Elevated CO₂ enhances plant photosynthetic rates, resulting in greater production of plant biomass or yield (Ainsworth and Long, 2005; Cure and Acock, 1986). The growth-enhancing effects of elevated CO₂ levels typically increase with rising temperature, and the optimum growth temperatures of several plants have been shown to rise substantially with increasing levels of atmospheric CO₂ (Berry and Bjorkman, 1980; McMurtrie and Wang, 1993; McMurtrie et al., 1992; Stuhlfauth and Fock, 1990). The pool of total soluble sugars and starch in plant leaves also increases at elevated CO₂ levels (Ye et al., 2010). While there is abundant literature describing the effects of temperature on virus accumulation and symptom expression, little is known regarding the responses of crops to virus infection under elevated CO₂ conditions, and few studies have investigated the interactions between plants and viruses under elevated CO₂ conditions.

Potato leaf roll virus (PLRV) is transmitted by Myzus persicae in a circulative, nonpropagative manner in which the virus is acquired through gut tissue into the aphid hemocoel and then exits through salivary tissues (Gray and Banerjee, 1999). In this study, the effects of increased temperature and CO₂ levels on PLRV infection and PLRV RNA content in both the aphid vector and plant tissues were examined using Solanum tuberosum as a plant host.

To examine the effects of atmospheric temperature on PLRV infection at CO₂ levels of 400 ± 50 ppm and 940 ± 50 ppm, 30-59 S. tuberosum cv. ‘Chubaek’ plants (7-8 cm in height) were inoculated with PLRV using M. persicae, and maintained in a 20 ± 2°C growth chamber. After 3 days, aphids were killed by chemical spray and the plants were moved to growth chambers where the temperature was maintained at 10-30 ± 2°C (in 5°C increments). PLRV RNA was quantified using real-time quantitative reverse transcription polymerase chain reaction (qPCR).

To examine the effects of atmospheric CO₂ concentration on PLRV infection, 30-48 plants (7-8 cm in height) were inoculated with PLRV using M. persicae, and grown in a growth chamber at 20 ± 2°C. After 3 days, aphids were killed by chemical spray and plants were moved to 20 ± 2°C growth chambers where CO₂ levels were adjusted to 400, 540, and 940 ppm (± 50 ppm). PLRV RNA was quantified using qPCR. We also examined the effects of atmospheric CO₂ on PLRV acquisition by the natural vector M. persicae. We allowed aphids to acquire the virus by feeding on PLRV-infected source plants (Physalis floridiana) at different CO₂ concentrations within the range 400-940 ppm (± 50 ppm) in a growth chamber, and then transferred single virus-loaded aphids onto individual test plants (P. floridiana). The plants were maintained at 20 ± 2°C until virus infection was determined 10 days later. Approximately 46-56 plants were used to test the transmission ability of the aphids. The M. persicae PLRV RNA content was determined using qPCR with 30 aphids, and total RNA extracted from individual aphids was determined using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany).

Stem cuttings approximately 7-8 cm in length were rooted in soil and used for PLRV infection. P. floridiana seedlings were used to test the transmission efficiency of PLRV by M. persicae. The plants were grown in ‘Barok’ commercial soil in an insect-proof glasshouse.

The PLRV inocula were field isolates of PLRV-infected plants obtained from B. Fenton (James Hutton Institute, Dundee, Scotland), as previously described (Chung and Palukaitis, 2011). Source plants for aphid transmission were PLRV-infected P. floridana plants, which were maintained in insect-proof cages and renewed every month.

Virus-free M. persicae cultures were maintained on Nicotiana tabacum cv. Samsun plants in insect cages in a 20°C growth chamber. To transmit the virus, single M. persicae of the second or third instar were used for each plant. No pre-acquisition starvation period was given for PLRV, and the virus acquisition period was 3 days. The duration of feeding on test plants was 3 days. Then the aphids were killed by spraying pesticide.

