A leu-rich repeat receptor-like protein kinase, FaRIPK1, interacts with the ABA receptor, FaABAR, to regulate fruit ripening in strawberry.

Strawberry (Fragaria×ananassa) is a model plant for studying non-climacteric fruit ripening regulated by abscisic acid (ABA); however, its exact molecular mechanisms are yet not fully understood. In this study, a predicted leu-rich repeat (LRR) receptor-like kinase in strawberry, red-initial protein kinase 1 (FaRIPK1), was screened and, using a yeast two-hybrid assay, was shown to interact with a putative ABA receptor, FaABAR. This association was confirmed by bimolecular fluorescence complementation and co-immunoprecipitation assays, and shown to occur in the nucleus. Expression analysis by real-time PCR showed that FaRIPK1 is expressed in roots, stems, leaves, flowers, and fruit, with a particularly high expression in white fruit at the onset of coloration. Down-regulation of FaRIPK1 expression in strawberry fruit, using Tobacco rattle virus-induced gene silencing, inhibited ripening, as evidenced by suppression of ripening-related physiological changes and reduced expression of several genes involved in softening, sugar content, pigmentation, and ABA biosynthesis and signaling. The yeast-expressed LRR and STK (serine/threonine protein kinase) domains of FaRIPK1 bound ABA and showed kinase activity, respectively. A fruit disc-incubation test revealed that FaRIPK1 expression was induced by ABA and ethylene. The synergistic action of FaRIPK1 with FaABAR in regulation of strawberry fruit ripening is discussed.

Introduction

The processes of ripening in fleshy fruit involve complex changes, including changes in sugar content, texture, color, flavor and aroma, many of which have been shown to be controlled by hormones, notably by ethylene in climacteric fruit and by abscisic acid (ABA) in non-climacteric fruit (Cherian, ; Kumar ; Shen and Rose, 2014). The ripening of strawberry (Fragaria×ananassa), which is a model plant for studying non-climacteric fruit ripening, is not only controlled by ABA (Chai ; Jia , 2013; Li , 2013; Daminato ; Han ) but is also known to involve several other hormones, including ethylene (Merchante ), jasmonate (JA; Concha ), brassinosteroids (BR; Chai ), and gibberellic acid (Csukasi ). Although much progress has been made in understanding strawberry fruit ripening, many aspects of the underlying molecular mechanisms are not well understood. Here, we focus on characterization of ABA-associated signaling in the regulation of ripening.

Several kinase receptors have been identified in plants, such as histidine kinase (HK)-type receptors including ethylene receptors (Chang ; Schaller and Bleecker, 1995) and cytokinin receptors (Hwang and Sheen, 2001; Inoue ; Yamada ), the brassinosteroid (BR) receptor BRI1 (He ; Wang ), and the candidate ABA receptor, RPK1 (a leucine-rich repeat receptor-like kinase; Osakabe ). It has previously been reported that the leu-rich repeat receptor-like protein kinases (LRR-RLKs) are localized to the plasma membrane, and contain an extracellular LRR and an intracellular Ser/Thr kinase (STK) domain (Friedrichsen ; Shiu and Bleecker, 2001; Tichtinsky ; Napier, 2004). They are involved in many signaling processes, such as CLV1-mediated apical meristem development (Clark ; Fletcher ; Rojo ), FLS2-mediated flagellin perception and signaling (Gómez-Gómez ), systemin and phytosulfokine receptors (Matsubayashi ; Scheer and Ryan, 2002), SRK-mediated self-incompatibility (Takasaki ), and BR signaling via the BRI1 receptor (Wang ).

Among the LRR-RLKs, BR signaling is particularly well characterized (Li and Chory, 1997; He ; Tichtinsky ; Kinoshita ; Kim and Wang, 2010; Clouse, 2011; Jiang ). The binding of BR to BRI1 at the cell surface activates BRI1 and its co-receptor the BRI1-associated receptor kinase 1 (BAK1) complex via a set of phosphorylation reactions, ultimately regulating the expression of BR-responsive genes (Tichtinsky ; Kim and Wang, 2010; Clouse, 2011; Jiang ). It has also been shown that a 70-amino-acid segment between LRR21 and LRR22 together with LRR22 are required for the binding (Li and Chory, 1997; He ; Kinoshita ). It is of note that, although RPK1, as a LRR-RLK, was first isolated from Arabidopsis thaliana 20 years ago (Hong ) and a subsequent study suggested that RPK1 may function in early ABA signaling possibly through LRR perception (Osakabe ), it is yet not clear that whether RPK1 proteins bind to ABA.

Much progress has been made in the identification of ABA receptors (Guo ), and several Arabidopsis receptors have been identified with different subcellular localizations, including the plasma membrane (GPCR-type G proteins, GTGs), the cytosol (START-domain PYR/PYL/RCAR), and the chloroplast (ABAR/CHLH, putative ABA receptor/Mg-chelatase H subunit) (Shen ; Fujii ; Ma ; Melcher ; Miyazono ; Nishimura ; Pandey ; Santiago ; Cutler ; Umezawa ). In addition, two core ABA signaling pathways have been proposed: the ‘ABA-PYR/PYL/RCAR-PP2C-SnRK2’ pathway (Fujii ), and the ‘ABA-ABAR-WRKY40-ABI5’ pathway (Shang ). In the first pathway, ABA promotes the interaction of PYR1 and PP2C, resulting in inhibition of PP2C, activation of SnRK2, and promotion of the expression of downstream factors, such as AREB/ABF, ion channels, and NADPH oxidases (Fujii ; Ma ; Park ; Santiago ; Cutler ; Umezawa ).

