Kiwifruit (Actinidia chinensis cv. Hongyang) were harvested from a commercial orchard in Dujiangyan City (Sichuan, China) at a commercial mature stage. Fruits of similar size and the absence of physical injuries or decays were selected, and randomly divided into 6 lots with 100 fruits each. Three lots were stored at low temperature (0°C) for 3 months (90 days), and other three were kept at room temperature (25°C) for 9 days. All fruits were enclosed with polyethylene film bag (0.02-mm thickness) to maintain relative humidity of about 95%. As replicates for each treatment (room- and cold-storage), three lots of 15 fruits were sampled in every interval of 3 or 30 days storage at room- or low-temperature, respectively. Fruit firmness and the soluble solids content (SSC) were measured as described (Xia et al., 2016). The kiwifruit sampled at the paired time points (3 days at 25°C vs. 30 days at 0°C, 6 days at 25°C vs. 60 days at 0°C, 9 days at 25°C vs. 90 days at 0°C) showed similar SSC and firmness. For further molecular and biochemical analysis, inner pericarp tissue collected from the equatorial region (1.5-cm thickness) of each fruit was cut into pieces, frozen in liquid nitrogen and stored at -80°C.

Anthocyanin was extracted and analyzed according to the following procedure. Briefly, 10 g of frozen samples were ground and extracted in 30 ml of methanol containing 2% (v/v) formic acid. After vortex, the homogenates were sonic-treated in ice-cold water for 20 min and then inoculated at 4°C overnight, followed by centrifugation at 3000g for 10 min. The extract was then filtered through 0.22 μm polytetrafluoroethylene filters and retained for component analysis. The anthocyanin analysis was performed on an Agilent 1290 Infinity UHPLC system coupled to Agilent 6540 UHD Accurate-Mass Q-TOF mass spectrometer (Agilent Technologies, USA). A 10 μl aliquot of sample was injected on C18 (150 mm × 4.6 mm) column and separations were carried out using a binary solvent system of Solvent A (water +0.1% formic acid) and Solvent B (acetonitrile +0.1% formic acid) at a flow of 300 μl min^-1. The initial mobile phase was 95%A 5%B, increased linearly to 30%A 70% B in 20 min, and held for 5 min before resetting to 95%A 5%B ready for the next injection. The column temperature was set to 30°C and anthocyanins were monitored at 520 nm. Quantification of anthocyanins was conducted by comparison with authenticated standard solution of cyanidin 3-O-galactoside (Sigma). Anthocyanin peaks were further identified by ESI-MS and the spectral data were obtained in positive ion mode over the range m/z 100–1000. The ESI voltage, capillary temperature, and capillary voltages was 39 V, 300°C and 7 μV, respectively.

Dihydroflavonol 4-reductase, ANS, and UDP-flavonoid 3-O-galactosyl transferase (UFGT) genes from A. chinensis ‘Hongyang’ were initially cloned using gene specific primers (Supplementary Table S1) designed on DFR (EST FG410069.1), ANS (EST FG407400.1) and UFGT (EST FG405592.1) sequences from an available kiwifruit EST library (Crowhurst et al., 2008). Full-length cDNA were cloned using a Rapid-amplification of cDNA Ends (RACE) kit (Takara). Putative MYB transcription factors were identified from the kiwifruit genome database (Huang et al., 2013), and four full-length MYB genes highly expressed in fruit tissue were selected. The open reading frames (ORFs) were predicted using GENSCAN, and protein sequences were aligned using ClustalX2.1. Phylogenetic tree of MYB proteins were constructed by neighbor-joining matrix with 1,000 bootstrap replicates using MEGA6. Names of all cloned genes in this study were according to the previous report by Li et al. (2015), except MYB4a which was named according to the phylogenic analysis.

