Columbia WT and T-DNA insertions originally obtained from ABRC/TAIR were utilized (Alonso et al., 2003). Null mutant insertions for Pirin1 (PRN1; At3g59220; SALK_006939) were utilized as controls in some experiments, and T-DNA insertion mutant line sig6-1 (AT2G36990; SAIL_893_C09) was used in the RNA-seq experiment. Null ADT mutants were tested in genetic and developmental assays, including ADT1 (At1g11790; SALK_138343; SALK_124232), ADT3 (At2g027820; SALK_029949; SALK_071907), ADT4 (At3g44720; SALK_065483; SALK_123367), ADT5 (At5g22630; SALK_088171; SALK_028611), and ADT6 (At1g08250; SALK_030329; SALK_109552). All mutant phenotype responses were confirmed in two accessions; the more common null mutant utilized in research (accession listed first) was featured in the data shown. Bulk seed stocks were grown as previously reported (Orozco-Nunnelly et al., 2014). T-DNA insertion phyA and phyB mutants used in these studies were previously isolated and described (Mayfield et al., 2007; Ruckle et al., 2007). The phyAphyB mutant was obtained from a genetic cross of the single mutants as described (Oh and Montgomery, 2013). Seed lines completely null for ADT2 (At3g07630) have not been available. Recently, it has been proven that ADT2 null mutation is lethal because ADT2 is essential for proper seed development (El-Azaz et al., 2018), and hence, ADT2 is not studied herein.

Seeds were surface sterilized and plated on 1X MS containing 1% sucrose and 0.7% Phytoblend (Caisson Laboratories, UT) essentially as previously described (Warnasooriya and Montgomery, 2009). Plated, sterilized seeds were cold stratified during imbibition in darkness at 4°C for 4 days. All experiments were conducted at constant temperature and humidity under continuous light in controlled-environment chambers. Light sources were those previously described (Warnasooriya and Montgomery, 2009) with continuous red, R (Rc; λ_max ~ 670 nm), continuous far-red, FR (FRc; λ_max ~ 735 nm), and continuous blue, B (Bc; λ_max ~ 470 nm) at the indicated fluences in µmol m⁻² s⁻¹. Continuous darkness is signified by Dc or 7D0L.

For light-dependent hypocotyl elongation experiments, hypocotyl lengths of seedlings grown under defined light conditions were measured by scanning the seedling images and quantifying lengths using ImageJ software (Plug-in used to stitch pictures that required two images to get whole seedling). We measured 25 seedlings in each of three independent biological replicates, resulting in 75 measured seedlings per line. For vertical plate assays, hypocotyl and root lengths were compared to a scanned ruler image taken at the same time by scanning and photography on day 7, post-stratification. After hypocotyl and root lengths were measured, the mean ratio of hypocotyl length:root length was determined.

Seeds were surface sterilized and plated on 0.5X MS (pH 5.8) with no sucrose and no added vitamins as described (Voll et al., 2004) on square (positioned vertically) petri plates or phytatrays, as detailed herein. For germination assays on phytatrays, sterilized seeds were mixed with 0.8% low melt agarose (0.5X MS; top agarose) with the addition of 1 μM ACC or 500 nM ABA or 1 μM IAA or 1 μM GA or 0.5X MS (control); then, a 100-μl aliquot of seed and top agarose was spotted onto a 0.5X MS phytatray base plate, where germination assays were performed similarly to Kim et al. (2019). Plated, sterilized phytatrays were placed into light-tight black Plexiglas boxes, sealed with aluminum foil, and then were cold stratified in a dark cold room at 4°C for 48 h (Kim et al., 2019). After 48 h, the plates were moved to one of several locations: to a 24-h, 20°C dark environmental room, or a white light environmental room (Orozco-Nunnelly et al., 2014) for varying numbers of days, ranging from 0 to 7 days, depending on the experiment. White light sources, as well as dim green light for handling dark-grown seedlings, have been described (Voll et al., 2004); red and blue Plexiglas and filters used for supporting (low fluence) work have been described (Warpeha et al., 1989), where the fluence rate received by the vertical plates determined by a LiCor meter was 3 (Rc) and 7 (Bc) μmol m⁻² s⁻¹, respectively. All accessions of mutants were tested in these experiments; one representative accession is featured in the figures.