The second pathway is triggered by the putative ABA receptor ABAR, which is also named the magnesium-chelatase subunit H protein (CHLH; Shen ). The Arabidopsis ABAR/CHLH specifically binds ABA via the C-terminal domain (Wu ), and ABA binds to CHLH but not to the other Mg-chelatase components/subunits, including CHLI (I subunit), CHLD (D subunit), and GUN4 (Du ). In the ABA-ABAR-WRKY40-ABI5 core signaling complex, WRKY40 acts to suppress ABA signaling, and high levels of ABA promote ABAR–WRKY40 interactions, then activate expression of ABA-responsive genes, including the ABA-responsive transcription factors ABI5 and ABI4 (Shang ; Adhikari ; Liu ; Yan ). A chloroplast co-chaperonin 20 (CPN20) has been identified to interact with ABAR, and negatively regulates ABA signaling at upstream of the WRKY40 transcription repressor (Zhang , 2014). A pentatricopeptide repeat (PPR) protein, SOAR1 (a suppressor of ABAR-overexpression 1), relays the core ABA signaling downstream of ABAR and upstream of ABI5, and mediates crosstalk between the PYR/PYL/RCAR- and ABAR-mediated signaling pathways (Mei ). Finally, SnRK2.6/OST1 has also been shown to be an interaction partner between ABAR and PYR/PYL/RCAR in Arabidopsis guard cells (Liang ). Taken together, these reports suggest that ABAR mediates a central ABA signaling pathway in Arabidopsis.

In addition to Arabidopsis (Shen ; Legnaioli ; Tsuzuki , 2013; Ibata ), the involvement of ABAR in ABA signaling has also been found in leaf guard cells of tobacco (Nicotiana benthamiana) and peach (Prunus persica) (Jia ; Du ), and ABAR promotes ripening in strawberry fruit in response to ABA (Jia ; Li ; Kadomura-Ishikawa ). Using Tobacco rattle virus-induced gene silencing (VIGS) in strawberry fruit, down-regulation of FaABAR expression was shown to inhibit ripening, and the white phenotype of these RNAi fruits could not be rescued by addition of exogenous ABA, demonstrating that ABAR positively regulates ripening under normal conditions (Jia ). Importantly, Kadomura-Ishikawa found that FaMYB10 plays a role as a signal transduction mediator from ABAR perception of ABA to anthocyanin synthesis during strawberry fruit ripening, and it was recently reported that FaABAR can bind to ABA (Zhang ).

However, despite such insights having been made, the downstream component relaying the FaABAR signaling in ripening is not known. Here, we describe the identification of a strawberry LRR-RLK gene, FaRIPK1, through the use of a yeast two-hybrid assay with FaABAR as bait. We present the results of bioinformatic, physiological, biochemical, and molecular analyses that collectively suggest that FaRIPK1 is localized to the plasma membrane and nucleus, and can bind to ABA to regulate strawberry fruit ripening.

Materials and methods

Plant material

Octaploid strawberry (Fragaria×ananassa ‘Beinongxiang’) plants were grown in a greenhouse (15–25 °C, relative humidity 60–90%, and 16/8 h light/dark cycles) during the spring seasons in 2015 and 2016. After anthesis, 200 small green fruit from 40 plants were tagged. Fruit were collected at seven stages: small green (SG), large green (LG), de-greening (DG), white (Wt), initial red (IR), partial red (PR), and full red (FR),. Six uniformly sized fruit per plant from three replicate plants were sampled at each stage and frozen rapidly in liquid nitrogen, and were then stored at –80 °C until further use.

RNA isolation, cDNA library construction, and yeast two-hybrid screening

Three white fruit were randomly selected from the frozen samples for RNA isolation and cDNA synthesis. After removal of achenes, the receptacles (pulps) were ground into a powder in liquid N2. Total RNA was extracted from 5 g of powder using an EASYspin RNA extraction kit (Biomed, China) according to the manufacturer’s instructions. Genomic DNA was removed by a 15-min incubation at 37 °C with RNase-Free DNase (TaKaRa, China) followed by an RNA Clean Purification Kit (BioTeke, China). High-quality total RNA was assessed by an OD260/OD280 of 2.0 and an OD260/OD230 of 2.0 in a spectrophotometer, and a 2:1 ratio of 28S to 18S rRNA bands using gel electrophoresis.

Total RNA (3 μg) from white fruit was reverse-transcribed to synthesize the first-strand cDNA using the SMARTTM cDNA Library Construct Kit (TaKaRa). The double-stranded cDNA (dscDNA) was then synthesized by long-distance (LD)-PCR with the 5′ PCR primer (12 μM, 5′-AAGCAGTGG TATCAACGCAGA GT-3′). The amplified dscDNA was digested with proteinase K and SfiI, and then purified using CHROMA SPIN-400 columns (Clontech, USA) to remove fragments ≤200 bp. The large dscDNA was packaged using a phage packaging system, ligated into a blunted Trip Ex2 vector (TaKaRa), transformed into Escherichia XL1-Blue cells, and then plated onto Luria–Bertani (LB) agar plates containing 30 μg ml–1 chloramphenicol. Twenty white clones were randomly selected from the cDNA library and PCR-amplified to assess the recombinant rate and the size of the inserted cDNA fragments. The purified dscDNA and pGADT-Rec plasmids were co-transformed into competent cells of yeast strain Y187 (Pichia pastoris) and cultivated on SD/–Leu medium (TaKaRa).

