Fruit maturation is an important developmental stage because the set of biochemical pathways involved in ripening make the fruit attractive, perfumed, and edible (Bouzayen et al., 2010). In addition, the ripening process also helps seed dispersal (Wang et al., 2020a). There is considerable ongoing research to help understand the complex regulation of this important process (Bouzayen et al., 2010). Based on their ripening behavior, fleshy fruits have been divided into two groups: climacteric and non-climacteric (McMurchie et al., 1972). Climacteric fruits such as tomato are characterized by an ethylene burst accompanied by an increase in respiration at the onset of ripening. In contrast, non-climacteric fruits such as orange lack this ethylene-associated respiratory peak (Paul et al., 2012). Usually, climacteric fruits have a shorter shelf life than non-climacteric fruits (Hiwasa-Tanase and Ezura, 2014), and breeding programs for fruit crops are oriented to increase shelf life to minimize postharvest losses (Payasi and Sanwal, 2010).

Ethylene plays a primary role in initiating climacteric fruit ripening (McMurchie et al., 1972). Its production is low at the pre-climacteric stage, while there is a massive auto-stimulated ethylene production at the onset of the ripening stage. Exogenous ethylene treatment can also induce the ethylene burst at the pre-climacteric stage of climacteric fruits, thereby advancing the ripening process (Hiwasa-Tanase and Ezura, 2014). In contrast, treatment with the ethylene inhibitor 1-methylcyclopropene (1-MCP) delays fruit ripening (Blankenship and Dole, 2003; Watkins, 2006). Antisense-induced repression of ethylene synthesis enzymes in tomato also delays fruit maturation (Hamilton et al., 1990). These results demonstrate the key role of ethylene in regulating ripening in climacteric fruits. Interestingly, some reports have shown that ethylene can also play a role in non-climacteric fruit ripening (Katz et al., 2004; Bouzayen et al., 2010), although in non-climacteric fruit, a low level of ethylene during the whole developmental process is found (Paul et al., 2012). In melon, some fruit ripening processes are independent of ethylene, such as flesh softening, sugar accumulation, or flesh color, that do not change in ethylene-suppressed melon fruit (Flores et al., 2001; Pech et al., 2008), confirming that the control of fruit ripening is a complex trait.

Genes involved in fruit ripening have been largely studied in either climacteric or non-climacteric species (Gapper et al., 2013; Osorio et al., 2013; Lü et al., 2018) and tomato has emerged as a prime model of climacteric fruit ripening (Alexander and Grierson, 2002). Genetic characterization of several ripening-related mutants in tomato has advanced our knowledge of the mechanisms that regulate fruit ripening (Giovannoni, 2007). The ripening-inhibitor (rin), non-ripening (nor), and Colorless non-ripening (Cnr) mutations have been useful to understand the transcriptional regulation of fruit ripening (Vrebalov et al., 2002; Manning et al., 2006; Giovannoni, 2007; Wang et al., 2020b). However, the milder ripening phenotypes observed for the CRISPR/Cas9 knockouts of these three genes have resulted in a re-evaluation of their original proposed role as master regulators of fruit ripening, suggesting that a network of partially redundant components exists to regulate this important biological process (Ito et al., 2017; Gao et al., 2019, 2020; Wang et al., 2019, 2020a). Additional transcription factors involved in the regulation of fruit ripening have also been identified in tomato. The MADS-box transcription factor TOMATO AGAMOUS-LIKE1 (TAGL1; Itkin et al., 2009; Vrebalov et al., 2009) is highly expressed during fruit ripening. TAGL1-silenced fruit did not ripe normally, with reduced levels of carotenoids and ethylene. FRUITFULL homologues (TDR4/FUL1 and MBP7/FUL2) are also MADS-box transcription factors involved in fruit ripening in an ethylene-independent manner, having redundant functions in cell wall modification (Bemer et al., 2012). FUL1/2 and TAGL1 may regulate different subsets of the known RIN targets. APETALA2a (AP2a) is a negative regulator of tomato fruit ripening, with its silencing causing elevated ethylene production and early fruit ripening (Chung et al., 2010; Karlova et al., 2011). Other NAC proteins have also been found to be involved in regulating ripening, among them SlNAC1, SlNAC3, SlNAC4, and SlNAM1, suggesting that a complex regulatory network of fruit ripening exists (reviewed in Liu et al., 2022). For non-climacteric fruit, strawberry is one of the most studied plants (Osorio et al., 2013) and several genes involved in strawberry fruit ripening have recently been identified, including FaPYR1 (Chai et al., 2011), FaExp2 (Civello et al., 1999), FaASR (Chen et al., 2011), FaABI1 (Jia et al., 2013), and FaRIF (Martín-Pizarro et al., 2021). These studies have provided valuable information on gene function related to regulation of fruit ripening.

