OVERWINTERING OF APLANOBACTER STEWARTI

A Real-Time PCR Differentiating Pantoea stewartii subsp. stewartii From P. stewartii subsp. indologenes in Corn Seed

Stewart's wilt of corn caused by the bacterium Pantoea stewartii subsp. stewartii is a seed-borne disease of major phytosanitary importance. Many countries have imposed restrictions on corn seed imports from regions where the disease occurs to prevent the potential introduction of the pathogen. Current laboratory testing methods (enzyme-linked immunosorbent assay [ELISA] and polymerase chain reaction [PCR]) cannot readily distinguish P. stewartii subsp. stewartii from the closely related subspecies Pantoea stewartii subsp. indologenes. However, P. stewartii subsp. indologenes, a nonpathogen on corn, is occasionally found on corn seed as part of the resident bacterial population and can yield false positive test results. A real-time PCR targeting the cpsAB intergenic sequence was developed to specifically detect P. stewartii subsp. stewartii from corn seeds and distinguish it from P. stewartii subsp. indologenes. The assay successfully detected P. stewartii subsp. stewartii from corn seed, and P. stewartii subsp. indologenes-contaminated seed lots, which previously yielded false positives by ELISA and published PCR methods, were negative. The absence of P. stewartii subsp. stewartii and the presence of P. stewartii subsp. indologenes in this seed were confirmed by size differentiation of the cpsAB amplicons in a conventional PCR. By distinguishing the two subspecies, the assays described would avoid false positive results and help prevent unnecessary restrictions on international movement of corn seed.

The bacterium Pantoea stewartii uses two different type III secretion systems to colonize its plant host and insect vector

Plant- and animal-pathogenic bacteria utilize phylogenetically distinct type III secretion systems (T3SS) that produce needle-like injectisomes or pili for the delivery of effector proteins into host cells. Pantoea stewartii subsp. stewartii (herein referred to as P. stewartii), the causative agent of Stewart's bacterial wilt and leaf blight of maize, carries phylogenetically distinct T3SSs. In addition to an Hrc-Hrp T3SS, known to be essential for maize pathogenesis, P. stewartii has a second T3SS (Pantoea secretion island 2 [PSI-2]) that is required for persistence in its flea beetle vector, Chaetocnema pulicaria (Melsh). PSI-2 belongs to the Inv-Mxi-Spa T3SS family, typically found in animal pathogens. Mutagenesis of the PSI-2 psaN gene, which encodes an ATPase essential for secretion of T3SS effectors by the injectisome, greatly reduces both the persistence of P. stewartii in flea beetle guts and the beetle's ability to transmit P. stewartii to maize. Ectopic expression of the psaN gene complements these phenotypes. In addition, the PSI-2 psaN gene is not required for P. stewartii pathogenesis of maize and is transcriptionally upregulated in insects compared to maize tissues. Thus, the Hrp and PSI-2 T3SSs play different roles in the life cycle of P. stewartii as it alternates between its insect vector and plant host.

Influence of Host Resistance on Stewart's Wilt Forecasts and Probability of Exceeding Thresholds for Use of Seed-Treatment Insecticides on Sweet Corn

Many sweet corn (Zea mays) hybrids commercially available today have higher levels of resistance to Stewart's disease (caused by Pantoea stewartii subsp. stewartii) than the cultivars from which Stevens developed the first forecast of this disease in the 1930s. Incorporating levels of host resistance into forecasts of the seedling wilt phase of Stewart's disease (i.e., Stewart's wilt) could improve control decisions for sweet corn which are made prior to planting. Incidence of systemic infection of seedlings was assessed on 27 sweet corn hybrids with a range of reactions to P. stewartii. In total, 741 observations were collected from 1998 to 2009 in 79 field trials at 15 locations throughout Illinois and one each in Kentucky and Delaware. Relative frequency distributions of the incidence of systemic Stewart's wilt were developed for combinations of hybrids with different levels of resistance and ranges of winter temperature from Stewart's wilt forecasts. The probability of exceeding thresholds of 1 or 5% incidence that warrant the use of seed-treatment insecticides on sweet corn grown for fresh market or processing, respectively, was determined from these frequency distributions. Levels of host resistance affected the incidence of systemic seedling wilt within ranges of winter temperatures used by Stewart's wilt forecasts. For moderate and resistant hybrids, frequency distributions of Stewart's wilt incidence and mean incidence ranging from 0.7 to 1.8% did not differ among three winter temperature ranges above -2.8°C. Conversely, distributions of Stewart's wilt incidence on susceptible hybrids differed among each of the four ranges of winter temperature from the Stevens-Boewe forecast (i.e., >0.6, -1.1 to 0.6, -2.8 to -1.1, and <-2.8°C), with mean incidence ranging from 0.5 to 8.5%. Occurrence of Stewart's wilt also differed among trials varying in number of winter months above -4.4°C, the criterion used by the Iowa State forecast of this disease. Levels of host resistance to P. stewartii also affected the occurrence of Stewart's wilt as predicted by the Iowa State method. The probability of exceeding economic thresholds of 1 or 5% incidence of systemic Stewart's wilt depended on levels of host resistance and winter temperature. Stewart's wilt is unlikely to exceed economic thresholds when the mean winter temperature is below -4.4°C. When mean winter temperature was above -2.8°C, the probability of exceeding 1% incidence of systemic Stewart's wilt was 0.59 for susceptible sweet corn hybrids and 0.28 for moderate and resistant hybrids. When mean winter temperature was below -2.8°C, the probability of exceeding 1% incidence of systemic Stewart's wilt was 0.22 for susceptible hybrids and 0.04 for moderate or resistant sweet corn hybrids. The probability of exceeding 5% incidence was less than 0.1, except when the mean winter temperature was above -2.8°C and susceptible hybrids were grown.

