Pleiotropic properties of GOLDEN2-LIKE transcription factors for crop improvement

Abstract

Crop improvement can be affected by enhancing the efficiency of photosynthesis-associated bioprocesses such as chlorophyll biosynthesis, chloroplast biogenesis, the functioning of photosystems including light-harvesting complexes, and carbon fixation. To achieve this, the GOLDEN2-LIKE (GLK) transcription factors represent promising targets since they play a positive role for greening traits in diverse plants. To scrutinize the pleiotropic impact of GLKs, we summarized all phenotypic traits reported in functional studies that used transgenic approaches to lose or gain gene functions. Additionally, we also discussed altered plant phenotypes with respect to their physiological–biochemical aspects and environmental stress responses. From these results, we conclude that GLKs consistently increase chlorophyll biosynthesis, enhance chloroplast division, and increase photosynthetic rate. They individually influence other traits including yield, phytochemical accumulation, and biotic and abiotic stress resistance. Collectively, GLKs have potential as key regulators to effect increases in overall agricultural quality across plant species. This suggests that they may be among the most promising target genes for future agro-biotechnology applications.

Introduction

To increase crop yield and quality, the greening trait during plant vegetative stages, has been used to enhance chlorophyll biosynthesis, chloroplast development, photosynthetic efficiency, and resistance against biotic and abiotic stresses. More precisely, direct regulation of target genes involved in greening-related mechanisms is the primary strategy. To achieve this, several transcription factors (TFs) have received significant attention as key regulators. These include two positive GOLDEN2-LIKE (GLK) and ELONGATED HYPOCOTYL5 (HY5) TFs as well as negative TFs such as PHYTOCHROME INTERACTING FACTORS (PIFs). GLKs are also known as master regulators of chloroplast biogenesis by positively coordinating the expression of photosynthesis-associated nuclear genes (PhANGs) in many plants, including Arabidopsis (Arabidopsis thaliana), moss (Physcomitrella patens), rice (Oryza sativa), tomato (Solanum lycopersicum), hybrid birch (Betula platyphylla × B. pendula), peanut (Arachis hypogaea), and liverwort (Marchantia polymorpha) [1,2,3,4,5,6,7,8,9]. From these results, the influence of GLKs on the transcriptional levels of PhANGs appears to be widespread among plant species.

Historically, in 1912, the name “GOLDEN (G)” originated from the golden color phenotype of maize (Zea mays) mutant that behaved recessively according to simple Mendelian inheritance [10]. This first mutant was named as golden1 (g1) following discovery of a second mutant, golden2 (g2), which also showed a golden plant color phenotype in maize [11]. Seventy years later, G2 was cloned as a novel transcriptional regulator for chloroplast development and its hypothesized function was to maintain C₄ development due to bundle sheath defects found in the leaves of maize g2 mutants [12]. Three years later, the name “GOLDEN 2-LIKE (GLK)” was given to the new member of the same family in maize, ZmGLK1, which was then distinguished from G2, as well as two rice orthologs, OsGLK1 and OsGLK2, which both showed higher amino acid sequence relatedness to ZmGLK1/ZmG1 than G2 [13]. This finding supported the notion that GLKs function redundantly in mesophyll cells for common C₃ photosynthesis but G2 acts in a bundle sheath cell-specific manner for C₄ photosynthesis.

GLKs, which belong to a subgroup of the GARP (GOLDEN2 in Zea mays; ARR-B, in Arabidopsis thaliana; Psr1 in Chlamydomonas reinhardtii) superfamily, show strong structural conservation. Each consists of a nuclear localization signal, a DNA-binding domain (DBD), AREAEAA hexapeptide, a proline-rich domain, and a GLK/C-terminal (GCT) box [13, 14]. The DBD is a MYB-related B domain (~ 60 amino acids in length) that conserved in GARP superfamily with three α-helices. The consensus sequence of the AREAEAA hexapeptide motif, which immediately follows the third helix of the B domain confers specific characteristics to GLKs that distinguish them from other GAPR members [1, 15]. In addition, the GCT is also present in other members of the GARP family including PSEUDO RESPONSE REGULATOR2 proteins [13, 16]. Interestingly, ancestral GLK orthologs emerged in some green algae species in which the GCT box was absent. Later, GLKs with a GCT box diversified with one, two, or more copies in most land plants from bryophytes to angiosperms, with the exception of the obligate parasite Sapria himalayana, which completely lacks photosynthetic ability [9, 15, 17]. Taken together, these evidences strongly support the key function of GLKs for chloroplast development.

To date, evidence supports the notion that GLKs coordinate internal transcriptional networks to ultimately incite chloroplast development by sensing and reacting to signals from diverse sources, including light, plastid retrograde signals, phytohormones, senescence, and biotic and abiotic stresses [15, 17]. Moreover, the direct regulation of gene expression by interaction by GLKs with the promoters of target genes has been confirmed for chlorophyll biosynthesis and photosynthetic apparatus in Arabidopsis, peanut, and hybrid birch [4, 7, 8, 18]. Furthermore, other cases of direct promoter binding by GLKs has been reported for Camellia sinensis with respect to two catechin biosynthetic genes and in Arabidopsis for three B-BOX domain proteins related to flowering time as well as four ABA-signaling genes [19,20,21]. Meanwhile, miscellaneous protein partners that are known to interact with GLKs have been reported for a wide range of traits in various plants: tomato for senescence and fruit development, peach (Prunus persica) for auxin signaling, apple (Malus domestica) for chloroplast development, peanut for water stress, Arabidopsis for cell death, brassinosteroid signaling, cotyledon opening, and photosynthesis, and radish (Raphanus sativus) for chlorophyll biosynthesis [8, 22,23,24,25,26,27,28,29,30,31]. The fact that GLKs are involved in diverse molecular mechanisms supports the idea that they may improve overall photosynthetic performance and other important traits such as development, growth, and environmental response.

