Seasonal Regulation of Petal Number

Abstract

Four petals characterize the flowers of most species in the Brassicaceae family, and this phenotype is generally robust to genetic and environmental variation. A variable petal number distinguishes the flowers of Cardamine hirsuta from those of its close relative Arabidopsis (Arabidopsis thaliana), and allelic variation at many loci contribute to this trait. However, it is less clear whether C. hirsuta petal number varies in response to seasonal changes in environment. To address this question, we assessed whether petal number responds to a suite of environmental and endogenous cues that regulate flowering time in C. hirsuta We found that petal number showed seasonal variation in C. hirsuta, such that spring flowering plants developed more petals than those flowering in summer. Conditions associated with spring flowering, including cool ambient temperature, short photoperiod, and vernalization, all increased petal number in C. hirsuta Cool temperature caused the strongest increase in petal number and lengthened the time interval over which floral meristems matured. We performed live imaging of early flower development and showed that floral buds developed more slowly at 15°C versus 20°C. This extended phase of floral meristem formation, coupled with slower growth of sepals at 15°C, produced larger intersepal regions with more space available for petal initiation. In summary, the growth and maturation of floral buds is associated with variable petal number in C. hirsuta and responds to seasonal changes in ambient temperature.

Figure 1.

Petal number varies according to season and age in C. hirsuta. A, Petal number varies between zero (0 p) and four (4 p) in the flowers of a single inflorescence in C. hirsuta. B, Box plot of floral organ counts per flower in C. hirsuta. C, Longitudinal section through a stage 3 floral bud shows growth of the abaxial (ab) sepal is advanced compared to the adaxial (ad) sepal. D, Scanning electron micrograph of a stage 4 floral bud shows outgrowth of all four sepals and early petal and stamen primordia. E and F, Longitudinal section (E) and scanning electron micrograph (F) of stage 5 floral buds show sepals growing over the floral meristem, prominent stamen primordia, and small petal primordia. G and H, Longitudinal section (G) and scanning electron micrograph (H) of stage 6 flowers show sepals enclosing the bud with their tips folding within the bud. Section shows developing stamen and carpel primordia but no petal primordia. I and J, Longitudinal section (I) and scanning electron micrograph (J) of older floral buds show that petal primordia are either present or absent (arrow, J) in whorl two. K, Box plot of petals per flower in spring and summer flowering populations of C. hirsuta, n = 419 flowers. L, Box plot of petals per flower in the C. hirsuta Ox accession grown in late spring and summer field conditions, n = 468 flowers. M, Average petal number per flower at consecutive floral nodes in C. hirsuta Ox plants grown in summer (white circles) and late spring (black circles) field conditions, n = 10 plants in each condition, error bars show se. The first floral node is labeled as 1. Pairwise comparisons by Mann-Whitney U test. Significance levels: ***P < 0.001, **P < 0.01, *P < 0.05. Box plots show 25th to 75th percentiles; whiskers extend down to 10th and up to 90th percentiles; black line shows median; red line shows mean. Numbers in C to H indicate floral stage; st, stamen; p, petal; c, carpel. Scale bars = 20 µm (C–J).

Figure 2.

C. hirsuta flowering is accelerated by long days, vernalization, cool ambient temperature, and simulated shade. A, C. hirsuta vegetative rosette. B to D, Scanning electron micrographs of C. hirsuta plants following SD to LD transfer. B, shoot apical meristem (SAM) initiating leaf primordia (L) 4 d after transfer; C, inflorescence meristem (IM) initiating floral meristem primordia (FM) 5 d after transfer with the oldest primordium at stage 3; D, IM initiating FM 6 d after transfer with the oldest primordium at stage 5. E, C. hirsuta flowering raceme. F, Line graph shows cumulative floral bud production starting 5 d after SD to LD transfer. G, Relative expression levels of ChAP1 and ChSVP mRNA in vegetative SAM, early IM (5–10 floral buds), and IM at anthesis, shown as fold change of vegetative levels. Expression is compared pairwise to vegetative samples. H, Flowering time, measured as number of rosette leaves and number of days after germination when the first floral buds are visible, and floral maturation interval, measured as the number of days between visible floral buds and anthesis, were compared between different photoperiods, ambient temperatures, and light qualities. ND, Neutral days (12:12); SD, short days (8:16); LD, long days (16:8); WL, white light; BL, blue light; FRL, far-red light. I, Flowering time, measured as number of rosette leaves (left y axis) and visible floral buds (right y axis), in response to vernalization at ∼4°C. Pairwise comparisons by Mann-Whitney U test. Significance levels: ***P < 0.001, **P < 0.01, *P < 0.05. n = 26–28 (ND, 20°C, WL); n = 7–26 (LD, 20°C, WL); n = 26–29 (SD, 20°C, WL); n = 12–31 (ND, 15°C, WL); n = 29–31 (ND, 20°C, FRL); n = 22–31 (ND, 20°C, BL); n = 12 (vernalization). Error bars show se. Scale bars = 1 cm (A and E) and 100 µm (B–D).