Interaction screening was performed using a Gal4-based yeast two-hybrid (Y2H) system (Clontech) according to the manufacturer’s instructions. Based on the coding sequence of FaABAR (4146 bp with 1381 amino acids; GenBank No GQ201451), the two halves consisting of N691 (1–2073 bp) and C690 (2074–4146 bp) were separately cloned into pGBKT7 to construct the bait vectors. The primer sequences are listed in Supplementary Table S1 at JXB online. Both vectors were transformed into yeast Y2HGold and grown on both SD/–Trp and SD/–Trp +X-gal plates, and positive blue clones (≥2 mm in diameter) were selected; Y2HGold strains carrying the bait vectors were hybridized with Y187 strains containing the cDNA library, and cultured for 20–24 h on SD/–Trp/–Leu/–His/–Ade plates to screen positive clones for sequencing.

Confirmation of the FaABAR–FaRIPK1 interaction

Y2H assays were performed using the Matchmaker Gal4-based two-hybrid system (Clontech, Palo Alto, CA, USA). The cDNA of FaABARC690 and FaRIPK1 (GenBank no. KM874830) were inserted into the pGADT7 and pGBKT7 vectors, respectively. Primer sequences are listed in Supplementary Table S1. The constructs were transformed into yeast strain AH109 using a lithium acetate method (Shen ). Yeast cells were cultured on –Leu/–Trp medium according to the manufacturer’s instructions. Transformed colonies were plated onto –Leu/–Trp/–His/–Ade medium with or without β-galactosidase to test for possible interactions.

Expression and purification of the RIPK1 and ABARC690 proteins

The coding sequences of strawberry RIPK1 and ABARC690 were amplified by PCR and cloned into pET-28a and pGEX-4T-1 vectors (BioTeke, China) to create a RIPK1-His and ABARC690-GST (glutathione S-transferase) fusion proteins, respectively. Primer sequences are listed in Supplementary Table S1. Both constructs were transformed into E. coli BL21 (DE3) cells. The E. coli strains containing the expression plasmids were grown at 37 °C in 1 l of LB medium until the optical density of the cultures at 600 nm was 0.6–0.8. Protein expression was induced by the addition of 0.5 mM IPTG (isopropy-β-D-thiogalactoside) at 16 °C with shaking at 160 rotations per min. After 16 h, the cells were lysed and proteins were purified with Ni-sepharose and glutathione-sepharose as described in the manufacturer’s protocols (BioTeke, China). The purified proteins were dialysed against PBS (phosphate buffer solution).

GST pull-down assays

For the GST pull-down assays, purified ABARC690-GST (glutathione-sepharose resin; BioTeke, China) and FaRIPK1-His proteins were combined in PBS buffer for 90 min at room temperature (RT) and 20 μl glutathione-sepharose resin was added. The resin and reaction mixture was incubated for 30 min at RT with gentle shaking at 10 min intervals. The resin was washed five times with PBS. After the final wash, the bound protein was eluted and boiled for 5 min with 6× loading buffer, then 10 μl of eluate was analysed using 10% SDS-PAGE and detected with anti-GST and anti-His antibodies (Abmart) as described in the manufacturer’s protocols.

BiFC assays

To determine whether the interaction between strawberry ABARC690 and RIPK1 takes place in planta, a bimolecular fluorescence complementation (BiFC) assay was used. The RIPK1 and ABARC690 cDNAs were separately cloned into pSPYNE and pSPYCE (Walter ), respectively. Primer sequences are listed in Supplementary Table S1. Plasmids containing YFPN-RIPK1 and YFPC-ABARC690 were introduced into Agrobacterium tumefaciens GV3101 and transformed into Nicotiana benthamiana leaves. At 3 d after infiltration, the YFP (yellow fluorescent protein) signal was detected using a Zeiss LSM 710 META confocal microscope (LSM 710 META; Zeiss, Germany).

Real-time PCR

For real-time PCR, the reaction mixtures (20 μl) contained 10 μl of SYBR Premix Ex Taq (TaKaRa), 0.4 μl of 10 μM forward-specific primer, 0.4 μl of 10 μM reverse-specific primer, and 2 μl of 0.1 μM cDNA templates. The mixture was placed in a CFX Sequence Detector (Bio-Rad, Hercules, CA, USA), and DNA amplification was conducted. Actin was used as a reference gene. Relative gene expression was analysed by the 2−ΔΔCT method (Livak and Schmittgen, 2001). All primer sequences are listed in Supplementary Table S1. The experiment was repeated three times.

Construction of VIGS vectors, and agro-infiltration

For virus-induced gene silencing (VIGS), the pTRV1 and pTRV2 (Tobacco rattle virus) vectors (Liu ) were donated by Dr Liu Yu-le (Qinghua University, China). A 680-bp FaRIPK1cDNA fragment was amplified using the primers listed in Supplementary Table S1. The amplified fragment was cloned into a pMD19-T vector (TaKaRa) digested with Xba1 and KpnI, and subsequently cloned into the viral vector pTRV2 that had been cleaved with Xba1 and KpnI. Agrobacterium tumefaciens strain GV3101 containing pTRV1, pTRV2, or pTRV2-FaRIPK1 in a 1:1 ratio was syringe-infiltrated into strawberry fruit as described by Fu . Twenty de-greening fruits attached to plants were used for this treatment, and the experiment was conducted twice. Phenotypes were observed approximately 30 d after anthesis.

Determination of fruit firmness, soluble solids, anthocyanins, and ABA content

Fruit firmness was measured using a fruit hardness tester (FHM-5; Takemura Electric Work Ltd., Japan) after removing the skin. The soluble solids content of the flesh was measured using a handheld sugar measurement instrument (MASTER-100H; ATAGO, Japan). Anthocyanin and ABA contents were measured as described by Jia ). Three uniformly sized fruit were used to determine each parameter, and the experiment was repeated three times.

