Uncovering the molecular signature underlying the light intensity-dependent root development in Arabidopsis thaliana

Background Root morphology is known to be affected by light quality, quantity and direction. Light signal is perceived at the shoot, translocated to roots through vasculature and further modulates the root development. Photoreceptors are differentially expressed in both shoot and root cells. The light irradiation to the root affects shoot morphology as well as whole plant development. The current work aims to understand the white light intensity dependent changes in root patterning and correlate that with the global gene expression profile. Results Different fluence of white light (WL) regulate overall root development via modulating the expression of a specific set of genes. Phytochrome A deficient Arabidopsis thaliana (phyA-211) showed shorter primary root compared to phytochrome B deficient (phyB-9) and wild type (WT) seedlings at a lower light intensity. However, at higher intensity, both mutants showed shorter primary root in comparison to WT. The lateral root number was observed to be lowest in phyA-211 at intensities of 38 and 75 μmol m − 2 s − 1. The number of adventitious roots was significantly lower in phyA-211 as compared to WT and phyB-9 under all light intensities tested. With the root phenotypic data, microarray was performed for four different intensities of WL light in WT. Here, we identified ~ 5243 differentially expressed genes (DEGs) under all light intensities. Gene ontology-based analysis indicated that different intensities of WL predominantly affect a subset of genes having catalytic activity and localized to the cytoplasm and membrane. Furthermore, when root is irradiated with different intensities of WL, several key genes involved in hormone, light signaling and clock-regulated pathways are differentially expressed. Conclusion Using genome wide microarray-based approach, we have identified candidate genes in Arabidopsis root that responded to the changes in light intensities. Alteration in expression of genes such as PIF4, COL9, EPR1, CIP1, ARF18, ARR6, SAUR9, TOC1 etc. which are involved in light, hormone and clock pathway was validated by qRT-PCR. This indicates their potential role in light intensity mediated root development. Electronic supplementary material The online version of this article (10.1186/s12864-019-5933-5) contains supplementary material, which is available to authorized users.