In this review, we briefly summarize phenotypic influences of GLKs on diverse plants as identified by transgenic knockout, functional complementation, or overexpression approaches. In doing so we focus on the physiological–biochemical aspects of plants. Moreover, we also consider the effects of GLKs on resistance to biotic and abiotic stress following loss- or gain-of-function experiments. This review will advance our understanding of the useful features of GLKs and unveil their great potentials with pleiotropic impacts for crop improvement. Finally, we suggest prospects for their application to substantial crop improvement projects.

Greening trait improvement

Plants are autotrophic organisms and therefore depend on light to sustain life functions. Photosynthesis requires a chlorophyll pigment and a chloroplast organelle to perform photosynthesis at scale when exposed to light. Nevertheless, their regulation may occur by specific regulatory mechanisms within the plant body. Here, we categorized recent research of the GLK's impacts on plant greening traits into three major factors, chlorophyll biosynthesis, chloroplast development, and photosynthesis.

Chlorophyll biosynthesis

Chlorophylls, which include both chlorophyll a and chlorophyll b, are photosynthetic pigments that cause plants to appear green. In general, chlorophyll levels are the strongest indicator of plant greening traits. Therefore, chlorophyll content has been considered as a major characteristic during reverse genetic studies of GLKs.

Since a paler yellow–green (“golden”) color was observed in a maize g2 mutant plant [1, 11, 12, 18, 32], loss-of-functional studies involving diverse plant GLKs have been performed using diverse biotechnologies including transposon insertion, T-DNA insertion, RNA interference, CRISPR-Cas9-based genome editing, homologous recombination, virus-induced gene silencing, and gene cosuppression (Table 1). Knock-out (KO) phenotypes of GLKs were first reported in Arabidopsis, which has two GLK genes, and these were found to display a normal green leaf color in both single KO mutants but a pale green in a double KO mutant [1]. Correlated with these color traits, chlorophyll content was significantly decreased in the leaves, siliques, and seedlings in the double mutant but not in either single mutant, thereby suggesting the functional redundancy of two GLKs in most photosynthetic tissues [4, 18, 32,33,34]. In cases of moss and rice, both of which have two GLK copies, greening-related plant colors and chlorophyll levels of wild type and single mutant plants were indistinguishable, but the double mutant was paler green and showed decreased chlorophyll content; this finding supports the same functional redundancy in green tissues such as moss protonema, gametophores, and rice leaves [2, 35]. Furthermore, single knock-down (KD) and KO mutations have been used to decrease the chlorophyll content of the target organs of several plants, including the hairy roots of peanut, the leaves of hybrid birch and hybrid poplar, the gemmae of liverwort, and the leaves and unripe fruits of peach and tomato. In each case, paler green color traits were observed in GLK-suppressed transgenic plants than the wild type, which suggests that GLKs have a substantial impact on chlorophyll biosynthesis [6,7,8,9, 23, 29, 36].

Table 1 Physiological and biochemical phenotypes of loss-of-function GLK mutants and functional complementation results in a double KO Arabidopsis GLK mutant

To assess the role of GLKs from different plants (i.e., Arabidopsis, moss, peanut, apple, and peach), complementation assays have been performed with the pale green Arabidopsis double mutant (atglk1atglk2). For these tests, the chlorophyll content was measured to determine whether other GLKs could functionally rescue native GLK function (Table 1). First, the chlorophyll deficiency of an atglk1atglk2 leaves was found to be complemented fully by AtGLK1 and partially by PpGLK1 (moss), suggesting that there is both functional conservation and divergence between these two plant GLKs [2]. Individual complementation of AtGLK1 and AtGLK2 also showed complete rescue of the double mutant phenotype in leaves and siliques overexpressed using the 35S promoter [4, 32]. In this study, changing the promoter to facilitate differential tissue-specific GLK expression resulted in different degrees of complementation associated with different leaf phenotypes. For example, full rescue was achieved using the photosynthetic cell-specific RbcS promoter, partial complementation was achieved using the phloem-specific AtSUC2 promoter, and no complementation was achieved using the epidermis-specific FDH promoter. Moreover, phenotypic rescues of an atglk1atglk2 were successfully performed using overexpression of heterogeneous GLKs under the 35S promoter; this was observed in leaves by AhGLK1 (peanut), in leaves and seedlings by PpGLK1 (peach), and in seedlings and siliques by MdGLK1 (apple) [18, 23, 24]. Additionally, a T-DNA yellow-green leaf (yl) mutant of hybrid birch was successfully rescued with respect to its greening color and chlorophyll level via constitutive overexpression of BpGLK1 (birch) [7]. Taken together, these results confirm that GLKs can affect chlorophyll content when expressed in photosynthetic cells, and that this mainly occurs in a cell-autonomous manner.