Subcellular localization

For transient expression of subcellular localization (Subach ), the FaRIPK1 coding sequence was cloned into a pSuper1300-GFP vector at the SpeI and KpnI restriction sites. Then, FaRIPK1–GFP/AtPIP2-mCherry (a marker for an intrinsic plasma membrane) was co-transformed into Arabidopsis mesophyll protoplasts. The fluorescence signals were observed using a confocal laser-scanning microscope (LSM 710 META; Zeiss, Germany) after incubation at 22 °C for 16 h.

In vitro incubation of fruit discs

In vitro treatments were performed on three tissue discs from three fruit using 100 µM ABA and 50 µM ethephon (BioTeke, China), and ABA-free buffer incubation was used as the control. The incubation was done as described by Song . After a 2-h incubation followed by washing with double-distilled water, the discs were frozen in liquid nitrogen and kept at –80 °C until use. The experiment was performed with three replicates.

Expression and purification of LRR283 and STK319 proteins

The sequences encoding the leucine-rich repeated domain (from bp 1 to 283, named LRR283) and the serine/threonine-protein kinase domain (from bp 826 to 1145, named STK319) were amplified, expressed, and purified using the EasySelect™ Pichia Expression Kit and ProBond™ Purification System (Invitrogen, USA). Primer sequences are listed in Supplementary Table S1. The sequences were cloned into the pPICZ B expression vector in-frame with the C-terminal His tag, and transformed into E. coli to select transformants on low-salt LB plates containing 25 μg ml–1 Zeocin™. The purified and linearized recombinant plasmids were transformed into Pichia pastoris to select the Zeocin™-resistant yeast transformants on YPDS (yeast peptone dextrose sorbitol) plates containing the appropriate concentration of Zeocin™. The two fusion proteins were expressed and purified by Ni-NTA His Bind Resins according to the manufacturer’s protocols (Invitrogen, USA).

Isothermal calorimetric analysis

The purified LRR283 was concentrated using an Amicon Ultra-4 centrifugal 3-kDa filter (Millipore, USA) at 3000 g for 20 min in a swing bucket rotor (Sigma, USA). The protein was then desalted for buffer exchange using a HiTrap desalting column (GE Healthcare, USA). The HiTrap desalting column was filled with isothermal titration calorimetry (ITC) buffer (20 mM phosphate buffer, pH 7.4, 150 mM NaCl, 20 mM KCl) to remove ethanol completely and equilibrate the column. The final concentration of the LRR fusion protein was adjusted to 100 µM. The protein samples were titrated using a 2-μl injection with 1 mM (+)ABA every 4 min in a calorimeter (ITC200; Microcal) at 25 °C. ABA was titrated into BSA (bovine serum albumin) protein as a negative control. The experiment was repeated three times. Data fitting was performed using the ORIGIN 7.0 software supplied with the instrument.

In vitro phosphorylation assays

Concentration of SKT319 was performed as described above. The purified proteins were dialysed against kinase buffer (50 mM Tris-HCI, pH 7.6, 10 mM MgCI2, 1 mM MnCl2, 0.2 mM ATP, 1 mM DTT) at 4 °C overnight. According to a previous report on detection of protein kinase activity (Novikova ), kinase reactions using 3 μg purified STK319 were incubated for 30 min at room temperature in 20 μl of kinase buffer containing 3 μg of myelin basic protein (MBP) and 2 μCi of γ-32P-ATP, and the reaction was stopped by adding SDS-PAGE loading buffer, and the phosphorylation of MBP was analysed by autoradiography after 12% SDS-PAGE.

Results

Screening for FaABAR-interacting proteins from strawberry fruit at the white stage

It is thought that ABA perception sites may be tissue-, organ-, or even cell-specific, as well as specific to the developmental stage (Wang and Zhang, 2008). Strawberry FaABAR is an ABA receptor that serves an inducer of fruit ripening (Jia ; Zhang ), and identification of FaABAR-interacting proteins can therefore provide insights into the action of ABA in ripening. We used strawberry fruit at the white stage, which is a distinct stage related to ripening (Wang ), to generate a high-quality cDNA library that could be used as a basis for subsequent Y2H screening. The library which was confirmed by gel electrophoresis (Supplementary Fig. S1A, B) and PCR (Supplementary Fig. S1C).

Given that the C-terminal half of the Arabidopsis ABAR is known to play an important role in ABA binding and interactions with other proteins (Wu ; Shang ), we cloned two separate parts of strawberry FaABAR (GenBank GQ201451) into the pGBKT7 vector, N691 (from bp 1 to 2073) and C690 (from bp 2074 to 4146). The recombinant vectors were transformed into competent yeast Y2HGold cells and screened on SD/–Trp/X-α-GAL and SD/–Trp/AbA/X-α-GAL media for identification of possible self-activation, and these showed that the recombinant pGBKT7 -N691 and pGBKT7-C690 vectors had no auto-activation or toxicity in yeast cells, meaning that ABAR was suitable for Y2H screening. Finally, the Y2H Gold (pGBKT7-N691) and Y2HGold (pGBKT7-C690) cells were hybridized with the Y187 cDNA-library (Supplementary Fig. S1D) and cultivated separately on SD/–Trp/–Leu/–His/–Ade medium. After 16 d, no positive clone was found for the Y2HGold (pGBKT7-N691) hybrid (data not shown); in contrast, 73 positive clones were found for the Y2HGold (pGBKT7-C690) hybrid (Supplementary Fig. S1E), and 28 positive clones were identified by back-crossing (Supplementary Fig. S1F). Based on sequencing of the 28 clones and homolog analysis using BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi), a red-initial protein kinase 1 with a 3465-bp sequence encoding a 1154-amino-acid (aa) protein was found (GenBank KM874830), and was named FaRIPK1.