Background
Light is an essential parameter for the optimal growth and survival of plants. The quality, quantity, direction and duration of light are important factors required for various aspects of plant development [1]. Root development comprises of different aspects such as primary root elongation, lateral root elongation, lateral root branching, root geotropism, root hair formation etc. Root patterning beneath the soil plays a crucial role in penetration, anchorage and gravitropism leading to absorption of water and nutrient. To perceive light, plants have evolved with many canonical photoreceptors such as phytochromes (PHYs), cryptochromes (CRYs), phototropins (PHOTs) and UVB-resistance 8 (UVR8) [2,3]. Light regulates the patterning of shoot as well as root system [4]. It has been shown to regulate all the aforesaid aspects of root development at different stages of plant life-cycle [5][6][7][8]. Light signal from shoot can translocate through phloem to root and alters light-mediated responses [9]. Although, roots grown beneath the soil generally don't experience direct light, still few parts of the root are exposed to some amount of light seeping through the cracks, pores of the soil and affects the overall root development. Light percolates through the soil and reaches the root which leads to the production of reactive oxygen species (ROS) and promotes root growth when present in an optimal level. Ha et.al., 2018 has shown the correlation of photoreceptors with ROSmediated root growth. The shoot localized PHYs have been reported to mediate this response through ROS accumulation in roots. A shoot-localized abscisic acid (ABA) signaling component is also shown to be involved in PHYB-mediated primary root elongation. It has been shown that when roots are exposed to light, shoot PHYs induce ABA biosynthesis and signaling mediator which further promote primary root growth. PHY controls the translocation of ABA signals from shoot to root which increase the expression of ABA INSENSITIVE 5 (ABI5) gene, which activates PEROXIDASE 1 (PER1) in the root. ABI5 encodes a basic leucine-rich zipper transcription factor while PER1 encodes a peroxidase that detoxifies ROS. When root is exposed to light, activation of PER1 leads to detoxification of ROS and maintains its level, which promotes root growth. This showed that PHYB promotes primary root growth through regulating ABI5 and PER1 activity and ROS accumulation [10]. ELON-GATED HYPOCOTYL 5 (HY5), one of the major transcription factors, downstream to PHYs has been shown to be a mobile signal. It has been documented that, HY5 is activated in shoot, translocate to roots and regulates the root architecture [11]. Intrinsic component such as phytohormones are also involved in regulating root patterning. Auxin is one of the important phytohormones that plays a major role in root development. HY5 has also been shown to be a mediator of PHY and auxin signaling. Along with phytohormones, nutrients such as sugar also plays an important role in root development. The cross-talk of sugar and phytohormone such as auxin has been shown to modulate the root growth and development [12]. MEDIATOR (MED) complex is one of the most important candidate that couples sugar and auxin signalling pathways in root development. In support of this, MED12 and MED13 genes have been shown to promote primary root length, root hair number and root hair length by enhancing the cell elongation, cell division and auxin response. Addition of sucrose compliments the root defect in med12 and med13 mutants. Thus, MED12 and MED13 are the important candidate genes which link auxin signaling and nutritional status of the root [13]. On the contrary, med18 mutant has been reported to show shorter primary root, lesser number of lateral roots with longer and denser root hairs. The alteration of root architecture in med18 mutant is because of altered auxin response and its distribution in primary root. Although in natural condition, roots grow under relative darkness still there is always some communication and signal translocation between shoot to root, historically known as light piping. In med18 mutant, it has been shown that shoot perceives the light and causes the death of root meristem cells. However, cell death at root meristem occurred irrespective of direct light exposure to the roots, which suggested a long-distance communication between shoot and root is plausible. This indicated that root growth is affected in similar fashion irrespective of its light irradiation [14]. There are very few available reports which document the light and hormone cross-talk in root development [15][16][17][18]. The light quality and quantity both affect the root patterning [19]. Although, the effect of light intensity on regulation of root development has not been well investigated at molecular level and needs a systematic study.
Nicotiana tabacum plants grown under variable intensity of WL for different durations, showed altered root growth, leaf biomass, sugar content and chlorophyll level. It has been reported that, plants constantly grown either for 14 or 18 days under 60 μmol m − 2 s − 1 (condition A) light, they have lesser fresh weight of root and shoot as compared to plants grown under 300 μmol m − 2 s − 1 (condition B) light. However, plants grown for first 14 days under 60 μmol m − 2 s − 1 light and then subsequent 4 days under 300 μmol m − 2 s − 1 light (condition C), after a total of 18 days of light treatment, the fresh weight of root and shoot were observed to be intermediate of that of continuous 18 days under 60 μmol m − 2 s − 1 and 300 μmol m − 2 s − 1 light intensity. This suggested an additive effect of light fluence and duration. Further, glucose, fructose and sucrose levels were observed to be highest in plants grown under condition C compared to condition A and B. Then, it was concluded that the promotion of root growth under higher light intensity is due to enhanced carbohydrate transport from shoot to root and independent of sugar content [20]. In tomato, Solanum lycopersicum cyclophilins (SlCyp1) are peptidyl-prolyl cis/trans isomerases which play an important role in plant development. It is transported from shoot to root as a phloem mobile signal. The trafficking of SlCyp1 is enhanced with increasing light intensity, leading to profound root growth. [21]. It has been reported that, in gymnosperms such as Pinus sylvestris L. (Scots pine), plants grown in the presence of different intensities of red (R) and far-red (FR) light (1, 10, 25 and 100 μmol m − 2 s − 1 ) showed variation in root growth. Irradiation of R light had no significant effect on root length whereas low intensity of FR light (1 and 10 μmol m − 2 s − 1 ) has resulted shorter root in comparison to the seedlings grown under complete darkness. However, under highest intensity of FR light, root length was significantly higher as compared to other light conditions. Therefore, it can be concluded that equal intensity of monochromatic light didn't show similar type of effect on plant development [22]. In a recent report, Arabidopsis thaliana seedlings grown in two different experimental setups, in one the roots were directly exposed to WL while in another, root was covered and the shoot was exposed to light (D-root system). In this report, it has been shown that the spatial expression of the photoreceptor genes such as UVR8, CRY1, CRY2, PHOT1, PHOT2, PHYA and PHYB varied in different segments of roots under the above mentioned experimental setups [23]. Although light intensity has been shown to influence the root development, how it correlates with the gene expression, leading to root patterning needs to be examined in detail. In the present work whole seedlings were grown under 38, 75, 112 and 150 μmol m − 2 s − 1 intensities of WL with 16 h light and 8 h dark. The root patterning was analysed in WT and PHY mutants. Using microarray-based approach, we analysed the variation in the global transcription profile of the root tissue of WT seedlings. Further, we evaluated the expression pattern of various genes that play important role in light signaling, hormone signaling and clock-regulated pathways.