Quantifying the Feeding Periods Required by Corn Flea Beetles to Acquire and Transmit Pantoea stewartii

The feeding periods required by corn flea beetles to acquire and transmit Pantoea stewartii were investigated in the Stewart's disease of corn pathosystem. To quantify the effect of acquisition feeding period on percentage of acquisition, field-collected corn beetles were allowed to feed for 6, 12, 24 36, 48, and 72 h on corn seedlings previously inoculated with a rifampicin- and nalidixic acid-restraint strain of P. stewartii. Acquisition of P. stewartii by corn flea beetles was considered positive if the rifampicin- and nalidixic acid-marked strain was recovered on selective media. To quantity the effect of transmission feeding period on percent transmission of P. stewartii by corn flea beetles, P. stewartii- infested corn flea beetles were allowed to feed on healthy corn seedlings for periods of 3, 6, 12, 24, 36, 48, and 72 h. After the appropriate transmission feeding period, leaf tissues surrounding the sites of feeding scars were cultured for the presence of the P. stewartii-marked strain. Transmission of P. stewartii was considered positive if the marked strain was recovered on selective media. Acquisition of P. stewartii occurred within 6 h and the percentage of corn flea beetles that had acquired P. stewartii after 72 h ranged from 68 to 94%. The change in P. stewartii acquisition by corn flea beetles (Y) with respect to acquisition feeding period (X) was best described by the Gompertz model, with R² values ranging from 91 to 99%. The mean time for acquisition by 50% of the corn flea beetles was 36.5 ± 11.6 h. The minimum transmission feeding time required for corn flea beetles to transmit P. stewartii following a 48-h acquisition feeding period was less than 3 h. The percent transmission of P. stewartii by corn flea beetles was nearly 100% after a 48-h transmission feeding period and was 100% by 72 h. Among population growth models evaluated, the monomolecular model best described the relationship between percent transmission (Y) and transmission feeding periods (X), with R ² values of up to 84%. However, a nonlinear form of the monomolecular model better quantified the relationship between percent transmission and transmission feeding period, because pseudo-R² values ranged between 98.1 and 99.5%. The predicted transmission feeding time required for 50% of P. stewartii-infested corn flea beetles to transmit the pathogen was 7.6 ± 0.87 h. These results suggest that the corn flea beetle is a highly efficient vector that can quickly acquire and transmit P. stewartii, thereby requiring insecticide seed treatments and foliar insecticides that act quickly to prevent corn flea beetles from acquiring and transmitting P. stewartii to corn plants.

