Isolation and Expression of NAC Genes during Persimmon Fruit Postharvest Astringency Removal

NAC genes have been characterized in numerous plants, where they are involved in responses to biotic and abiotic stress, including low oxygen stress. High concentration of CO2 is one of the most effective treatments to remove astringency of persimmon fruit owing to the action of the accumulated anoxia metabolite acetaldehyde. In model plants, NAC genes have been identified as being responsive to low oxygen. However, the possible relationship between NAC transcription factors and persimmon astringency removal remains unexplored. In the present research, treatment with a high concentration of CO2 (95%) effectively removed astringency of “Mopan” persimmon fruit by causing decreases in soluble tannin. Acetaldehyde content increased in response to CO2 treatment concomitantly with astringency removal. Using RNA-seq and Rapid amplification of cDNA ends (RACE), six DkNAC genes were isolated and studied. Transcriptional analysis indicated DkNAC genes responded differentially to CO2 treatment; DkNAC1, DkNAC3, DkNAC5 and DkNAC6 were transiently up-regulated, DkNAC2 was abundantly expressed 3 days after treatment, while the DkNAC4 was suppressed during astringency removal. It is proposed that DkNAC1/3/5/6 could be important candidates as regulators of persimmon astringency removal and the roles of other member are also discussed.


Introduction
Proanthocyanidins (PAs) are secondary metabolites and phenolic oligomers that result from the condensation of flavan-3-ol units. PAs have a variety of uses and are beneficial for human health owing to their antioxidant properties, which include anti-allergic, anti-tumor, anti-aging properties, and the prevention of cardiovascular disease [1][2][3]. Moreover, they can be used to detoxicate snake venom and have been used to adsorb radioactive compounds from nuclear waste and recycle the heavy metals from electronic waste [4][5][6][7]. Persimmon (Diospyros kaki) is a unique fruit that accumulates PAs in abundance [8]. Persimmon fruit can be divided into astringent and nonastringent types, and most of the commercial cultivars in China are of the astringent type [9]. Astringent persimmons accumulate abundant PAs in the fruit flesh even at maturity and soluble PAs cause astringency, which severely affects the industry and consumer acceptance [9,10]. Thus, astringency removal is critical for the persimmon industry.
Due to the importance of astringency removal for commercial production, a range of technologies has been developed, such as CO2 and N2 [11][12][13]; C2H4 and alternate freezing and thawing [14,15]. Among these, high concentration CO2 treatment is very commonly used and its underlying physiological and molecular mechanisms of astringency removal are much more clearly understood than for the other treatments. It has been suggested that acetaldehyde, which accumulates during anaerobic respiration induced by high CO2, causes soluble tannins to become insoluble during the treatment, thus reducing astringency [16]. Alcohol dehydrogenase (ADH) and Pyruvate decarboxylase (PDC) enzyme activities increased during deastringency and this was associated with acetaldehyde biosynthesis [17]. Our previous results indicated that ADH and PDC enzyme were both induced by CO2 and C2H4 treatments, and DkADH1, DkPDC1, DkPDC2 and DkPDC3 expression was up-regulated in response to both ethylene and high CO2 treatments [18]. When DkPDC2 was transiently over-expressed in persimmon leaves, they showed a lower level of soluble tannins, compared with leaves infiltrated with an empty vector, suggesting that DkPDC2 is a key gene regulating soluble tannins and astringency removal [18]. These results suggested CO2 driven astringency removal requires acetaldehyde metabolism, brought about by modulation of ADH and PDC enzymes and genes. However, understanding of the molecular basis of persimmon astringency removal is still limited.
In model plants, a few transcription factors have been reported to be involved in the hypoxia response, such as NAC and ethylene response factors (ERF) [19]. As mentioned above, persimmon fruit astringency removal by CO2 treatment is considered to operate mainly via the hypoxia fermentation pathway. In persimmon fruit, the only transcription factors analyzed during persimmon fruit astringency removal have been the ERF gene family. Recently, four ERF genes, DkERF9, DkERF10, DkERF19 and DkERF22, were shown to manifest trans-activation of DkADH and DkPDC genes [20]. However, the relations between other transcription factors and astringency removal have rarely been reported. In addition to ERFs, NAC genes are the main transcription factors reported to be involved in the plant hypoxia response, which indicated their potential to be involved in persimmon astringency removal. The NAC designation is based on highly conserved consensus sequences in the N-terminal region of a Petunia gene (NAM), Arabidopsis ATAF1/2 and CUC2 proteins [21]. More than 100 NAC genes have been identified and characterized in Arabidopsis [22]. Among them, ANAC019, ANAC055 and ANAC072 enhanced tolerance to drought stress in Arabidopsis [23], and ANAC2 is involved in response to plant hormones [24]. In addition, ANAC102 was shown to be induced by low oxygen (0.1%), and overexpression of ANAC102 up-regulated ADH expression [19]. Since treatment with a high concentration of CO2 triggered expression of ADH/PDC genes and acetaldehyde metabolism in persimmon, this suggested that the Arabidopsis low oxygen responsive transcription factors such as ANAC102 had the potential to be involved in astringency removal, but experimental evidence for this was lacking in persimmon.
In the present research, six DkNAC genes were isolated from "Mopan" persimmon fruit based on differentially expressed genes (DEG) in an RNA-seq database using Rapid amplification of cDNA ends (RACE) technology. High CO2 (95%) treatment was applied to the fruit and DkNAC transcripts were analyzed during the deastringency process. Some DkNAC genes were found to be positively correlated with persimmon fruit deastringency, and the possible roles of these and other DkNAC genes are discussed.