ADT construct development and subsequent plant transformation occurred as has been described (Orozco-Nunnelly et al., 2014). Standard molecular biology techniques and the Gateway system (Invitrogen) were used in the cloning procedures of the ADT family members described herein. Purified WT Arabidopsis cDNA was used to generate the ADT full ORF fragments for each ADT member (except ADT2 as aforementioned), where each ORF was represented by ADT1 [1,179 bp], ADT4 [1,275 bp], ADT5 [1,278 bp], and ADT6 [1,242 bp], and where ADT3 cloning of the ORF was described previously (Para et al., 2016). Each ADT ORF fragment was cloned into the Invitrogen pENTR/D-TOPO vector. LR reactions (Invitrogen) were then performed with verified (i.e., by sequencing in both directions) entry clones to obtain expression clones. The 35S promoter was cloned as described (Orozco-Nunnelly et al., 2014), and the dpGreen binary vector derivative containing a NOS terminator with a C-terminal GFP fusion and spectinomycin and BASTA resistance genes was used, resulting in each ADT construct (35S::ADTX-GFP), except for ADT3, which was cloned for an earlier study as described with the native promoter (Para et al., 2016), and separate construct utilizing the 35S promoter. All constructs were confirmed via restriction enzyme digest, PCR, and sequencing. Verified expression clones were transformed in adt mutant backgrounds via floral dip as described (Para et al., 2016). 35S::ADTX-GFP was introduced into null mutants (ADT1 [At1g11790; SALK_138343]; ADT3 [At2g27820; SALK_029949]; ADT4 [At3g44720; SALK_065483]; ADT5 [At5g22630; SALK_088171]; and ADT6 (At1g08250; SALK_030329]) via the floral dip method as described, and screened by BASTA resistance in each generation until homozygous (T3) for experimentation (Orozco-Nunnelly et al., 2014). For each ADT, the four most vigorous T3 lines were selected for experiments. Living seedlings were mounted in sterile water on slides at the same time (~9 a.m.) in the day cycle, then viewed on a Zeiss Observer.Z1 deconvoluting microscope fitted with a high-resolution camera (Axiovision 503; Zeiss, Oberkochen, Germany), using the 20× objective to view expression fluorescence, sectioned by optical apotome (1-μm optical slices, no bleed through) illuminated by XCite 120 LED (Lumen Dynamics, Waltham, MA USA) DAPI, FITC, and Texas Red LEDs. Images were managed by Zeiss Zen pro software (2012). At least 30 seedlings were viewed per experimental replicate, with three biological replicates performed. Merged images of DAPI, FITC, and Texas Red LEDs are shown.

Germination was scored as complete emergence of the radicle as described (Kim et al., 2019). Phenotypic responses on untreated (control) and experimental hormone plates were determined by comparison of mutants to Columbia WT on a Zeiss Stereo Discovery V.8 microscope at 1× using Axiovision with images recorded, and entire plates were photographed on day 4 after stratification using Nikon Coolpix on a white light box or black background. At least three sets of 30 seeds were scored for germination on phytatrays at 0 to 72 h every 4 h, post-stratification.

Anthocyanins were extracted and quantified from 5-day-old seedlings as previously detailed (Montgomery et al., 1999; Warnasooriya and Montgomery, 2009). At least 100 seedlings per line from four biological replicates were used.

Heat maps were constructed using AtGenExpress public Arabidopsis microarray datasets with mean-normalized values (www.weigelworld.org ) and visualization tools from BAR Heatmapper Plus (http://bar.utoronto.ca ). For tissue-specific data, seedlings were 7 days old grown in continuous white light (Wc). Light-pulsed seedlings were 4 days old.

The far-red block of greening (FR-BOG) experiment was performed according to Alameldin et al. (2020); Arabidopsis seeds were surface sterilized with 35% (v/v) commercial bleach containing 0.025% (v/v) SDS for 15 min. Seeds were rinsed five times with sterile distilled water and planted on growth medium containing 0.5X MS salts (Caisson Laboratories, Smithfield, Utah, USA) and 0.7% (w/v) Phytoblend (Caisson Laboratories). Seeds on solid medium were stratified for 4 days at 4°C in the dark and were then divided into two groups. One group was incubated in a Percival LED (light-emitting diode)-equipped growth chamber (model E-30LED; Percival, Perry, IA, USA) at 22°C under constant far-red LED (FRc; λ_max ~ 735 nm) light at 5 μmol m⁻² s⁻¹ for 5 days, then in a Percival environmental chamber model no. CU36LA under Wc at 100 μmol m⁻² s⁻¹ for 5 days. The second group was kept in the dark (D) for 5 days at 22°C as the control treatment, then grown under 100 μmol m⁻² s⁻¹ Wc for 5 days.

