Azacytidine arrests ripening in cultivated strawberry (Fragaria × ananassa) by repressing key genes and altering hormone contents

Abstract

Background: Strawberry ripening involves a number of irreversible biochemical reactions that cause sensory changes through accumulation of sugars, acids and other compounds responsible for fruit color and flavor. The process, which is strongly dependent on methylation marks in other fruits such as tomatoes and oranges, is highly controlled and coordinated in strawberry.

Results: Repeated injections of the hypomethylating compound 5-azacytidine (AZA) into green and unripe Fragaria × ananassa receptacles fully arrested the ripening of the fruit. The process, however, was reversible since treated fruit parts reached full maturity within a few days after AZA treatment was stopped. Transcriptomic analyses showed that key genes responsible for the biosynthesis of anthocyanins, phenylpropanoids, and hormones such as abscisic acid (ABA) were affected by the AZA treatment. In fact, AZA downregulated genes associated with ABA biosynthetic genes but upregulated genes associated with its degradation. AZA treatment additionally downregulated a number of essential transcription factors associated with the regulation and control of ripening. Metabolic analyses revealed a marked imbalance in hormone levels, with treated parts accumulating auxins, gibberellins and ABA degradation products, as well as metabolites associated with unripe fruits.

Conclusions: AZA completely halted strawberry ripening by altering the hormone balance, and the expression of genes involves in hormone biosynthesis and degradation processes. These results contradict those previously obtained in other climacteric and fleshly fruits, where AZA led to premature ripening. In any case, our results suggests that the strawberry ripening process is governed by methylation marks.

Keywords: Abscisic acid; Azacytidine; Demethylation; Ripening; m5-cytosine.

Fig. 1

A Halves of AZA treated strawberry fruit receptacles. B Treated fruits after the AZA treatment was stopped. Fruits were initially collected at the green G3 stage (fully developed green receptacles) and kept with pedicels immersed in a Murashige and Skoog media solution. Then 0.5 mL of a 1 mM solution of AZA was injected into fruit halves every other day. The treated parts remained green or white. The other halves were injected with water and used as controls. They ripened at a regular pace and developed a normal red color

Fig. 2

Relative hormone abundance at different stages of strawberry receptacle development and ripening

Fig. 3

ABA biosynthetic and degradation pathways. The enzymes were down regulated and upregulated by the AZA treatment are shown in blue and red, respectively. Boxplots show the levels of metabolites in control and treated receptacles, the y-axis representing the means ± SE of normalized area for each compound as determined for internal standard isovitexin in LC–MS analysis. Three biological replicates were used. ZEP zeaxanthin epoxidase; NSY neoxanthin synthase; NCED 9-cis-epoxycarotenoid dioxygenase; SDR Short chain dehydrogenase reductase; AAO Abscisic aldehyde oxidase. Significant differences as determined with Student’s t-test analysis (**) p < 0.05

Fig. 4

Scheme of the phenylpropanoid, flavonoid and anthocyanin pathways. The enzymes down-regulated by the AZA treatment are shown in blue. Boxplots show the levels of metabolites in control and treated receptacles, with the y-axis representing the normalized area of each compound normalized to internal standard isovitexin and fresh weight in LC–MS analysis. Three biological replicates were used. PAL phenylammonia lyase; 4CL 4-coumaryl-CoA ligase; ANS anthocyanidin synthase; C4H cinnamic acid 4-hydroxylase; CHI chalcone isomerase; CHS chalcone synthase; DFR dihydroflavonol reductase; F3H flavanone 3-hydroxylase; F3’H flavonoid3’-hydroxylase; GT1 anthocyanin 3-glucosyltransferase; MT malonyl transferase; UF3GT UDP flavonol 3–glucosyltransferase. Significant differences as determined with Student’s t-test (**) p < 0.05