Results
Phytochrome mutants showed variation in root patterning under different intensities of white light Arabidopsis seedlings, WT, phyB-9 and phyA-211 showed variable root growth responses under different light intensities of 38, 75, 112 and 150 μmol m − 2 s − 1 WL (Fig. 1a). Primary root length in 6-days old seedlings was found to be slightly shorter in case of phyA-211 as compared to phyB-9 and WT under 38 μmol m − 2 s − 1 light intensity. Under 150 and 112 μmol m − 2 s − 1 light, the primary root length was observed to be marginally shorter in both the PHY mutants as compared to WT (Fig. 1b). It was also observed that lateral and Fig. 1 Light intensity dependent root growth varies in phytochrome mutants. a Phenotypic differences in root architecture. b Primary root length (mm) for WT, phyB-9 and phyA-211 under four different WL intensities of 38, 75, 112 and 150 μmol m − 2 s − 1 . The analysis was done in 6-days old seedlings. Error bars represent SE. The means were compared using Tukey test with p ≤ 0.05. Same letters denote statistical significance. There were five technical replicates, each replicate consists of 20 seedlings. Scale bar = 10 mm adventitious root growth varied in PHY mutants under different light intensity in comparison to WT (Fig. 2a). The zoomed image for qualitative details of lateral root growth has been shown in Additional file 1: Figure S1. When light intensity increased from 38 to 150 μmol m − 2 s − 1 , the number of lateral roots approximately increased by two times in case of WT, phyB-9 and three times in phyA-211. However, under 112 μmol m − 2 s − 1 light, phyB-9 has significantly lesser number of lateral root and about 16.7 and 30.3% reduction were observed in comparison to 75 and 150 μmol m − 2 s − 1 light intensity respectively. The lateral root number in phyA-211 decreased by 12.5 and 35.7% as compared to WT under 38 μmol m − 2 s − 1 and 75 μmol m − 2 s − 1 light respectively (Fig. 2b). Interestingly, under 112 μmol m − 2 s − 1 light, the lateral root number in case of phyB-9 was reduced by 39.5 and 24.5% in comparison to WT and phyA-211 respectively. At higher intensity of 150 μmol m − 2 s − 1 , no significant difference was found in lateral root number as all the genotypes showed~14-16 rootlets (Fig. 2b).
The number of adventitious roots was shown to increase by increasing light intensity till 75 μmol m − 2 s − 1 and apparently saturates in between 112 μmol m − 2 s − 1 and 150 μmol m − 2 s − 1 . It has been observed that the adventitious root number was least in case of phyA-211 as compared to WT and phyB-9 under all light intensities (Fig. 2c). The adventitious root number in case of phyA-211 was least under 38 μmol m − 2 s − 1 as compared to other light intensities. The microscope images of lateral root growth for qualitative visualization in 6days old seedling have been shown in Additional file 2: Figure S2.