Next, to reveal the effects of GLKs on plant physiological characteristics, overexpression studies have used GLKs from diverse plants (Table 2). First, ectopic expression of AtGLK1 or AtGLK2 was performed using four different promoters in a uniform ripening (u)/u tomato variety that exhibits a light green fruit phenotype. Here, a defect in U encoding SlGLK2 resulted in a normal green color for all transgenic tomato plants [5]. Interestingly, in unripe fruits, promoters expressed before ripening—including 35S, RbcS, and LTP—resulted in increased chlorophyll levels and a dark green color. However, a PDS promoter expressed later during fruit development showed no change. This confirmed that the influence of GLKs is apparent only while greening processes are operating. In addition, AtGLK1 or AtGLK2 have been constitutively overexpressed in Arabidopsis using the 35S promoter in multiple studies [33, 34, 37, 38]. In these studies, transgenic Arabidopsis plants exhibited slightly darker green color throughout the plant and showed increased chlorophyll content in examined target organs, including roots and seedlings.

Table 2 Summary of altered physiological and biochemical phenotypes following overexpression of GLKs

Meanwhile, ectopic overexpression of native GLKs has been performed in diverse plant systems. A rice FOX (i.e., a full-length cDNA overexpressor)-mutant designed to activate OsGLK1 showed a good correlation between greening levels and chlorophyll content in calli, roots, and young seedlings while appearing a normal green color in photosynthetic organs [3]. Additionally, overexpression of two tomato genes, SlGLK1 and SlGLK2, increased the chlorophyll content of dark green immature fruits present on normal green tomato plants [6, 29]. Furthermore, overexpression studies of native GLK genes in hybrid birch (BpGLK1), peanut (AhGLK1), hybrid poplar (PabGLK5), and liverwort (MpGLK) all resulted in enhanced chlorophyll levels in target organs, including leaves, hairy roots, leaves, and gemmae, respectively [7,8,9, 36]. In addition, four GLK genes of peach (PpGLK1), kiwifruit (AchGLK1), and tea plant (CsGLK1 and CsGLK2) have been constitutively overexpressed using the 35S promoter in tomato host plants. All transgenic fruits exhibited a darker green color and increased chlorophyll content in unripe stages [23, 39]. Moreover, two tea genes were found to be able to increase chlorophyll levels even in leaves [20]. In cereal crops, similar results were found for two maize GLK genes that were individually overexpressed in rice plants under constitutive promoters such as maize Ubi (ZmUbi) and 35S, as well as both individually and together under an endogenous promoter (i.e., ZmG1 and ZmG2, respectively) [40, 41]. For example, all transgenic rice plants showed an increase in chlorophyll content along with a darker green color in their leaves and florets. This evidence strongly supports the potential of maize GLKs to improve the greening traits of rice, which may result in higher productivity. As mentioned so far, GLK overexpression generally leads to increased chlorophyll content when expressed at the time and place where the photosynthetic mechanism operates; this can be coordinated by using an appropriate promoter. These results clearly indicate that the greatest impact of GLKs may be on chlorophyll biosynthesis.

Chloroplast development

Chloroplasts are plant-specific organelles that perform photosynthesis using thylakoids, a novel structure equipped with photosynthetic apparatus as well as pigments such as chlorophylls and carotenoids [42]. Since GLKs have been found to regulate chloroplast development in Arabidopsis [1, 32], the biogenesis and development of chloroplasts has been considered as a key trait in loss-of-functional studies of GLKs (Table 1). In Arabidopsis leaves, the disordered chloroplast thylakoid ultrastructure of KO mutants was first observed chloroplast phenotype in photosynthetic mesophyll or bundle sheath cells. Here, the authors demonstrated functional redundancy between AtGLK1 and AtGlLK2, both which were less abundant in the thylakoids of a double mutant. However, this was not observed in either single mutant [1, 32]. In the double mutant, the chloroplast number per cell was normal but the size was reduced. Notably, the authors also identified rudimentary thylakoid lamellae as shown on the chloroplasts of the maize g2 mutant [33]. Later, a single KO mutant, atglk2, showed a slightly smaller chloroplast size, lower staking degree, and lower abundance of thylakoid components in Arabidopsis seedlings [34]. Furthermore, much reduced thylakoid-membrane networks and grana stacking in the leaf chloroplasts were reported in T-DNA insertional KO and RNAi-mediated KD mutants of BpGLK1 in hybrid birch plants [7]. A CRISPR/Cas9-mediated KO mutant (MpGLK) in liverwort gemmae showed slightly increased chloroplast number but significantly decreased chloroplast size and scarce thylakoid stacking, suggesting overall inhibition of chloroplast development [9]. Similarly, in moss and rice, both of which have two GLK genes, a double mutant but neither single mutant showed impairment of granal formation in the protonema and gametophore of moss and rudimentary thylakoid development in the leaves of rice [2, 35].