Total RNA was extracted according to the methods reported by Zhu et al. (2013) from 1 g of kiwifruit tissues. Using a Takara PrimeScript^® RT reagent kit with gDNA eraser, the first strand cDNA was synthesized by reverse transcription following manufacturer’s instruction. Quantitative real-time PCR was performed on StepOnePlus Real-Time PCR System (Applied Biosystems) using SYBR^®Premix Ex Taq^TM (TliRNaseH Plus) (Takara, Japan). Gene-specific primers were designed with Primer Express software 3.0 and presented in Supplementary Table S2. The PCR program was conducted as follows: 15 s at 95°C; 40 cycles of 5 s at 95°C, and 30 s at 60°C; followed by an automatic melting curve analysis. Three independent biological replicates were measured for each sample. The relative expression level for each gene was calculated using the comparative Ct method (2^-ΔCt method) with a kiwifruit actin as a reference.

The CDS of AcMYB5-1, AcMYB5-2, and AcMYBA1-1 were amplified using specific primers (Supplementary Table S3) containing KpnI and BamHI restriction enzymatic site, respectively. The PCR product was recombined with the linearized vector pCAMBIA2300 (In-Fusion HD Cloning Kit; Clontech). The resulting construct were sequenced and introduced into Agrobacterium tumefaciens strain GV3101. Agrobacterium were cultured on LB agar plates with kanamycin and incubated at 28°C. The freshly grown Agrobacterium were sedimented by centrifugation for 5 min at 6000 g, resuspended in infiltration buffer, and incubated at room temperature for 2 h before infiltration. Tobacco plants were grown for five to six weeks and the young leaves were syringe-infiltrated with A. tumefaciens suspension in abaxial side of the leaf. Control was infiltrated with Agrobacterium containing pCAMBIA2300 (empty vector) at the same time and the transient expression was assayed 5 days after infiltration.

All statistical analyses were carried out using SPSS 13.0 software (SPSS Inc., USA). Data from assay of anthocyanin content or gene expression were analyzed by one-way analysis of variance (ANOVA), and the mean comparison was performed by Duncan’s multiple range tests. Differences at P < 0.05 were considered as significant.

Red-flesh is the distinguishing feature of ‘Hongyang’ kiwifruit due to the accumulation of anthocyanin in its inner pericarp during ripening. To identify the main anthocyanin in kiwifruit during storage, UHPLC Q-TOF-MS was applied to analyze the anthocyanin composition from the extract of fruit. Our results showed that all samples of ‘Hongyang’ kiwifruit presented two peaks that were identified as cyanidin 3-O-xylo(1-2)-galactoside and 3-O-galactoside by comparing both retention times and ESI-MS data (Supplementary Figure S1). The dominant anthocyanin peak had a molecular ion of m/z 581, while another peak had a molecular ion of m/z 499. The results indicated that cyanidin 3-O-xylo(1-2)-galactoside was the major anthocyanin present in ‘Hongyang’ fruit during storage, which is consistent with previous report in pre-harvest ‘Hongyang’ fruit (Man et al., 2015b).

As shown in Figure 1A, after 90 days of storage at low temperature, the red coloration of fruit inner pericarp was more notable than that of fruit at room temperature for 9 days. To investigate the influence of low temperature on anthocyanin content in ‘Hongyang’ fruit, the accumulation of major anthocyanin during the storage was investigated. The 3-O-xylo(1-2)-galactoside content in inner pericarp increased gradually with prolonged storage time (Figure 1B). In room temperature-stored fruit, it increased rapidly from 11.88 at 0 day to 18.87 μg⋅g^-1 FW at 9 days. By contrast, in low temperature-stored fruit, it accumulated gradually and reached a peak value of 23.87 μg⋅g^-1 FW after 90 days. At the same time, fruit firmness decreased obviously with storage time prolonging (Figure 1C), and SSC increased gradually (Figure 1D) in kiwifruit both stored at 25 and 0°C. The firmness and SSC of fruit stored at room temperature for 9 days and low temperature for 90 days showed no significant difference.

