End-Of-Day LED Lightings Inﬂuence the Leaf Color, Growth and Phytochemicals in Two Cultivars of Lettuce

: Four light treatments (W: white light; EOD-B: end-of-day enhanced blue light; EOD-FR: end-of-day supplementary far-red light; EOD-UV: end-of-day supplementary ultraviolet-A light) were designed to explore the e ﬀ ects of end-of-day (EOD) lightings (30 min before dark period) on leaf color, biomass and phytochemicals accumulation in two lettuce cultivars ( Lactuca sativa cv. ‘Red butter’ and ‘Green butter’) in artiﬁcial light plant factory. EOD-FR stimulated the plant and shoot biomass of two cultivars, and EOD-B suppressed the growth of ‘Red butter’ but induced higher biomass in ‘Green butter’. EOD lightings generated brighter, greener and yellower leaf in ‘Red butter’ at harvest, but the highest lightness and the deepest redness of ‘Green butter’ leaf were observed in the middle growth stage. ‘Red butter’ had prominent higher contents of chlorophylls and carotenoids, while these pigments showed less sensitivity to the interaction of cultivars and EOD lightings. EOD lightings impeded the accumulation of anthocyanin in two cultivars, except EOD-UV slightly increased the anthocyanin contents in ‘Green butter’. EOD-UV strengthened the antioxidant capability of ‘Green butter’, but EOD lightings had di ﬀ erent e ﬀ ects on the antioxidant and nutritional compound contents in two lettuce cultivars.


Introduction
The increasing world population combined with decreasing arable land provide an excellent opportunity for the development of plant factory in urban areas. In addition, the aroused concern of food safety requires precision in environmental control to balance the yield production, agronomic characteristics, and the nutritional qualities of vegetable crops. Lettuce (Latuca sativa L.) is largely consumed in the world, and it is arguably one of the most common crops in plant factories with artificial light (PFALs). Lettuce is rich in natural pigments like anthocyanins, carotenoids and chlorophylls, and nutritional compounds such as vitamin C, proteins, and phenolics [1,2]. These phytochemicals not only contribute to the leaf color and/or the flavor of lettuce plant, but also benefit human health.
Light functions as the energy source to drive photosynthesis and as a signal that directly or indirectly regulates the plant morphology and physiology [3][4][5]. Plants perceive light via a complex array of photoreceptors (phytochromes, cryptochromes, phototropins and UVR8), which are defined by the absorbed light wavelength. When the light environment (light quality, quantity, directionality, and photoperiod) changes, these photoreceptors can switch forms or be activated in distinct manners [6,7]. Consequently, they transduce diverse light signals to modulate the core

Growth Measurements
Sample collecting and growth measurements were carried out at 24 days after light treatments ( Figure 2). Five lettuce plants in each treatment were randomly selected and weighed by an analytical balance immediately after harvest. Then the samples were oven dried at 75 • C for 72 h to determine the dry weight and calculate the moisture content.

Color Measurements
The leaf color of lettuce was non-destructively determined using a colorimeter (CR-10 plus, Konica Minolta Inc., Tokyo, Japan) at 3, 6,9,12,15,18,21, and 24 days after light treatments (D3-D24), respectively. The value L* means lightness; a* represents the color from green to red; b* suggests the color from blue to yellow.

Chlorophyll and Carotenoids Measurements
The contents of chlorophyll and carotenoids were determined according to ethanol test [30]. Fresh samples of lettuce (0.2 g) were homogenized with 8 mL ethanol until the tissue turned white. The leach liquor absorbance was measured at 440 nm, 645 nm, and 663 nm by a UV-spectrophotometer (Shimadzu UV-16A, Shimadzu Corporation, Kyoto, Japan), respectively. The contents of chlorophyll and carotenoids were calculated as the following equations: where V is the volume of extract solution (8 mL), and W is the fresh weight (0.2 g) of the sample.

