Tocopherols are lipid-soluble isoprenoids of the tocochromanol family which are mainly synthesized in photosynthetic organisms (Fritsche et al., 2017; Mène-Saffrané, 2017). Their chemical structure consists of a polar chromanol ring, originated from homogentisate (HGA), and a lipophilic isoprenoid side chain derived from a specific prenyl pyrophosphate donor. While HGA is the common precursor for all tocochromanols, the polyprenyl precursor varies according to the type of tocochromanol and is phytyl pyrophosphate (PPP) for tocopherol synthesis (Mène-Saffrané, 2017). Additionally, according to the position and degree of methylation of the chromanol ring, four natural sub-forms can exist: δ- (one methyl group), β- and γ- (two methyl groups), and α-tocopherol (three methyl groups). Tocopherols are of great importance because, together with tocotrienols, they are the only natural compounds exhibiting vitamin E activity in animal cells and are essential as dietary nutrients (Traber and Sies, 1996). In plants, tocopherols play diverse physiological functions, of which their role as antioxidants, either scavenging peroxyl radicals or quenching reactive oxygen species, is probable the most notable (Havaux et al., 2005; Mène-Saffrané et al., 2010). Nonetheless, other functions have been recently described for tocopherols in plants, including their involvement in photo-assimilate transport, carbohydrate metabolism, cellular signaling and plant’s response to biotic and abiotic stresses (Falk and Munné-Bosch, 2010; Muñoz and Munné-Bosch, 2019; Ma et al., 2020b).

The tocopherol biosynthesis pathway (Figure 1) has been well-characterized in the last decades, with all the vitamin E biosynthetic genes (VTE genes) encoding the enzymes catalyzing the core steps of tocopherol synthesis identified (Fritsche et al., 2017; Mène-Saffrané, 2017; Muñoz and Munné-Bosch, 2019). Tocopherol synthesis is initiated by the condensation of HGA with PPP, a reaction regulated by homogentisate phytyl transferase (HPT; VTE2) that results in the formation of 2-methyl-6-phytyl-1,4-benzoquinol (MPBQ) (Collakova and DellaPenna, 2001). After this step, the pathway can split into two branches depending on the subsequent reaction of MPBQ, leading toward the synthesis of δ- and β-tocopherol or γ- and α-tocopherol. Then, MPBQ can either be directly cyclized into δ-tocopherol, by the tocopherol cyclase (TC; VTE1), or it can be first methylated into 2,3-dimethyl-6-phytyl-1,4-benzoquinol (DMPBQ), by a MPBQ methyltransferase (MPBQ-MT; VTE3). The resulting product DMPBQ is then converted into γ-tocopherol by the same TC (VTE1) mentioned before. The final step of tocopherol synthesis is the methylation of δ- and γ-tocopherol into β- and α-tocopherol, respectively, which is catalyzed by the γ-tocopherol methyltransferase (γ-TMT; VTE4). These steps represent the tocopherol-core pathway, and of these enzymes only HPT (VTE2) is considered exclusive to tocopherol synthesis, while MPBQ-MT (VTE3), TC (VTE1), and γ-TMT (VTE4) are also involved in the synthesis of the other tocochromanols (Fritsche et al., 2017; Mène-Saffrané, 2017). VTE2 has been reported to be a limiting step in tocopherol synthesis in seeds and leaves of Arabidopsis and other model plants (Savidge et al., 2002; Collakova and DellaPenna, 2003a). However, it does not seem to limit tocopherol synthesis in fruit of species such as tomato, olive, and mandarin (Quadrana et al., 2013; Georgiadou et al., 2019; Rey et al., 2021a), where VTE3 appears to play a more important role regulating tocopherol content (Quadrana et al., 2013; Rey et al., 2021a).

