Abstract:
Diuraphis noxia Kurdjumov (Hemiptera: Aphididae: Macrosiphini) is a major agricultural pest that causes extensive economic losses to the wheat and barley industries. Resistant cultivars were relatively successful in controlling this pest until the recent development of new D. noxia (Russian wheat aphid, RWA) biotypes. The aim was to investigate the role of the aphid endosymbiont, Buchnera aphidicola, in the RWA-host interaction. It was hypothesized that variations in the endosymbiont’s key essential amino acid biosynthetic pathway genes, their copy numbers, and/or expression levels, maybe a determining factor influence the RWA’s success in the aphid-host interaction. Aphid symbiont species content, key essential amino acid biosynthetic gene variation, plasmid copy numbers and expression levels of ten different RWA biotypes were determined, using DGGE, RT-PCR, RT-qPCR, 5’-RACE and sequencing. The RWA biotypes were shown to be monosymbiotic, with plasmid copy numbers varying between biotypes. Only a single CCC-insert in a non-coding region of the leucine plasmid differed between the biotypes. Similar variations were identified in the family Aphididae, suggesting a regulatory function for this region. The presence of this CCC-insert in a plasmid led to an increase in the leader sequence length of the leuA gene. The insert may also have a functional role through gene regulation, since it increased the expression levels of subsequent genes (leuA and leuB). An endosymbiont that upgrade the host’s diet with the required essential amino acids will be beneficial to RWAs when feeding on resistant wheat cultivars as it will enhance aphid fitness. This suggests selective pressure of resistant wheat cultivars on the aphid, i.e. the incapability to change resistant cultivar essential amino acid content, could select for individuals with beneficial endosymbionts. B. aphidicola could therefore play a role in the development of RWA biotypes. The influences that statistical normalization methods have on the final identification of differentially regulated Affymetrix probe sets in RWA-plant interactions were also investigated. The hypothesis was that a subset of the probe sets determined as differentially regulated would be consistent, regardless of the normalization and background method utilized, if all the other analyses are kept constant. This subset would be normalization-method-independent. The data of two Affymetrix RWA-plant interaction experiments were analyzed with five different normalization and background correcting methods and at three different confidence levels, with the results subjected to FDR and FWER correction algorithms. The results showed that on average a third of the regulated genes were only selected after normalization by a single method and that the total number of genes deemed regulated was highly normalization method dependent. Normalization-method-biases could also not be countered by increased confidence levels and these biases eventually determined the probe sets deemed differentially regulated, even after FDR and FWER corrections. Both these strategies actually increased normalization-method-biases and these could only be corrected by using multiple normalization methods to identify the normalization-method-biases-independent probe set subset.