Identification of differentially expressed genes under four different intensities of white light in WT root
In all the genotypes root architectural differences were observed to follow a pattern when grown under different Fig. 2 Phytochrome mutants show variability in light intensity dependent lateral and adventitious root growth. a Lateral and adventitious root architectural differences. The lateral and adventitious root have been indicated by arrowheads and arrows respectively. b Phytochrome mutants show variation in lateral root number. c Adventitious root number varies among the genotypes under all four WL intensity. The analysis has been done in 6-days old seedlings. Error bars represent SE. The means were compared using Tukey test with p ≤ 0.05. Same letters denote statistical significance There were three technical replicates, each replicate consists of 20 seedlings. Scale bar is 10 mm intensities of WL. With this information, microarray was carried out with root samples of 5-days old WT seedlings grown under four different WL intensities (mentioned earlier). Differential expression analysis was performed to identify the differentially expressed genes (DEGs) from microarray data. The DEGs having minimum fold change (FC) of~1. 2  When all light intensities were investigated with respect to 75 μmol m − 2 s − 1 light, 65 common DEGs were found. The total number of common DEGs was 136, when 150 vs 38, 112 vs 38, 75 vs 38 μmol m − 2 s − 1 light intensity were taken into consideration. CYTOCHROME P450 81F2 (CYP81F2), At2g39445, At5g22555 and At2g44130 genes have been found to be common and altered under all six comparative light conditions. CYP81F2 is a membrane-localized protein, known to play a role in indole glucosinolate biosynthesis and offers resistance to fungus named Plectosphaerella cucumerina [24]. At2g39445 encodes for phosphatidylinositol n-acetylglucosaminyltransferase, At5g22555 is a putative transmembrane protein while At2g44130 encodes a F-box protein KISS ME DEADLY 3 (KMD3). KMD3 has been reported to be induced by Meloidogyne incognita (root-knot nematode) and makes the plant susceptible towards this nematode [25]. The genes which were unique and common in two, three, four or five comparative light conditions have been summarized in supplementary excel files (Additional files 3 and 4).
Gene ontology enrichment analysis of over-represented differentially expressed genes Gene ontology (GO) enrichment analysis was performed to understand how light intensity affects different biological phenomena and processes in roots. GO analysis is based on the gene products and the related functions at molecular and cellular level with available literature databases. The analysis type used was PANTHER Overrepresentation Test and annotation version was GO Ontology Database [26,27]. The categorization has been done on the basis of Fisher's exact with FDR multiple test correction type. GO (Fig. 5f ) light intensities. The composite figure describing detailed GO analysis of DEGs has been presented in Additional file 5: Figure S3.
All the GO categories and terms which were highly enriched or over-represented under different light conditions have been summarized in Table 2. In case of 150 vs 112 μmol m − 2 s − 1 light condition, significant DEGs were found only under CC category, plasma membrane and cell periphery were the most affected GO terms. When the DEGs were analysed for 150 vs 75 μmol m − 2 s − 1 light, the enriched DEGs were categorized under CC and MF domains. Under this light condition, the most enriched terms were cell, cytoplasm and membrane under CC and Kyoto encyclopedia of genes and genomes colour pathway analysis of DEGs to investigate the gene functionality Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis represents the gene function and their utilities in biological, cellular and molecular process [28]. This analysis is based on the present database consisting of various genome sequencing, bioinformatic information etc. The DEGs obtained from microarray were further subjected to KEGG colour pathway analysis. The total number of DEGs under 150 vs 112 μmol m − 2 s − 1 light condition was 1014 and only 184 DEGs were identified with KEGG analysis. Under 150 vs 75 μmol m − 2 s − 1 light, 1424 DEGs were obtained from microarray data. In this light condition, the number of DEGs identified through KEGG colour pathway was 290. In case of 150 vs 38 μmol m − 2 s − 1 , 420 genes were identified by KEGG out of 1789 DEGs obtained from microarray analysis. Under 112 vs 75 μmol m − 2 s − 1 light, 1163 DEGs were found from microarray, out of these, 249 genes were identified through KEGG analysis. In case of 112 vs   The detailed KEGG analysis of DEGs has been presented in Table 3. This analysis showed that the metabolic pathway was the major ones affected with largest number of genes under all comparative light intensities.

Validation of microarray result by qRT-PCR for DEGs involved in hormonal, light signaling and clock regulated pathways
We further validated the microarray result by qRT-PCR of few selected genes. Selection of genes for qRT-PCR validation was based on their i) expression level data obtained from microarray analysis and ii) significance and potential role in root development as per the KEGG pathway analysis. Therefore, the genes invoved in hormone signaling, light signaling and clock-regulated pathways were chosen for qRT-PCR validation. The root development is affected by various factors such as light, phytohormones, circadian clock etc. Photoreceptors such as PHYA and PHYB play major role in root positive phototropism [7]. PHYs have also been shown to be involved in root growth as well as its gravitropic response [5]. The phytohormones also regulate the lateral root initiation in a dose dependent manner [29]. They have also been shown to control root hair initiation and its growth [30]. High intensity of light and exogenous auxin stimulate adventitious rooting in Eucalyptus [31]. On the other hand, circadian clock genes have also been shown to regulate lateral root emergence [32].  Genes such as AUXIN RESPONSE FACTOR 2 (ARF2), AUXIN RESPONSE FACTOR 4 (ARF4) and AUXIN RE-SPONSE FACTOR 18 (ARF18), LIKE AUX 2 (LAX2), SMALL AUXIN UPREGULATED RNA 9 (SAUR9), SMALL AUXIN UPREGULATED RNA 26 (SAUR26) and IAA7 were selected to evaluate the effect of WL intensity on auxin homeostasis, while, TYPE A RESPONSE REGULATOR 6 (ARR6), KISS ME DEADLY 1(KMD1) and COP1 INTERACTING PROTEIN 1 (CIP1) were analysed to understand the impact of WL intensity on cytokinin and ABA signaling, respectively. Genes involved in light signaling such as PHYTOCHROME RAPIDLY REG-ULATED 2 (PAR2), HY5, PHYTOCHROME INTERACT-ING FACTOR 4 (PIF4), EARLY PHYTOCHROME RESPONSE 1 (EPR1), CONSTANS LIKE 3 (COL3), CON-STANS LIKE 9 (COL9), COP9 SIGNALOSOME COM-PLEX SUBUNIT 6A (CSN6A) and COP9 SIGNALOSOME COMPLEX SUBUNIT 6B (CSN6B) were selected to investigate the correlation of WL intensities with expression level of downstream light signaling components. Three clock-associated genes such as CIRCADIAN CLOCK-ASSOCIATED 1 (CCA1), TIMING OF CAB EXPRESSION 1 (TOC1) and PSEUDO RESPONSE REGULATOR 9 (PRR9) were also chosen to understand the response of light quantity on circadian clock regulation. The possible functions of all the above-mentioned genes and respective primers have been compiled in Table 4 (Figs. 6, 7, 8, 9, 10 and 11). The FC of gene expression data has been summarized in Table 5.
In comparison to 75 μmol m − 2 s − 1 under low light intensity of 38 μmol m − 2 s − 1 , the expression of CSN6A, CSN6B, EPR1, HY5 and COL9 genes was higher Negative regulator of shade avoidance syndrome responses and transcriptional repressor of SAUR15 and SAUR68 [33].