The primer sets used for RT-PCR and qPCR have previously been described (Chung et al., 2016). RT-PCR was used to determine virus infection in plants after inoculation. Total RNA used for RT-PCR or qPCR was prepared with an RNeasy Mini Kit. The RT-PCR reaction was performed as previously described (Chung and Palukaitis, 2011). To quantify the absolute copy numbers of PLRV, we constructed a standard curve employing known concentrations of in vitro transcripts. Synthesis of RNA transcripts was performed using mMESSAGE mMACHINE T7 (Ambion, Austin, TX, USA) according to the manufacturer’s instructions. Clones of PLRV 627/pGEM-T plasmid DNA were linearized with the SalI restriction enzyme and treated with T7 RNA polymerase. Absolute levels of viral RNA are expressed as the number of viral copies per nanogram of total RNA. For qPCR, 40 ng DNAse-treated total RNA (Ambion) was reverse-transcribed (Promega ImProm-II RT, Madison, WI, USA) using a gene-specific primer. We added 8 μL cDNA mix prior to subsequent qPCR. All of the reactions proceeded in a C1000 Touch Thermal cycler (Bio-Rad, Hercules, CA, USA) using the SYBR-Green method (Universal SYBR-Green Supermix, Bio-Rad) according to the following protocol: 1 cycle at 95°C for 30 s; and 39 cycles at 95°C for 10 s, 60°C for 30 s, and 65-95°C in increments of 0.5°C and at intervals of 5 s. We used the qPCR analysis program CFX Manager 3.1 (Bio-Rad).

The SAS 4.2 statistical package (SAS Inc., Cary, NC, USA) was used for data analysis.

We determined the effects of atmospheric temperature (10-30 ± 2°C in 5°C increments) and CO₂ concentration (400 ± 50 ppm, 540 ± 50 ppm, and 940 ± 50 ppm) on PLRV infection in a growth chamber. The results showed that PLRV infection was influenced by both temperature and CO₂ concentration. Below CO₂ levels of 400 ± 50 ppm, both the PLRV infection rate and PLRV RNA content increased as the temperature increased to 20 ± 2°C; however, both of these factors declined at temperatures higher than 20°C (Table 1, Fig. 1). When CO₂ concentration was increased to 940 ± 50 ppm, comparatively more plants were infected by PLRV at 30 ± 2°C than at 20 or 25 ± 2°C. PLRV RNA content was unchanged over a temperature range of 20-30 ± 2°C (Fig. 2).

The effects of atmospheric CO₂ concentration on PLRV acquisition by M. persicae and PLRV RNA content in plant tissues was investigated in a growth chamber at 20 ± 2°C. M. persicae PLRV RNA content increased slightly at elevated CO₂ levels (940 ± 50 ppm); however, this increase was not statistically significant (Fig. 3). The transmission rates of PLRV in P. floridana increased as CO₂ concentration increased (Table 2). More PLRV RNA accumulated in tissues of plants maintained at 540 or 940 ± 50 ppm than in those maintained at 400 ± 50 ppm (Fig. 3).

PLRV infection in S. tuberosum plants was influenced by both temperature and CO₂ concentration. The PLRV infection rate and PLRV RNA content were highest in plants maintained at 15-20 ± 2°C after inoculation in a growth chamber at 400 ± 50 ppm CO₂. The highest virus titer was observed at 18-20°C, which is the optimum temperature for S. tuberosum plant growth (Haverkort, 1990). In another study using P. floridana, the PLRV infection rate was highest at 25 ± 2°C, and greater PLRV levels were evident in plants maintained at 20-25°C (Chung et al., 2016).

