Different sets of TaCKX genes affect yield-related traits in wheat plants grown in a controlled environment and in field conditions

Background TaCKX wheat gene family members (GFMs) encode the enzyme cytokinin oxidase/dehydrogenase (CKX), which irreversibly degrades cytokinins. The genes are important regulators of cytokinin content and take part in growth and development, with a major impact on yield-related traits. The goal of this research was to test whether these genes might be differentially expressed in the field compared to laboratory conditions and consequently differently affect plant development and yield. Results We compared expression and crosstalk of the TaCKX GFMs and TaNAC2-5A gene in modern varieties grown in a growth chamber (GC) and in the field and looked for differences in their impact on yield-related traits. The TaNAC2-5A gene was included in the research since it was expected to play an important role in co-regulation of these genes. The range of relative expression levels of TaCKX GFMs and TaNAC2-5A gene among tested cultivars was from 5 for TaCKX8 to more than 100 for TaCKX9 in the GC and from 6 for TaCKX8 to 275 for TaCKX10 in the field. The range was similar for four of them in the GC, but was much higher for seven others and TaNAC2-5A in the field. The TaCKX GFMs and TaNAC2-5A form co-expression groups, which differ depending on growth conditions. Consequently, the genes also differently regulate yield-related traits in the GC and in the field. TaNAC2-5A took part in negative regulation of tiller number and CKX activity in seedling roots only in controlled GC conditions. Grain number and grain yield were negatively regulated by TaCKX10 in the GC but positively by TaCKX8 and others in the field. Some of the genes, which were expressed in seedling roots, negatively influenced tiller number and positively regulated seedling root weight, CKX activity in the spikes, thousand grain weight (TGW) as well as formation of semi-empty spikes. Conclusions We have documented that: 1) natural variation in expression levels of tested genes in both environments is very high, indicating the possibility of selection of beneficial genotypes for breeding purposes, 2) to create a model of an ideotype for breeding, we need to take into consideration the natural environment.