Figure 1.
Transcript analysis of CO1 and CO2 via RT-PCR in field-grown P. deltoides.
(A) Average monthly high and low temperatures and daylength in Mississippi (USA) and year-round expression of CO1 and CO2 in mature P. deltoides. CO1 and CO2 graphs show the relative fold change in expression levels normalized against March expression level. Dashed lines indicate missing samples. Error bars show standard deviation about the mean. (B) CO1 and CO2 transcripts were most abundant in leaf tissues of mature P. deltoides sampled in May, but least abundant in the shoot apex. Poplar UBQ was used as an internal control. Letters above the bars showing the abundance of CO1 or CO2 transcripts indicate statistically significant (P≤0.001) differences. Error bars indicate SD about the mean. (C) CO1 transcripts were expressed abundantly in juvenile and mature trees (April). However, CO2 transcripts were significantly more abundant in mature and juvenile trees. Letters above the bars showing the abundance of CO1 or CO2 transcripts indicate statistically significant (P≤0.001) differences. Error bars indicate SD about the mean. (D) Transcript abundance of CO1 and CO2 did not significantly (P>0.8) fluctuate in leaves sampled in February. LHY transcripts were significantly (P≤0.05) abundant at mid-day in the same tissues. (E) Transcript levels of CO1 and CO2 were significantly (P≤0.001) higher at dawn in leaves sampled in May, whereas LHY was significantly (P≤0.001) more abundant in the morning. *≤0.05 and **≤0.005, statistical significance within time points.
Figure 2.
In situ expression analysis of CO1 and CO2 in leaf, reproductive bud, and shoot apex collected during the growing season from mature P. deltoides.
Panels in the first two columns were results from the bright-field image of in situ hybridization and colorimetric detection of CO1 or CO2 transcripts. The antisense probe generated positive signals (dark blue) if present, while the sense probe served as negative control. The third column (schematic drawing) illustrates leaf cross-sections and longitudinal reproductive bud and shoot apex sections where CO1 and CO2 transcripts (pink color) were located, based on visual observations, as well as captured images. Scale bar, 100 µm.
Figure 3.
Environmental regulation of CO1 and CO2 transcription in mature P. deltoides.
(A) Abundance of CO1 transcripts increased (P≤0.0002) under short days (SD) in leaves of field-grown trees (two genotypes), but showed no significant (P>0.05) difference under SD in growth rooms. CO2 expression did not change significantly under SD and long days (LD). FT2 transcripts were significantly (P≤0.0001) less abundant under SD in trees grown both in the field and growth room. Samples were collected 2 h after sunrise or the beginning of the light period. (B) Reduced ambient light intensity did not significantly (P>0.05) affect CO1 and CO2 transcription in field-grown trees. Conversely, transcript abundance of FT2 was significantly reduced (P≤0.008) at the lower light intensity. (C) Temperatures of 38, 25, and 4°C did not significantly (P>0.05) influence the abundance of CO1 and CO2 transcripts under LD, except that CO2 transcripts were significantly (P≤0.05) fewer at 38°C. FT2 transcription was significantly (P≤0.0005) repressed by 38°C and 4°C. (D) While FT2 transcription decreased significantly (P≤0.05) under low, medium, and severe water stress (predawn leaf water potential in MPa), the levels CO1 and CO2 transcripts were not significantly (P>0.05) affected when compared with controls. *≤0.05, **≤0.005, and ***≤0.0005, statistical significance. Error bars represent standard deviation about the mean.
Figure 4.
Ectopic expression of CO1 and CO2 individually (Pro35S:CO1 or Pro35S:CO2) or together (Pro35S:CO1/CO2) in poplar (P. tremula × P alba).
(A) When compared with controls at age 5, Pro35S:CO1 or Pro35S:CO2 trees did not differ in reproductive onset, spring reproductive and vegetative bud break, and fall bud set. Pro35S:FT2 trees showed year-round active growth. Red arrows denote the emerging inflorescence in the spring, whereas black arrows point the dormant terminal vegetative bud in the fall. Unlike wild-type and vector controls, Pro35S:CO1 or Pro35S:CO2 trees (1, 2, and 3) significantly overproduced CO1 or CO2 transcripts when analyzed via qRT-PCR in leaves sampled in April. While the expression of FT1 was undetectable, that of FT2 fluctuated with no clear trend in controls and CO1- or CO2-overexpressing trees. Letters above the bars showing the abundance of CO1 or CO2 transcripts indicate statistically significant (P≤0.001) differences. Error bars indicate SD about the mean. (B) When Pro35S:CO1 and Pro35S:CO2 were co-expressed in the same trees, no difference between the transformants and controls was observed in spring bud break and fall bud set in two years. However, Pro35S:FT2 trees showed a non-dormant phenotype. Black arrows indicate the terminal vegetative bud, whereas purple arrows point to the axillary vegetative bud. The axillary vegetative buds were opening and preformed leaves were emerging from the control and co-expressing transgenic trees on March 23. Unlike wild-type and vector-control plants, co-expressing transgenic trees (1, 2, 3, and 4) significantly overproduced CO1 and CO2 transcripts in leaves sampled in April. While the expression of FT1 was undetectable, that of FT2 fluctuated with no clear trend in controls and CO1/CO2 overexpressing trees. Letters above the bars showing the abundance of CO1 or CO2 transcripts indicate statistically significant (P≤0.001) differences. Error bars indicate SD about the mean.
Table 1.
Field-grown Pro35S:CO1, Pro35S:CO2, and control trees were observed for the onset of reproduction for five years, evaluated for the number of reproductive buds or catkins at age 5, and measured for height, diameter, and shoot growth at age 5.
Figure 5.
Transcripts downstream of CO1 and CO2 and their year-round transcript levels were identified in mature P. deltoides via microarray.
Log2 fold-change of each time point relative to the baseline time point (September or Sep) was calculated. Clusters to the left of the heatmaps represent modules and the columns to right of the heatmaps represent the up- (red) and down-regulation (blue) of downstream genes. Months relative to September are above the heatmaps. The pie charts to the right of each heatmap show the functional categorization of GO Biological Process terms. N = number of genes. The Venn diagram shows the number of genes that were common to both CO1 and CO2 (CO1/CO2) datasets, and the pie chart below the diagram shows the GO categorization of CO1/CO2 transcripts. Up (↑) and down (↓) arrows represent partitioning of overall percentage in each pie. “**” denotes the GO term is significantly (P≤0.001, except “development” for genes downstream of CO1 P≤0.006) over-represented in the microarray data when a hypergeometric test was conducted.