Ten-day-old seedlings were soaked in 200 µl of N,N-dimethylformamide per milligram fresh mass and kept in the dark for 24 h at 4°C (Moran, 1982). The absorbance of samples was then measured at 647 nm, 664 nm, and 900 nm (baseline control) using an Agilent 8453E UV-visible spectrophotometer (Santa Clara, CA, USA). Chlorophyll content was calculated based on previously described equations (Inskeep and Bloom, 1985). All assays were done with at least three biological replicates.

Total RNA was extracted from whole seedlings (FR BOG-treated and controls) grown at 22°C using the OMEGA E.Z.N.A Plant RNA kit (catalog no. R6827). cDNA was synthesized using a Reverse Transcription System (Quanta bio) qScript cDNA SuperMix (Omega Bio-tek, Norcross, GA, USA) and the instructions of the manufacturer. Libraries were prepared using the Illumina Stranded mRNA Library Kit and Ligation with IDT for Illumina RNA UD Indexes following the manufacturer’s recommendations except that half volume reactions were used. Completed libraries were quality checked and quantified using a combination of Qubit dsDNA HS and Agilent 4200 TapeStation HS DNA1000 assays. All libraries were normalized down to the lowest concentration and equal volumes of these normalized libraries were pooled. The pool was quantified using the Invitrogen Collibri Quantification qPCR kit. The library pool was loaded onto one lane of an Illumina NovaSeq 6000 S4 flow cell and sequencing was performed in a 2 × 150-bp paired-end format using a NovaSeq v1.5 300 cycle reagent cartridge. Base calling was done by Illumina Real Time Analysis (RTA) v3.4.4, and output of RTA was demultiplexed and converted to FastQ format with Illumina Bcl2fastq v2.20.0.

An initial quality check of the RNA reads was performed using FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ ). Trimmomatic v0.32 (Bolger et al., 2014) was used to filter the RNA reads to remove adaptors and low-quality reads. A sliding window method was used to scan the reads with 4-base wide and cut when the base quality was below a threshold of 2. The minimum read length cutoff was 100 bp. Data quality was explored after filtering with FastQC. STAR/2.6.0c (Spliced Transcripts Alignment to a Reference; Dobin et al., 2013) was used to map the RNAseq reads to the TAIR10.1 Arabidopsis thaliana genome (RefSeq assembly accession: GCF_000001735.4) with the default settings of the twopassMode Basic option with intron size 21–6,000 nt. In all samples, >90% of the RNAseq reads were mapped to the reference genome. RNA‐Seq data have been deposited to the NCBI Gene Expression Omnibus database (BioProject accession number: PRJNA991474).

The HTseq-count function in HTseq (High-Throughput sequencing) v0.6.1 (Anders et al., 2015) was used in the default mode and stranded=yes for generating read counts. HTseq-count output was fed into DESeq2 (Love et al., 2017) for differential expression analysis using the standard steps represented in the DESeq function (Love et al., 2014). A gene was considered differentially expressed if the adjusted p-value <0.05 and the |log fold change| > 1 and had a Transcript Per Million (TPM) > 1 in at least one condition. The p-value was adjusted with a q-value false discovery rate (Benjamini and Hochberg, 1995).

There were different kinds of data collected, including hypocotyl measurements, numbers of germinated seeds, pigment content, and expression data. Appropriate statistical tests were selected based on the experiment. Germination data were plotted in GraphPad Prism showing mean and error based on SD with error bars indicated. Data were considered significant when p was <0.05. For chlorophyll content and light-dependent hypocotyl elongation experiments in R, FR, and B light, data were analyzed using one-way ANOVA, with the post-hoc Tukey HSD test applied where relevant (https://astatsa.com/OneWay_Anova_with_TukeyHSD/ ). For anthocyanin content, two-tailed, unpaired Mann–Whitney U-test analyses were performed to compare the means of anthocyanin contents per milligram of fresh weight for a particular mutant line relative to WT in either darkness or light. For the 0- to 7-day seedlings grown in darkness (7D 0L), white light (0D 7L), or 3D4L, and the low-fluence 3-day seedlings ( Supplementary Figures 2, 3 ), at least three replicates of 10–20 seedlings each were analyzed where the unpaired t-test was used; for some samples, applying Welch’s correction was appropriate (GraphPad Prism, v.9). Statistical analyses used for differential gene expression and clustering analyses are described in section 2.12.