Tocopherol content in plant cells is also highly dependent on the availability of the precursors HGA and PPP (Mène-Saffrané, 2017; Pellaud and Mène-Saffrané, 2017). In plants, the precursor HGA originates in a two-step reaction from the degradation of the amino acid L-tyrosine (L-Tyr) synthetized in the shikimate (SK) pathway. L-Tyr is converted into 4-hydroxyphenylpyruvate (HPP) by a tyrosine aminotransferase (TATs), and then transformed into HGA by a HPP dioxygenase (HPPD) (Figure 1). The loss of function of either TAT or HPPD in Arabidopsis has resulted in a reduction of tocopherol levels (Norris et al., 1998; Riewe et al., 2012), while the overexpression of these genes only modestly increased total tocopherols in leaves and seeds (Tsegaye et al., 2002; Karunanandaa et al., 2005). In fleshy fruit, such as tomato and mango, the role of HPPD in tocopherol accumulation has been reinforced, as higher accumulation of HPPD transcripts has been associated with genotypes containing high tocopherols levels (Quadrana et al., 2013; Singh et al., 2017). Recently, it has been shown that engineering the chorismate–tyrosine pathway in tomato fruit to produce HPP in combination with the overexpression of Arabidopsis HPPD resulted in a moderate increment in tocopherols (Burgos et al., 2021). The other precursor necessary for tocopherol synthesis, PPP, can be derived from geranylgeranyl pyrophosphate (GGPP), produced in the MEP pathway, or alternatively from the recycling of phytol formed in the degradation of chlorophylls (Figure 1). Two enzymes are involved in the supply of PPP via the recycling of phytol: phytol kinase (VTE5) and phytyl phosphate kinase (VTE6) (Valentin et al., 2006; vom Dorp et al., 2015). VTE5 appears to be a key enzyme regulating tocopherol content in fruits of tomato and olive, controlling the supply of PPP toward the tocopherol-core pathway (Georgiadou et al., 2015; Almeida et al., 2016). Nonetheless, in ripe tomato and mandarin fruits, tocopherol accumulation appears to be more influenced by the up-regulation of genes of the MEP pathway, such as DXS and GGDR (Quadrana et al., 2013; Almeida et al., 2015; Gramegna et al., 2019; Rey et al., 2021a). While the gene DXS regulates the influx into the MEP pathway, GGDR controls the final reduction of GGPP into PPP, and therefore its final availability for condensation with HGA (Estévez et al., 2001; Pellaud and Mène-Saffrané, 2017).

Tocopherol accumulation has been mainly studied in leaves and seeds, but they have also been detected in fruits, stems, roots, flowers, and other plant tissues (Horvath et al., 2006a). In general, α-tocopherol is the main form found in leaves, while γ-tocopherol is the predominant in seeds of most species (Horvath et al., 2006a). In fruits, α-tocopherol seems to be the main isoform accumulated (Chun et al., 2006; Horvath et al., 2006a), with variable contents depending on the species and through fruit ripening (Osuna-García et al., 1998; Almeida et al., 2011; Quadrana et al., 2013; Georgiadou et al., 2015, 2019). In recent years, great advances have been made into the regulation of tocopherol biosynthesis in fruit of different species, such as tomato (Almeida et al., 2011, 2016; Quadrana et al., 2013; Gramegna et al., 2019; Burgos et al., 2021), pepper (Arango and Heise, 1998; Koch et al., 2002), olive (Georgiadou et al., 2015, 2016, 2019), and mango (Singh et al., 2017). These studies concluded that tocopherol accumulation in fruit is mainly transcriptionally regulated, and that tocopherol biosynthetic genes are modulated in a temporal manner, and also influenced by environmental factors (Almeida et al., 2011, 2020; Quadrana et al., 2013; Georgiadou et al., 2016, 2019; Singh et al., 2017; Gramegna et al., 2019).

Citrus is one of the world’s most important fruit crops, being highly demanded for fresh consumption and juice processing (Spreen et al., 2020). This genus is characterized for its genotypic and phenotypic diversity (Wu et al., 2018), with a vast range of fruits with a different composition of nutrients and bioactive compounds (Rodrigo and Zacarías, 2006; Cano et al., 2008; Ma et al., 2020a). Current information about tocopherol contents in Citrus fruit is very limited. Tocopherols accumulate in the flavedo (external colored part of the peel) of mandarin, orange, lemon, grapefruit, and other less known species in the range of 65–130 μg g^–1, and mainly in the forms of α- and γ-tocopherol, with composition varying according to the specie (Assefa et al., 2017; Rey et al., 2021a, b). In the pulp, tocopherols have also been detected in fruit of mandarin, grapefruit, and orange in the form of α-tocopherol and at concentrations lower than in the flavedo (1.6–25 μg g^–1) (Chun et al., 2006; Cardeñosa et al., 2015). Recently, genes involved in the tocopherol-core pathway as well as genes regulating the supply of the precursors HGA and PPP have been identified in the Citrus genome. Analysis of their transcriptional profiling in the flavedo of mandarins and grapefruits in response to cold stress suggested candidate genes and mechanisms regulating tocopherol accumulation in each specie (Rey et al., 2021a, b). However, the temporal changes of tocopherols and the regulation of their synthesis during maturation of Citrus fruit have not been explored yet. Therefore, the aim of the present work was to perform a comparative study of tocopherol accumulation during on-tree fruit maturation in different Citrus genotypes belonging to the main horticultural Citrus groups: orange, mandarin, lemon, and grapefruit. The relationship between tocopherol accumulation and the expression of tocopherol-biosynthetic genes in the flavedo and pulp of fruit from the selected species and ripening stages was also investigated.