Total Anthocyanins Measurement
The contents of total anthocyanins (TA) was measured according to pH differential method [31]. Two fresh samples of lettuce (1.0 g) were homogenized with pH 1.0 potassium chloride buffer (50 mM KCl and 150 mM HCl) and pH 4.5 sodium acetate buffer (400 mM CH 3 COONa and 240 mM HCl), respectively. After 5 min 14,000 g centrifuging at 4 • C, the supernatants were measured at 510 nm by a UV-spectrophotometer.
where A 1 and A 2 are the absorbances of the sample extracted from pH 1.0 buffer and pH 4.5 buffer, respectively. The number 484.8 is the molecular weight of cyaniding-3-glucoside chloride. The number 24.825 is the absorption coefficient at 510 nm. The dilution factor in this measurement is 1.

Phytochemical Measurement
The DPPH radical inhibition percentage (DPPH) measurement was based on the method of Musa et al. [32]. Fresh samples of lettuce (0.5 g) were homogenized with 8 mL ethanol for 30 min in darkness. After 15 min of being centrifuged at 3000 rpm, the supernatant was used to prepare three types of mixture (Ai: Supernatant of 2 mL mixed with 2 mL 0.2 µM DPPH; Aj: Supernatant of 2 mL mixed with 2 mL ethanol; Ac: 0.2 µM DPPH mixed with 2 mL ethanol). These mixtures were determined at 517 nm by the UV-spectrophotometer. The DPPH radical inhibition percentage was calculated as follows: The ferric ion-reducing antioxidant power (FRAP) measurement was according to Tadolini et al. [33]. Fresh samples of lettuce (0.5 g) were homogenized with 8 mL ethanol for 30 min in darkness. After 15 min of being centrifuged at 3000 rpm, the supernatant (0.4 mL) was added to the FRAP reagent (3.7 mL), and the mixture was preserved in a 37 • C water bath for 10 min. The FRAP reagent was prepared by mixing 300 mM acetate buffer (pH 3.6), 20 mM ferric chloride, and 10 mM 2,4,6-tripyridyl-S-triazine (TPTZ) in 40 mM HCl in the proportion of 10:1:1 (v:v:v). The absorbance was then determined at Agronomy 2020, 10, 1475 6 of 19 593 nm by the UV-spectrophotometer. FeSO 4 ·7H 2 O was used as the standard, and the results were expressed as mmol·g −1 FW.
The total phenolic compounds (TPC) measurement was conducted as stated by Tadolini et al. [33]. Fresh samples of lettuce (0.5 g) were homogenized with 8 mL ethanol for 30 min in darkness. After 15 min of being centrifuged at 3000 rpm, the supernatant (1.0 mL) was mixed with 0.5 mL of Folin-ciocalteu' ultra-pure water reagent (1:1, v:v) and 1.5 mL 26.7% Na 2 CO 3 solution (w:v). The mixture was diluted to a total volume of 10 mL with ultra-pure water. After 2 h reaction, the absorbance was recorded at 760 nm with the UV-spectrophotometer. TPC values were calculated from the gallic acid standard curve, and the results were expressed as mg gallic acid equivalent fresh weight (mg GAE·g −1 FW).
The total flavonoids (TF) measurement was in accordance with the method by Sánchez-Rangel et al. [34]. The supernatant (1 mL) was mixed with 30% ethanol (10 mL, w:v) and 5% NaNO 2 solution (0.7 mL, w:v). After 5 min, 10% Al(NO 3 ) 3 solution (0.7 mL, w:v) was added in for 5 min reaction. Then, 5% NaOH solution (5 mL, w:v) and 30% ethanol (8.6 mL, v:v) were added. The absorbance was determined at 510 nm with the UV-spectrophotometer. Rutin hydrate was used as the standard, and the results were expressed as mg·g −1 FW.
The contents of soluble proteins (SP) were determined according to Blakesley and Boezi [35]. Fresh samples of lettuce (1.0 g) were ground with 8 mL distilled water. After being centrifuged (8000 rpm, 4 • C) for 10 min, the supernatant (1 mL) was mixed with Coomassie brilliant blue G-250 solution (5 mL, 0.1 g·L −1 ). The absorbance was measured at 595 nm by a UV-spectrophotometer.
The contents of soluble sugars (SS) measurement were performed according to Kohyama and Nishinari [36]. Fresh samples of lettuce (1.0 g) were extracted with 80% ethanol (10 mL, v:v) and then homogenized with activated carbon powder (10 mg) and water bath (80 • C) for 40 min. The extract was diluted to a total volume of 25 mL with 80% ethanol (v:v). Then the filter liquor (0.2 mL) was mixed with diluted water (0.8 mL) and of sulfuric acid anthrone reagent (5 mL). After 10 min water bath (100 • C), the absorbance at 625 nm was detected by the UV-spectrophotometer.
The contents of nitrates were determined with the method proposed by Cataldo et al. [38]. Fresh samples of lettuce (1.0 g) were homogenized with 10 mL deionized water and water bath (100 • C) for 30 min. The filtrate was diluted with deionized water to a total volume of 25 mL. Then the extract (0.1 mL) was mixed with 5% salicylic acid-H 2 SO 4 reagent (0.4 mL, w:v). After 10 min incubation, 8% NaOH (9.5 mL, w:v) was added and the absorbance was measured at 410 nm with a UV-spectrophotometer.