ARR6
Type-A response regulator, acts as negative regulator of cytokinin signaling [34].

HY5
Transcription factor playing downstream to photoreceptor signaling, promotes photomorphogenesis. It is also involved in root greening and gravitropism [15,35].

KMD1
Type-A response regulator which targets type-B response regulator and acts as negative regulator of cytokinin signaling [39].

EPR1
Transcriptional factor playing role in PHYA-mediated cotyledon opening and it is regulated by circadian clock [42].

COL3
Zn finger protein which is a positive regulator of red light signaling and photomorphogenesis. Also regulates root and shoot development [43].

CSN6A
Component of COP 9 signalosome complex regulating ubiquitin conjugation pathway and plays role in repression of photomorphogenesis in dark [48].

CSN6B
Component of COP 9 signalosome complex regulating ubiquitin conjugation pathway and plays role in repression of photomorphogenesis in dark [48].

CCA1
Morning loop gene which functions along with LHY1 and it represses TOC1 [49].

TOC1
Evening complex gene, regulates plant fitness by controlling the clock output through repressing morning and evening loop of circadian clock [50]. ( Fig. 11.1a, .1b, .1c, .1d and .1e) and hormone pathway genes such as ARF18, CIP1, KMD1 and LAX2 were also upregulated ( Fig. 11.2a, .2b, .2c and .2d). CCA1 expression was also shown to be upregulated under this low light condition (Fig. 11.3). The genes which were induced or repressed under low light intensity in comparison to high intensity of light have been summarized in Table 6. Further to validate the transcriptomics data, we performed the qRT-PCR for ARR6 and CIP1 genes in  (Fig. 6.4a and .4b) whereas ARR6 has been shown to be significantly upregulated in case of 35S::PhyBGFP (Fig. 6.4b). When the expression of these two genes were analysed in presence of 75 μmol m − 2 s − 1 in comparison to 150 μmol m − 2 s −1 light intensity, ARR6 was downregulated in all three genotypes (Fig. 7.4a) and CIP1 showed downregulation only in case of phyB-5 genotype (Fig. 7.4b). Similarly, under low light of 38 μmol m − 2 s − 1 intensity in comparison to 150 μmol m − 2 s − 1 , both these genes were downregulated. However, CIP1 did show a significant downregulation only in case of Ler and phyB-5 (Fig. 8.4a and .4b).

