Transgenerational accumulation of methylome changes discovered in commercially reared honey bee (Apis mellifera) queens

Published: 24-04-2020| Version 2 | DOI: 10.17632/wzfmyz3rp8.2
Xu Jiang He,
Andrew Barron,
Yi Bo Liu,
Zi Long Wang,
Wei Yu Yan,
Zhi-jiang Zeng


Whether a female honey bee (Apis mellifera) develops into a worker or a queen depends on her nutrition during development, which changes the epigenome to alter the developmental trajectory. Beekeepers typically exploit this developmental plasticity to produce queen bees by transplanting worker larvae into queen cells to be reared as queens, thus redirecting a worker developmental pathway to a queen developmental pathway. We studied the consequences of this manipulation for the queen phenotype and methylome over four generations. Queens reared from worker larvae consistently had fewer ovarioles than queens reared from eggs. Over four generations the methylomes of lines of queens reared from eggs and worker larvae diverged, accumulating increasing differences in exons of genes related to caste differentiation, growth and immunity. We discuss the consequences of these cryptic changes to the honey bee epigenome for the health and viability of honey bee stocks.


Steps to reproduce

G1E were generation 1 queens reared from eggs transferred to queen cells on the 2nd day after laying. G1L1 were G1 queens reared from one-day old larvae transferred to queen cells on the 4th day after laying. G1L2 were G1 queens reared from two-day old larvae transferred to queen cells on the 5th day after laying. Eggs from the G1E queen were transferred as eggs to queen cells on the 2nd day post laying to create the G2E group. Eggs from the G1L1 queens were transferred to queen cells on the 4th day after laying to create the G2L1 group. Eggs from the G1L2 queens were transferred to queen cells on the 5th day after laying to create the G2L2 group (Fig. 1). Queen cells of the G2 groups were treated the same way as the G1 queen cells. The emerging G2 queens were reared and sampled as for the G1 groups. We repeated this process to create the 3rd and 4th generation queen groups: G3E, G3L1, G3L2 and G4E, G4L1 and G4L2. Three types (E, L1 and L2) of queens were sampled in each generation (G1-G4) for methylation and ovariole anlaysis. Queens sampled for methylation analysis were flash frozen in liquid nitrogen when collected after emergence and stored in a -80 ℃ refrigerator. Of the six queens sampled for methylation analysis, three or four queens were used for genome-wide methylation testing. The brain, ovary and thorax of each queen was dissected over ice as one sample. Tissues from each queen were pooled for genomic DNA extraction. In total, 12 queen groups were sampled for meythylation analysis. DNA samples were then sent for whole-genome bisulfite sequencing analysis by the Novogene Bioinformatics Technology Co., Ltd/ Of all samples, one sample in G2L2 group did not meet our quality control requirements for genomic sequencing, consequently we lost one G2L2 sample.