An explanatory model of temperature influence on flowering through whole-plant accumulation of FLOWERING LOCUS T in Arabidopsis thaliana

Abstract

We assessed mechanistic temperature influence on flowering by incorporating temperature-responsive flowering mechanisms across developmental age into an existing model. Temperature influences the leaf production rate as well as expression of FLOWERING LOCUS T (FT), a photoperiodic flowering regulator that is expressed in leaves. The Arabidopsis Framework Model incorporated temperature influence on leaf growth but ignored the consequences of leaf growth on and direct temperature influence of FT expression. We measured FT production in differently aged leaves and modified the model, adding mechanistic temperature influence on FT transcription, and causing whole-plant FT to accumulate with leaf growth. Our simulations suggest that in long days, the developmental stage (leaf number) at which the reproductive transition occurs is influenced by day length and temperature through FT, while temperature influences the rate of leaf production and the time (in days) the transition occurs. Further, we demonstrate that FT is mainly produced in the first 10 leaves in the Columbia (Col-0) accession, and that FT accumulation alone cannot explain flowering in conditions in which flowering is delayed. Our simulations supported our hypotheses that: (i) temperature regulation of FT, accumulated with leaf growth, is a component of thermal time, and (ii) incorporating mechanistic temperature regulation of FT can improve model predictions when temperatures change over time.

Keywords: Arabidopsis; Arabidopsis thaliana; FT; Framework Model; crop simulation model; flowering time; mathematical model; phenology; photoperiodic flowering; thermal time.

Figure 1.

Overview of FM-v1.5. (A) Summary of how temperature influence on FT was modeled in FM-v1.0 and in FM-v1.5 (blue text) as well as the steps involved in modifying FM-v1.5 (gray text). (B) Schematic of Model FM-v1.5. Temperature (through SVP, FLM, and CO), day length, and the circadian clock regulate expression of FT in the Photoperiodism and Phenology modules per unit tissue. The leaf number and relative leaf age, outputs of the Functional Structural Plant module, are used to determine the capacity of each leaf to express FT, and leaf area is used to determine the amount of leaf tissue present. FT is summed across all leaves in a plant and added to the whole-plant FT from the previous time step. The model ceases leaf production and determines the days to bolt (DtB) when FT reaches a pre-set threshold set by using the leaf number for plants grown in long days (LD, 16-h light, 8-h dark) at 22°C, similarly to the way in which thermal units accumulate to a consistent value before a developmental transition occurs. Red illustrates where adjustments were made to the original model (FM-v1.0). The bold, italic numerals correspond to the numbers in the model description in text.

Figure 2.

FM-v1.5 mimics general behaviours of CO and FT in response to temperature and can accommodate the overall change in amount across treatments. Observed (A, B) and predicted (C, D) diurnal patterns of FT (A, C) and CO (B, D) gene expression in warm-day, cool-night temperature-cycle (22/12 °C-night) treatments and in conditions in which the temperature dropped from 22 °C to 12 °C at dawn, then remained at the cooler temperature (22/12 °C-day) relative to the 22 °C-constant temperature control. The y-axis (A–D) is in nmol. The x-axis (A–D) is in zeitgeber time (ZT), and represents hours after dawn. The white and black bars represent light and dark periods, respectively. Error bars = 1 SE. If error bars are not visible, the SE is smaller than the height of the symbol. Correlation between predicted and observed results for FT (E) and CO (F), as calculated as the AUC 4 days after temperature treatments are imposed. Treatments include warm-day, cool-night cycles, drops to cooler temperatures at dawn, and growth from seed at constant temperatures. All treatment groups include 12, 17 and 22 °C. Dotted lines = correlation, solid lines = one-to-one line. Open circles are drop from 22 °C to 17 °C at dawn (C) growth from seed at 12 °C (F). Data from Kinmonth-Schultz et al. (2016) pooled and compiled in A and B (license to reproduce data obtained from John Wiley and Sons, #4601440806514).

Figure 3.