Melon (Cucumis melo L.) is a suitable model to study fruit ripening, because there are both climacteric and non-climacteric genotypes (Ezura and Owino, 2008). Genetic analysis of a biparental population of the cantaloupe type Védrantais (VED, climacteric) × PI 161375 (SC, non-climacteric) inbred lines indicated that ethylene production and fruit abscission are controlled by two independent loci, Al-3 and Al-4 (Périn et al., 2002). In recent studies, a near isogenic line SC3-5-1 derived from the non-climacteric parental lines SC and the inodorus type Piel de Sapo (PS) had a climacteric ripening phenotype, and two QTLs, ETHQB3.5 and ETHQV6.3, were found to be involved in the regulation of climacteric ripening in SC3-5-1 (Eduardo et al., 2005; Moreno et al., 2008; Vegas et al., 2013). Previously, ETHQV6.3 was found to be encoded by a NAC transcription factor CmNAC-NOR (MELO3C016540.2), phylogenetically related to the tomato SlNAC-NOR (Ríos et al., 2017). TILLING mutants containing non-synonymous mutations in the coding region of CmNAC-NOR had a delayed ripening phenotype, suggesting that CmNAC-NOR is an important regulator of climacteric ripening in melon. To further investigate the CmNAC-NOR function, in this study, we generated and phenotyped CRISPR/Cas9 mutants with different disruption levels.

NAC transcription factors are a large gene family involved in plant development and the environment stress response (Guo and Gan, 2006; Nuruzzaman et al., 2013; Hernández and Sanan-Mishra, 2017). Among them, the SlNAC-NOR gene from tomato is well known as a key regulator in fruit ripening (Giovannoni, 2007). In previous studies, we have characterized melon CmNAC-NOR, a close homologue of SlNAC-NOR according to a phylogenetic analysis with NAC genes from several species, and demonstrated its involvement in fruit ripening with the delayed ripening phenotype of two TILLING mutant lines (Ríos et al., 2017). RNA-seq expression analysis of CmNAC-NOR in Ved (climacteric) and PS (non-climacteric) at different ripening stages showed that both lines have a similar expression profile during ripening (Supplementary Figure S2). In this study, we obtained two diploid CmNAC-NOR-disrupted melon plants with different edited sites (nor-1 and nor-3) by using the CRISPR/Cas9 system (Figure 1). nor-1 is a complete knock-out mutant where climacteric ripening behavior is almost absent, without ethylene or aroma production, a volatile profile similar to unripe fruit, absence of abscission layer, with no external color change (Figures 2, 4) and with low expression of several ripening-related genes (Supplementary Figure S1), suggesting that CmNAC-NOR plays a significant role in regulating fruit ripening in melon. Nor-3 is a knock-down mutant with one amino acid deletion at the NAC (NAM, ATAF1,2, CUC2) domain, which causes around 8 days delay in ripening, but without affecting either fruit quality (Figure 3), nor gene expression of several ripening-related genes at harvest (Supplementary Figure S1), suggesting a potential way to control fruit ripening in melon by disrupting CmNAC-NOR. Moreover, in the heterozygous nor-1, there was also a 20 day delay in ethylene production (Figure 2A), suggesting that nor-1 might have potential for extending the shelf life of fruit in melon breeding programs.

