Evaluation of the summer and winter bee transcriptomes
Through RNA-sequencing we obtained on average 17.3 million pseudo-alignments on the honey bee reference transcriptome for each biological replicate of nurse (N), forager (F) and winter (W). For the protein coding genes in the honey bee genome (Amel_HAv3.1) (Wallberg, A. et al., 2019), 92.38% (9178/9935) passed our low count filter.
The number of genes significantly differentially expressed (DEGs) between nurse, forager, and winter bees in each tissue are listed in Table 1. In fat body tissue, we found 683 genes differentially expressed between nurses and foragers, 1617 between winter bees and foragers, and 934 between winter bees and nurses. In flight muscle flight muscle tissue, we found 2673 DEGs between nurses and foragers, 1176 between winter bees and foragers, and 2855 between winter bees and nurses. Interestingly, there were many more DEGs in flight muscle tissues (2763) versus abdominal tissue (693) in nurses versus foragers. (Table S3-S7)
We performed chi-squared tests to determine, for each tissue, whether the number of DEGs between winter bees and nurses is significantly different than the number of DEGs between winter bees and foragers. The results are displayed in Table 2. There are significantly more genes differentially expressed between winter bees and forager bees (versus nurse bees) in the fat body tissues (\(\chi^{2}=211.76\),\(p<{2.26*10}^{-16}\)), suggesting that winter bees are more similar to nurse bees in this tissue. In the flight muscle, there are significantly more genes differentially expressed between winter bees and nurse bees (relative to forager bees; \(\chi^{2}=895.06\),\(p<{2.26*10}^{-16}\)), suggesting that winter bees are more similar to forager bees in this tissue.
To further evaluate how the expression patterns of winter bees correspond to those of summer bees, we conducted hierarchical clustering (Figure 1) and k -means clustering analyses (calling k = 2 clusters), using all 9178 genes (Figure 2). Both analyses demonstrated that, in fat body tissue, winter bee and nurse samples form a cluster independently of foragers, while in flight muscle tissue, winter bee and forager samples form a cluster independently of nurses. When clustering with only genes that were differentially expressed between nurses and foragers (n = 2763 flight muscle tissue DEGs; n = 683 fat body tissue DEGs) (Figure S3), the same clustering is observed.