To obtain insight into tissue localization and regulation of expression of the ADT genes, we constructed heat maps using Arabidopsis AtGenExpress public microarray datasets of young Co-0 WT seedlings (www.weigelworld.org ; Figure 1 ). ADT genes are expressed at relatively low levels in most tissues, and in a tissue-specific manner in 7-day seedlings grown in Wc ( Figure 1A ). Previously reported data using qRT-PCR analyses (Rippert et al., 2009) indicated some similarities, and some major differences with these data, but age and growth conditions including light levels were distinct for the respective information. Given the prior association of some ADT genes with light regulation (Warpeha et al., 2006), we assessed the impact of distinct light qualities in 4-day-old seedlings. Each ADT member indicated responsiveness to all light qualities ( Figure 1B ). Notably, ADT genes were expressed at higher levels with a longer irradiation, the exception being ADT2, which expresses at higher levels regardless of light condition, with the exception of UV-A, UV-A/B, and R pulse where a longer irradiation (or in darkness, a longer treatment) increased induction. Most ADTs expressed at higher levels in longer irradiations of UV-A/B, B, and FR.

In planta tissue localization of fluorescently tagged ADT proteins was localized by exploring each tissue of the seedlings by microscopy. Green fluorescence above background indicated expression comparable to the transcript analysis results ( Figure 1A ) for 7-day plants. Each ADT gene family member GFP reporter construct (35S promoter, except for ADT3 that utilized the native promoter due to a larger study (Para et al., 2016) was transformed into a corresponding null adt mutant. Seedlings were grown under basta selection, under a long-day (16-h light:8-h dark; 16:8) growth condition, then live seedlings were viewed at day 7 post-planting. ADT protein, which may accumulate distinctly based on protein stability and/or turnover relative to 35S promoter-driven gene expression, was observed by microscopy as generally low throughout seedlings, with specific, predominating tissue-specific patterns for each ADT member ( Supplementary Figure S1 ). ADT1-GFP and ADT6-GFP accumulated in cotyledon mesophyll cells, particularly in the chloroplasts (confirmed by fluorescence pattern [21]), but where ADT1 was sporadic throughout the mesophyll (from individual chloroplasts to many chloroplasts), ADT6 was detected in most mesophyll cells in all sections through the cotyledon. Although ADT3-GFP exhibited low expression in newly divided cells and SAM, it was most discernable in the cotyledon epidermal layer, especially at the plasma membrane of pavement cells and in immature guard cells, which confirms reported results for young dark-grown seedlings (Para et al., 2016). ADT4-GFP was most expressed in the upper hypocotyl, occurring largely in plastids. ADT5-GFP was the only ADT with a distinct expression in the root. Null adt2 is embryonic lethal (El-Azaz et al., 2018), so ADT2 was not studied in this context.

To elucidate the role of ADT-mediated Phe synthesis during the earliest stage of the seed-to-seedling transition, we investigated the role of ADT in the seed germination process. Given the recognized role of hormones in modulating the timing of seed germination, seeds of adt mutants were grown in complete darkness in the absence or presence of a specific hormone in the medium ( Figure 2 ). All adt seeds germinated similarly to WT (~100%) by 72 h post-stratification in the absence of exogenous hormone (No Hormone [NH]; Figure 2A ). However, adt5 exhibited a delay in germination compared to WT, whereas adt6 and adt3 germinated before WT by 24 h. adt1 and adt4 demonstrated germination patterns similar to WT. GA at 1 μM had unique effects on the germination of the adt mutants ( Figure 2B ). adt6 seeds germinated more quickly than WT. adt4 germinated similarly to WT, whereas adt3 and adt1 germination was delayed but within range of WT responses. Germination of all adt mutants in the presence of GA was similar to WT (~100%) by 72 h, except adt5, which indicated delay in any germination up to ~36 h, then gradual germination increase over the 72-h period. In response to ACC, adt6 germinated rapidly within 24 h, similar to the NH condition ( Figure 2C ). adt1, adt3, and adt4 germination were delayed by growth on ACC similarly to WT, with approximately 50% of seeds germinating by 36 h, but between 36 and 72 h, adt1 germination continued in a linear fashion, behind WT and adt4, with 100% germination by 72 h. adt3 seed germination decreased from 36 to 72 h with just over 60% fully germinating. adt5 like NH indicated a slow to start germination process, but then exhibited linear germination over time, with only ~60% germinated by 72 h. ABA at 500 nM caused an expected delay in WT relative to untreated (NH) seeds, with, adt1, adt3, and adt4 indicating similar delays, where adt3 did not fully recover after 48 h ( Figure 2D ). ABA blocked adt5 germination, even by 72 h, post-stratification. adt6 appeared completely insensitive to ABA, germinating at the same rate as NH. In response to IAA, WT was slightly delayed, and adt1, adt3, and adt5 were inhibited in increasing effect (i.e., adt5 was the most inhibited, where the overall profile resembled responses to GA and ACC) ( Figure 2E ). Considering germination overall with or without hormones, adt4 had near-identical germination responses as WT. adt6 was uniquely unresponsive to all hormones, and adt5 was uniquely delayed in germination in response to all hormones, where ABA appeared to completely shut down germination.

