setosa with PCR/RT-PCR. Composite samples of top, middle and basal leaves were used. Removal of shoot tips encouraged lateral shoot growth. Subsequent testing for reversion was performed on these lateral shoots, selecting one shoot every round of testing. Symptom development on I. setosa was monitored for 6 weeks and confirmation done with PCR/RT-PCR.
For the evaluation of reversion in roots, vines (~250 mm long with 3-5 nodes) of SPFMV-reverted plants of each of the five cultivars were grown for 16 weeks in a screenhouse in basins of~25 kg of soil mixture (as above), with five vines per basin. Plants were harvested and three mature storage roots from each basin were sprouted for up to 6 weeks in pots containing~1 kg of soil mix, with one root per pot.
Three leaves (top, middle and basal) were taken weekly from one sprout per bucket (other sprouts were removed) and tested for    T A B L E 3 Reversion from a range of viruses for sweetpotato cultivars tested using Ipomoea setosa and PCR/RT-PCR in the screenhouse at MUARIK, Uganda Note: The experiment was performed in a growth room at the ARO. Confirmation of the lack of viral presence was confirmed by weekly testing using RT-PCR for a period of 10 weeks after sap inoculation. Plants were pruned after 4 weeks, and the new stems were subsequently tested for SPVC infection. Roots were also tested to verify that the whole plant had reverted.

| Reversion from sweetpotato viruses in environments with different temperatures
The field experiments were conducted at MUARIK and at a farm in Each plot consisted of two rows per cultivar (one SPFMV-infected and one healthy control). Maize was planted between the blocks as a barrier. The presence of SPFMV was tested at 2, 7 and 12 weeks after planting by grafting shoot tips~50 mm long onto I. setosa. Symptom development on I. setosa was observed for 6 weeks. planting the vines, soil amendments were added at the ratios above.

| Reversion from SPFMV for plants grown in soil with different enhancements
Reversion from SPFMV was evaluated seven times over a period of 3 months using RT-PCR, one shoot per test, taking a composite sample of the top, middle and basal leaves.
To evaluate the composition and quality of the soil and amendments before use, a blind analysis, as described by Okalebo, Gathua, and Woomer (2002), was performed for three samples each of the soil, manure and NPK fertiliser to obtain an average composition ( Table 2). The analysis was done at the soil laboratory of the College of Agricultural and Environmental Sciences, Makerere University.
T A B L E 5 Virus detection in plants co-infected with Sweet potato feathery mottle virus (SPFMV) and Sweet potato chlorotic stunt virus (SPCSV), which together cause sweet potato viral disease (SPVD)

T A B L E 6 Effect of temperature on incidence of reversion from infections of Sweet potato leaf curl Uganda virus (SPLCUV) or Sweet potato feathery mottle virus (SPFMV)
No. of plants that reverted from

SPLCUV infection SPFMV infection
Cultivar 20 C 3 0 C 2 0 C 3 0 C Resisto 2/6 4/6 New Kawogo 1/6 3/6 5/12 10/12 Note: Six plants each of cvs. Resisto and New Kawogo were graft-inoculated with SPLCUV, or six plants of cv. New Kawogo were graft-inoculated with SPFMV, and the infection status of the plants was tested using PCR/RT-PCR after 4 weeks at either 20 or 30 C. Different categories (i.e., for each cultivar at the different temperatures) above were statistically significant at 95% with a chi-squared value equal to 0.079 (df = 1).  (Table 3 and Figure 2). Infection of cv. NASPOT 11 by SPFMV was not observed, and this cultivar was only infrequently infected with SPCSV, SPMMV or SPLCUV (Table 3 and Figure 2). The cv. Resisto reverted effectively from infection by SPLCUV (better than any cultivar except for NASPOT 11), but it reverted poorly (and least effectively of all cultivars) from infection with all other viruses (  SPVC was not detected in any mock-inoculated plants.