Fruit Deastringency
The soluble tannin content remained almost constant during storage in the control fruit, whereas in contrast, the CO2 treatment caused a rapid decrease in the concentration of soluble tannin from 0.917% at day 0% to 0.229% at day 1 ( Figure 1). Soluble tannins content was measured both by the Folin phenol method, and also visualized by tannin printing, using filter paper soaked with 5% FeCl2. The extent, and location, of soluble tannin content was revealed by the tissue printing, and was substantially lower in CO2 treated fruit, compared with control fruit ( Figure 2). Acetaldehyde, the main compound responsible for insolublization of soluble tannin, was also analyzed. The results indicated that there was a very obvious increase (approx. five fold) in acetaldehyde accumulation in CO2 treated fruit one day after the start of treatment. However, such differences were abolished, following cessation of the treatment ( Figure 3).
By utilizing soluble tannin and tanning printing assays, 95% CO2 treatment was shown to be effective in inducing deastringency in postharvest "Mopan" cultivar persimmon. These results confirm and extend our previous findings [18] and other reports using various persimmon cultivars [11,12,18]. A burst of acetaldehyde production during deastringency treatment has been widely reported in various persimmon cultivars, such as "Rojo Brillante", "Kaki Tipo", "Lycopersicon" [25,26]. Thus, the transient accumulation of acetaldehyde further supports the role and effectiveness of 95% CO2 treatment on persimmon fruit deastringency.

NAC Genes Isolation and Sequence Analysis
Six novel putative NAC genes were isolated from persimmon fruit, five of which were full-length sequences and one partial CDS, and designated as DkNAC1-6 (GenBank accession nos. KP222303-KP222308). DkNAC1-5 were predicted to encode proteins of 303, 266, 242, 214 and 194 amino acid (aa), respectively, however, the protein encoded by DkNAC6 could not be predicted from the partial CDS.
Alignment analysis with the deduced proteins encoded by the DkNAC genes showed that they contained the NAC conserved domain in their N-terminal regions, which was divided into five subdomains (A-E) [27] (Figure 4). However, the C-terminal regions showed less similarity between different members of the NAC family ( Figure 4).
To examine the phylogenetic relationship between the NAC proteins in persimmon and Arabidopsis, a phylogenetic tree was constructed based on their translated amino acid sequences. As shown in Figure 5, DkNAC1 and DkNAC6 were homologous with, and closely related to ANAC102, which is involved in the viability of Arabidopsis seeds following low-oxygen treatment [19]; DkNAC2 and DkNAC3 share the same branch, and are closely related to ANAC029, which was reported to play an important role in leaf senescence in Arabidopsis [28]; DkNAC4 was clustered with ANAC013, which has been reported to be involved in mitochondrial retrograde regulation of the oxidative stress response in Arabidopsis [29]; DkNAC5 was similar to ANAC104, which was identified as part of the leaf senescence transcriptome [30]. Collectively, these data suggested that DkNAC1-6 may exhibit diverse functions. In order to investigate the possible roles of NAC transcription factors in persimmon fruit deastringency, six differentially expressed DkNAC genes were identified by RNA-seq data [20]. Alignment of the six DkNAC proteins showed that they shared a highly conserved N-terminal region, containing five consensus subdomains, which are termed the NAC domains [27]. NAC genes have been classified previously into two groups based on similarities of NAC domain structures in Arabidopsis thaliana and rice [27]. Group I proteins are composed of 14 subgroups, including TERN, ONAC022, SENU5, NAP, AtNAC3, ATAF, OsNAC3, NAC2, ANAC011, TIP, OsNAC8, OsNAC7, NAC1 and NAM, and group II can be divided into four subgroups, namely ANAC011, ONAC003, ONAC001 and ANAC063 [27]. DkNAC1 and DkNAC6 belong to group I (ATAF subfamily). ANAC102 belongs to the ATAF subgroup [19]. DkNAC2-4 also belongs to group I, whereas DkNAC2 and DkNAC3 belong to the NAP subfamily and DkNAC4 belongs to the NAC2 subfamily and DkNAC5 belongs to group II. Thus, phylogenetic and structural analyses indicated that DkNAC1 and DkNAC6 were mostly likely to be involved in persimmon deastringency, as indicated by the high similarity with ANAC102.