FT expression declines in later produced leaves. Leaves of short-day-grown plants, that had not yet transitioned to flowering, aged 2 (A), 4 (B) and 6 (C) weeks old were exposed to long days or short days (D) for 3 days, then harvested at 16 h after dawn on the third day to determine FT amount per leaf. The colours in (D) correspond to the colours and ages in panels (A–C). FT levels were determined by absolute copy number and normalized within a replicate. The simulated proportion of FT per unit leaf tissue (cm⁻², solid lines) for each plant age is shown. This value was used in FM-v1.5 as a modifier to adjust the amount of FT produced by each leaf. Percent of the leaf area showing staining in pFT:GUS plants (E). For all, the two cotyledons and first two true leaves were pooled for each sample as they emerge in pairs. The youngest leaves, just emerging at the apex (1–2 mm in length) were also pooled. Older leaves in the 6-week-old plants failed to yield 2 μg total RNA and were excluded. For each plant inset, asterisk indicates one of each cotyledon pair (*). The 10th and 11th leaves to emerge are labelled. The shading of the bar graphs (light to dark) indicates leaf age (oldest, first to emerge, to youngest) and corresponds to the shading in the plant insets. Scale bars = 0.5 cm.

Figure 4.

(A, B) Whole-plant FT accumulation influenced by temperature in changing and constant cool-temperature conditions, differs more strongly from the 22 °C control than does accumulated MPTUs. Total FT accumulated in constant and changing temperature conditions relative to 22 °C constant temperatures (indicated by arrowheads) 9 days post emergence, equivalent to 1 week in changing temperature treatments. (A) LTP+GE: FT accumulation in full FM-v1.5 model, i.e. temperature affects FT gene expression though CO and SVP/FLM as well as through leaf tissue production; LTP: FT accumulation only with leaf tissue production as influenced by temperature, temperature influence on FT gene expression excluded; MPTU: accumulated Modified Photothermal Units from FM-v1.0. Here, daytime and nighttime temperatures are given equal weight. (B) GE: FT accumulation considering only influence of temperature on FT gene expression, decoupled from leaf production. 22/12 or 17 °C-night indicates warm-day, cool, night cycles, 22/12 or 17 °C-day indicates treatments in which the temperature drop occurred at dawn, then remained cool for the duration of the experiment, constant indicates temperatures remained constant from seed.

Figure 5.

FT accumulation as influenced through CO and SVP/FLM and leaf tissue production can improve model predictions in changing temperature conditions compared to MPTUs. (A) Comparison of simulated (lines, FM-v1.5 LTP+GE) and observed (symbols) leaf number by week in Col-0 in constant 22 °C conditions (22 °C-constant) and in 22 °C-day, 12 °C-night temperature cycles (22/12 °C-night). (B) Final leaf number of Col-0 at bolt as observed (obs.) and predicted (pred.) by incorporating temperature influence on FT though leaf tissue production (LTP) and FT gene expression (GE) (FM-v1.5 LTP+GE), leaf tissue production only (FM-v1.5 LTP), and through traditional MPTUs (FM-v1.0). (C, D) The difference between predicted and observed days to bolt in Col-0 and Landsberg erecta (Ler) using FM-v1.5 LTP+GE (C) and MPTUs in FM-v1.0 (D). (E) Observed and predicted final leaf number and (F) the difference between predicted and observed results using MPTUs in FM-v1.0, adjusted so that daytime and nighttime temperatures are given equal weight. (B–F) Plotted over three nighttime temperatures. Daytime temperature was 22 °C. (C, D, F) Horizontal line at zero is the position in which there is no difference between predicted and observed results. Error bars = 1 SD. If error bars are not visible, the SD is smaller than the height of the symbol. Observed leaf number and days to bolt pooled and compiled from Kinmonth-Schultz et al. (2016) (license to reproduce data obtained from John Wiley and Sons, #4601440806514).

Figure 6.

FT fails to accumulate to a threshold in some cool-temperature conditions. Plants grown at constant cool (12 °C) temperatures from seed (constant) or after 1 week at 22 °C (22/12 °C-day) do not accumulate FT to a threshold set using 22 °C constant temperatures in long days (thick black line). Altering the threshold to decline with developmental time (thick grey line) improves the predictive capacity of FM-v1.5, as we propose in the discussion.

Figure 7.

Growth is slowed and flowering is delayed in plants exposed to 12 °C for 2, 4 or 6 days, then returned to warm temperatures (24 °C), relative to control plants grown continuously in warm temperatures. (A) Average leaf number of plants recorded at dawn after 2, 4 or 6 days in 24 °C (control) or 12 °C temperature conditions. (B) Relative seedling sizes on dawn of Day 7, after completion of all cool-temperature treatments (scale bars = 1 cm, 0 = control). Individual images cropped from the same photograph and scaled together (see original image, Supporting Information—Fig. S11). (C) Relative flowering progression 3 days after appearance of last floral stem (bolt) in plants exposed to 12 °C for 2, 4 or 6 days relative to 24 °C control (0, scale bar = 5 cm).