In tomato, fruit ripening is affected in both natural and CRISPR/Cas9 knock out mutants of SlNAC-NOR (Gao et al., 2020). However, the phenotype of fruit of SlNAC-NOR mutations (CR-NOR) induced by CRISPR/Cas9 has been found to be much less severe than the natural mutant slnac-nor (Wang et al., 2019; Gao et al., 2020). Mature fruit of the slnac-nor mutant does not produce an ethylene burst (Giovannoni, 2007; Adaskaveg et al., 2021) and has little carotenoid content (Giovannoni et al., 1995; Kumar et al., 2018), but CR-NOR fruits can produce ethylene and synthesize much more carotenoids than slnac-nor at the ripe stage (Gao et al., 2020). Therefore, slnac-nor was reported as a gain-of-function mutation (Gao et al., 2020) suggesting that SlNAC-NOR does not act as a master regulator but as a major gene controlling the ripening process (Wang et al., 2020a). In the VED climacteric melon, fruit ripening occurred around 35 ~ 40 DAP (Figure 2A) and was associated with a transient increase in autocatalytic ethylene production, accompanied by changes in rind and flesh color, flesh firmness, sugar content, and aroma production (Pereira et al., 2018, 2020; Mayobre et al., 2021). In the nor-3 knock down mutant, the mutation was located within the DNA/protein binding region of the NAC-NOR transcription factor, so we could expect a reduction of its binding affinity, affecting its regulation of ripening-related genes. As it was expected, we observed a delay in ripening (ethylene and aroma production), suggesting that the mutated protein was still functional but less efficient (Figure 2). This phenotype agreed with Ríos et al. (2017), where non-synonymous mutations in the conserved NAC domain region caused a delay in fruit ripening. Concerning fruit quality, we demonstrated that even though the fruit nor-3 ripened later, it was able to produce a peak of ethylene similar to VED, which probably allowed the nor-3 fruit attain the same quality parameters as VED such as VOCs profile or flesh firmness (Figure 3), that are ethylene-dependent or partially dependent (Pech et al., 2008). This was confirmed by the transcription analysis of genes involved in ethylene biosynthesis, ester production, and fruit softening, where nor-3 showed similar expression data as VED at harvest. In the complete knock-out mutant nor-1, ethylene production was blocked (Figure 2A), the expression of CmACS1 and CmACO1 was repressed (Supplementary Figure S1), the aroma component changed (Figure 2C), CmAAT1 and CmADH2 were downregulated (Supplementary Figure S1), the rind color did not change from green to yellow (Figure 4B), no abscission layer formed, the flesh was firmer than VED (Figure 4F), and two CmPGs were downregulated (Supplementary Figure S1). The flesh color of nor-1 seemed visually less orange than VED (Figure 4B), although we did not detect significant differences in carotenoid content, nor in CmOr expression (Supplementary Figure S1). It remains to be tested if the composition of individual carotenoid compounds is altered in the mutant, without modifying the total carotenoid content. Our findings suggest that nor-1 resembles the tomato phenotypes of slnac-nor and CR-NOR mutants. Unlike the tomato CRISPR mutant CR-NOR, the main climacteric ripening components were almost blocked in nor-1 melon. In addition, nor-1 was insensitive to external ethylene treatment (Figure 4), except for abscission layer formation, suggesting that CmNAC-NOR gene is a major key regulator of fruit ripening in melon.

The different phenotypes of tomato and melon CRISPR NAC-NOR mutants might be explained by their different editing patterns. Although they lose the transcriptional regulation region, their NAC domain is altered at different levels. The NAC domain contains five subdomains (A-E) that play an important role in DNA-binding (Kikuchi et al., 2000; Ernst et al., 2004). The slnac-nor mutant has been found to contain a complete NAC domain, resulting in a gain-of-function mutation, while the CR-NOR mutant produced a truncated protein of 47 aa, which lost NAC subdomains B-E, but still had the NAC subdomain A (Gao et al., 2020). Here, the editing of nor-1 started from the NAC subdomain A (Figure 1E), so the whole NAC domain was affected in nor-1, resulting in a loss-of-function mutation, whereas nor-3 lost a single amino acid at NAC subdomain A, resulting in a delay of ripening.

Fruit flavor is an important trait as it affects consumer preferences. Volatile esters are major contributors to fruit flavor giving the fruity aroma to climacteric melons (El Hadi et al., 2013). Compared to the controls, the esters content was dramatically reduced in nor-1 and the content of aldehydes was increased, which explain the green, fresh aroma of these fruits. The nor-1 VOCs profile was more similar to unripe melons or to non-climacteric melons such as inodorus types (Mayobre et al., 2021) than to the VED profile. These results suggest that CmNAC-NOR could be involved in the regulation of the AAT genes, which are known to be ethylene-dependent and are responsible for volatile ester formation (El-Sharkawy et al., 2005; Cao et al., 2021). As expected, when we measured the expression of CmAAT1, we observed that it was downregulated in the nor-1 mutant, compared to the control lines and the nor-3 mutant. This is also consistent with a recent study that reported that the NAC transcription factor PpNAC1 (with homology to SlNAC-NOR) regulates fruit flavor ester biosynthesis in peach by activating PpAAT1 expression (Cao et al., 2021). However, to demonstrate that CmNAC-NOR directly binds CmAAT1, further experiments are needed.

