Four transgenic cowpea lines (named IT97K-T, IT98K-T, Gourgou-T and Nafi-T) and their respective non-transgenic lines, i.e. conventional lines (IT97K, IT98K, Gourgou and Nafi) were used. Each transgenic line shared similar genetic make-ups with its conventional counterpart except the presence of the cry1Ab gene in the transgenic line. All seeds were taken from the National biosafety laboratory seed collection.

Seeds from each cowpea line were soaked for about 15 hours and transferred on wet blotting paper in Petri dishes until germination started. Small portions of about 50 mg were removed from the cotyledon without disturbing the embryo for the detection of Cry1Ab protein. Three pre-germinated seeds with known status were sown per cowpea line in 10-liter buckets in a biosafety containment greenhouse. The buckets were arranged following a complete randomized design with three replications per treatment.

To ascertain the presence or absence of Cry1Ab protein in transgenic and conventional cowpea, respectively, seed stocks were analyzed by enzyme-linked immunosorbent assay (ELISA) using the QualiPlate kit for Cry1Ab/Cry1Ac (Envirologix Inc., USA). Seed portions collected as indicated above, were placed in 1.5 Eppendorf tubes, and homogenized in 300 µL of extraction buffer using disposable pipette tips. The antigen extraction buffer was composed of phosphate buffer saline, pH 7.4 containing 2% of polyvinylpyrrolidone (PVP-40). After centrifugation of the extracts at 10,000 xg for 10 min the supernatants were incubated in duplicates. Non-transgenic seeds (negative controls) were tested in triplicates and the threshold for Cry1Ab detection was determined as the average of their absorbance readings plus three times the standard deviation [25] .

Cowpea aphid-borne mosaic virus isolate BF72 of our collection was first propagated on susceptible cowpea cultivar “Gorom local” by mechanical inoculation. To prepare the inoculum, leaves showing symptoms were collected and 1g was ground in 5 ml of inoculation buffer (a mixture of 0.1 M NaH₂PO₄ and 0.1 M Na₂HPO₄, adjusted to pH 7.4). The leaf extract was filtered on muslin cloth and carborundum 600 mesh was added. The inoculum was rubbed with fingers onto the leaves of germinated cowpea seedlings 7 days after sowing (DAS). Inoculated plants were monitored daily for symptom appearance.

CABMV accumulation was assessed by double antibody sandwich ELISA [26] . Polyclonal rabbit antibodies raised against CABMV particles were used as coating antibodies. The same antibodies linked to alkaline phosphatase were used as conjugate. Last fully expanded leaves of inoculated plants were collected and pooled by cowpea line. Plant extracts were prepared by grinding 1g of leaf in 10 mL of antigen extraction buffer. Extracts were centrifuged as indicated above and tested in triplicates.

For each cowpea line, the appearance of CABMV-induced symptoms was monitored daily. Symptom severity was assessed three weeks post-inoculation using the 1 - 5 scale proposed by [27] where 1 represents no visible symptom and 5 indicates very severe symptoms (mosaic associated with leaf deformation and reduction of plant growth). The number of days to 50% flowering and virus multiplication in inoculated plants were also assessed.

All data were collected using Microsoft Excel and then imported into R software [17] for statistical analysis. The data were expressed as mean ± standard deviation of three replicates. Assumption for normality was assessed through QQ-plots and the Shapiro-Wilk test, while homogeneity of variances was assessed using the Levene test. Analysis of variance (ANOVA) was used to compare data measurements across cowpea lines. Post-hoc tests for the separation of the means were conducted using Fisher’s LSD with the significance level set at p < 0.05.

The results of the analysis of the presence of the Cry1Ab protein in the seeds of transgenic and conventional cowpea lines are summarized in Table 1 . The presence of the Cry1Ab protein was confirmed in all transgenic varieties at varying levels between 59% and 72%. Statistical analysis of the results using the chi-square test did not reveal significant differences (χ² = 3.26; p = 0.35). The presence of the Cry1Ab protein at rates less than 100% indicates that parts of the seeds (27.9% - 41%) of all the transgenic seed lots do not contain the Bt toxin. No transgenic seed was found in conventional cowpea seed lots.

^aThe names of transgenic cowpea lines end with “-T” to distinguish them from their respective conventional counterparts.

In response to CABMV inoculation, the plants showed various mosaic symptoms ( Figure 1 ). In the yellow mosaic type was characterized the leaves were predominantly yellow indicating severe chlorosis. The vein-banding pattern was also associated with chlorosis leaving areas with chlorophyll along the veins. A green mosaic type also occurred with less obvious chlorosis but associated with leaf deformation. The different types of symptoms were observed indistinctively on the various cowpea cultivars.

Table 2 summarizes three parameters of the development of symptoms. All plants inoculated with the virus extract developed clear symptoms during the first- and second-week post inoculation. The time of the onset of the symptoms ranged between 7- and 11-day post-inoculation (dpi). Symptoms appeared earlier in cowpea lines Nafi-T with disease incidence of 66% at 7 dpi. Most inoculated plants showed clear symptoms up to 10 dpi. Symptoms appeared at 11 dpi were observed on 1 out of 9 plants only, in cowpea lines Gourgou and IT97K-T. One-way ANOVA indicated that not all cowpea lines developed symptoms in the same time frame (F_7,64 = 3.09; p = 0.007).

^aThe names of transgenic cowpea lines end with “-T” to distinguish them from their respective conventional near-isogenic lines. ^bdpi: days post inoculation. Means followed by the same letter(s) within the symptom onset time column are not significantly different. Cowpea lines did not differ significantly in symptom severity. ^cAUDPC: area under disease progress curve.

In comparing transgenic cowpea lines with their non-transgenic counterparts, the onset of symptoms was significantly different (p = 0.002) in the pair Nafi/Nafi-T only. All other pairs showed similar durations for the appearance of symptoms. The earliness of the development of symptoms in Nafi-T was also associated with a higher area under disease progress curve (AUDPC).

The symptom severity scores were comprised in a very narrow range between 3.3 and 3.7. No statistical difference was found among mean severity scores (F_7,64 = 0.594; p = 0.76).

The average absorbance readings reflecting the levels of viral accumulation in plants varied between 1.213 ± 0.015 in cowpea line Gourgou-T and 1.597 ± 0.019 in Nafi-T.

There was a significant cowpea line effect (F_7,16 = 34.62; p < 0.001) ( Figure 2 ). Interestingly, on the one hand, no significant differences were observed between pairs of transgenic/non-transgenic line IT97K/IT97K-T (p = 0.556), IT98K/IT98K-T (p = 0.238), Gourgou/Gourgou-T (p = 0.065), and Nafi/Nafi-T (p = 0.162). On the other hand, marked differences were found between most of the other combinations of cowpea line. Notably, each of the two lines Gourgou and Gourgou-T significantly differed from any other cowpea line (p < 0.001) by their lower absorbance readings.

The results of the assessment of days to 50% flowering (DF50) of cowpea lines are shown in Figure 3 . Taken together, the DF50 varied on average between 34.3 ± 1.5 days and 42.0 ± 1.7 days. As indicated by one-way ANOVA, differences in DF50 across the cowpea lines were significant (F_7,16 = 8.17; p = 0.0003).

Comparisons between the transgenic cowpea lines and their corresponding non-transgenic parents showed a significant difference between Nafi and Nafi-T only (p = 0.0037). No significant differences were found between cowpea lines in pairs IT97K/IT97K-T (p = 0.243), IT98K/IT98K-T (p=0.634) and Gourgou/Gourgou-T (p = 0.243).