Confirmation of the FaRIPK1–FaABAR interaction

To confirm the interaction of FaABAR with FaRIPK1, Y2H, GST pull-down, and BiFC assays were carried out. In the Y2H assay, it was found that FaABAR interacted directly with FaRIPK1, since all yeast cells grew better in –Leu/–Trp (–LT) medium, and co-transformation of expected negative pairs (ABARC690-AD and BD, AD and RIPK1-BD, AD and ABARC690-BD, RIPK1-AD and BD, AD and BD) did not result in growth on –Leu/–Trp/–His/–Ade (–LTHA) medium. In contrast, the ABARC690-AD and RIPK1-BD, RIPK1-AD and ABARC690-BD pairs grew better on the –LTHA medium (Fig. 1A).

Fig. 1.

Confirmation of FaABAR interaction with FaRIPK. (A) FaRIPK1 interacts with FaABARC690 in yeast. AD, activation domain; BD, binding domain. (B) A GST (glutathione S-transferase) pull-down assay shows that FaRIPK1 interacts with FaABARC690in vitro. (C) BiFC (bimolecular fluorescence complementation) shows the interaction of FaRIPK1 and FaABARC690in vivo. eYFP, enhanced yellow fluorescent protein. YFPN and YFPC stand for the N- and C-terminus of eYFP, respectively.

To determine whether strawberry ABARC690 interacts with RIPK1 in vitro, ABARC690-GST and FaRIPK1-His proteins were used in a pull-down assay (Supplementary Fig. S2). Western blot analysis showed that ABARC690 interacted with RIPK1 in vitro while the control did not (Fig. 1B). To confirm this, A. tumefaciens carrying strawberry ABARC690-YFPN and RIPK1-YFPC plasmids or RIPK1-YFPN and ABARC690-YFPC plasmids were co-injected into the lower epidermis of tobacco leaves. After incubation for 48 h, a strong YFP fluorescence signal was detected in the nuclei. In contrast, the negative controls ABARC690-YFPN and YFPC, YFPN and RIPK1-YFPC, YFPN and ABARC690-YFPC, RIPK1-YFPN and YFPC, or YFPN and YFPC did not have detectable YFP fluorescence, confirming that ABAR and RIPK1 interacted in the nuclei in vivo (Fig. 1C). Taken together, the results showed that FaABAR interacted with FaRIPK.

Bioinformatics analysis of the FaRIPK1 protein

Bioinformatics analysis of the FaRIPK1 protein revealed several conserved domains, including leucine-rich repeats (LRRs, from aa 81 to 276), SEEEED (serine-rich region of AP3B1, clathrin-adaptor complex, from aa 344 to 463), EDR1 (ethylene-responsive protein kinase Le-CTR1, from aa 544 to 776), and serine/threonine-protein kinase(STK, from aa 837 to 1045; Supplementary Fig. S3), strongly suggesting that FaRIPK1 is a leu-rich repeat receptor-like protein kinase (LRR-RLK). The protein sequence was used for a BLAST search in the NCBI database: the orthologs with FaRIPK1 ≥50% identity were selected (Supplementary Fig. S4), and the orthologs that have previously been characterized in plants were selected to build a phylogenetic tree using the Neighbor-Joining method (Saitou and Nei, 1987). FaRIPK1 and its orthologs were classified into two large groups, with Fragaria separated from the others, and Fragaria ananassa FaRIPK1 sharing 100% identity with the Fragaria vesca LOC101307783 (GenBank, XP_004298363.1) protein sequence (Fig. 2A, black arrow). In the other sub-cluster, FaRIPK1 showed 56% identity with to the protein sequence from Fragaria vesca LOC101301574 (GenBank, XP_004296590.1; Fig. 2A, grey arrow), indicating that there are two contigs (FaRIPK1 and FaRIPK2) in the Fragaria plants. However, in the LOC101301574 contig, we found no SEEEED-conserved domain (Supplementary Fig. S5). These results indicated that FaRIPK1/LOC101307783 is a unique gene in the Fragaria species. This is further suggested by the BLAST search for the FaRIPK1 nucleotide in the NCBI database, where hits with scores ≥200 were only found in Fragaria species (Supplementary Fig. S6, highlighted in red). Further analysis using the ExPASy database (http://www.expasy.org/ proteomics) predicted that FaRIPK1 is an acidic (pI 5.13) and 127-kDa transmembrane protein with at least three transmembrane regions (Fig. 2B).

Fig. 2.

Phylogenetic tree and in silico transmembrane analysis. (A) A phylogenetic tree of FaRIPK1 was analysed using the neighbor-joining method (Saitou and Nei 1987). The numbers at the nodes represent bootstrap values, and the scale bar represents 0.2 substitutions per nucleotide position. The confidence values represent bootstrap values from 1000 replicates. (B) Prediction of the potential transmembrane domains in FaRIPK1. The prediction was performed using the ‘TMpred Prediction of Transmembrane Regions Server’ (http://www.ch.embnet.org/software/TMPRED_form.html). There are three potential transmembrane sites at the amino acid residues 158–161, 632–637, and 1047–1050 under a strict cut-off with a score of 2.2.

