- This dissertation is comprised of four studies on different elements of Populus biotechnology and genomic science: 1) growth improvement by recombinant DNA (rDNA) modification of gibberellin (GA) metabolism; 2) gene editing using zinc-finger nucleases (ZFNs); 3) induction of floral sterility by RNA interference (RNAi) of AGAMOUS (AG) gene homologs; and 4) transcriptome analysis during natural leaf senescence.
The genus Populus includes a number of environmentally and commercially important tree species. The breeding of Populus has emphasized improved stress resistance, increased wood production, and tailored wood properties for diverse uses. Genetic engineering appears to be able to complement or facilitate Populus breeding in many ways, such as imparting improved or novel traits and accelerating the breeding process. However, the efficacy of traits shown in laboratory or greenhouse trials with model plants and trees may not translate to the field in their effects or stability under plantation conditions. In addition, the scientific knowledge for advanced engineering of useful traits such as bioproduct synthesis are inadequate, and public and regulatory concerns over transgene dispersal from these plants may hinder their delivery of benefits unless such gene movement is mitigated. We conducted several studies to address these research needs.
In study 1, we examined the growth and hormonal consequences of modified GA metabolism and signaling in a model poplar genotype (Populus tremula x P. alba). We tested a total eight genes (i.e., rDNA constructs), two cisgenes, four intragenes, and two transgenes, in greenhouse, and six of them under field conditions. Greenhouse studies showed that four constructs significantly increased stem volume, and two constructs significantly modified the proportion of root or shoot biomass. One of the cisgenic constructs elevated the concentrations of bioactive GA1 and its direct precursor GA20. In field, however, only one intragenic construct led to significant growth improvement. The inconsistent growth performance between greenhouse and field tests was possibly to due to cross-talk among the GA pathway and other stress response pathways, or interactions between the cis- and intragenes with similar endogenes. Our results emphasize the need for extensive field trials in evaluating the value and pleiotropic effects of GA-modifying genes and other applied biotechnology research.
In study 2, we evaluated the efficiency of ZFN-induced mutagenesis in poplar organogenic systems. We created four pairs of heat-inducible ZFN systems, two targeting the floral meristem identity gene LEAFY (LFY), and two targeting the floral homeotic gene AG, into two hybrid poplar clones. Despite that over a total of 21,000 explants were transformed, we were only able to obtain 391 transgenic events. We found two of these events had 7-bp deletions in one allele of the AG2 gene. We did not observe mutations in the LFY or AG1 genes. The ZFN constructs, together with heat induction of their expression, also seemed to depress transformation rates. These results suggested a mutation rate of zero to 0.3% per explant per allele, among the lowest reported for ZFN mutagenesis in plants, and the need for further technical improvements towards efficient use of ZFNs in poplar.
In study 3, we examined floral changes as a result of RNAi-induced suppression of two AG orthologues in field-grown poplar. Among a total of 34 flowering RNAi-events, 17 of them showed altered phenotypes, including altered floral organ identity, loss of floral organ determinacy, and impaired ovule and seed-hair development that was associated with the degree of AG suppression. In addition, five seedless events showed strong suppression of two SEEDSTICK (STK) paralogues. All RNAi-events appeared normal in their vegetative morphology and growth, and alterations in floral phenotypes were stable over multiple years. These results suggested that RNAi suppression of AG and STK genes appears to be a safe and effective means of genetic containment in poplar.
In study 4, we examined transcriptome changes during leaf senescence in field-grown Populus trichocarpa ‘Nisqually-1’, which is the basis of the reference genome. Using monthly (from May to October) transcriptomes from three years (2012, 2015, and 2016) containing 36,007 expressed Populus gene models, we found 17,974 genes that were differentially expressed during the study period. In addition, we assigned 2015 and 2016 collections into four major developmental phases and identified 14,415 DEGs were directly related to transitions between the developmental phases. Gene ontology (GO) enrichment revealed that GOs related to catalytic activity, kinase activity, and enzyme inhibitor activity were up-regulated during leaf senescence, while those related to photosynthesis were down-regulated. A total of 881 TF genes from 54 TF families, notably bHLH, MYB, ERF, MYB-related, NAC, and WRKY family members, were active during leaf senescence. Fuzzy C-means clustering analysis sorted the differentially expressed genes into five groups based on their gene expression patterns. Two groups of genes showed constantly increasing expression during senescence, and a number of upstream sequence motifs, including five with high-confidence, were identified associated with these groups. This study provides a robust resource for investigating leaf senescence and related developmental traits such as programmed cell death and stress tolerance in perennial species. It also could inform efforts to engineer the synthesis of coproducts in leaves in association with senescence onset to mitigate its costs to net photosynthesis.