NAC Genes Expression
The six NAC genes had different expression patterns. Transcripts of DkNAC1, DkNAC5 and DkNAC6 were transiently induced by the CO2 treatment, peaking after one day, and the expression of DkNAC3 peaked after two days. DkNAC5 was the most strongly up-regulated, with its mRNA increasing in abundance approximately 1200-fold following the treatment. Unlike DkNAC1/3/5/6, transcripts of DkNAC2 increased at four days (three days after treatment was stopped). Unlike the other DkNAC genes, the expression of DkNAC4 was decreased during CO2 treatment and then recovered by day 4 ( Figure 6). One of most interesting results was that DkNAC1, DkNAC3, DkNAC5 and DkNAC6 were up-regulated by CO2 treatment, and that this occurred concomitantly with the decrease in soluble tannin. This indicates that in persimmons, expression of DkNAC1, DkNAC3, DkNAC5 and DkNAC6 is highly correlated with fruit deastringency. The expression pattern of DkNAC1 and DkNAC6 further supported the prediction from phylogenetic and structural analyses, that ANAC102 homologs might be important regulators of the hypoxia response, which benefit deastringency in persimmon fruit. Besides DkNAC1 and DkNAC6, DkNAC3 and DkNAC5 were also observed to increase in expression in response to 95% CO2 treatment; however, DkNAC3 and DkNAC5 belong to the NAP subfamily and group II, respectively. These results may indicate that more subfamilies might be involved in the deastringency response in persimmon fruit and more generally in the hypoxia response in plants. Although hypoxia-responsive NAC genes have rarely been reported, the results from the studies on ANAC102 also indicated that additional NAC genes might exist for the hypoxia response, as ANAC102 knockout lines did not show altered ADH gene transcription in Arabidopsis [19]. Taken together, results from persimmon and previous finding in Arabidopsis indicate that additional NAC genes are potentially involved in the hypoxia response in plants.
In addition to the four above-mentioned DkNAC genes, DkNAC2 and DkNAC4 expression also exhibited interesting features. Unlike the positively regulated NAC genes, DkNAC4 mRNA levels were negatively correlated with astringency removal. Nevertheless, no repressive domain, such as an EAR motif, was observed within the coding region of DkNAC4. The increase in DkNAC2 mRNA occurred on the third day after treatment was removed, after the fruit had lost their astringency and started to senescence. These results indicated that DkNAC2 might be involved in fruit ripening or senescence-related processes, but not astringency removal. Similar results were also reported in other fruit, such as banana, where transcripts of MaNAC1 and MaNAC2 were up-regulated in banana post-harvest ripening, and it was suggested that MaNAC genes may be involved in banana fruit ripening via interaction with ethylene signalling components [31]. It was also reported in Arabidopsis siliques that AtNAP, a NAC gene whose expression increased with the progression of silique senescence, played a key role in its senescence [32].

Materials and Treatments
The "Mopan" (astringent cultivar) persimmon fruit were harvested from a commercial orchard (Fangshan, Beijing) in 2013, with mean firmness of 47 N, and then transported to the laboratory the next day. Fruits with no defects and uniform size were chosen for treatment. CO2 (95%) and control treatments were performed according to our previous report [18] (Min et al., 2012). After firmness measurement at each sampling point, fruit flesh was removed from the peel and shredded, then frozen with liquid nitrogen and kept in −80 °C refrigerator until required.