FaRIPK1 localization and expression analysis

To elucidate the expression of FaRIPK1, total RNA was extracted from the roots, stems, leaves, flowers, and fruit of the 1-year-old strawberry plants, as well as fruit at seven different stages: small green (SG), large green (LG), de-greening (DG), white (Wt), initial red (IR), partial red (PR), and full red (FR). Real-time PCR showed that FaRIPK1 was expressed throughout the plant, including roots, stems, leaves, flowers, and fruits (IR stage), with the highest expression in roots followed by stems, leaves, flowers, and fruit (Fig. 3A). The expression levels of FaRIPK1 increased slightly from the SG to LG stages, then increased significantly at the DG stage, then increased sharply at the Wt stage, where it reached its maximum level (Fig. 3B). At the red coloration stages, FaRIPK1 transcription levels declined gradually (Fig. 3B). These results suggested that FaRIPK1 is associated with the onset of coloration, and thus it was named a red-initial protein kinase 1 (FaRIPK1; GenBank KM874830).

Fig. 3.

Expression patterns and localization of FaRIPK1. (A) Relative expression levels in roots, stem, leaf, flower, and fruit as determined by RT-PCR. (B) Relative expression levels during different stages of fruit ripening as determined by RT-PCR: SG, small green; LG, large green; DG, de-greening; Wt, white; IR, initial red; PR, partial red; FR, full red. Actin mRNA was used as the internal control. Data are means +SE (n=3). Different letters indicate statistically significant differences at P<0.05 as determined by Duncan’s test. (C) Localization of FaRIPK1 in Arabidopsis thaliana protoplasts. AtPIP2-mCherry is a marker for an intrinsic plasma membrane protein. DAPI (diamidino phenylindole) is a nuclear staining indicator. The GFP (green fluorescent protein), RFP (red fluorescent protein), and DAPI signals were visualized under confocal microscopy after a 16-h incubation period at 22 °C. Scale bars are 20 µm.

To examine the function of the FaRIPK1 protein, a subcellular localization analysis using the GFP–FaRIPK1 fusion protein in Arabidopsis protoplasts was performed, which revealed accumulation in both the plasma membrane and nucleus, as evidenced by co-localization with AtPIP2-mCherry (a marker for an intrinsic plasma membrane) and DAPI (nuclear staining), respectively (Fig. 3C). Taken together, these results suggested that FaRIPK1 is a plasma membrane- and nucleus-localized protein that may be involved in the onset of coloration in strawberry fruit.

Silencing of FaRIPK1 by VIGS affects strawberry fruit development processes

Several studies have demonstrated that TRV-mediated virus-induced gene silencing (VIGS) is a useful tool in studying strawberry fruit ripening (Chai ; Jia ; Li ). Here, we silenced the FaRIPK1 gene in 2-week-old strawberry fruit at the DG stage, using a mixture of Agrobacterium cultures containing pTRV1 and pTRV2 carrying a 560-bp fragment of the FaPIPK1 gene (pTRV2-FaRIPK1), with control fruit being infiltrated with the empty TRV vectors. At 2 weeks after infiltration, the control fruit had all turned fully red (Fig. 4A). In contrast, the VIGS fruits showed various chimeric symptoms with the co-existence of normal red tissue and TRV-infiltrated transgenic uncolored tissues within one fruit (Fig. 4B). The FaRIPK1 transcript levels were down-regulated 60–90% in these areas, as determined by real-time PCR (Fig. 4C).

Fig. 4.

TRV VIGS-induced silencing of FaRIPK1 in developing strawberry fruit. (A) De-greening (DG) fruit attached to the plant were infiltrated with Agrobacterium containing TRV alone (control fruit). (B) TRV carrying a fragment of FaPIPK1 (RNAi fruit). Photographs of the infiltrated fruit were taken 2 weeks after infiltration. (C) RT-PCR analysis of the FaPIPK1 transcript levels in the receptacles in the control and various RNAi fruit (VIGS 1–3). Actin mRNA was used as an internal control. Data are means +SE (n=3). Asterisks indicate statistically significant differences compared with the control at P<0.05 as determined by Duncan’s test.

Silencing FaRIPK1 expression by VIGS affects ripening-related processes and corresponding gene expression

To further understand the role of FaRIPK1 in ripening, we analysed several ripening-related phenomena, including changes in fruit firmness, soluble solid concentrations (representing sugars), anthocyanin content, and ABA levels, in the transgenic fruit, in which FaRIPK1 transcripts were down-regulated by more than 70% (chimeric fruit, n=3) compared to the control (full-red fruit). We also assessed the transcript levels of various genes associated with these processes: PG1 (polygalacturonase) and PL1 (pectate lyase) for fruit firmness; SPS (sucrose phosphate synthase) and SUT1 (sucrose transporter 1) for sugar content; CHS (chalcone synthase) and DFR (dihydroflavonol 4-reductase) for anthocyanin content; ABAR (ABA receptor), ABI1 (ABA insensitive1), ABI3, ABI4, ABI5, and NCED1 (9-cis-epoxycarotenoid dioxygenase) for ABA content and signaling (Chai ; Jia , 2013). Compared to the control, fruit firmness was higher in VIGS fruit (Fig. 5A), soluble solids (Fig. 5B) and anthocyanin (Fig. 5C) contents were lower, and ABA content was higher (Fig. 5D). Real-time analysis showed that, compared to the control, the VIGS fruit had higher levels of ABAR, ABI4, and NCED1 expression, but lower expression of PG1, PL1, SUT1, CHS, DFR, ABI3, and ABI5 (Fig. 5E). The changes in transcript levels associated with genes related to firmness (PG1 and PL1), sugar (SPS, and SUT1), and pigments (CHS and DFR) in VIGS fruit were consistent with FaRIPK1 playing a role as an inducer of ripening.

Fig. 5.