Taking advantage of the genetic and phenotypical diversity of the genus Citrus, the aim of this work was to investigate the changes in tocopherol contents and their regulation during maturation of fruit of four Citrus species belonging to the main horticultural groups: grapefruit, lemon, sweet orange, and mandarin. Tocopherols were identified in all the selected Citrus species throughout maturation, with contents varying between tissues, maturation stages, and genotypes (Figures 2, 3 and Supplementary Tables 3, 4). The tocopherol profile was in agreement with previous reports, with α- and γ-tocopherol being the main forms detected in Citrus fruit (Assefa et al., 2017; Rey et al., 2021a, b). However, predominance of one form or another seems to be dependent on the Citrus specie, as similar or higher γ-tocopherol contents have been detected in less common Korean Citrus genotypes (Assefa et al., 2017). In fruit of other species, the prevalence of α- or γ-tocopherol is also specie-specific and, while α-tocopherol is the main form in tomato, pepper, mango, grape, olive, and avocado (Koch et al., 2002; Horvath et al., 2006b; Quadrana et al., 2013; Singh et al., 2017; Georgiadou et al., 2019; Vincent et al., 2020), γ-tocopherol accumulates at higher concentrations in zucchini and raspberry fruit (Carvalho et al., 2013; Rodov et al., 2020). In our experimental conditions, β-tocopherol was not detected, and δ-tocopherol was only identified in some samples but at levels below the limit of quantification, indicating that the δ-/β-tocopherol branch may not have a significant contribution to the tocopherol pool in citrus fruit.

Many physiological and metabolic processes are independently regulated in the flavedo and pulp of Citrus fruit (Tadeo et al., 2020), and this seems to be the case for tocopherol metabolism as contents were higher in the flavedo than the pulp (Figures 2, 3). Similarly, higher concentrations of other bioactive compounds have been reported in the flavedo than in the pulp of citrus (Alquézar et al., 2008, 2013; Alós et al., 2014; Assefa et al., 2017). A possible explanation for this is that the direct exposure of the flavedo to environmental stresses could lead to a higher demand for antioxidants to cope with these adverse conditions. In relation to this, light could play a relevant role either as a stress factor or by its direct impact in the regulation of tocopherol biosynthesis. Tocopherols are involved in the photo-protection of plants (Spicher et al., 2017; Ma et al., 2020b), and increases in contents have been reported in response to high light stress (Collakova and DellaPenna, 2003b; Havaux et al., 2005). Furthermore, a positive role of light in the transcriptional regulation of certain tocopherol biosynthetic genes has been proposed in grapefruit (Rey et al., 2021b) and tomato (Gramegna et al., 2019).

The differences detected between fruit tissues were not only quantitative but also in the pattern of accumulation during maturation (Figures 2, 3), suggesting that different mechanisms may operate in the regulation of tocopherol synthesis in the flavedo and pulp of Citrus fruit. The differential accumulation of tocopherols between citrus fruit tissues may also be related to their chlorophyll contents, as an alternative source of PPP is through the recycling of free phytol formed during the degradation of chlorophylls (Figure 1; Valentin et al., 2006; Georgiadou et al., 2015; vom Dorp et al., 2015; Almeida et al., 2016). While chlorophyll content in the pulp of Citrus fruit is negligible or only detected in IG fruit, concentrations in the flavedo are high throughout development and decrease notably with the transition of chloroplasts to chromoplasts (Alquézar et al., 2008, 2013; Lado et al., 2015, 2019). Therefore, the higher concentration of chlorophylls in the flavedo at immature stages and their degradation as ripening progresses could lead to a higher availability of free phytol to form PPP, higher influx into the tocopherol-core pathway and an enhancement of tocopherol contents. Accumulation of tocopherols concomitantly with chlorophylls breakdown during color change has been also described in fruit of other species like pepper (Osuna-García et al., 1998; Koch et al., 2002) and olive (Georgiadou et al., 2016).