| Reversion from co-infection with SPFMV and SPCSV
At MUARIK, five sweetpotato cultivars were graft-inoculated with SPFMV and SPCSV using SPVD-infected I. setosa leaves. Sweetpotato plants developed typical SPVD symptoms with cv. NASPOT 11 having milder symptoms by 11 weeks. Virus detection using I. setosa revealed differential infection of SPFMV or SPCSV among cultivars over a period of 11 weeks. Within the first 3 weeks, both SPFMV and SPCSV were Note: Values shown indicate the number of reverted plants (out of an initial 20) following graft inoculation with SPFMV, when grown on soil with different soil amendments. The virus was detected using RT-PCR. The start time (0 days) for this experiment was 2 weeks after grafting, when the plants were found to be virus-infected using RT-PCR. Effects of soil amendments on reversion of sweetpotato plants from infection with SPFMV. detected in all plants of cvs. Beauregard, NASPOT 1 and New Kawogo.
However, SPFMV and SPCSV were detected in only three and six out of seven plants of cvs. NASPOT 11 and Resisto, respectively. All other plants had single infections of either SPFMV or SPCSV (Table 5). Variable expression of viral symptoms was also observed 5-11 wpi for all cultivars of sweetpotato and I. setosa (Table 5). Thus, no cultivar fully reverted from SPVD (Table 5). At the ARO, no reversion from SPVD was observed in cv. Beauregard at 12 wpi (data not shown).

| Effect of temperature on reversion
Generally, more plants of cv. Resisto than cv. New Kawogo reverted from SPLCUV at both 20 and 30 C (Table 6

| Reversion of sweetpotato plants grown in soils with varied enhancements
Generally, viral reversion was more prevalent in the treatments in which nutrients were added (Table 8). When soil was amended, more reversion was observed within the first 6 days, especially for cvs.
NASPOT 11 and NASPOT 1. The NPK amendment led to the most reversion for all cultivars, followed by the NPK + manure treatment.
The addition of amendments, especially NPK or NPK + manure, increased the proportion of cv. Resisto plants that reverted by 3 months compared with the control. All plants of cv. NASPOT 11 and most of cv. NASPOT 1 had reverted from SPFMV by 3 months for all treatments, but fewer cv. Resisto plants had reverted (Table 8).