Phenotypes related to chloroplast development have been observed in few cases of functional complementation studies (Table 1). For example, a defective granal phenotype of an atglk1atglk2 leaves was fully restored by RbcS::AtGLK1 expression and was partially restored by AtSUC2::AtGLK2 expression [32]. This suggests GLKs can act cell-autonomously in photosynthetic cells but also act non-cell-autonomously to a limited level of non-photosynthetic phloem companion cells to induce chloroplast development in adjacent mesophyll cells. The same atglk1atglk2 mutant was completely rescued to wild type levels as measured by the number of granal thylakoids in leaves and siliques by 35S::PpGLK1 expression [23]. In addition, the hybrid birch y mutant, which shows a much reduced thylakoid-membrane network was fully rescued by 35S::BpGLK1 expression, showing a well-developed chloroplast with closely stacked grana [7]. As summarized in Table 1, all phenotypic results from loss of function and functional complementation studies suggest that GLKs play a leading role in inducing chloroplast development in photosynthetic tissues and show functional redundancy in cases where two copies are present.

Overexpression studies of GLKs have been performed in tomato, Arabidopsis, hybrid birch, and rice (Table 2). All ectopic expressions of GLK genes in tomato were constitutively expressed using the 35S promoter in an u/u variety for AtGLK1 and AtGLK2 and common tomato for SlGLK1, SlGLK2, PpGLK1, AchGLK, CsGLK1, and CsGLK2 [5, 6, 20, 23, 39]. Transgenic tomato plants commonly showed enhanced chloroplast development in immature/unripe fruits, with increases in number, size, and density of chloroplasts as well as high accumulation in the grana stacking of thylakoids. Similarly, in Arabidopsis, increased chloroplast number and a highly stacked grana structure was formed by overexpression of either AtGLK1 or AtGLK2 in roots or AtGLK2 in seedlings [34, 37]. Meanwhile, constitutive overexpression of native GLK genes in hybrid birch showed little difference from the wild type in the granal stacks of leaf chloroplasts as well as a slight increase in chlorophyll levels, thereby suggesting a somewhat weak effect of BpGLK1 relative to other cases [7]. In rice, OsGLK1-FOX plants ectopically induced chloroplast biogenesis with well-developed thylakoids membranes and grana structures in calli (under light conditions), coleoptiles, and leaf sheaths of young leaves but in mature plants [3]. In contrast, two maize GLK genes, ZmG1 and ZmG2, were found to enhance chloroplast development, causing increased size, number, and significantly enhanced stacking of thylakoid membranes per chloroplast unit area in rice plants [41]. This finding confirmed that—unlike OsGLK1—both maize genes can promote chloroplast development even in mature rice plants. Thus, these results show that GLKs are key regulators of chloroplast biogenesis as well as chlorophyll biosynthesis.

Photosynthesis

Photosynthesis is a comprehensive biochemical reaction that takes place in chloroplasts. These organelles are equipped with photosynthetic apparatus, including photosystems and photosynthetic pigments such as chlorophyll [43]. Photosynthetic efficiency has traditionally been considered as a major feature for increasing crop yield. To assess whether GLKs may be useful for this purpose, in what follows we consider plant phenotypes related to photosynthetic efficiency.

As summarized in Table 1, the effect of GLKs on photosynthesis following loss of function was first observed in an Arabidopsis double mutant, atglk1atglk2. These plants showed a lower electron transport rate and PSII photochemical yield in the light. These phenotypes were restored comparable to the wild type levels when AhGLK1 was complementally overexpressed [18]. In hybrid birch, a defect in BpGLK1 was confirmed to affect photosynthesis in the yl mutant. This defect resulted in a significant decrease in net photosynthetic rate (Pn), stomatal conductance, transpiration rate, and initial fluorescence, but showed similar values with respect to maximal quantum efficiency of PSII (Fv/Fm), photochemical quantum yield of PSII (ΦPSII), and photochemical quenching. This result suggests that BpGLK1-KO can cause impairment in photosynthetic capacity but does not influence photosynthetic efficiency. Meanwhile, in hybrid poplar, the PaGLK RNAi lines have been found to show significantly higher Pn without changes in Fv/Fm relative to the WT [36]. This indicates that photosynthetic efficiency was not affected by suppression of PaGLK. Furthermore, in liverwort the photosynthetic apparatus was found to remain functional without changes in Fv/Fm in the gemmae of the gene-edited mpglk mutant despite a significant reduction in chlorophyll as well as the appearance of smaller chloroplasts with an altered ultrastructure [9]. Overall, these findings suggest that the influence of GLKs on photosynthesis more strongly determines the structural formation and maintenance of the photosynthetic apparatus than the efficiency itself.

The effects of GLKs on photosynthesis have been observed in numerous overexpression studies (Table 2). In Arabidopsis, ectopic expression of either AtGLK or AtGLK2 improved phototropic activity, the photoautotrophic growth of roots, and caused an increase in CO₂ fixation but not in leaves [37]. However, the Fv/Fm levels of these mutants were decreased in roots despite being unchanged in leaves. This finding suggests that the imbalance in the photosynthetic machinery decreases the efficiency of light utilization in the root chloroplasts, which have lower photochemical efficiency than the leaf chloroplasts. In another study, the overexpression of native PaGLK5 in hybrid poplar showed a significant reduction in Pn without changes in Fv/Fm relative to the WT [36]. The discrepancy between the decrease in Pn and the increase in chlorophyll content indicated that PaGLK5 does not directly affect photosynthetic efficiency. In rice, OsGLK1-FOX, which corresponds to the overexpression of OsGLK1, exhibited higher photosynthetic activity in unusually green calli [3]. This suggests that ectopic expression of native OsGLK1 enables photosynthesis in non-green (ivory-yellow) cells but is not able to increase greening traits including photosynthetic activity in the mature stages of photosynthetic parts of rice plants. However, individual constitutive overexpression of two maize GLK genes in field-grown plants led to improved light-harvesting efficiency via photosystem II with an accompanying increase in values such as stomatal conductance, intercellular CO₂ concentration, nP, ΦPSII, and Fv/Fm [40]. These results were further reproduced using two constitutive promoters (i.e., ZmUbi and 35S) and using two endogenous maize promoters (i.e., ZmG1 and ZmG2) in rice plants, respectively [41]. All transgenic rice plants—either ZmG1, ZmG2, or ZmG1/ZmG2—had higher photosynthetic rates and higher nP, stomatal conductance, mesophyll conductance, and light use efficiency in mature flag leaves of 4-month-old stages compared to wild type rice. Accordingly, it was proposed that these two maize GLK genes could synergistically improve the photosynthetic performance of rice even when under the control of native promoters.