Data Analysis
Data were analyzed by a one-way analysis of variance (ANOVA), using SPSS 25.0 software. Significance at p < 0.05 was performed by the Tukey's test. XLSTAT 2019 software was used for statistical computing and multivariate principal component analysis (PCA). TBtools software [39] was used for visualizing the transformed data into a cluster heatmap.

Growth and Biomass
The agronomic traits of lettuce cultivars were governed by the genetic as well as the interaction between genotype and environment. Most of the growth and biomass parameters were significantly influenced by cultivars, EOD light treatments, and the interaction between the two factors at harvest  Tables 2 and 3). The highest fresh weight, dry weight, and moisture contents of plant and root were recorded in 'Green butter' × EOD-FR, while the lowest fresh and dry mass were observed in 'Red butter' × EOD-B (Figures 3 and 4). Concerning EOD lightings, EOD-FR stimulated the biomass of two lettuce cultivars ( Figure 3). In 'Red butter', the fresh weights of total plant (28.04%) and shoot (31.34%) exhibited remarkably increases under EOD-FR, and the parallel trend was found in dry weights (14.21% and 18.37%) (Figure 3a,b,d,e). Similar in 'Green butter', the fresh weights of plant (35.68%), shoot (37.09%), and root (19.95%) were significantly increased by EOD-FR, as well as dry weights of plant and shoot (31.50% and 39.89%) (Figure 3a-e). Consequently, EOD-FR significantly decreased the root-shoot ratio in two lettuce cultivars (Figure 4d). Moreover, the growth responses to EOD-B and EOD-UV varied in terms of cultivars (Tables 2 and 3). In 'Red butter', EOD-B suppressed the fresh weights of plant (15.30%) and shoot (16.66%), as well as the homologous dry weights (13.57% and 15.92%) (Figure 3a,b,d,e). However, EOD-B led to significantly higher plant fresh weight (13.15%) in 'Green butter' (Figure 3a). Regarding EOD-UV, the fresh weights of 'Red butter' plant (15.80%) and shoot (17.44%) were increased (Figure 3a

Leaf Color Transformation and Pigment Content
Two lettuce cultivars presented differential coloration during the growth ( Figure 5). On the whole, 'Red butter' leaf was darker, redder, and yellower than 'Green butter' leaf. Regarding the EOD lightings, 'Red butter' had brighter, greener and more yellow leaf under all EOD lightings, while 'Green butter' leaf was darker under EOD-B and EOD-UV, and the leaf was redder and more yellow under all EOD lightings, respectively compared with W ( Figure 5). In 'Red butter', the lowest redness (a* = −10.83), the highest yellowness (b* = 26.57) and lightness (L* = 39.18) were recorded at the 24 th day under lighting treatment (D24), suggesting a fading coloration during growth. Whereas in 'Green butter', the highest lightness (L* = 55.26) and the deepest redness (a* = -8.42) were observed at D12 and D9, respectively. Pigment contents and ratios were involved in lettuce leaf color. 'Red butter' possessed significantly higher contents of chlorophyll a, chlorophyll b, chlorophyll (a + b), carotenoids, and TA contents than 'Green butter' (Figure 6), characterizing two cultivars from red leaf and green leaf ( Table 4). Except for carotenoids, different EOD lightings markedly influenced the pigment contents (Figure 6a-e). Interestingly, the interaction of cultivars and lightings had insignificant effects on most of the pigment contents and pigment ratios, while significant differences were observed in TA content and the ratio of chlorophyll (a + b) and TA (Table 4). In 'Red butter', TA content was greatly reduced by EOD-FR (67.41%), EOD-B (58.04%), and EOD-UV (41.07%), compared to W (Figure 6e). Consequently, the chlorophyll (a + b)/TA ratio was in the following order: EOD-FR > EOD-B > EOD-UV > W (Figure 6h). Differently in 'Green butter', the TA content was decreased by EOD-FR (83.33%) and EOD-B (81.25%) but increased by EOD-UV (30.43%) (Figure 6e). As a result, the chlorophyll (a + b)/TA ratios increased significantly under EOD-FR and EOD-B (Figure 6h).