Figure 3.

Petal number in C. hirsuta responds to photoperiod, ambient temperature, light quality, and vernalization. A and B, Box plots compare petal number per flower between different photoperiods, ambient temperature, and light quality (A), and in response to vernalization at ∼4°C (B). C and D, Average petal number per flower over successive floral nodes in response to ambient temperature (C) and to photoperiod (D); the first floral node is labeled as 1; error bars show se. Petals/flower are scored in the first 25 floral nodes in A and B. Pairwise comparisons by Mann-Whitney U test. Significance levels: ***P < 0.001, **P < 0.01, *P < 0.05. ND, Neutral days (12:12); SD, short days (8:16); LD, long days (16:8); WL, white light; BL, blue light; FRL, far-red light. n = 26–28 (ND, 20°C, WL); n = 7–26 (LD, 20°C, WL); n = 26–29 (SD, 20°C, WL); n = 12–31 (ND, 15°C, WL); n = 29–31 (ND, 20°C, FRL); n = 22–31 (ND, 20°C, BL); n = 12 (vernalization). Box plots show 25th to 75th percentiles, whiskers extend down to 10th and up to 90th percentiles, black line shows median, and red line shows mean.

Figure 4.

Ambient temperature affects the size of intersepal regions in C. hirsuta floral buds. A, Early floral development in C. hirsuta was staged by the size and shape of sepals according to Supplemental Table S2. M, Floral meristem; LS lateral sepals; AbS, abaxial sepal; AdS, adaxial sepal. All subsequent floral buds are oriented this way. Green dashed line indicates position of optical sections shown in D. B and C, Local curvature computed with MorphoGraphX software in stage 4 (B) and stage 5 (C) floral buds of C. hirsuta at 20°C (top) and 15°C (bottom). Heat map indicates tissue curvature in 10⁻² μm⁻¹, where flat is white, negative curvature is blue, and positive curvature is red. Arrows indicate flat regions between sepals. Dashed blue lines approximate the sepal-meristem boundaries when these are obstructed by sepal growth in C. D, Optical cross sections of stage 4 floral buds of C. hirsuta at 20°C (top) and 15°C (bottom). Two measures of floral meristem size are indicated: the direct distance (yellow arrow) and the curvilinear distance (red arrow) between lateral sepal boundaries. E, Bar plot of curvilinear distance across the floral meristem (white) and epidermal cell area in the floral meristem (gray) of stage 4 floral buds of C. hirsuta at 20°C and 15°C. Curvilinear distance is significantly different between temperatures (P = 0.02; Student’s t test). F, Areal growth computed with MorphoGraphX software in time-lapse series of C. hirsuta floral buds at 20°C (top) and 15°C (bottom). Arrows indicate regions of slower growth between sepals. Heat map indicates cell areal growth in percentage over intervals of 24 h (20°C, top) or 36 h (15°C, bottom). Stage 3 is marked in yellow to highlight the considerable delay in development through this stage at 15°C. Note the 15°C series is composed of two different time-lapse series (Supplemental Fig. S11), and the heat map is saturated for high growth in the sepals in order to discriminate between growth in the floral meristem and intersepal regions. Scale bars = 20 μm.

Figure 5.