Soluble Tannins Assay and Tannin Printing
The content of soluble tannins, the most important index for astringency, was measured using Folin-Ciocalteu reagent according to the method described in our previous report [14] (Yin et al., 2012).
The printing method, a convenient way of identifying persimmon astringency location and loss, was also used for the examination of soluble tannins in the fruit flesh according to [33] Hu et al., (2013). Filter papers were soaked with 5% FeCl2 for about 5-10 min, then moved to a 60 °C oven until dry. The whole fruit was cut into two parts and printed on the processed filter paper. The black color indicated soluble tannins and the intensity of black reflected the soluble tannins content.

Acetaldehyde Determination
Acetaldehyde production was determined with a gas chromatograph Instrument (Agilent 6890N, Folsom, CA, USA) with a FID column (HP-INNOWAX, 0.25 mm, 30 m, 0.25 μm, Agilent J&W, Folsom, CA, USA) according to the method described previously [20] (Min et al., 2014) with some modification. The injector, detector and oven temperatures were 150, 160, and 100 °C, respectively. Sec-butyl alcohol was added to each vial as an internal control. The results were calculated using standard curves for acetaldehyde and ethanol, respectively.

RNA Extraction and cDNA Synthesis
Total RNA was prepared according to the method used previously [18]  . Traces of contaminating genomic DNA in total RNA were removed with TURBO DNase (Life Technologies, Gaithersburg, MD, USA). One microgram DNA-free RNA was used for cDNA synthesis by iScript cDNA Synthesis kit (Bio-Rad, Hercules, CA, USA) following the manufacturer's protocol. For each sampling point, three biological replicates were used for RNA extraction.

Gene Isolation and Sequence Analysis
The NAC genes were isolated based on RNA-Seq data. RNA-Seq was performed by the Beijing Genome Institute (BGI) (Shenzhen, China) according to our previous paper [20] (Min et al., 2014). Rapid amplification of cDNA ends (RACE) was chosen to amplify the 3' and/or 5' end Untranslated region (UTR) of NAC, using a SMART™ RACE cDNA Amplification Kit (Clontech, Palo Alto, CA, USA). The sequences of primers used for RACE are listed in Table 1. The gene sequences were translated with online software (http://web.expasy.org/translate/) and were confirmed with the BLAST methods in Genbank. A phylogenetic tree of NAC genes was generated using ClustalX (v 1.81) [34] and calculated using Figtree (v1.3.1, University of Edinburgh, Edinburgh, UK). The deduced amino acid sequences of homologous genes of Arabidopsis were obtained from TAIR (The Arabidopsis Information Resource).

Oligonucleotide Primers and Real-Time PCR
Oligonucleotide primers for real-time PCR analysis were designed with primer3 (v. 0.4.0, http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). The specificity of primers was determined by melting curves and PCR products resequencing as described in [35] Yin et al. (2008). The sequences of oligonucleotide primers are listed in Table 2. Table 2. Sequences of the primers used for real-time PCR.

Gene Primary PCR (5' to 3') Secondary PCR (5' to 3') DkNAC1
GGTAGCATCATAAGCGTTAATCTG ACGCAGAGTACATGGGGAAT ACGGAGAAGCCTGCATAGC Real-time PCR was carried out with Ssofast EvaGreen Supermix kit (Bio-Rad) and CFX96 instrument (Bio-Rad) for gene expression studies according to our previous report [18]  . The relative abundance of each gene was calibrated with samples from day 0 fruit. Actin, a housekeeping gene, was used as the internal control [9] (Akagi et al., 2009).

Statistical Analysis
Origin 8.0 (OriginLab, Northampton, MA, USA) was used to prepare the figures. Statistical significance of differences was calculated with least significant difference (LSD) using DPS software (v. 3.11, Zhejiang University, Hangzhou, China).

Conclusions
In conclusion, soluble tannins and acetaldehyde assay results showed that a high concentration of CO2 treatment was effective in inducing deastringency of "Mopan" persimmon fruit. Six DkNAC genes were isolated and the transcriptional level of DkNAC1, DkNAC3, DkNAC5 and DkNAC6 were highly correlated with fruit deastringency. Finally, DkNAC1/3/5/6 are identified as prime candidates for further transcriptional regulatory analysis in persimmon fruit to understand the molecular control of astringency removal.