Leaf Color Transformation and Pigment Content
Two lettuce cultivars presented differential coloration during the growth ( Figure 5). On the whole, 'Red butter' leaf was darker, redder, and yellower than 'Green butter' leaf. Regarding the EOD lightings, 'Red butter' had brighter, greener and more yellow leaf under all EOD lightings, while 'Green butter' leaf was darker under EOD-B and EOD-UV, and the leaf was redder and more yellow under all EOD lightings, respectively compared with W ( Figure 5). In 'Red butter', the lowest redness (a* = −10.83), the highest yellowness (b* = 26.57) and lightness (L* = 39.18) were recorded at the 24th day under lighting treatment (D24), suggesting a fading coloration during growth. Whereas in 'Green butter', the highest lightness (L* = 55.26) and the deepest redness (a* = −8.42) were observed at D12 and D9, respectively.
Pigment contents and ratios were involved in lettuce leaf color. 'Red butter' possessed significantly higher contents of chlorophyll a, chlorophyll b, chlorophyll (a + b), carotenoids, and TA contents than 'Green butter' (Figure 6), characterizing two cultivars from red leaf and green leaf (Table 4). Except for carotenoids, different EOD lightings markedly influenced the pigment contents (Figure 6a-e). Interestingly, the interaction of cultivars and lightings had insignificant effects on most of the pigment contents and pigment ratios, while significant differences were observed in TA content and the ratio of chlorophyll (a + b) and TA (Table 4). In 'Red butter', TA content was greatly reduced by EOD-FR (67.41%), EOD-B (58.04%), and EOD-UV (41.07%), compared to W (Figure 6e). Consequently, the chlorophyll (a + b)/TA ratio was in the following order: EOD-FR > EOD-B > EOD-UV > W (Figure 6h). Differently in 'Green butter', the TA content was decreased by EOD-FR (83.33%) and EOD-B (81.25%) but increased by EOD-UV (30.43%) (Figure 6e). As a result, the chlorophyll (a + b)/TA ratios increased significantly under EOD-FR and EOD-B (Figure 6h).

Phytochemical Profiles
The contents of phytochemicals differed between two lettuce cultivars and among light treatments. The superior antioxidant capacity (DPPH and FRAP) and antioxidant compounds contents (TPC, TF, and VC) were observed in 'Red butter' (Figure 7 and Table 5). Concerning the interaction of cultivars and lightings, most of the antioxidant-related indices showed insignificant difference ( Table 5), except that EOD-UV (17.96%) and EOD-B (16.04%) increased the DPPH compared to W in 'Green butter' (Figure 7a). EOD-FR obviously decreased the contents of SP (21.21%) and SS (27.46%) in 'Red butter', while EOD-B significantly increased SP (15.82%) and SS (6.37%) contents in 'Green butter', respectively, compared with W (Figure 7f,g). The most abundant SS content was observed in 'Green butter' × EOD-FR with the increase of 18.02%, as compared to 'Green butter' × W (Figure 7g). The lowest nitrates content was recorded in 'Red butter' × EOD-UV (0.56 mg·g −1 FW), while the highest content was observed in 'Green butter' × EOD-UV (0.87 mg·g −1 FW) (Figure 7h), indicating a differential accumulation of nitrates in two lettuce cultivars in response to EOD ultraviolet-A lighting. NS, *, **, *** represent non-significant or significant at p < 0.05, 0.01, and 0.001, respectively, according to two-way ANOVA, Tukey's honest significant difference tests. C = cultivars, L = lightings. Chl = chlorophyll, TA = total anthocyanins, Caro = carotenoids.