The transcriptional analysis of tocopherol genes revealed that differences in tocopherol concentrations between tissues were associated with a generalized higher expression in the flavedo than in the pulp (Figures 4–7), which reinforces the hypothesis of the differential regulation of tocopherols accumulation between tissues. Moreover, transcriptional patterns during fruit maturation and correlation analysis pointed specific candidate genes that may regulate tocopherol levels in each tissue (Figures 4–8). In the flavedo, TAT1 and VTE4 showed a significant positive correlation to total, α- and γ-tocopherol contents (Figures 8A,B) suggesting the importance of these genes in regulating tocopherol accumulation in this tissue. TAT1 plays a major role in tocopherol synthesis by regulating HGA availability (Riewe et al., 2012), and an induction of this gene during senescence has been previously reported in Arabidopsis leaves, and associated with an increase in α- and γ-tocopherol content (Holländer-Czytko et al., 2005). On the other hand, the gene VTE4, which encodes for γ-TMT, plays a role in shaping tocopherol composition rather than increasing tocopherol contents (Bergmüller et al., 2003). Modifications in the expression of VTE4 have been successful in increasing α-tocopherol but in detriment of γ-tocopherol contents, and thus not altering total contents (Bergmüller et al., 2003; Collakova and DellaPenna, 2003a; Maeda et al., 2006). Therefore, in the flavedo of citrus the combined up-regulation of TAT1 and VTE4 during maturation could indicate a higher influx into the tocopherol core-pathway, due to a higher availability of HGA, which afterward is mostly converted into α-tocopherol by the increase in downstream VTE4. Still, it is important to keep in mind that many substrates and enzymes involved in tocopherol synthesis also participate in other metabolic pathways, such as HGA, and therefore the possible channeling to other pathways should be considered. Additionally, other genes, such as HPPD or GGPPS1, were also induced during ripening in specific citrus species and may also contribute to the increase of tocopherols, although they were not significantly linked to tocopherol contents in the network analysis (Figure 8B).

The expression of most genes involved in the regulation of PPP production tended to decrease in the flavedo during maturation with no apparent effect on tocopherol contents (Figures 2, 4A,B). Mutant plants of Arabidopsis in some of these genes (vte6, dxs, and ggpps11) have resulted in a reduction in tocopherol levels (Estévez et al., 2001; vom Dorp et al., 2015; Ruiz-Sola et al., 2016). Nonetheless, the expression of GGPPS6, the citrus orthologous gene of Arabidopsis GGPPS11, was significantly and negatively correlated with total and α-tocopherol levels, whereas VTE6 and DXS2 were negatively correlated with γ-tocopherol (Figure 8B). A down-regulation of DXS has been previously reported in the flavedo of grapefruit, orange, and mandarin linked to the onset of chlorophylls degradation at color break, rather than to the accumulation of other MEP-derived isoprenoids (Alós et al., 2006; Alquézar et al., 2008, 2013; Lado et al., 2015, 2019). Collectively, these results suggest that the supply of PPP by the recycling of free phytol or by de novo synthesis through the MEP pathway may not constrain tocopherol synthesis in the flavedo of Citrus during maturation.