Yield increase

As global food demand rapidly grows, yield increases are becoming more and more important. Over the past few decades, crop biotechnologists have tried to increase agricultural yield through technological advancements involving molecular biology and plant transformation. In maize, the contribution of genetic improvement was estimated at only 13% compared to 39% derived from agronomic improvement [44].

In Table 2, data related to the yield potential of GLKs is summarized. Rice is a major cereal crop along with corn and wheat and is consumed daily by about 50% of the global population [45]. Two studies reported evidence of yield increases in rice following ectopic overexpression of two maize GLKs [41]. One study showed a 30–40% increase in both vegetative biomass and grain yield. This was accompanied by increases in several agronomic traits such as panicle length, panicle weight, and seed number per panicle when a constitutive promoter was used [40]. This finding verified that these increases may be caused by enhanced photosynthetic capacity in field-grown transgenic plants but was not affected by the levels of hormones such as auxin, cytokinin, and gibberellin due to their statistical difference relative to the WT. The other study showed clear positive effects of maize GLKs on grain yields in transgenic rice plants that overexpressed either ZmG1 or ZmG2 (or both) using constitutive promoters (i.e., ZmUbi or 35S) or the maize ZmG1 and ZmG2 promoters [41]. Most rice plants showed increases in shoot biomass and grain yield traits including total panicle number and total grain weight per plant. In particular, the highest grain yields occurred when both maize GLKs were simultaneously expressed under endogenous promoters, and resulted in a 70% increase in grain yield. However, one rice plant overexpressing ZmG1 under the ZmUbi promoter produced smaller seeds without showing an apparent yield increase. These findings prompt us to suggest caution; very high levels of expression using strong constitutive promoters may lead to negative effects on seed development. However, this finding did not agree with those of a previous study, which showed that overexpression of either ZmG1 or ZmG2 using the same ZmUbi promoter resulted in similarly high grain yields (i.e., best performance increases of 56% and 118% relative to the wild type) during a single season in Hainan. This study was performed using a japonica-type Kitaake cultivar but the latter study used a japonica-type TNG67 cultivar. This difference suggests the possibility that different results may be obtained depending on the rice host variety [40, 41]. Meanwhile, endosperm-specific overexpression of OsGLK1 has been found to increase grain yield by approximately 20% as well as increase the values of several components including filled grain per panicle, the seed-setting rate, and grain yield per plant in the conventional rice variety ZH11 [46]. However, this mutant also showed deterioration in rice grain quality, showing seriously increased chalkiness due to spherical, loosely packed, and irregularly polyhedron-shaped grains.

Phytochemical effects

Phytochemicals refer to organic compounds that are naturally produced by plants. They confer self-defense functions against disease and environmental change and supply important health benefits for many heterotrophic organisms. Phytochemicals mainly include terpenoids such as carotenoids, phytosterols, and saponins, as well as polyphenols such as anthocyanins and other flavonoids [47]. Of these, the levels of carotenoids and flavonoids have been reported to be influenced by GLKs. Here, we comprehensively describe the contributions of GLKs on phytochemical content, focusing on the target organs of diverse plants when examined through reverse genetic approaches that examine loss or gain of GLK functions.

Carotenoid biosynthesis

Carotenoids are natural pigments that belong to the terpenoid group. In plants, they perform photosynthesis along with chlorophylls and participate in various biological processes as photo-protectants, antioxidants, precursors of plant hormones. Carotenoids also play roles as color attractants in plants, since they are naturally colored and have diverse color hues (e.g., yellow, orange, and red). This also makes them suitable for use in the production of food and animal fodder. Moreover, diverse carotenoids such as β-carotene, zeaxanthin, astaxanthin, and capsanthin have received attention as major metabolites involved in plant biofortification, since they provide nutritionally relevant provitamin A components. Moreover, carotenoids can also provide health-promoting ingredients such as antioxidants and other free-radical scavengers [48, 49].

The effects of GLKs on carotenoid content have been reported in many plants, including hybrid birch, Arabidopsis, hybrid poplar, tomato, and rice. As shown in Table 1, decreased carotenoid levels following functional suppression of GLKs was first demonstrated in two hybrid birch mutants, yl and the BpGLK1-reduced expression line [7]. The former was fully rescued by constitutive overexpression of native BpGLK1 in the leaves of hybrid birch plants (C-yl). Similarly, a decrease in carotenoid content was observed in the leaves of the PaGLK RNAi-hybrid poplar mutant and in the calli of the Arabidopsis atglk1atglk2 mutant [36, 38]. Taken together, these results reveal a positive correlation between GLK expression and carotenoid levels and an especially close relationship with chlorophyll levels in photosynthetic leaves.