Discussion
Although earlier studies have investigated the effect of light intensity as well as light quality on root development, yet the molecular players that fine tune the output response are poorly understood [1,17]. Light signaling mediated through phytohormones modulate plant development [52]. Under shade, where the R:FR ratio is very low, the light is primarily sensed by PHYs and their downstream factors [53]. This leads to change in plant phenotype such as elongated hypocotyl, longer petiole, small leaves, apical hook formation, etc. On the other hand, phytohormones such as gibberellin (GA), auxin (IAA), ethylene (ET) and brassinosteroid (BR) have been shown to regulate plant development under relative dark and shade conditions. GA promotes hypocotyl elongation and suppresses other photomorphogenic features, these phenotypic differences appear through inactivation of HY5 and other light signaling transcription factors such as PIFs. Light controls GA synthesis by downregulating GA biosynthetic enzymes and stimulating GAinactivation enzymes [54]. It has also been reported that shade (low R:FR) induces hypocotyl growth via enhanced TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOP-SIS 1 (TAA1)-dependent auxin biosynthesis [53]. PIF5 promotes ethylene-mediated apical hook formation, one of the characteristics of shade avoidance syndrome in dark [55]. Along with light, circadian clock also regulates phytohormone signaling and synthesis. Ethylene production is regulated by circadian clock genes such as TOC1 and CCA1, similarly ACC SYNTHASE 8 (ACS8) is also controlled by light and circadian clock [56]. Auxin signaling and responses are also regulated by circadian clock [57]. It has been reported that cytokinin t-zeatin treatment induces expression of TOC1, GIGANTEA (GI) in morning and CCA1 in the evening. Active NAA or ABA treatment has been shown to downregulate CCA1 expression in morning [58]. TOC1 and other circadian clock-associated genes have been reported to control the expression of genes involved in GA biosynthesis [59]. Clock regulates the transcript level of PIF4 and PIF5 genes [60]. In the present work, we have shown that different WL intensities (150, 112, 75 and 38 μmol m − 2 s − 1 ) show variable effects on the root development in case of WT and PHY mutants. Light quantity influenced the root architecture as well as the expression profile of the genes playing role in light signaling, phytohormone related and clock-regulated pathways. It has been reported that, although only aerial shoot portion perceive light, some amount of light gets translocated down through light piping and influences root patterning [9]. In the current work, we explored the candidate genes involved in light intensity-based root patterning over direct exposure of seedlings to light.

Primary, lateral and adventitious roots are affected by different intensities of white light
Different intensities of light as well as their quality have been shown to affect root development [20][21][22]17]. PHYs sense the quality as well quantity of light and have been shown to influence phototropism, gravitropism and elongation of root [5]. Under shade condition, where the ratio of R:FR is reduced, the hypocotyl elongates, however, the effect of shade on root development is poorly understood [61]. We have shown here that under different light intensity, the root patterning changes significantly. Primary root elongation was more influenced in case of PHY mutants. Higher light intensity (150 μmol m − 2 s − 1 ) induced larger number of lateral as well as adventitious roots. PHYB predominantly affected the adventitious root growth whereas PHYA has been shown to be