Given the impact of ABA on adt5 germination and that ABA impacts transcription of chloroplast genes by a PP2C-dependent activation of nuclear encoded sigma factors (SIGs) (Woodson et al., 2013; Yamburenko et al., 2015), we examined interplay between ADT family members and plastid development. As phyA is known to regulate sigma factors, especially Sig6 (Oh and Montgomery, 2013), and they are both involved in the plastid development-related FR-BOG response in plants (Alameldin et al., 2020), we assessed FR-BOG-dependent gene expression in a sig6 mutant ( Figure 3 ).

Our RNA-seq data of the sig6 mutant line under FR-BOG conditions indicated that ADT5 was the only ADT gene that was downregulated in the sig6 mutant line under FR-BOG treatment compared to growth in white light ( Figure 3B ), suggesting a potential role of ADT5 in the FR-BOG response. To examine phenotypes that may be predicted by the alteration of ADT5 mRNA levels in a sig6 mutant under FR-BOG compared to wild type, we tested the greening response of the adt5 line under FR-BOG. Our results indicated that, like phyA and sig6 that have been previously shown to have an altered greening response in FR-BOG (Alameldin et al., 2020), the adt5 mutant similarly accumulates chlorophyll under FR-BOG ( Figure 4 ). This result supports a potential role of ADT5 in the chloroplast development process.

To further understand the role of ADT5 in the ABA-dependent chloroplast development process, we investigated ABA-related genes in our RNA-seq dataset. Our data indicate the misregulation of numerous ABA response genes ( Figure 3B ). Multiple members of the PP2C protein family, known to regulate ABA signaling negatively, are downregulated in sig6 under FR-BOG conditions [56, 57. 58].

We detected a downregulation of two phenylalanine ammonia-lyase (PAL) genes in a sig6 mutant under FR-BOG ( Figure 3 ); PAL catalyzes the conversion of phenylalanine to trans-cinnamic acid, the first step of phenylpropanoid biosynthesis leading to diverse plant metabolites reactions (Truman et al., 2007; Widjaja et al., 2009). PAL is also involved in the ABA signaling pathway, where PAL was misregulated in aba3 mutants under oxidative stress. Our data also indicated the downregulation of the Cinnamate 4-hydroxylase 4CL1 gene that controls the last step of the general phenylpropanoid pathway (Li et al., 2015).

Additionally, we detected a downregulation of flavanone 3-hydroxylase (F3H) that is coordinately expressed with chalcone synthase and chalcone isomerases and is involved in flavonoid biosynthesis (Shih et al., 2008). Flavonol synthase 3 (FLS3), involved in the last step of flavonol biosynthesis (Pelletier et al., 1999), was also downregulated in the sig6 mutant under FR-BOG. Furthermore, five of the UDP-Glycosyltransferase superfamily members’ mRNA levels were significantly reduced in the sig6 mutant under FR-BOG. UDP-Glycosyltransferase superfamily proteins play a central role in the last step of the anthocyanin biosynthesis pathway (Yonekura-Sakakibara et al., 2014). Additionally, the GLUTATHIONE S-TRANSFERASE PHI 12 (GSTF12), which functions as a carrier to transport anthocyanin from the cytosol to tonoplasts (Wagner et al., 2002), was also downregulated in the sig6 mutant under FR-BOG. Since adt5 was delayed in germination in response to ABA, and given the fact that ADT5 regulates phenylalanine biosynthesis, our RNA-seq data analysis suggests that ADT5 might play an important role in the anthocyanin biosynthesis pathway.

After emergence of the seedling from the seed, and prior to emergence into light (presumably under soil or leaf litter in natural contexts), seedling development initiates with rapid stem elongation in darkness. Following emergence into a light environment, photoreceptors perceive ambient light, then initiate the inhibition of stem elongation and promotion of cotyledon expansion, plastid development, and root development. Until the seedling is fully autotrophic, these processes are supported by metabolites stored in the seed, and made in the emerging seedling. We investigated the role of ADTs in darkness, continuous light, or as a result of light-to-dark-transition, from 0 to 7 days.