As shown in Fig. 1 and Table 2, the effects of GLK overexpression on carotenoid content were first reported in the u/u (slglk2) tomato variety via constitutive ectopic expression of either AtGLK1 or AtGLK2 using the 35S promoter [5, 20]. Both genes individually increased lycopene content by 10–60% in ripe fruit, supporting the notion that higher GLK expression during early stages could contribute to higher carotenoid content in the later stages of fruit development. In this study, the general selection of u for uniform fruit ripeness was determined to be an inadvertent compromise by a tomato breeding program for ripe fruit quality in exchange for more desirable carotenoid traits. The same Arabidopsis genes were constitutively overexpressed in Arabidopsis to increase carotenoid levels in non-photosynthetic roots, seed-derived calli, and etiolated seedlings [37, 38]. In this study, the molecular mechanism responsible for the direct effects of Arabidopsis GLKs on carotenoid accumulation was hypothesized as follows: GLKs form nuclear condensates with G-BOX BINDING FACTOR which then binds to the PSY promoter region, thereby resulting transcriptional upregulation of PSY [38]. In tomato, several GLK genes, SlGLK1, SlGLK2, PpGLK1, AchGLK, CsGLK1, and CsGLK2, were individually overexpressed using the same 35S promoter [6, 20, 23, 39]. All transgenic tomato plants exhibited enhanced carotenoid and chlorophyll levels in green fruits. This overexpression ultimately resulted in higher total carotenoid content in ripe red fruits, suggesting that the unripe green fruit stage is source of photosynthate, in contrast to the ripe fruit stages. In addition, the homologous overexpression of two native tomato genes, SlGLK1 and SlGLK2, increased β-carotene and lutein content more than lycopene content in the skin of red fruit [6]. Moreover, heterologous overexpression of tea genes, CsGLK1 and CsGLK2, increased the carotenoid content of leaves [20].

Fig. 1

© 2016 DBCLS TogoTV, CC-BY-4.0 https://creativecommons.org/licenses/by/4.0/deed.ja.)

Positive effects of GLKs on various plant species. The vertical axis is the publication year, and the horizontal axis is displayed separately by trait types to improve crops. Picture source: Arabidopsis, Peanut, Birch, Poplar: craiyon.com; Moss: The image of Moss is from TogoTV (

In cereal crops such as rice, leaf carotenoids were significantly increased (along with chlorophyll levels) when two maize GLKs were constitutively overexpressed [40]. Grain carotenoid levels were also enhanced relative to a wild type ZH variety following endosperm-specific overexpression of OsGLK1 (G) [46]. The GLK impact of carotenoid levels was confirmed by co-expression with three carotenogenic genes (i.e., tHMG1, ZmPSY1, and PaCrtI, which encode the enzymes truncated 3-hydroxy-3-methylglutaryl-CoA reductase, phytoene synthase, and phytoene desaturase, respectively) together (GHPC). This revealed that OsGLK1 could increase total carotenoid content by nearly threefold relative to HPC rice endosperm. Taken together, introduction of GLKs appears to have a major impact on carotenoid production in both source and sink organs of important crops such as a rice.

Flavonoid biosynthesis

Flavonoids, a major class of phenolic compounds, comprises several subclasses such as anthocyanins, flavonols, flavones, isoflavones, and flavanones. In plants, they play essential roles in mediating plant responses to biotic and abiotic environmental factors. In addition, in humans they confer health benefits since they scavenge free radicals, thereby contributing protection against broad spectrum of diseases [47]. Of the flavonoids, anthocyanins are water-soluble natural pigments that are responsible for the red, blue, and purple colors found in fruit and floral tissues. Catechins are flavonol-type polyphenols that are accumulated in tea plants [20, 33, 34].

Reduced anthocyanin levels following loss-of-function mutations have been reported only in Arabidopsis (Table 1). Moreover, the degree of reduction differed among Arabidopsis mutants: the lowest was observed in the double mutant, the next lowest in the atglk2 single mutant, and no effect was observed in the atglk1 single mutant [33]. Furthermore, decreased anthocyanin content following the disruption of AtGLK2 was re-confirmed in Arabidopsis seedlings [34]. In these studies, increased anthocyanin levels following gain-of-function manipulations have also been reported in Arabidopsis (Table 2). Independent overexpression of either AtGLK1 or AtGLK2 resulted in a significant increase in anthocyanin accumulation in seedlings [33, 34]. Interestingly, both studies also suggested putative molecular mechanisms by which AtGLK1 may regulate anthocyanin biosynthesis. The former suggested that AtGLK1 positively regulates sucrose-induced anthocyanin biosynthesis via action upstream of the MYB-LIKE2 (MYBL2) TF, which is a key negative regulator of anthocyanin biosynthesis [33]. The latter study proposed that AtGLK2 positively regulates anthocyanin biosynthesis via ELONGATED HYPOCOTYL 5-mediated light signaling in Arabidopsis as well as by antagonistic function between AtGLK2 and AtMYBL2 [34]. Meanwhile, overexpression of two tea GLKs (i.e., CsGLK1 or CsGLK2) increased total flavonoid content—especially catechin by 4.49- and 4.05-fold in tomato fruits, respectively [20]. This study also proposed that CsGLKs are positively regulators of light-regulated catechin accumulation in tea plants by activating R2R3-type CsMYB5b. Finally, they speculated that GLKs may have great potential for simultaneous enhancement of the accumulation of catechins and carotenoids in the fruits of horticultural crops such as tomato.