Ability of an ELISA-Based Seed Health Test to Detect Erwinia stewartii in Maize Seed Treated with Fungicides and Insecticides

Two sets of experiments were done to examine whether seed-treatment chemicals affected the ability of an enzyme-linked immunosorbent assay (ELISA)-based seed health test to detect Erwinia stewartii. The chemicals evaluated included Actellic, Apron, Captan, Cruiser, Gaucho, Maxim, Poncho, Thiram, and Vitavax in 11 seed-treatment combinations. In one experiment, seed-treatment chemicals were evaluated quantitatively in a critical region of ELISA absorbance values near 0.5 using maize seed that were spiked with uniform quantities of a liquid suspension of E. stewartii. The number of bacteria in each sample was estimated from ELISA absorbance values using standard curves. Log CFU of E. stewartii per sample were not significantly different among the untreated control and the 11 seed treatments compared with Tukey's Studentized Range Test (P = 0.05). Means of log CFU/ml for all treatments were tightly clustered around 5.70 which corresponded to an absorbance value of 0.440 and a bacterial population of about 500,000 CFU/ml. In a second set of experiments, seed treatment chemicals were evaluated based on qualitative decisions that resulted from the ELISA-based seed health test of seed lots of Jubilee and A632 infected with E. stewartii. The number of negative samples was not substantially greater than expected based on binomial probabilities except for samples of Captan/Vitavax-treated A632, which we considered to be a type I error. The mean absorbance values of positive samples ranged from 1.42 to 1.72 for A632 and from 1.51 to 1.91 for Jubilee and did not differ significantly among the seed treatments. There was no consistent evidence from these experiments that fungicide or insecticide seed treatments interfered with the sensitivity of the ELISA or altered low (e.g., 0.5) or high (e.g. 1.4 to 1.9) absorbance values. The ability of the ELISA-based seed health test to detect E. stewartii in maize seed was not affected by these seed treatments.

Rates of Transmitting Erwinia stewartii from Seed to Seedlings of a Sweet Corn Hybrid Susceptible to Stewart's Wilt

Rates of transmitting Erwinia stewartii from seed to seedlings were estimated from field grow-outs of seedlings grown from seed infected with E. stewartii. Infected seed were produced in 1998, 1999, and 2000 on a Stewart's wilt-susceptible sweet corn hybrid, Jubilee. Seedlings were inoculated repeatedly with pinprick inoculators and suspensions of E. stewartii were injected into ear shanks of the primary ears of each adult plant. Seed from inoculated plants were harvested and bulked. Single kernels were assayed for E. stewartii to estimate the proportion of kernels infected with E. stewartii. Estimates of E. stewartii-infection were 15.6 ± 4.3, 49.4 ± 3.9, and 12.5 ± 2.4% for seed produced in 1998, 1999, and 2000, respectively. Approximately 61,800 seedlings were grown in DeKalb, IL in 1999 and 83,400 and 60,000 seedlings were grown in Plover WI in 2000 and 2001, respectively, from infected seed lots produced the previous year. Approximately 10,000, 12,200, and 29,400 seedlings of susceptible sweet corn hybrids also were grown each year from commercial seed produced in Idaho where Stewart's wilt does not occur. Based on estimates of kernel infection in each seed lot and plant populations in each grow-out trial, about 9,600, 41,200, and 7,500 seedlings were grown from infected kernels in 1999, 2000, and 2001, respectively. Seedlings at the two- to three-leaf stage were examined for symptoms of Stewart's wilt. Infected plants were confirmed by microscopic observations of bacterial ooze and by enzyme-linked immunosorbent assay. When data were combined from all three trials, 59 of approximately 58,300 seedlings grown from infected seed were infected with E. stewartii based on symptoms of Stewart's wilt and E. stewartii-positive leaf tissue samples. Of these 59 seedlings, 22 probably were infected from seed-to-seedling transmission of E. stewartii and 37 probably were the result of natural infection due to the presence of flea beetles in DeKalb in 1999. Twenty-two infected seedlings from 58,300 infected kernels corresponds to a seed-to-seedling transmission rate of 0.038%. This rate of seed-to-seedling transmission of E. stewartii is substantially lower than seed transmission rates reported in the first half of the twentieth century; however, it is similar to seed-to-seedling transmission rates reported from other recent research.

Transmission of Erwinia stewartii from Plants to Kernels and Reactions of Corn Hybrids to Stewart's Wilt

