Effect of regulated overexpression of the MADS domain factor AGL15 on flower senescence and fruit maturation

Abstract

We have examined the effect of regulated overexpression of AGL15, a member of the MADS domain family of regulatory factors, on reproductive tissues. Using molecular and physiological markers, we show that constitutive overexpression of AGL15 in Arabidopsis leads to delay and down-regulation of senescence programs in perianth organs and developing fruits and alters the process of seed desiccation. Through genetic crosses, we show that the rate of water loss in the maturing seeds is dictated by the genetic composition and physiological state of the maternal tissue, rather than the embryo. To define the developmental time and/or place when senescence programs are most affected by elevated AGL15 levels, we expressed AGL15 under the control of various promoters. Expression during senescence or in abscission zone cells did not produce delays in floral organ senescence or abscission. Using a glucocorticoid-inducible expression system, we show that an increase in AGL15 levels around the time of flower opening is necessary to delay senescence and increase floral organ longevity.

Figure 1

GUS activity in flowers of transgenic plants carrying SAG12:GUS, a senescence reporter construct. A and B, GUS activity at 3 (A) and 4 (B) DAP in flowers of plants that do not overexpress AGL15. Perianth organs abscise around 4 DAP in these plants. C and D, Little or no GUS activity can be detected at 3 (C) or 4 (D) DAP in the flowers of plants that carry the 35S:gAGL15 transgene and constitutively overexpress AGL15. Perianth organs are retained for long periods in these plants.

Figure 2

Analysis of fruit maturation in plants that constitutively overexpress AGL15. A, Chlorophyll content of maturing siliques of wild-type (wt) plants versus plants that constitutively overexpress AGL15 (35S:AGL15). Each data point represents the mean of three experimental replicas. Bars indicate the sd of the mean. B through D, GUS activity in the silique tissues of plants carrying the senescence reporter SAG12:GUS. B, Quantitative assays. Siliques were collected at 17 DAP from plants that do not overexpress AGL15 (gray bar) and at 17, 23, or 25 DAP from plants that overexpress AGL15 (black bars). C and D, Histochemical staining. The siliques on the left in each image were collected at 17 DAP from plants that do not overexpress AGL15. The siliques on the right in each image were collected at 17 (C) or 23 (D) DAP from plants that overexpress AGL15. SAG12 promoter activity was down-regulated and delayed in plants that overexpress AGL15.

Figure 3

Analysis of seed desiccation in plants that overexpress AGL15. Seeds were excised from maturing siliques at various times (DAP) and seed water content was calculated from measurements of the fresh and dry weight of each sample. A, The water content of seeds of self-pollinated wild-type plants (black line) declines more rapidly than the water content of seeds of self-pollinated transgenic plants that overexpress AGL15. The transgenic plants carried one copy of either of two transgenes (35S:AGL15, blue line; or 35S:gAGL15, red line) in the hemizygous condition. B, The water content of seeds of self-pollinated wild-type plants (black line) was compared with the water content of seeds derived from reciprocal crosses between wild-type plants and plants that overexpress AGL15. The green line shows the water content of seeds derived from crosses where wild-type plants were pollinated with transgenic pollen. The pink line shows the water content of seeds derived from crosses where plants that overexpress AGL15 were pollinated with wild-type pollen.

Figure 4

Analysis of perianth organ senescence and abscission in transgenic plants that overexpress AGL15 in abscission zones. A through D, Immunolocalization of AGL15 in perianth organ abscission zones. No AGL15 accumulation can be detected in cells in the abscission zones in wild-type plants (A and C). AGL15 accumulates at elevated levels in the nuclei of cells in the abscission zones in transgenic plants carrying the Chi:gAGL15 construct (B and D). The location of the abscission zones shown at higher magnification in C and D are indicated by the boxes in A and B. az, Abscission zone; c, carpel; p, petal; r, receptacle; s, sepal. Bars = 20 μm. E through G, Inflorescences of wild-type (E) and transgenic plants carrying either the Chi:gAGL15 transgene (F) or 35S:AGL15 transgene (G). Overexpression of AGL15 in the abscission zones does not produce the visible delay in perianth organ abscission and senescence associated with constitutive overexpression. H, Analysis of the force needed to remove petals in progressively older flowers (in different positions relative to the youngest open flower) in wild-type plants and two lines of transgenic plants carrying the Chi:gAGL15 transgene. Each data point represents the mean of eight to 12 measurements. Bars indicate the sd.

Figure 5

Analysis of perianth organ senescence and abscission in wild-type plants and transgenic plants that overexpress AGL15 during senescence. A, Gel-blot analysis of total RNA samples (5 μg lane⁻¹) isolated from leaves of wild-type or transgenic plants and hybridized with probes specific for AGL15 or 18S rRNA sequences. AGL15 transcripts accumulate in a senescence-dependent manner in transgenic plants carrying SAG12:gAGL15 constructs. AGL15 transcripts can be detected in non-senescing, green leaves (GL) of plants that constitutively overexpress AGL15 (35S:AGL15), but cannot be detected in green leaves of either wild-type plants (wt) or transgenic plants that carry SAG12:AGL15 constructs (SAG12:AGL15). AGL15 transcripts accumulate at higher levels in the yellow, senescing leaves (YL) of plants carrying SAG12:gAGL15 constructs. Hybridization with 18S rRNA probes indicates equal loading of samples in each lane. B and C. The inflorescences of wild-type plants (B) and plants that carry SAG12:gAGL15 transgenes (C) are indistinguishable in terms of floral organ senescence and abscission.

Figure 6

Effects of glucocorticoid-induced overexpression of AGL15. A, Gel-blot analysis of total RNA samples (5 μg lane⁻¹) isolated from plants in which AGL15 overexpression can be induced by treatment with dexamethasone (+DEX) or controls, hybridized with probes specific for AGL15 or 18S rRNA sequences. Plants carrying the induction system were either not treated (AGL15 [−DEX]) or were treated after bolting (AGL15 [+DEX] ab) or continuously (AGL15 [+DEX] c) with dexamethasone. AGL15 mRNAs were particularly abundant in lines that showed strong phenotypic changes after treatment ([+DEX] st). AGL15 transcripts were less abundant in plants that constitutively overexpress AGL15 (35S:AGL15) and were undetectable in wild-type plants (wt). Hybridization with 18S rRNA probes indicates equal loading of samples in each lane. B and C, Inflorescence (B) and individual flower (C) after induction of AGL15 overexpression by spraying with dexamethasone. D, Inflorescence of a similar plant sprayed with control solution lacking dexamethasone.