Geometry of the intersepal regions in Arabidopsis floral buds differs from C. hirsuta. A, Early floral development in Arabidopsis was staged by the size and shape of sepals according to Smyth et al. (1990). M, Floral meristem; LS, lateral sepals; AbS, abaxial sepal; AdS, adaxial sepal. All subsequent floral buds are oriented this way. Green dashed line indicates position of optical section shown in D. B and C, Local curvature computed with MorphoGraphX software in stage 4 (B) and stage 5 (C) floral buds of Arabidopsis at 20°C. Heat map indicates tissue curvature in 10⁻² μm⁻¹, where flat is white, negative curvature is blue, and positive curvature is red. Arrows indicate intersepal regions. Dashed blue lines approximate the sepal-meristem boundaries when these are obstructed by sepal growth in C. D, Optical cross section of stage 4 floral bud showing two measures of floral meristem size: the direct distance (yellow arrow) and the curvilinear distance (red arrow) between lateral sepal boundaries. E, Areal growth computed with MorphoGraphX software in time-lapse series of Arabidopsis floral buds at 20°C. Arrow indicates intersepal region with slower growth. Heat map indicates cell areal growth in percentage over intervals of 24 h. Note the heat map is saturated for high growth in the sepals in order to discriminate between growth in the floral meristem and intersepal regions. Scale bars = 20 μm.

Figure 6.

Variation in petal number is correlated with sepal trichomes in C. hirsuta. A, Average petal number (black circles) and sepal trichome number (white circles) over consecutive floral nodes are correlated (R² = 0.72); the first floral node is labeled as 1. B, Box plot compares sepal trichome number per flower between different photoperiods, ambient temperature, and light quality. C and D, Average sepal trichome number per flower over successive floral nodes in response to ambient temperature (C) and to photoperiod (D). Sepal trichomes/flower are averaged over the first 25 floral nodes in B. Pairwise comparisons by Mann-Whitney U test. Significance levels: ***P < 0.001, **P < 0.01, *P < 0.05. ND, Neutral days (12:12); SD, short days (8:16); LD, long days (16:8); WL, white light; BL, blue light; FRL, far-red light. n = 26–28 (ND, 20°C, WL); n = 7–26 (LD, 20°C, WL); n = 26–29 (SD, 20°C, WL); n = 12–31 (ND, 15°C, WL); n = 29–31 (ND, 20°C, FRL); n = 22–31 (ND, 20°C, BL). Error bars show se. Box plots show 25th to 75th percentiles, whiskers extend down to 10th and up to 90th percentiles, black line shows median, and red line shows mean.

Figure 7.

Petal number variation is insensitive to accelerated ageing in C. hirsuta. A, Relative expression of ChSPL3, ChSPL9, and ChSPL15 transcripts in vegetative SAM (black), early IM with 5 to 10 floral buds (light gray), and IM at anthesis (dark gray). Pairwise differences in expression are significant for ChSPL3 between early IM and anthesis, and for ChSPL9/15 between vegetative SAM and anthesis. B and C, Flowering time, measured by rosette leaf number (B) and time to anthesis (C) in transgenic lines expressing miRNA-resistant versions of SPL9 and SPL10. D, Average petal number per flower (gray) and sepal trichomes per flower (white) in pSPL9:rSPL9 and pSPL10:rSPL10 lines. E, Average sepal trichome number per flower is lower in all floral nodes in pSPL10:rSPL10 lines. F, Average petal number per flower is similar at successive floral nodes in wild-type, pSPL9::rSPL9, and pSPL10:rSPL10 plants. Pairwise comparisons by Mann-Whitney U test. Significance levels: ***P < 0.001, **P < 0.01, *P < 0.05. Petals and sepal trichomes/flower are averages of the first 25 floral nodes. Error bars show se.

Figure 8.

Petal number in C. hirsuta responds to elevated GA during long-day photoperiods. A, Flowering time, measured by rosette (dark gray), and cauline (light gray) leaf number, and number of days after germination when the first floral buds are visible, responds to increasing GA₃ concentrations (0, 0.1, and 10 µm) in ND. B, Average number of petals per flower (dark gray) responds to increasing GA₃ concentrations (0, 0.1, and 10 µm) in LD but not ND. Sepal trichome number (light gray) was decreased by10 µm GA₃ treatment in LD and ND. Multiple comparisons by Kruskal-Wallis test with post hoc analysis using the Tukey Kramer method. Significance levels: ***P < 0.001, **P < 0.01, *P < 0.05. ND, neutral days (12:12); LD, long days (16:8). Petals and sepal trichomes/flower are averages of the first 25 floral nodes. Error bars show se.