Owing to the responses of adt3, adt5, and adt6 seedlings in variable light conditions, we measured hypocotyl lengths of these mutants in specific R, FR, or B monochromatic light conditions. Plants were grown at two different fluence rates of continuous R (Rc; 50 and 100 μmol m ^{− 2} s ^{− 1}), FR (FRc; 5 and 25 μmol m ^{− 2} s ^{− 1}), or B (Bc; 25 and 50 μmol m ^{− 2} s ^{− 1}) light with a control set grown in Dc (black bars) for each light condition. A phyA mutant was used as a positive control for FR and B light, and a phyB mutant was used as a positive control for R light experiments. Of note, adt3 has been previously shown to have a B-dependent role in Phe synthesis (Warpeha et al., 2006; Warpeha et al., 2008). An insertion mutant of Pirin1, prn1, was also included as it is involved in accumulation of phenylpropanoids, and in B- and ABA hormone-dependent signaling (Orozco-Nunnelly et al., 2014). These data are shown in Figure 6 (representative seedlings for conditions are shown Figures 6A–E ).

Under an Rc illumination of 50 μmol m ^{− 2} s ^{− 1}, adt5 and prn1 were longer (p < 0.01 for both) than WT, similar to the phyB mutant (p < 0.01), though not to the same magnitude ( Figures 6C, F ). At a higher fluence of Rc (i.e., 100 μmol m ^{− 2} s ^{− 1}), no adt mutants were significantly longer than WT ( Figure 6I ), suggesting a light intensity-dependent role. Under an FRc of 5 μmol m ^{− 2} s ^{− 1}, the adt6 (p < 0.01) mutants were longer than WT, which was similar but to a much lesser degree than the FR-light insensitive phyA mutant (p < 0.01; Figures 6D, G ). Similar to Rc data, a higher fluence of FRc light did not indicate any significant differences of adt mutants and WT ( Figure 6J ). Under Bc at 25 μmol m ^{− 2} s ^{− 1}, adt3 was longer than WT (p < 0.01), similar to the positive control phyA (p < 0.01; Figures 6E, H ). adt5 was shorter than WT (p < 0.01) under these conditions ( Figures 6E, H ). Markedly, at 50 μmol m ^{− 2} s ^{− 1} of Bc, all of the mutants tested were significantly longer than WT ( Figure 6K ), similar to phyA. Notably, among the lines tested, adt6 exhibited a consistently longer hypocotyl than WT in complete darkness on sucrose-containing medium ( Figure 6 ). These results confirm the input of multiple wavelengths of lights into ADT function during seedling development, as predicted from the gene expression data ( Figure 1 ). As Rc and Bc illumination impacted adt5 and adt6 development in hypocotyl inhibition assays on sucrose-containing medium at 7 days, we also examined seedlings during the early post-germination period under a lower fluence rate of Rc or Bc on sucrose-free medium, assessed at 3 days (72 h) post-stratification ( Supplementary Figure S3 ). Shoots and roots of adt6 were longer than WT under Rc and Bc, and achieved significance for all measurements (p < 0.001); adt5 was significant for all measurements (Shoots, Bc p < 0.001; Rc p < 0.01) except Rc roots (n.s.).

The accumulation of anthocyanins is promoted by light, and anthocyanins are synthesized from the Phe-dependent phenylpropanoid pathway (Chen et al., 2016). Thus, we assessed anthocyanin accumulation in specific single adt mutants grown under Wc conditions at the early seedling stage. Anthocyanin levels in the control phyAphyB mutant were reduced to 63% of WT levels ( Supplementary Figure S4 ), and the prn1 mutant that overproduces specific pigments in response to light (Orozco-Nunnelly et al., 2014) was also used as a control. Notably, the adt5 mutant accumulated only ~70% of WT anthocyanin levels ( Supplementary Figure S4 ). No other mutant had a light-dependent defect in anthocyanin accumulation under our conditions, though adt5 and prn1 accumulated less anthocyanins in Dc than that observed for WT ( Supplementary Figure S4 ).