Stress resistance

Plants face diverse biotic and abiotic stresses and in order to survive, have evolved mechanisms of detecting and responding to stresses on the physiological, biochemical, and molecular levels. To consider whether GLKs have resistance to either biotic or abiotic stresses, we compiled all possible phenotypes reported when GLKs were suppressed or overexpressed using transgenic approaches (Table 3).

Table 3 Summary of the influence on biotic and abiotic stress resistance following suppression and overexpression of GLKs in diverse plants

Biotic stress

Plants are constantly exposed to various pathogens such as fungi, bacteria, and viruses which reduce in crop productivity [50]. Therefore, enhancing pathogen resistance is an important target trait for many types of agriculture. In the previous section, we described the positive effects of GLKs on plant growth, agricultural yield, and phytochemical. Here, we will consider the biotechnological applicability of GLKs on disease resistance (Table 3).

The first role GLKs play in plant defense was demonstrated in Arabidopsis by constitutive overexpression of AtGLK1. This resulted in resistance against to Fusarium graminearum, a fungal pathogen responsible for major losses in cereal crops such as wheat and maize, where it causes Fusarium head blight and gibberella ear mold, respectively [51]. It was suggested that overexpression of AtGLK1 may reprogram gene expression networks to upregulate defense-related genes as a resistance mechanism. Further elucidation of the GLK roles in plant defense involved examining two different pathogens, Hyaloperonospora arabidopsidis (Hpa) Noco2 (a biotrophic oomycete) and necrotrophic Botrytis cinerea (a necrotrophic fungus) [52]. Interestingly, when taken together these results showed conflicting patterns of resistance associated with GLKs; namely, atglk1atglk2 plants exhibited higher resistance against Hpa Noco2 while 35::AtGLK1 plants showed heightened resistance against B. cinerea. Ultimately, AtGLKs were confirmed to act on jasmonic acid (JA)-dependent susceptibility to Hpa Noco2 and JA-independent immunity to B. cinerea. Another dual resistance study involving two biotic pathogens, Sclerotinia sclerotiorum (a necrotrophic fungal pathogen) and Pseudomonas syringae pv. tomato DC3000 (a phytopathogenic bacterium), was reported in Arabidopsis plants that showed constitutive overexpression of peanut AhGLK1b [53]. Here, the suggested function of AhGLK1b in plant disease resistance involved the upregulation of defense-related genes including a multidrug and toxin extrusion efflux protein, PR10, and Phox/Bem 1, which is involved in multiple disease resistance.

Furthermore, the involvement of GLKs in plant resistance to viruses was first investigated in Arabidopsis in response to infection by Cucumber mosaic virus (CMV) [54]. Researchers found that atglk1atglk2 plants were more susceptible than Atglk1 and Atglk2 to CMV and suffered more serious damage. This included higher oxidative damage, more compromised PSII photochemistry, and more reactive oxygen species accumulation [54]. Taken together, these findings suggested that AtGLK1 and AtGLK2 might play redundant roles in CMV resistance in Arabidopsis. A second case was reported in tobacco plants that overexpressed tobacco NbGLK1 and were then subjected to Potato virus X (PVX) infection [55]. This study found that NbGLK1 can act as an immune activating protein to reduce susceptibility to PVX by activation (i.e., de-repression) via activated Retinal homeobox gene 1, an intracellular immune receptor, following virus perception. Rice is also damaged by viruses, the most destructive of which is rice black-streaked dwarf virus (RBSDV) [56]. Quantitative trait loci (QTL) analysis using the RBSDV-resistant Wuke variety explored three genes associated with RBSDV resistance. Interestingly, OsGLK1 was identified as a major resistance QTL. This finding was reinforced by an empirical test of whether resistance against RBSDV was supported by OsGLK1; this involved both KD and the stable overexpression of OsGLK1 in rice plants.

Abiotic stress

Due to their sessile life history, plants experience diverse environmental stresses including ultraviolet radiation, drought, cold, heat, and salinity [50]. A variety of abiotic stresses affect the growth and development of plants and ultimately cause reduced agricultural productivity. To further explore the influence of GLKs on abiotic stress resistance, we reviewed the results of phenotypic changes (Table 3).

High concentrations of ozone (O₃), the most phytotoxic air pollutant in the troposphere, is known to cause oxidative stress in plants and is associated with crop losses. One study in Arabidopsis showed that a strong tolerance to ozone was associated with a closed-stomata phenotype caused by suppression of either AtGLK1 or AtGLK2 via fusion of the SRDX, the ERF-associated amphiphilic repression motif repression domain [57]. In contrast, the overexpression of either AtGLK1 or AtGLK2 caused plants to exhibit higher sensitivity to ozone when plants showed an open-stomata phenotype. Taken together, these findings suggest that AtGLKs may play a positive role in regulating the expression of genes related to K + ion channels. It has also been proposed that AtGLKs may be effective tools for conferring resistance to air pollutants by controlling stomatal movement. For instance, another study of Arabidopsis atglk1atglk2 mutant plants revealed osmotic stress resistance during seedling development accompanied by abscisic acid (ABA)-hypersensitivity [19]. Exactly contrary to this, the overexpression of these two genes showed hypersensitivity to osmotic and salt stresses and an ABA-hyposensitive phenotype, respectively. This evidence suggests that AtGLKs are associated with resistance to osmotic and dehydration stress in Arabidopsis.

