Genomic analysis in the sting-2 quantitative trait locus for defensive behavior in the honey bee, Apis mellifera

Abstract

We have sequenced an 81-kb genomic region from the honey bee, Apis mellifera, associated with a quantitative trait locus (QTL) sting-2 for aggressive behavior. This sequence represents the first extensive study of the honey-bee genome structure encompassing putative genes in a QTL for a behavioral trait. Expression of 13 putative genes, as well as two transcripts that were present in a honey-bee EST database, was confirmed through reverse transcription analysis of mRNA from the honey-bee head. Whereas most transcripts exhibited little or no variation between European and Africanized honey-bee alleles, one transcript demonstrated significant nonsynonymous substitutions, deletions, and insertions. All 13 putative genes lacked similarity to known invertebrate or vertebrate proteins or transcripts. This observation may be reflective of the processes that determine the genomic evolution of an insect with social behavior and/or haplo-diploidy and are an indication of the unique nature of the honey-bee genome. These results make this sequence an invaluable research tool for the ongoing honey-bee whole-genome sequencing effort.

Figure 1

Annotation of Am36L17. The Am36L17 BAC clone was assembled into a single sequence with a GC content of ∼40%. All putative transcripts determined by gene-prediction methods (orange) are named numerically (36L17 1-15); 36L17.EST1 and 36L17.EST2 (red) were identified by their presence in the honey-bee brain EST database. The genetic marker linked to aggressive behavior in honey bees and used to isolate this BAC clone is a11-.31. The microsatellite sequence is AME509514. The position of simple repeats and the 371-bp repeat are identified (black bars and triangles, respectively).

Figure 2

Reverse transcription analysis of putative transcripts. RT-PCR was performed on RNA obtained from European worker (EW), European drone (ED), and Africanized worker (AW) heads. Primers were designed to an exon in each transcript. The putative transcripts from 36L17.1 to 36L17.15, 36L17.EST1, and 36L17.EST2 were visualized after 20 cycles. Positive control primers to the Apis glutamate transporter (Am-EAAT +ve control) produced the expected size fragment of 525 bp. The negative control using Taq Polymerase (Am-EAAT-No RT control) confirmed the lack of DNA contamination. With the exception of 36L17.12 and 36L1714, all predictions, including the segments corresponding to the EST hits, generated single RT products of the expected sizes in Africanized and European workers. 36L17.12 and 36L17.14 produced multiple RT-PCR products of varying sizes, indicating a possible failure of the RT-PCR reaction, nonspecific reaction products, or an absence of expression of that particular transcript in the bee head. With the exception of 36L17.7, all products seen in European workers were seen in European drones. RT-PCR products from most transcripts appeared to be present at a lower level in drones in comparison with workers in several cases. In contrast, 36L17.1 appeared to have a slightly elevated level of transcript in the drone sample in comparison with that of the workers. All transcripts seen in European worker heads appear to be expressed in their Africanized counterparts at similar levels.