Stewart's wilt reactions of 98 food-grade, white corn hybrids, 3 yellow dent corn hybrids, and 23 sweet corn hybrids and infection of kernels by E. stewartii were evaluated in 1998, 1999, and 2000. Stewart's wilt symptoms were rated from 1 (no appreciable spread of symptoms) to 9 (dead plants) following inoculation. The mean Stewart's wilt ratings for the food-grade, white corn and yellow dent corn hybrids were 1.9, 2.4, and 2.9 in 1998, 1999, and 2000, respectively. The mean Stewart's wilt ratings for the sweet corn hybrids were 3.8, 4.2, and 4.6 in 1998, 1999, and 2000, respectively. Hybrids with ratings less than 3 were classified as resistant. Hybrids with ratings between 3 and 4.5 were classified as moderate. Hybrids with ratings greater than 4.5 were classified as susceptible. Ears harvested from each row in 1998 and 1999 were assayed for E. stewartii using an enzyme-linked immunosorbent assay (ELISA)-based seed health test. Kernels from 16 hybrids were positive for E. stewartii in 1998. Kernels from 11 hybrids were positive for E. stewartii in 1999. Kernel infection by E. stewartii was affected considerably by the level of host resistance (i.e., reactions of seed parent plants). For hybrids classified as resistant, estimates of kernel infection were 0.024 and 0.0007% in 1998 and 1999, respectively. For hybrids with moderate reactions to Stewart's wilt, estimates of kernel infection were 0.19 and 0.07% in 1998 and 1999, respectively. For hybrids with susceptible reactions to Stewart's wilt, estimates of kernel infection were 11.6 and 7.8% in 1998 and 1999, respectively. Based on high levels of Stewart's wilt resistance in food-grade, white corn hybrids, and low rates of kernel infection by E. stewartii in resistant and moderate hybrids, there is an exceedingly low probability of introducing E. stewartii to areas where it does not occur by transmitting the bacterium in grain of the food-grade, white corn hybrids evaluated in this study. Although all of the kernels harvested in these experiments were produced as grain on open-pollinated F1 hybrids, the rates of kernel infection observed for hybrids with resistant, moderate and susceptible reactions to Stewart's wilt are applicable to seed produced on inbred lines with equivalent Stewart's wilt reactions.

Seed Transmission of Pantoea stewartii in Field and Sweet Corn

Seed transmission of Pantoea stewartii was evaluated by assays of more than 76,000 plants in greenhouse and field grow-out trials. Fourteen P. stewartii-infected seed lots were obtained from two dent corn inbreds and two sweet corn cultivars that were inoculated with either a rifampicin and nalidixic acid-resistant strain (rif-9A) or a wild-type strain (SS104) of P. stewartii. Four additional seed lots were collected from naturally infected inbreds. Percentages of infected kernels ranged from 0.8 to 72%, as determined by agar plating or by individual-kernel enzyme-linked immunosorbent assay (ELISA). Plants grown from this seed were assayed by a stem-printing technique that consisted of cutting and pressing a cross-section of each stem onto agar media. Prints were examined for development of P. stewartii colonies after 24 and 48 h. The transmission rate from seed produced on the inoculated plants was 0.066% (28 of 42,206 plants), based on all seedlings assayed. Transmission was estimated to be 0.14% from infected kernels. The transmission rate from seed produced on naturally infected plants was 0.0029% (1 of 34,924 plants), based on all seedlings, and 0.022% from infected kernels. Seed transmission occurred significantly less often (P = 0.034) from seed produced on naturally infected plants than from seed produced on inoculated plants, probably due to greater kernel damage caused by ear shank inoculation. The rarity of seed transmission of P. stewartii from heavily infected seed lots that would ordinarily be rejected due to poor germination suggests that the likelihood of seed transmission from good quality commercial seed corn is virtually nonexistent.

Herbaceous Weeds Are Not Ecologically Important Reservoirs of Erwinia tracheiphila

The potential of herbaceous weeds commonly growing in or adjacent to cucurbit crops to serve as alternate hosts and overwintering reservoirs of Erwinia tracheiphila, a causal agent of cucurbit wilt, was investigated. Methods for isolation, maintenance, long-term storage, and detection of E. tracheiphila from infected plants were developed. E. tracheiphila was consistently detected by enzyme-linked immunosorbent assay (ELISA) and reisolated from infected, susceptible, cucurbit species. When six common herbaceous weed species were inoculated, E. tracheiphila was detected in 49% (combined species) of the plants by ELISA 3 weeks after inoculation. However, we were unable to reisolate E. tracheiphila from these plants by standard techniques. Immunoaffinity isolation with a sensitivity of 2 CFU per sample also failed to recover E. tracheiphila from weed species. Comparisons of cucumber and goldenrod inoculated with live or formaldehyde-killed E. tracheiphila indicated that immunoassays could detect nonviable E. tracheiphila systemically spread in plants 3 weeks post-inoculation. In these tests, the pathogen was reisolated only from cucumber plants inoculated with live E. tracheiphila. Although we could reproduce serological evidence of E. tracheiphila antigen in the weeds investigated, our results do not support the hypothesis that E. tracheiphila can infect, survive in, or overwinter in the weed species tested.