Multivariate Principal Component Analysis
To compare the correlation of all growth and quality traits in two lettuce cultivars' response to different EOD lightings, the principal component analysis (PCA) was performed ( Table 6 and Figure 8). The first seven principal components (F1-F6) were associated with eigen values > 1, and account for approximately 87.14% and 90.37% of the cumulative variance in 'Red butter' and 'Green butter', respectively (Table 6). The first two factors (F1 vs. F2) of the PCA were presented in the correlation circle and scatterplot (Figure 8), and explained 50.35% of the total data variance of 'Red butter' and 62.89% for 'Green butter'. The correlation circle (Figure 8a,c) illustrated the relationships among growth parameters, antioxidants, and nutrient components, by identifying the angle between two vectors (0 • < positively correlated < 90 • ; uncorrelated: = 90 • ; 90 • < negatively correlated < 180 • ) and the distance from the center of the circle (r > 0.5 means relative higher correlation). From the results of 'Red butter', strong positive correlations were found between carotenoids (Caro) and nitrates contents, carotenoids and the leaf redness (a*), total flavonoids (TF) and soluble sugar (SS) contents, and among chlorophylls, respectively, while negative correlations were observed between carotenoids and the chlorophylls, carotenoids and nitrates, chlorophylls and a*, and between TPC contents and the plant and shoot fresh weights (P-FW and S-FW), respectively (Figure 8a). In 'Green butter', positive correlations were identified between nitrates and chlorophylls, among DPPH, FRAP, TF, and TA, while these indices were negatively correlated with the SS content, the leaf lightness (L*), and the leaf yellowness (b*) (Figure 8c). Interestingly, TPC contents had an insignificant correlation with the chlorophyll contents in 'Red butter', whereas a strong negative correlation in 'Green butter' was shown (Figure 8a,c). The scatterplot (Figure 8b

Heatmap Analysis
A heatmap synthesizing the response of the measured parameters provided an integrated view of the effect of different EOD lightings on the growth and quality of lettuce (Figure 9).
Regarding 'Red butter', the EOD-UV and EOD-FR clusters were the closest to each other in terms of measured parameter responses, and they were equidistant from cluster W (Figure 9a). The EOD-UV and EOD-FR clusters both showed brighter and yellower leaf, as well as higher fresh yield and dry weight (Figure 9a). Meanwhile, cluster EOD-B was considerably separated from the other three clusters: EOD-B decreased the fresh and dry weight of plant and shoot; increased the root-shoot ratio; and led to higher contents of VC, carotenoids, nitrates, TPC, and TF compared to W, EOD-FR, and EOD-UV, contributing to separate the EOD-B cluster from the others (Figure 9a). Moreover, the heatmap indicated an affinitive pattern in the leaf color parameters (L*, a*, and b*) and the plant and shoot water content, as well as the content of chlorophylls (Figure 9a).

Heatmap Analysis
A heatmap synthesizing the response of the measured parameters provided an integrated view of the effect of different EOD lightings on the growth and quality of lettuce ( Figure 9). from the other three clusters: EOD-FR elicited higher accumulations of VC, SS, carotenoids, and promoted the fresh and dry weight of lettuce (Figure 9b). At the same time, the EOD-UV cluster facilitated the accumulation of antioxidants (total anthocyanins, total flavonoids, DPPH, and FRAP) and chlorophylls (Figure 9b). Interestingly, the patterns of a* and L* in 'Green butter' were close to the plant and shoot water content, but b* was close to SS contents, which differed from 'Red butter' (Figure 9b).  Regarding 'Red butter', the EOD-UV and EOD-FR clusters were the closest to each other in terms of measured parameter responses, and they were equidistant from cluster W (Figure 9a). The EOD-UV and EOD-FR clusters both showed brighter and yellower leaf, as well as higher fresh yield and dry weight (Figure 9a). Meanwhile, cluster EOD-B was considerably separated from the other three clusters: EOD-B decreased the fresh and dry weight of plant and shoot; increased the root-shoot ratio; and led to higher contents of VC, carotenoids, nitrates, TPC, and TF compared to W, EOD-FR, dry mass of 'Red butter' (Figure 3). These might be due to the weaker and shorter EOD ultraviolet-A light in this study.