In the pulp, most of the genes exhibited a similar expression tendency to that of the flavedo (Figures 4–7), although tocopherol concentrations were lower and followed contrasting temporal patterns (Figures 2, 3). A similar pattern of expression among genotypes was detected in the pulp for the genes TAT1, HPPD, VTE6, GGPPS1, and GGDR (Figure 6), whereas changes in the expression of the other genes followed a distinct profile depending on the specie (Figures 6, 7). Interestingly, TAT1 was induced during maturation in the four species, similarly to the changes detected in the flavedo but opposite to the pattern of tocopherol accumulation in the pulp, and HPPD was induced in the pulp and in the flavedo of grapefruit, orange, and mandarin (Figures 4C, 6C). An induction of genes of the SK pathway (TAT1 and HPPD) has also been observed with maturation in tomato and olive fruit but, as in the citrus pulp, did not mirror the changes in tocopherol contents (Quadrana et al., 2013; Georgiadou et al., 2015). This suggests that the up-regulation of these genes could be developmentally modulated and not necessarily associated to tocopherol contents in the pulp or in fleshy fruits. Since both TAT1 and HPPD are involved in the synthesis of the precursor HGA, which is not specific of tocopherol synthesis, it is likely that their induction is related to the synthesis of other tocochromanols which also use HGA as a precursor (Szymańska and Kruk, 2010; Kruk et al., 2014; Burgos et al., 2021). In relation to the other genes involved in tocopherol synthesis, the decrease in tocopherol contents observed in the pulp of grapefruit, orange, and mandarin was synchronized with the down-regulation of the genes VTE6, DXS2, GGDR, and VTE2 (Figures 6, 7). These results suggest that transcriptional regulation of these genes play a role determining tocopherol content in the pulp, which was supported by the correlation network results (Figures 8C,D). These genes, in particular those involved in the supply of PPP, have been previously proposed as limiting steps in tocopherol accumulation in the flesh of tomato (Quadrana et al., 2013; Almeida et al., 2015). In the pulp of lemon, expression of these genes varied from the other species but still reflected the minor changes detected in tocopherol accumulation during ripening in this specie. In the pulp, genes TAT1, HPPD, DXS1, GGPPS1, VTE1, and VTE4 showed a negative correlation with total contents or a specific tocopherol form (Figures 8C,D). Then, in this tissue, the down-regulation of genes involved in regulating PPP availability (VTE6, DXS2, and GGDR) may constrain tocopherol synthesis, and the up-regulation of other genes of the SK pathway (TAT1 and HPPD) and tocopherol-core pathway (VTE1 and VTE4) is not enough to compensate this limitation.

The network analyses in the flavedo and pulp also revealed genes that are co-regulated in both tissues during fruit maturation but in the pulp a higher number of genes were co-regulated compared to flavedo (Figures 8B,D). The gene VTE6 was positively linked to DXS2 and GGDR, all involved in the supply of PPP in both tissues. Other genes that seemed to be co-expressed in both tissues were those of the SK pathway with the genes involved in the last steps of the tocopherol-core pathway (Figure 8).

Finally, the differences in the accumulation of tocopherols detected among genotypes in each tissue (Figures 2, 3) were not clearly related to the expression of any gene (Figures 4–7; Supplementary Tables 5, 6), although some trends were observed in the flavedo. Focusing on differences in gene expression at early maturation stages, when differences among genotypes were already evident (Supplementary Table 3), a higher expression of GGDR and VTE3b was detected in the flavedo of lemon and orange than in the other genotypes (Supplementary Table 5), which could explain the higher contents in these two species. Higher expression of these genes has been previously reported in Citrus genotypes accumulating higher tocopherol concentrations (Rey et al., 2021a) and also in other fruit species as tomato (Quadrana et al., 2013; Gramegna et al., 2019).

This study addresses for the first time a comparative analysis of tocopherols accumulation and transcriptional regulation of their biosynthetic genes during fruit maturation of four Citrus species. Differences in tocopherol contents were detected between the flavedo and pulp, and also during maturation and genotypes. Concentration of tocopherols in the flavedo were between 2 and 50 times higher than in the pulp, and this was associated with higher expression levels in the flavedo of most genes involved in the precursors PPP and HGA synthesis, and the tocopherol core-pathway. Moreover, tocopherols increased in the flavedo with maturation while they tended to decrease in the pulp, and correlation and network analysis allow to suggest candidate genes regulating tocopherol accumulation in each tissue. In the flavedo, the increase in tocopherol contents mirrored a marked up-regulation of TAT1 and VTE4, while contents in the pulp may be limited by the expression of VTE6, DXS2, and GGDR, regulating PPP availability. Furthermore, the genes TAT1 and VTE4, HPPD and VTE1, and VTE6 with DXS2 and GGDR, were co-regulated and shared a similar pattern during maturation in both tissues, suggesting they are developmentally modulated. Finally, the differences among genotypes were not clearly correlated with the expression of specific genes, indicating the involvement of other regulatory mechanisms modulating differences among species.

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

MR and LZ: conceptualization, supervision, and funding acquisition. FR, MR, and LZ: methodology and writing – review. FR: experimental work, formal analysis, data curation, and writing – original draft preparation. All authors have read and agreed to the published version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

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.