| DISCUSSION
Reversion was generally more frequent in East African than the United States sweetpotato cultivars. However, the rate of reversion from viruses of different families varied among the cultivars. For example, cv. New Kawogo reverted more often from SPFMV (potyvirus) than from SPLCUV (begomovirus) infection, whereas cv. Resisto reverted more from SPLCUV than SPFMV (Table 3 and Figure 2). Such variations are consistent with previous field observations in cv. New Kawogo plants by Wasswa et al. (2011). The greater prevalence of SPLCUV in Ugandan sweetpotato fields (Wasswa et al., 2011) is possibly influenced by the low rate of reversion from SPLCUV (a begomovirus) by cv. New Kawogo, despite its ability to revert from SPFMV, as observed here. Clark and Hoy (2006) (Table 3).
Reversion from SPFMV (Table 3) is consistent with observations by Gibson et al. (2014). Our study is the first to report evidence of reversion from SPMMV and SPLCUV. Plants of both East African and the United States varieties reverted more frequently from SPMMV and SPLCUV than from SPFMV, and this could explain the comparatively low field prevalence of SPMMV and SPLCUV observed in East Africa (Aritua, Legg, Smita, & Gibson, 1999) and the United States (Clark et al., 2012). Furthermore, shoots that regenerated from pruned reverted plants or root sprouts from reverted plants tested negative for the viruses, indicating complete reversion. This is the first report to show complete reversion from viruses in sweetpotato plants from the United States (Tables 3 and 4). Reversion was very effective in cv. Beauregard following sap inoculation (Table 4), which was much less aggressive than graft inoculation (Table 3). Reversion may be considered to be an extreme form of recovery (Basu et al., 2018;Gibson & Kreuze, 2015;Paudel & Sanfaçon, 2018).
None of the five tested cultivars (including cv. NASPOT 11) effectively reverted from infections involving SPCSV, including the SPFMV + SPCSV co-infection (Tables 3 and 5). The incomplete reversion from SPFMV + SPCSV combination was also observed by Mwanga, Yencho, Gibson, and Moyer (2013). In their study, some severely SPVD-affected genotypes developed localised symptoms from which they recovered. These recovered plants could only subsequently produce whole branches or individual shoots that had apparently reverted, being free of detectable virus when assayed by ELISA or grafted onto I. setosa.
Despite the failure of plants to effectively revert from SPCSV (and thus SPVD), this virus is rarer than SPFMV in the field, possibly for  (Milgram, Cohen, & Loebenstein, 1996). In a field in Brazil, 80% of experimental plants were SPFMV-infected by 10 weeks (Pozzer, Dusi, Silva, & Kitajima, 1994). In the United States, 100% infection by SPFMV can occur within 5-10 weeks (Bryan, 2002). Failure of plants to revert from SPCSV may be due to the presence of the p22 and RNase 3 silencing suppressor genes in the East African strain of SPCSV. These genes suppress the gene-silencing mechanism of the plant (Cuellar, Tairo, Kreuze, & Valkonen, 2008), and thus probably compromise plant defences to the extent that reversion cannot effectively occur and other viruses can easily infect (Kreuze, Savenkov, Cuellar, Li, & Valkonen, 2005), causing SPVD. This effect of SPCSV infection also may explain why plants failed to fully revert from a SPFMV + SPCSV co-infection (SPVD).
Under laboratory conditions, at an elevated temperature of 30 C, the reversion of the United States cv. Resisto from SPLCUV was greatly enhanced (Table 6 and Figure 3), and cv. New Kawogo also reverted effectively from two different virus families (Table 6 and Figure 3) at 30 C. Additionally, in different environments with temperature differences, field plants more frequently reverted when grown at higher (21-22 C) compared with lower temperatures (18-19.5 C) (Table 7), although we acknowledge that numerous other factors may have been at play in these fields. In the field experiments, reversion occurred more strongly for East African compared with the United States cultivars. These results together with the temperaturecontrolled experiments suggest that temperature may directly affect the reversion process. Similar results were observed by Rossel and Thottappilly (1985) in tropical root crops, Bertschinger, Keller, and Gessler (1995) in potato, Gibson and Otim-Nape (1997) in cassava and Aritua, Alicai, Adipala, Carey, and  in sweetpotato cvs. New Kawogo and Tanzania. These studies indicated that better rates of viral reversion are likely due to increased plant defence RNA interference mechanisms at high compared to low temperatures (Chellappan, Vanitharani, Ogbe, & Fauquet, 2005).

More plants reverted from SPFMV when grown in soil fertilised
with NPK or with a NPK-manure combination, but not in manure alone, compared with unamended soil (Table 8). It is probable that this occurred because the soil amendments (nutritional supplements) enhanced plant vigour and growth, as has been observed for radish (Sarker, Kashem, & Osman, 2012). Well-fertilised plants generally have more resources to fight infection. The use of soil nutrients to manage plant diseases caused by fungi and bacteria has been noted previously (Dordas, 2008); phosphorus specifically has been reported to enhance plant defences against viral disease (Huber & Graham, 1999). However, fertilisation is not a panacea for viral infection; Bhaduri, Rakshit, and Chakraborty (2014) found that in the presence of an external N supply (such as from NPK), visible symptoms (e.g., of SPFMV) are dependent upon competition for N between the virus and host cells.
The reversion from viral infection in sweetpotato observed here was virus-specific and affected by environmental factors. The contribution of genotype to the predisposition to revert was evident for sweetpotato cultivars from both East Africa and the United States, suggesting a generality to this phenomenon, which requires further study. Research on the mechanisms of reversion and the environmental conditions that improve reversion rates could provide additional methods to control sweetpotato viruses. Further investigation of the incidence of reversion in different sweetpotato cultivars could improve breeding and production of virus-free planting material without the use of expensive in vitro virus-elimination techniques.