There are two aspects of the phenotypes of adt mutants that appear particularly important to consider—the germination or early phenotype, where germination is typically completed for WT by 48–72 h post-stratification, and the seedling phenotype, observed herein at 4–7 days. During germination (0–3 days), the available pool of resources, including Phe, stored in the seed to support early growth appears critical, prior to assembly and function of the chloroplasts, where expression of ADTs is deemed important as reviewed (Perkowski and Warpeha, 2019). During seedling development, ADT5 and ADT6 have different actions, which may be based on the tissue in which they primarily occur. As observed in the analysis of germination ( Figure 2 ), it is clear that adt5 mutants have difficulty in germination, which may partly be explained in that ADT5 is expressed in the roots ( Figure 1 ; Supplementary S1 ) where roots emerge first from the seed. adt5 seeds may possess insufficient resources in the seed to support germination. The inability of adt5 mutants to overcome dormancy in the dark in response to germination-promoting hormones tested (ACC and GA) provides additional evidence of dependence on photoreceptors, perhaps reinforced by the fact that long irradiations of UV-A/B, blue, and far-red light increase the induced expression of the ADT5 transcript ( Figure 1 ). The phenotype of adt5 mutants under FR-BOG conditions mirrored the phenotype of phyA and sig6 ( Figure 4 ), which have previously been implicated in light-responsive, phytochrome-dependent plastid development (Alameldin et al., 2020), further highlighting a role for ADT5 function in light-dependent, photoreceptor-modulated processes. adt6 mutants exhibit a phenotype where no hormone had any influence on rapid germination of the mutant in darkness ( Figure 2 ). Primarily protein expression in planta appears to be in the cotyledon in the developing mesophyll ( Supplementary Figure S1 ). ADT6 protein may contribute to repressing germination until conditions, in general, are optimal. ADT6 expression is higher in seeds than ADT3 and ADT5 and more similar to ADT2 ( Figure 1A ) (Hossain et al., 2018), where its action might be in defense or other activities, which may be important in preparing the embryo for germination.

We explored further the relationship between complete darkness and white light as it impacted hypocotyl and root growth ( Figure 5 ). The observed responses for adt6 were distinct from those for adt5. The amount of darkness versus light affected hypocotyl and root elongation differently, with both Dc and Wc preventing adt5 from full (compared to WT) expansion in hypocotyls or roots. 3D4L indicated greater expansion of roots for adt5, exceeding WT. Any period of darkness (Dc or 3D4L) promoted greater root expansion of adt6 compared to WT and 3D4L, promoting longer adt6 hypocotyls. However, 7 days of white light reduced both shoot and root expansion compared to WT, a trend of reduced root+shoot expansion for all three adt mutants. adt3 responses were unique compared to adt5 and adt6, with longer shoots and roots in complete darkness (7D0L), reduced roots in light (0D7L), and increased shoots in 3D4L. Any darkness increased adt6 roots (7D0L and 3D4L), and any darkness increased adt3 shoots (7D0L and 3D4L). ADT proteins may have direct influence on auxin transport or directly affect flavonol concentrations in the root:shoot continuum, or in the meristems in different stages of development (Lewis et al., 2011; Liu et al., 2019). An increased exposure to FR or decreased R:FR impacts auxin transport, and is linked to changes in phenylpropanoid-related compounds, including a reduction in anthocyanins and increased lignin in maize (Afifi and Swanton, 2012). Adt5 mutants accumulated less anthocyanins in Wc ( Supplementary Figure S4 ) in our measurements. Chen et al.’s study of single and multiple ADT gene mutations indicated that ADTs contribute to anthocyanin accumulation, particularly ADT2 at an early stage (Chen et al., 2016).

ADT1 and ADT4, while not having dramatic phenotypes apparent in this age of plant development, did indicate unique transcript expression profiles ( Figure 1 ) that coordinated with the observed expression in the GFP fusions ( Supplementary Figure S1 ). ADT2 is important in seed formation and embryo for early survival (Chen et al., 2016). Indeed, both ADT2 and ADT3 (herein and (Para et al., 2016)) indicate impact on earlier processes, with ADT2 being required possibly mostly for seed development (Chen et al., 2016) and defense compounds (Bross et al., 2017) and ADT3 being more important immediately post-germination in development (Para et al., 2016). ADT2 and ADT3 perhaps provide Phe until ADT1, 4, 5, and 6 express in specific tissues of the seedling (photosynthesis in chloroplasts, and roots developing as storage organs). One major event that happens in the seed-to-seedling transition is chloroplast development in the developing leaf cells, where ADT2 and ADT3 may work to coordinate that development, given that different labs have shown that ADT2 (Bross et al., 2017; El-Azaz et al., 2018) and ADT3 (Para et al., 2016) are both required for processes preceding chloroplast development. Phe made via ADTs in general are also recently shown to have a number of specific impacts on photosynthesis and metabolism in older seedlings (~3 weeks old) (Höhner et al., 2018).