Next, two cases have identified a relationship between GLKs and drought resistance in crops that are greatly affected by drought. In peanut, overexpression of native AhGLK1 increased survival rates during recovery from drought in Arabidopsis [18]. AhGLK1 enables post-drought recovery by stimulating chlorophyll biosynthesis and photosynthesis via the unregulated expression of AhPORA. Moreover, cotton crops are also exposed to harsh environments during cultivation. Virus-induced silencing of GhGLK1 in cotton resulted in plants that were more vulnerable to drought and cold stress, whereas its overexpression in Arabidopsis showed greater adaptability following drought and cold treatments than the wild type [58]. Taken together, these findings suggest that GhGLK1 might be a key candidate gene for the simultaneously enhancement of cold and drought stress tolerance.

Current potential and future prospects for crop improvement

In this review, we evaluated the impact of GLKs regarding overall greening traits, yield, phytochemical accumulation, and stress resistance. We did so by examining different forms of plant transformation research, including gene knocking-out by insertion of transposons and T-DNA, CRISPR/Cas9-mediated genome editing, knocking-down by RNA interference, homologous recombination, virus-induction, and repressor domain fusion, and overexpression using diverse promoters. Through these studies, we identified positive effects of GLKs on chlorophyll biosynthesis, chloroplast development, photosynthesis, and carotenoid biosynthesis, as well as significant effects in particular cases on flavonoid biosynthesis and resistance to specific pathogens and forms of stress. As shown in Fig. 1, all positive results were listed chronologically. Moreover, our analysis shows that functional studies of GLKs have mainly focused on Arabidopsis for most traits. However, GLK studies have consistently attempted to improve agricultural traits in a diverse range of crops, including birch and poplar (wood), peanut (a legume), rice (a cereal), tobacco, and tomato (a vegetable). While chlorophyll and chloroplast traits were the primary consideration in most plant studies, photosynthesis and yield were a primary interest for rice studies, and differences in the abundance of two functional metabolites—i.e., carotenoids and flavonoids were identified in tomato fruits.

From the functional standpoint, several notable features are evident. First, there is evidence of function redundancy between GLK copies that rarely change greening traits in single KO mutants and show functional differences depending on promoter (e.g., 35S, Ubi, RbcS, AtSUC2, LTP, MpTBE2, ZmG1, and ZmG2, but not FDH and PDS). Thus, tests requiring photosynthetic cell expression may not be effective in nonphotosynthetic tissues such as roots, calli, dark-grown seedlings, ripe fruits, and seeds (Tables 1 and 2). Second, other finding strongly suggest that GLKs enhance photosynthetic performance (i.e., result in increased chlorophyll and carotenoid biosynthesis, increased chloroplast biogenesis, and more efficient photosynthetic machinery) in photosynthetic tissues. Moreover, enhanced photosynthates then contribute to improvements in yield and functionality via increased carotenoid and flavonoid content in both non-photosynthetic sink organs and photosynthetic source organs.

Furthermore, the influence of GLKs on biotic stress resistance was positively correlated for three fungi, including cereal pathogen F. graminearum as well as two necrotrophs B. cinerea and S. sclerotiorum, along with the phytopathogenic bacterium P. syringae pv. tomato DC3000. In Arabidopsis, AtGLK1 overexpression plant was susceptible against a biotrophic oomycete Hpa Noco2 compare to atglk1atglk2 [51, 52, 54]. GLK overexpression also conferred positive resistance against three pathogenic viruses, including CMV (i.e., AtGLKs in Arabidopsis), PVX (i.e., NbGLK1 in tobacco), and RBSDV (i.e., OsGLK1 in rice) [53, 55, 56]. Moreover, abiotic stress resistance was increased for drought and cold stress when AhGLK1 and GhGLK1 were individually overexpressed and for higher ozone concentration and osmotic stress when AtGLKs were knocked-out in Arabidopsis [18, 19, 57, 58]. Taken together, these results demonstrate that GLKs have great potential for improving resistance against diverse forms of stress (Fig. 1).

Given the above considerations, this review also shows that using an appropriate promoter is essential for applying GLKs to crop improvement. A late ripening-specific promoter (PDS) failed to generate a carotenoid increase in ripe tomato fruits [5]. However, a strong constitutive promoter (ZmUbi) for ZmG1- not ZmG2-overexpression led to negative impacts on seed development and endosperm-specific overexpression. Finally, GluB-1 can cause deteriorating grain quality due to increased chalkiness in rice [41, 46].

We currently face a food crisis caused by rapid population growth and global climate change. Consequently, the discovery of major genes influential enough to increase agricultural productivity may be the most important goal in plant biology. The use of GLKs to improve agricultural biotechnology has great potential for the improvement of numerous crops and may facilitate the sustainable survival of humanity via food security.