Leaf Color Responds to End-Of-Day Lightings
The additional FR radiation decreased the chlorophyll concentration per unit leaf area [42] and reduced the amount of chlorophyll (14%) and carotenoids (11%) [44] in lettuce. The tea (cv. Hangjinya) leaf under higher blue light ratio possessed higher b* value (yellower), which might be due to the lower contents of chlorophyll a, chlorophyll b, and chlorophyll (a + b) [52]. UV-A radiation elicited the anthocyanins accumulation in the hypocotyls in soybean sprouts, and the trend was consistent with the expression pattern of anthocyanin biosynthesis-related genes [53].
No similarity was found in this study; the color of lettuce roughly followed a similar change process throughout the growth stage, which presented a decrease in redness and an increase in yellowness ( Figure 5). The reason might be related to the dynamically varied relationship between the lettuce growth rate and pigment levels. The lightness (L*) of 'Red butter' increased from D6 to D24, and the highest value was recorded under EOD-UV at D24 (Figure 5a). From D9 to D 24, the redness (a*) of 'Red butter' leaf gradually faded while the yellowness (b*) increased (Figure 5b,c). However, a* and b* values of 'Red butter' seemed to be unaffected by EOD lightings (Figure 5b,c). Whereas, the leaf color of 'Green butter' exhibited fluctuation during the growth time (Figure 5d-f). The L* value under EOD-UV remained stable throughout the growth period, while the peak under other lightings was observed at D12 (Figure 5d). The peak of a* value was observed at D6 × EOD-UV and D9 × EOD-FR, which might be caused by the pigment biosynthesis in adaption to EOD UV-A and FR lights at early growth stage (Figure 5e). However, the color parameters of 'Green butter' at D24 were not significantly different among EOD lightings (Figure 3d-f). These results were consistent with the unchanged contents of chlorophylls and carotenoids (Figure 6a-d). Moreover, compared with chlorophylls contents (1.78 and 1.49 mg·g −1 ), the TA contents (0.33 and 0.08 mg·g −1 ) were not dominant either in 'Red butter' or 'Green butter' (Figure 6c,e). Therefore, the effects of significantly decreased TA content under EOD lightings could not be reflected in color parameters.
To conclude, EOD lightings of blue light, far-red light, and ultraviolet-A light could not significantly affect the leaf color of lettuce. The effects of EOD lightings might be to moderately aggravate/weaken the degree of color change or to accelerate the change trend, but it cannot reverse or eliminate these changes.