Based on the results of the growth studies under monochromatic light, the impact of ADT family proteins on Phe pools may contribute to the photoregulation of hypocotyl elongation. Rc, FRc, and Bc light impact hypocotyl elongation of distinct adt mutants relative to WT ( Figure 6 ). This impact may be due to direct regulation of ADT gene expression by light ( Figure 1B ), but could also be due to the general light-associated reduction of Phe pools on growth and development in these mutants. These results suggest an interaction between light and Phe homeostasis. This is further supported by the observed light-impacted differential expression of individual ADT genes ( Figure 1 ). In some cases, adt mutants and phytochrome mutants have similar light-dependent phenotypes. For example, phyA and adt6 have defects in FR- and B-dependent inhibition of hypocotyl elongation responses at 7 days ( Figure 6 ) and both phyB and adt6 under Rc were different from WT responses. These data suggest potential phytochrome regulation of the light-associated role for ADT6 and also indicated where adt6 hypocotyls are reduced in 0D7L, but had greater expansion in 7D0L.

The notable and differential defects of adt mutants to hormone treatment in germination assays and defects in light responsiveness suggest interplay between light and hormones in the modulation of Phe homeostasis. Light has been shown to regulate a number of hormonal pathways in plants. The integration of GA signaling and light-dependent phytochrome signaling has been previously shown to involve two PHYTOCHOME-INTERACTING FACTOR (PIF) proteins, PIF3 (Feng et al., 2008) and PIF4 (de Lucas et al., 2008). The transcription factor HY5, which is a key regulator of photomorphogenesis (Oyama et al., 1997), also has been linked to hormonal modulation. HY5 impacts GA (Alabadí et al., 2008), auxin (Cluis et al., 2004; Sibout et al., 2006), and ABA (Chen et al., 2008) signaling. Recent studies have also linked G-box-specific binding transcription factor bZIP16 to light, GA, and ABA signaling through regulating the expression of genes including PIL5 and RGL2 (Hsieh et al., 2012). PIL5 (Oh et al., 2006) and RGL2 (Lee et al., 2002) have both been linked to the regulation of seed germination, a phenotype disrupted in adt mutants. Light also has a complex regulation of the conversion of ACC to ethylene, dependent on fluence, spectral quality, age of seedlings, and responses to a pulse of light versus continuous irradiations, reviewed in Zdarska et al. (2015). The Rc, FRc, and Bc data of adt5 or adt6 compared to WT, as well as the FR-BOG phenotype of adt5, indicate that phytochromes and blue light photoreceptors play unique roles with ADT5 and ADT6 in development.

Development during the seed-to-seedling transition is an intricate multi-component process in higher plants, requiring energy allocation to rapid development and growth, while the emerging seedling also utilizes energy to sense and respond to environmental signals and to avoid and/or defend against potential detrimental environmental factors. Phe is the first committed precursor of the phenylpropanoid pathway, and genetic and biochemical data indicate that Phe is a concentration-limiting substrate (Margna, 1977; Lee et al., 2002; Rohde et al., 2004; Voll et al., 2004; Warpeha et al., 2008). Phe concentrations are likely coordinated with other phenylpropanoid enzymes, several of which have been shown to peak in expression in 3-day-old white light-grown seedlings (Margna et al., 1989a), a developmental age and growth condition utilized in our experiments. Expression of phenylalanine ammonia-lyase (PAL; EC 4.3.1.5), the enzyme catalyzing the first step of the phenylpropanoid pathway that deaminates L-phenylalanine to trans-cinnamic acid, is highly responsive to environmental and developmental cues and induced early in seedling development (Ohl et al., 1989; Kubasek et al., 1992). pal1 pal2 pal3 pal4 quadruple mutants of Arabidopsis were severely stunted in growth at the early seedling stage and sterile (Lamb et al., 1989). Depending on the adt mutant assessed, the growth of seedlings compared to WT was different, so the location of the Phe accumulation in the seedling (tissue) as well as location in cell types may matter. All data considered suggest that through either direct light-mediated signaling impacts on Phe levels or light-associated disruptions in the regulation of Phe synthesis resulting in reduced Phe pools, ADT-dependent Phe homeostasis impacts the coordination of the regulation of the existing energy budget with growth and development. The expression of each ADT appears finely tuned to environmental cues and available resources. Perhaps to prepare the seedling to produce the levels of metabolites and structures required, ADTs evolved as mainly type II isoforms with relaxed regulation to enable the massive production of Phe in the seed-to-seedling transition (El-Azaz et al., 2016; El-Azaz et al., 2022). Because of the uniqueness of responses for each ADT, Phe regulation is likely specific to tissues with respect to age and level of development and, hence, may be critical for promoting the successful establishment of seedlings in dynamic environments.