Altering FaRIPK1 expression affects ripening-related parameters and relative levels of gene transcripts. Transgenic fruit in which FaRRP1 was down-regulated by more than 70% and controls (full-red fruit) were used to determine (A) fruit firmness, (B) soluble solid concentrations (representing sugars), (C) anthocyanin content, and (D) abscisic acid (ABA) content. (E) The mRNA expression levels of a set of genes related to firmness (PG1 and PL1), sugar content (SUT1 and SPS), pigmentation (CHS and DFR), ABA biosynthesis (NCED1), and ABA signaling (ABAR, ABI1, ABI3, ABI4, and ABI5) were assessed in RNAi (VIGS) and control fruit. Actin mRNA was used as an internal control. Data are means +SE (n=3). Asterisks indicate statistically significant differences compared with the control at P<0.05 as determined by Duncan’s test. PG1, polygalacturonase 1; PG1, pectate lyase; CHS, chalcone synthase; DFR, dihydroflavonol 4-reductase; SUT1, sucrose transporter 1; SPS, sucrose phosphate synthase; NCED1, 9-cis-epoxycarotenoid dioxygenase; ABAR, ABA receptor; ABI1, ABI3, ABI4 and ABI5, ABA insensitive1, 3, 4, and 5.

Biochemical nature of the FaRIPK1 protein, and the effects of ABA and ethylene on expression of the FaRIPK1 gene

To investigate the potential ABA-binding and kinase activity FaRIPK1, the sequences encoding the leucine-rich domain (from aa 1 to 283, named LRR283) and serine/threonine-protein kinase (from aa 826 to 1145, named STK319) were separately amplified and expressed in Pichia pastoris. After a 24-h induction period, 35-kDa (LRR283) and 39-kDa (STK319) recombinant fusion proteins were purified (Supplementary Fig. S7).

First, an isothermal calorimetric analysis of the purified protein was performed in a calorimeter, in which the protein samples were titrated against ABA. Binding of ABA to LRR283 showed a saturation kinetic curve (Fig. 6A), while the control results obtained from ABA and BSA conjugation did not show a saturation curve (Fig. 6B). The average binding value of LRR283 to ABA showed a dissociation constant (Kd) at 15.1 ± 2.4 µM, with a stoichiometry (N) at a ratio of approximately 1:1.

Fig. 6.

The properties of FaRIPK1, and the effects of ABA and ethylene on FaRIPK1 transcript levels. (A) The binding affinity of (+)ABA for purified LRR283 protein was assessed using isothermal titration calorimetry (ITC) and produced a specific saturation curve. (B) The binding of BSA (bovine serum albumin) to ABA was used as a control in ITC and did not result in a saturation curve. (C) Detection of FaRIPK1 kinase activity. Lane 1 is an incubation of STK319 with myelin basic protein (MBP)-free buffer; Lane 2 is an incubation of STK319 with MBP. Phosphorylated protein was analysed by SDS-PAGE and autoradiography. (D) The effects of 100 μM ABA and 50 µM ethephon on FaRIPK1 transcript levels were assessed using incubation of fruit discs and real-time PCR. Actin mRNA was used as an internal control. Data are means +SE (n=3). Asterisks indicate statistically significant differences compared with the control at P<0.05 as determined by Duncan’s test.

Second, a FaRIPK1 kinase assay was performed in an incubation buffer including STK319, γ-32P-ATP, and MBP, with the control using MBP-free buffer. After SDS-PAGE and autoradiography, a radio-labeled band could be detected when STK319 was incubated with MBP, but not in the control (Fig. 6C), suggesting that FaRIPK1 has kinase activity.

Third, we measured the in vitro effects of ABA and ethylene on FaRIPK1 expression levels in fruit discs. Treatment with 100 µM exogenous ABA and 50 µM exogenous ethephon both promoted expression levels after 2 h incubation compared to the control incubated with buffer only (Fig. 6D), indicating that FaRIPK1 expression is induced by ABA and ethylene.

Discussion

FaRIPK1 is a unique protein present in strawberry fruit

It is thought that ABA perception sites may be tissue-, organ-, or even cell-specific, as well as specific to different developmental stages (Wang and Zhang, 2008), and the multiple ABA receptors that have been identified reflect the breadth and versatility of its physiological functions (Hirayama and Shinozaki, 2007; Guo ). An Arabidopsis ‘ABA-ABAR-WRKY40’ core signaling pathway has been shown to function in seed germination, post-germination growth, and stomatal movement (Shang ; Du ; Liu ; Yan ; Mei ; Liang ), and it is known that strawberry ABAR binds ABA and positively regulates fruit ripening (Jia ; Kadomura-Ishikawa ; Zhang ). Given that the C-terminus of ABAR interacts with WRKY transcription factors in Arabidopsis (Shang ), in our Y2H screen we used the C690-terminal region rather than the N691-terminus and identified the protein interacting with ABAR as FaRIPK1 (Supplementary Fig. S1, Fig. 1). The presence of the four predicted conserved domains (LRRs, SEEEED, EDR1, and STK) suggested that FaRIPK1 is a LRR-RLK (Supplementary Fig. S3). In Fragaria species, the FaRIPK1/LOC101307783 shared 56% identity with Fragaria vesca LOC101301574 (named FaRIPK2, without a SEEEED domain), which exists ubiquitously in planta (Supplementary Figs S4–6). The unique protein, FaRIPK1, was further revealed by phylogenetic analysis (Fig. 2).

FaRIPK1 was found to be localized to both the plasma membrane and nucleus (Fig. 3), and to interact with FaABAR in the nucleus (Fig. 1), suggesting that the FaABAR–FaRIPK1 interaction might play a role in the regulation of gene expression to control strawberry fruit development. Notably, the LRR domain was found to bind ABA, and the STK domain to have mitogen-activated protein kinase activity (Fig. 6), consistent with FaRIPK1 mediating ABA signaling via the LRR domains, and suggesting that it might act as a co-receptor of ABAR. Taken together, the results suggest that that FaRIPK1 is a plasma membrane- and nucleus-localized, ABA-binding, LRR-RLK protein.