Full-length cDNA sequence of genes involved in the anthocyanin biosynthesis pathway, were cloned from ripe ‘Hongyang’ kiwifruit using RT-PCR and RACE, annotated as DFR1 (Achn014341), DFR2 (Achn0135311), ANS1 (Achn002561), ANS2 (Achn361621), UFGT1 (Achn209671), UFGT2 (Achn017071) and UFGT3 (Achn321621) according to previous description by Li et al. (2015). Expression level of those genes, i.e., DFR, ANS, and UFGT, were analyzed in fruit stored at room- and low-temperature (Figure 2).

Significant differences in the expression pattern of these genes were revealed during storage. The transcript abundance of DFR1 and DFR2 was not changed in fruit stored at room temperature, whereas it was obviously increased in fruit stored at low temperature, with a higher level after 90 days, about 3.4 and 4.8 times that of the initial storage, respectively. ANS expression gradually increased in the early stage of storage, with the maximum value at 3 and 6 days, respectively, and then declined in room temperature-stored fruit. By contrast, fruits stored at low temperature remained a lower level of the expression of ANS1 and ANS2 in the first 60 days of storage, but exhibited a higher level after 90 days. The expression of UFGT1 and UFGT3 decreased in fruit stored either at room temperature or low temperature; however, UFGT2 expression firstly increased and then declined. Low temperature storage enhanced the UFGT2 expression from 30 to 90 days. These results indicated that the mRNA levels of specific anthocyanin biosynthesis genes were induced by low temperature.

From the draft kiwifruit genome, four full-length MYB-related sequences were cloned from the ‘Hongyang’ kiwifruit, annotated as MYBA1-1 (Achn104391), MYB4a (Achn020361), MYB5-1 (Achn148821) and MYB5-2 (Achn366791). Phylogenetic analysis using the predicted protein sequences of the four kiwifruit MYBs and other published MYBs sequences suggested that MYB1-1 was most closely related to VvMYBPA1, and MYB4a was clustered with AtMYB4. Moreover, MYB5-1 and MYB5-2 were included in a small group with VvMYB5a and VvMYB5b (Supplementary Figure S2).

Expression levels of the four MYBs were analyzed in room and low temperature-stored fruit, respectively. As shown from Figure 3A, mRNA levels of MYBA1-1 increased gradually in room temperature-stored fruit, while its level rapidly increased in low temperature-stored fruit. There was no significant difference in the expression of MYB4a during low temperature storage (Figure 3B). Interestingly, the most significant differences were observed in the expression patterns of MYB5-1 and MYB5-2 in kiwifruit. The expression of MYB5-1 had a constantly lower level in fruit stored at room temperature throughout the storage time; however, it was sharply increased in low temperature-stored fruit, being a five times higher than that of initial time point, after 90 days of storage (Figure 3C). On the contrary, the MYB5-2 expression in room temperature-stored fruits increased gradually and peaked at 9 days, however, it remained at a basal level till the end of storage in low temperature-stored fruit (Figure 3D). The results indicated that low temperature storage could induce and enhance the expression of certain members of MYBs gene family.

To further investigate the roles of AcMYB5-1/5-2/A1-1 in regulation of anthocyanin biosynthesis, transient transformation of genes encoding the three transcription factors was carried out in Nicotiana benthamiana leaves. The expression levels of NtANS, NtDFR, and NtUFGT were analyzed in transiently transformed tobacco leaves (Figure 4). Transiently overexpression of AcMYB5-1/5-2/A1-1 up-regulated expression levels of NtANS and NtDFR, but did not alter the expression of NtUFGT in tobacco leaves. A strong induction of the NtANS and the NtDFR genes was found in AcMYB5-2 agro-infiltrated leaves, about 7 and 13 times higher, respectively, than that in control leaves. The results suggested that overexpression of AcMYB5-1/5-2/A1-1 could activate the gene expression of NtANS and NtDFR in tobacco.