An unexpected phenotype of nor-1 was that seeds were not well developed (Figure 4C). This phenotype has not previously been reported in the tomato NOR mutant or in other NAC genes in species such as peach (Pirona et al., 2013), apple (Yeats et al., 2019), and strawberry (Martín-Pizarro et al., 2021). However, there are some reports suggesting that NAC transcription factors regulate seed development and play a role in seed germination (Kim et al., 2008; Park et al., 2011; Wang et al., 2021; Liu et al., 2022). In a recent study, knock out of the ClNAC68 gene in watermelon delayed seed maturation and germination, but the germination rate was not affected (Wang et al., 2021), suggesting that there are additional NAC genes with diverse functions that regulate seed development. In tomato, NOR-like1 has been shown to be a positive regulator of fruit ripening; CRISPR/Cas9 mutants of NOR-like1 delayed fruit ripening and seriously affected seed development, reducing the number and weight of seeds, which showed poor germination (Gao et al., 2018). However, the target genes associated with seed development are still unknown. Only a recent study in grape showed that the NAC domain gene VvNAC26, which positively regulates ethylene and ABA-related genes to influence seed and fruit development, interacts with the transcription factor VvMADS9 (Zhang et al., 2021).

Non-climacteric melon cultivars as Piel de Sapo (PS), a variety belonging to the inodorous group in the melo subspecies, produce low amount of ethylene, insufficient to trigger the climacteric response and do not abscise when ripe. Climacteric varieties as VED, a variety from the cantaloupensis group in the melo subspecies, show a typical climacteric fruit ripening behavior, with a sharp ethylene peak and noticeable related climacteric traits as abscission layer formation at around 35 DAP. There are no significant differences between both types in soluble solid content nor firmness of the flesh (Pereira et al., 2020), but the fruit volatiles produced by both types are different (Mayobre et al., 2021). Three climacteric QTLs involved in fruit ripening have been characterized in melon, ETHQB3.5, ETHQV6.3, and ETHQV8.1 (Vegas et al., 2013; Pereira et al., 2020). So far, only the causal gene for ETHQV6.3 has been identified (CmNAC-NOR, Ríos et al., 2017). Introgression lines carrying the climacteric allele of each of the three QTL in the non-climacteric PS background are able to induce a very mild climacteric response (Vegas et al., 2013; Pereira et al., 2020). However, when combined in pairs or the three of them together, they interact epistatically, producing a dramatic climacteric effect in the non-climacteric background (unpublished). These data suggest that the non-climacteric PS may be impaired in ethylene production due to variations in more than one gene, and that the combination of two or more genes is necessary to rescue the typical climacteric response. In addition, the non-climacteric allele of ETHQV8.1 in the VED background delays ripening but does not result in a strong non-climacteric phenotype (Pereira et al., 2020). A plausible hypothesis is that at least these three genes/QTL are responsible of the ripening differences between non-climacteric melons from the inodorus group and the climacteric cantaloupe type. However, the complex molecular mechanisms cannot be yet understood until the causal genes of the other two QTL are identified. Still, we cannot rule out that additional genes are responsible for conferring a non-climacteric response in other non-climacteric melon types phylogenetically distant from PS.

In this study, we provide evidence that supports CmNAC-NOR as a key player in regulating climacteric fruit ripening in melon. As a master regulator, CmNAC-NOR independently mediates many ripening-associated traits. Our findings also suggest that CmNAC-NOR can be a potential target in breeding programs to modulate fruit maturation and shelf life in melon.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

BL did the experimental work and data analysis and wrote the original draft of the manuscript. MS grew the plants, phenotyped the ripening behavior, and measured the ethylene content. CM performed the expression and volatile experiments. AM-H supervised the CRISPR-Cas9 and genetic transformation experiments. MP and JG-M designed and supervised the work and reviewed and edited the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by grants AGL2015–64625-C2–1-R and RTI2018-097665-B-C2 funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe,” the Severo Ochoa Programme for Centres of Excellence in R&D 2016–2010 (SEV-2015-0533) funded by MCIN/AEI/10.13039/501100011033, the CERCA Programme/Generalitat de Catalunya and 2017 SGR 1319 grant from the Generalitat de Catalunya to JG-M. BL was also supported by grants from Youth Project of National Natural Science Foundation of China (31902035), The International Postdoctoral Exchange Fellowship Program of China (20170053), and a postdoctoral grant from the Severo Ochoa Programme for Centres of Excellence in R&D 2016–2010 (SEV-2015-0533). MS was supported by a grant BES-2017-079956 funded by MCIN/AEI/ 10.13039/501100011033 and by “ESF Investing in your future”. CM was supported by FI grant from the Secretaria d’Universitats i Recerca del Departament d’Empresa i Coneixement de la Generalitat de Catalunya and the co-funding of the European Social Fund (ESF)—“ESF is investing in your future.”

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the research reported in this paper.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.