Previous studies have shown that virus infection is relatively good when the host plants are maintained at their optimum growth temperature. Singh et al. (1988) observed that high relative humidity and high temperature (25-30°C) increased virus transmission by 30-35% while the PLRV infection rate increased with higher (30°C) post-inoculation temperatures in P. floridana. Swenson (1968) similarly observed that more pea plants were infected with bean yellow mosaic virus when grown at 30°C than when grown at 24°C or 15°C after inoculation. The optimal temperature for viral RNA replication in cells infected by soil-borne wheat mosaic virus is 17°C (Ohsato et al., 2003). Mangrauthia et al. (2009) determined that the optimum temperature for symptom development in PRSV-infected papaya is 26-31°C, supporting their results by showing that purified PRSV-HC-Pro recombinant protein bound 21 nt ds miRNA duplexes more efficiently at ambient temperature (25°C) than at high (35-45°C) or low (15°C) temperatures.

Relatively low temperatures (15-20°C) tended to delay symptoms in Potato virus Y-O (PVY-O)- or Potato virus A (PVA)-infected N. benthamiana plants (Chung et al., 2016), and in PRSV-infected papaya (Mangrauthia et al., 2009). Relatively high temperatures (40-45°C) also delayed symptom development in PRSV-infected papaya plants (Mangrauthia et al., 2009). Similar results have been reported in other host-virus systems. Melon necrotic spot virus symptoms were observed to diminish as temperature increased from 20 to 25°C (Kido et al., 2008), and less severe banana streak virus symptoms and lower virus titers were detected in plants grown at 28-35°C than in those grown at 22°C (Dahal et al., 1998). PVY-O and PVA symptoms in N. benthamiana plants grown at 30°C were attenuated compared to those of plants grown at temperatures below 25°C (Chung et al., 2016).

In this study, we found that an increase in atmospheric CO₂ concentration (940 ± 50 ppm) resulted in the infection of more potato plants by PLRV at 30 ± 2°C than at 20 or 25 ± 2°C; PLRV RNA content was unchanged over a temperature range of 20-30 ± 2°C. M. persicae PLRV RNA content increased, but not significantly, as atmospheric CO₂ concentration increased during PLRV acquisition from source plants. Consequently, PLRV transmission rates in P. floridana plants increased. More PLRV RNA accumulated in potato plants maintained at 540 or 940 ± 50 ppm CO₂ than in those maintained at 400 ± 50 ppm CO₂ at 20 ± 2°C.

Our results are consistent with those of a recent study, in which barley yellow dwarf virus titer increased by 36.8% in the leaves of infected plants grown under higher CO₂ concentrations than ambient levels (Trezbicki et al., 2015). Del Toro et al. (2015) reported that at elevated CO₂ concentrations, viral titers for cytomegalovirus markedly increased as the proportion of total plant protein content increased in leaf disks, albeit less so for PVY and three Potato virus X (PVX) constructs. Similarly, elevated CO₂ concentrations alleviated damage to N. tabacum plants due to PVY infection, or delayed viral spread to some extent (Ye et al., 2010). Elevated CO₂ concentration also decreased the incidence of tomato yellow leaf curl virus (TYLCV) (by 14.6% in 2009 and 11.8% in 2010), disease severity (by 20.0% in 2009 and 10.4% in 2010), and levels of TYLCV coat protein in tomato leaves (Huang et al., 2012), reportedly due to an increase in salicylic acid and jasmonic acid at high CO₂ concentration.

Few studies have considered how insect fitness is altered by changes in host plant physiology induced by elevated CO₂ concentration. According to a recent report (Sun et al., 2015), plant stomatal closure improves aphid feeding at elevated CO₂ levels. Elevated CO₂ concentrations were also shown to upregulate an abscisic-acid-independent enzyme, carbonic anhydrase, which led to a further decrease in stomatal aperture in aphid-infested plants. These effects enhance phloem-feeding time. Based on the above report, we suggest that the accumulation of high amounts of PLRV RNA in M. persicae at elevated CO₂ concentrations (940 ± 50 ppm) observed in this study may have been due to increased feeding time. Further studies are needed to determine whether the increase in the number of PLRV-infected plants and in PLRV RNA content in plant tissue due to high temperature and CO₂ concentration significantly increases damage to potato crops.