Light intensity influences root development and also affects the gene expression profile in root
When light signal translocates through the shoot to the root or seedling directly exposed to light, the gene expression profile of root changes differently. It has been shown that the spatial expression profile of photoreceptors changes throughout the root when shoot is directly exposed to light [23]. With microarray analysis, we found about 5243 DEGs under four different light intensities (150, 112, 75 and 38 μmol m − 2 s − 1 ) and highest and lowest number of DEGs were detected under 150 vs 38 μmol m − 2 s − 1 light condition and 150 vs 112 μmol m − 2 s − 1 light intensity respectively. The change of light intensity from high to low, severely affected the gene expression pattern of root whereas the slight change of 150 to 112 μmol m − 2 s − 1 light didn't have significant effect on gene expression profile. Interestingly, CYP81F2 (Cytochrome P450 in indole glucosinolate biosynthesis), At2g39445 (phosphatidylinositol n-acetylglucosaminyltransferase), At5g22555 (transmembrane protein) and At2g44130 (F-box protein) genes were altered in all comparative light intensity. The functions of these candidate genes should be studied in detail as this can correlate the light intensity-based root development with glucosinolate biosynthesis, nucleotide sugar and glycolipid signaling and SCF ubiquitin-mediated pathways.
GO enrichment and KEGG colour pathway analysis of DEGs highlight the potential role of specific pathways involved in light intensity-based root development GO analysis suggested about functionality of the genes altered under variable light intensities, their specific locations and their involvement in the major biological, cellular and molecular processes. The change in light intensity affected the root development possibly by influencing the catalytic activities of the genes present in root. Different intensities of light also altered the expression of gene predominantly localized in cytoplasm, cell periphery and membrane part of the cell. KEGG analysis has shown that various pathways such as metabolic, secondary metabolite biosynthesis, ribosome-mediated, carbon metabolism, plant hormone signaling, protein processing in ER, MAPK signaling, amino acid biosynthesis, starch and sucrose metabolic, oxidative phosphorylation, purine metabolic, amino and nucleotide sugar metabolic, etc. were altered in roots on irradiation of different WL intensities.
Light intensity affects the expression pattern of genes involved in the light signaling, hormone related and clock-regulated pathways in roots Earlier reports showed that light affects root development through photoreceptors. PHYA promotes root elongation under R, FR, B while PHYB enhances root length under R light only [61,18]. Lateral root growth is promoted by PHYA, PHYB and PHYE whereas the high light sensor PHYD inhibits this phenomenon [4]. Root hair formation is also promoted by light where PHYA and PHYB have been shown to stimulate root hair initiation [62]. Therefore, few genes known to play major role in light signaling were analysed to investigate that These genes are most probable candidate for sensing the light quality along with quantity to affect the root architecture. Expression of PHYs didn't change, however, their downstream genes were found to be altered with different light intensities irradiated on root. Light affects the hormone synthesis, signaling and transport and known to be involved in root development [63]. Hence, few genes involved in auxin, cytokinin and ABA pathways were analysed in this current work. The expression level of auxin responsive factors such as ARF2 and ARF4 was observed to be maximum under 112 μmol m − 2 s − 1 light. Auxin influx carrier; LAX2, negative regulator of cytokinin signaling; ARR6 and a positive regulator of ABA signaling; CIP1 have shown maximum expression under low light intensity of 38 μmol m − 2 s − 1 . Specific light intensity may alter certain metabolic processes that might account for the observed output gene regulation. Early auxin-responsive genes such as SAUR9 and SAUR26 were upregulated with decreasing light intensity and their expression was maximum at 38 μmol m − 2 s −1 light. This differential gene expression profiling showed that various hormonal pathways are also influenced by different light quantity that further lead to the differences in root architecture. Circadian clock has a major role in plant development and it has been reported to affect root growth [64,32]. Clock genes such as CCA1, PRR9 and TOC1 were analyzed in this current work and it was observed that, the expression of CCA1 and TOC1 was maximum under 38 μmol m − 2 s − 1 and 150 μmol m − 2 s − 1 light intensity respectively. They are associated with evening and morning complex and make sense to peak at low light intensities normally represented as dawn and dusk.  photoreceptor at a given temperature can perhaps add an extra layer of complexity to generate a transcriptomics pattern as a read-out. The investigation of genes involved in other phytohormone signaling, synthesis and transport under variable light intensities could be carried out. Few of the light signaling genes such as PIF5, FAR-RED ELONGATED HYPOCOTYL 1 (FHY1), FHY1-LIKE (FHL), CONSTITU-TIVE PHOTOMORPHOGENESIS 1 (COP1) etc. could also be analysed with respect to light intensity based root growth. The expression profiling of the candidate genes playing role in shade avoidance syndrome like HY5, LONG HYPOCOTYL IN FAR-RED LIGHT 1 (HFR1), Arabidopsis thaliana HOMEOBOX PROTEIN 4 (ATHB4) etc. could also be studied under changing light intensity [53]. The expression of these genes could be analysed in the shoot as well as root under the four WL intensities mentioned or other suitable intensities. These studies could be extended by performing different sets of experiments with variable intensity of R, FR, B light and then analyzing the expression of genes involved in light, phytohormone signaling, clock regulation as well as other related pathways.

Conclusion
Our current work showed the effect of different WL intensity on the root phenotype of WT as well as phyA and phyB mutants. The primary root length was shorter in case of phyA-211 under all four variable light intensities, however under higher intensities of 112 and 150 μmol m − 2 s − 1 both the PHY mutants had shown shorter primary root in comparison to WT. phyB-9 had shown more adventitious roots in comparison to phyA-211 and WT under same light intensity. Under 38 and 75 μmol m − 2 s − 1 light intensity, phyA-211 had lesser number of lateral roots whereas under 112 μmol m − 2 s − 1 light, phyB-9 showed lesser number of lateral roots in comparison to phyA-211 and WT. Both adventitious as well as lateral root number were similar in all genotypes under 150 μmol m − 2 s − 1 . Higher intensity of WL such as 150 μmol m − 2 s − 1 and lower intensity of light such as 38 μmol m − 2 s − 1 had the most significant effects on the gene expression profile of root as the number of DEGs was found to be highest in case of 150 vs 38 μmol m − 2 s − 1 light. KEGG colour analysis pathway has shown that most of the genes found from microarray belonged to metabolic, secondary metabolite synthesis, carbon metabolic and plant hormone signaling pathway. Genes such as PIF4, EPR1, COL3, CIP1, TOC1 etc. were shown to be differentially expressed under almost all WL intensities. These genes may act as possible candidates in correlation of light intensity dependent root development. This study showed that different quantity of light affects the transcript abundance of light, hormone and clock pathway genes in the root tissues, indicating their potential role in light intensity mediated root development. Transcriptomics profiling using laser capture microdissection at single cell/tissue level could be valuable and would add in depth understanding of the regulatory network.