FaRIPK1 serves as positive regulator in strawberry fruit ripening

It has previously been reported that FaABAR acts as an inducer of strawberry fruit ripening (Jia ; Kadomura-Ishikawa ). In the present study, although the mRNA expression levels of FaRIPK1 were lower in the receptacle than in roots, stems, leaves, and flowers, the gene was expressed more abundantly at the white stage of fruit ripening (Fig. 3). Notably, a recent report found that a set of metabolic transitions in strawberry fruit occurs during the green-to-white-to-red stages, and the white stage is thought to represent the onset of fruit ripening (Wang ). Accordingly, the higher FaRIPK1 expression level during the white stage suggests that this gene may be involved in ripening.

We used TRV-based VIGS to identify functions of FaRIPK1 in the fruit, and provide evidence to demonstrate that FaRIPK1, similar to FaABAR (Jia ), serves as an inducer of strawberry fruit ripening, as evidenced by a set of physiological and molecular changes, including in the phenotype (inhibiting red coloration), firmness (PG1 and PL1), sugar content (SUT1 and SPS), pigmentation (CHS and DFR), ABA biosynthesis (NCED1), and ABA signaling (ABAR, ABI1, ABI3, ABI4, and ABI5; Figs 4, 5). Notably, among the investigated genes, only expression of ABAR, ABI4, and NCED1 was up-regulated in the VIGS fruit (Fig. 5). It was previously demonstrated that the down-regulation of FaNCED1 results in a significant decrease in ABA levels in uncolored fruit due to feedback regulation (Jia ). In the present study, the down-regulation of FaRIPK1 expression also led to a significant increase in NCED1 expression levels and, as a result, to promotion of ABA accumulation in the FaRIPK1-VIGS fruit (Fig. 5). Similar to the FaABAR-VIGS fruit, a feedback-regulatory increase in ABA content in the FaRIPK1-VIGS fruit was also found (Fig. 5). Similarly, the increased transcript levels of both ABAR and ABI4 in FaRIPK1-VIGS fruit may result from a feedback-regulatory mechanism. It is notable that the transcription factor ABI4 was recently demonstrated to serve as an inducer of strawberry fruit ripening (Chai and Shen, 2016). Based on these results, we propose that FaABAR, FaRIPK, and FaABI4 synergistically regulate strawberry fruit ripening through a linear link.

Understanding the mechanism of FaRIPK1 action in the regulation of strawberry fruit ripening

Fleshy fruit ripening involves in complex physiological, metabolic, and molecular mechanisms, and is often coupled, externally, with a green–white–red color transition, and, internally, with a chloroplast–leucoplast–chromoplast conversion (Schweiggert ). A recent study revealed that the white stage represented the onset of fruit ripening (Wang ). In the present study, the FaABAR/CHLH gene was highly expressed from the small green (SG) to de-greening (DG) stages, which resulted, at least in part, from its being responsible for chlorophyll biosynthesis in the green fruit; while after the white stage, the increased FaABAR/CHLH transcript levels may be related to its receptor function in the reddening fruit (Supplementary Fig. S8). In contrast, the FaRIPK1expression levels increased slightly in the developing green fruit, then increased sharply at the white stage, where it reached its maximum level, before declining gradually (Fig. 3). The maximal expression of FaRIPK1 at the white stage might be related to a subsequent increase of FaABAR expression, which could initiate the onset of coloration. We suggest that the peak in FaRIPK1 expression initiated FaABAR expression and ripening, and that the reddening fruits no longer require such high FaRIPK1 transcript levels (Fig. 3), but do need elevated FaABAR expression (Supplementary Fig. S8). The down-regulation of FaRIPK1 expression consistently promoted FaABAR expression (Fig. 5). We hypothesize that the synergetic expression of FaRIPK1 and FaABAR may be required for ABA-mediated fruit ripening: FaRIPK1 is responsible for the onset of ripening, while FaABAR further promotes ripening. This idea is supported by the co-localization of the two proteins in the nucleus (Fig. 3). Given the localization of FaABAR in the chloroplast (Shen ; Shang ) and the localization of FaRIPK1 in the plasma membrane and nucleus (Fig. 3), the FaABAR–FaRIPK1 interaction may contribute to integrating ABA into multiple signaling pathways, possibly through ABI4 at various subcellular levels, including the plasma membrane, nucleus, and chloroplast. We also noted the present of an EDR1domain in FaRIPK1 (Supplementary Fig. S3) and observed that FaRIPK1 is induced by ethylene treatment (Fig. 6), suggesting that the relationships between ABA, ethylene, and FaRIPK1 are interesting and worthy of further investigation.

Supplementary data

Supplementary data are available at JXB online.

Table S1. The primers used in this study.

Fig. S1. Construction of a cDNA library from white fruit and a yeast two-hybrid using FaABAR as bait.

Fig. S2. Expression and purification of FaABARC690 and FaRIPK1.

Fig. S3. Putative conserved domains in FaRIPK1.

Fig. S4. Hits of the FaRIPK1 protein in a Blast search in the NCBI database.

Fig. S5. Putative conserved domains in a homolog of RIPK2.

Fig. S6. Hits of FaRIPK1 nucleotides (3465 bp) in a Blast search in the in NCBI database.

Fig. S7. Expression and purification of LRR283 and STK319.

Fig. S8. FaABAR expression levels during the development of strawberry fruit.

Click here for additional data file.