Lettuce Phytochemical Profiles in Relation to End-Of-Day Lightings
Plant species and cultivars respond to light recipes in different ways, and genotypic effect is the principle quantitative and qualitative variation in vegetables metabolites contents [54]. Different light recipes can lead to remarkable changes in plant transcriptomic pathways, but the metabolic traits may behave differently in different genotypes [55]. 'Red butter' presented stronger antioxidant capacity (DPPH and FRAP) and higher contents of antioxidant compounds (TPC, TF, and VC) than 'Green butter' (Figure 7a-e). With respect to C × L, significant enhancement of DPPH was observed under EOD-B and EOD-UV in 'Green butter' (Figure 7a and Table 5). The DPPH increased more in lettuce (1.3 times), spinach (1.2 times), and kale (1.2 times) under 17% added blue light than 100% red light, while it was enhanced in basil (1.2 times) and sweet pepper (1.1 times) under 9% added blue light [50]. Whereas, the antioxidant-related metabolites in 'Red butter' were little affected by EOD lightings (Table 5). Analogously, neither supplemental FR (160.4 µmol·m −2 ·s −1 , 16 h) nor UV-A (20.9 µmol·m −2 ·s −1 , 16 h) affected the contents of phenolics and ascorbic acid in 'Red Cross' lettuce [40]. UV light is often regarded as an abiotic stress to plant that stimulates the accumulation of reactive oxygen species in plants [56] and activates the defense and disease-resistance mechanisms [57]. In 'Klee' lettuce, the antioxidant contents of total phenolic (17.78%), total flavonoids (48.33%), and ascorbic acid (61.04%) were greatly simulated under UV-A supplementation (30 µmol·m −2 ·s −1 , 16 h) [24]. The increased DPPH in lettuce might be a response to the slight stress caused by EOD-UV and EOD-B. On the contrary, UV solar exclusion (exclusion of more than 99% of UV-A) led to significantly higher DPPH (17.16%) and total phenolics content (51.54%) in the spicas of Prunella vulgaris L. than under solar control [58].
Soluble sugar content was greatly reduced by 27.46% in 'Red butter' × EOD-FR, while it increased by 16.76% in 'Green butter' × EOD-FR (Figure 7g). Compared to the corresponding C × W, the soluble proteins content was significantly lower in 'Red butter' × EOD-FR (21.21%), while it was higher in 'Green butter' × EOD-B (15.82%) (Figure 7f). Differential sugar and protein contents were also observed in other lettuce cultivars in response to UV-A lights. Supplemental UV-A (10, 20, and 30 µmol·m −2 ·s −1 , 16 h) stimulated the soluble sugar content (12.74-26.11%) and total soluble protein content (13.73-23.53%) in 'Klee' lettuce, and the 10 µmol·m −2 ·s −1 UV-A obtained the best promotion effects [24]. Supplemental UV-A light (6 µmol·m −2 ·s −1 , 16 h) resulted in 1.76 times higher maltose content and nearly no change in total proteins in red leaf 'Red cos' lettuce, but it had no effects on maltose and total proteins in 'Lobjoits green cos' lettuce [55]. These suggested that the synthesis and/or metabolic processes of sugar and proteins in response to end-of-day FR, B, and UV-A were different in two lettuce cultivars.
Nitrate was the main form of nitrogen uptake in plants; there was, directly or indirectly, a transformation among sugars, proteins and nitrates concerning the ratio of carbon and nitrogen [59]. Nitrates content was lower in 'Red butter' (0.63 mg·g −1 ) than 'Green butter' (0.77 mg·g −1 ). The UV exclusion was reported to cause a significant debility in nitrate reductase activity, retarding the catalyze process that nitrate transformed into nitrite [60]. However, an increased nitrates content was observed in 'Green butter' × EOD-UV (Figure 7h), suggesting a distinct responses of plant nitrate contents in response to day-UV light and EOD-UV light.
Both PCA analysis and heatmap analysis validated the differential growth and phytochemical profiles of two lettuce cultivars under different EOD lightings (Figures 4 and 5). There were strong positive correlations among the a* value, carotenoids content, and nitrate contents in 'Red butter'; and these three indices were negatively related to the contents of chlorophyll a, chlorophyll b, chlorophyll (a + b), b* value, and L* value (Figure 8a). Analogously in 'Green butter', the L* value was negative related to a* value, but it had insignificant correlation with carotenoids ( Figure 8c). Although PCA described the correlations among different parameters, the characteristics of each EOD lighting could not be represented visually. Thus, we used the heatmap to provide a global view of agronomic and metabolic traits and identify the phenotypic variation patterns associated with different EOD lightings (Figure 9). From the heatmap, similar and differential response patterns of two lettuce cultivars under EOD lightings were observed. In 'Red butter', the comparison of EOD-FR vs. EOD-UV had similar characteristics (Figure 9a). However, in 'Green butter', the close patterns were changed to the comparison of W vs. EOD-B (Figure 9b). Overall, these results verified that different EOD light treatments evoke specific growth and metabolic responses in 'Red butter' and 'Green butter'.

Conclusions
In this paper, we investigated the potential of end-of-day blue, far-red, and ultraviolet-A light to regulate the leaf color, growth and phytochemical accumulation of two lettuce cultivars. 'Green butter' showed higher fresh yield, while 'Red butter' possessed higher antioxidant and nutrient values. EOD far-red light performed by increasing plant and shoot fresh yield in two lettuce cultivars, while slightly decreasing the nutrition compounds. The application of EOD blue light and EOD UV-A light in lettuce required specific cultivars consideration.
Considering the low intensity and short duration of EOD light supplementation used in this study, further researches are needed, including different light intensity, supplemental periods, and switches of EOD light qualities. Moreover, the photosynthesis parameters, related enzyme contents, and gene expression patterns can lead to a better understanding of the signal and energy effects of EOD lightings.