Plant material and growth conditions
Columbia Wild type (WT), phytochrome mutants such as phyB-9 and phyA-211 seed lines of Arabidopsis thaliana were used for root architectural studies. For microarray analysis, WT root samples were used. The initial qRT-PCR validation was performed with WT and for further transcriptome validation, Landsberg erecta (Ler) wild type, overexpressor line, 35S::PhyBGFP (Landsberg erecta) and mutant line, phyB-5 (Landsberg erecta) of PHYB were used. These seed lines were a kind gift from Prof. Eberhard Schäfer, Albert Ludwigs University of Freiburg, Germany [65,66]. The seeds were surface sterilized with 70 and 100% ethanol sequentially and dried under sterile bench. The sterilized seeds were sown on half-strength Murashige and Skoog salt (Himedia, Cat. No. PT021) supplemented with 1% sucrose (w/v) (MP biomedical, Cat. No. 194018) and 0.8% agar (w/v) (Himedia, Cat. No. PCT0901) in square plates and then kept for 72 h of cold stratification at 4°C in dark. The plates were kept vertically at 90°and grown under four different WL intensities (Long day, 16 h light/8 h dark) in the range of 150, 112, 75 to 38 μmol m − 2 s − 1 light in Percival light cabinets (Model No. CU36L6) at 22°C with relative humidity of~70%. Root phenotyping and RNA extraction were done with 6-days old and 5-days old seedlings respectively. Sample harvesting was approximately done at ZT = 12. The germination induction resulted in uniform germination among all genetic lines tested and confirmed by light microscopy, after 2 days of transferring in light. The root growth comparison was performed in 6-days old seedlings.

RNA extraction and microarray analysis
RNA was extracted from the root tissues of 5-days old WT seedlings grown under different light intensities (38,75,112 and 150 μmol m − 2 s − 1 ) with TRIzol reagent (Thermofisher Scientific, Cat. No.15596026), following the manufacturer's protocol. RNA samples were subjected to DNaseI treatment with TURBO DNA free kit (Thermofisher Scientific, Cat. No. AM1907). For microarray analysis, the RNA quality was analysed with the Bioanalyzer (Agilent Bioanalyzer 2100 system). cDNA was prepared from RNA samples with RevertAid H Minus First Strand cDNA Synthesis Kit (Thermofisher Scientific, Cat. No. K1631). Microarray analysis was performed following Affymetrix GeneChip® WT PLUS Reagent Kit Manual Target Preparation for GeneChip® Whole Transcript (WT) Expression Arrays. Microarray experiment was done at ILS laboratory, Gurgaon, Haryana, India. Two independent biological replicates for each condition were used for microarray experiment in addition to this, three technical replicates were utilized. DEGs obtained from microarray were analysed with One-Way Between-Subject ANOVA (unpaired) test. Gene Level Differential Expression Analysis was performed and genes having minimum FC of 1.

Venn diagram, gene enrichment (GO) pathway analysis and KEGG analysis
Venn diagrams were prepared for overlapping genes considering a combination of three comparative light conditions at a time. Gene lists from the microarray data were sorted and used for plotting Venn diagram using a custom script for R [67] using packages gdata [68], Vennerable [69], gplots [70], RBGL [71] and graph [72]. The script is available online on GitHub https://github.com/debaduttapatra/Venn-Diagram. GO pathway analysis was performed using Panther Over-representation test and the annotation version is GO Ontology database. The FISHER test was used and the categorization of DEGs under different GO terms was done considering FDR < 0.05. The KEGG search and colour pathway analysis was performed considering only a small group of DEGs.

qRT-PCR analysis for DEGs
For quantitative qRT-PCR analysis, an aliquot of the cDNA samples was prepared from the same RNA used for microarray of WT. For validation of transcriptome data, qRT-PCR was performed with Ler, 35S::PhyBGFP and phyB-5 root samples. The qRT-PCR was performed using Biorad Evagreen kit (Bio-Rad SSoFast EvaGreen Supermix, Cat. No. 172-5203) and Bio-Rad C 1000 Touch™ thermal cycler was used for qRT-PCR (CFX384™ Real Time System). The total volume of the reaction was 10 μl with 25 ng cDNA for WT, 50 ng cDNA for Ler, 35S:: PhyBGFP and phyB-5 and each primer of 5 μM concentration was used. The software used for data analysis was Bio-Rad CFX Manager 3.81, the gene expression was analysed with FC of 1.2 and p-value < 0.05. The transcript level was normalized with a pair of primer specific for POLYUBIQ gene. qRT-PCR was done at least in triplicate and the data shown represent the mean + SE.
The forward and reverse primers are listed in the supplementary data (Additional file 6).