An ATP-Binding Cassette Transporter, ABCB19, Regulates Leaf Position and Morphology during Phototropin1-Mediated Blue Light Responses

Blue light regulates multiple processes that optimize light capture and gas exchange in plants, including chloroplast movement, changes in stomatal conductance, and altered organ positioning. In Arabidopsis ( Arabidopsis thaliana ), these processes are primarily modulated by the blue light phototropin photoreceptors phot1 and phot2. Changes in leaf positioning and shape involve several signaling components that include NON-PHOTOTROPIC HYPOCOTYL3, PHYTOCHROME KINASE SUBSTRATE, ROOT PHOTOTROPISM2, and alterations in localized auxin streams. Direct phosphorylation of the auxin transporter ATP-BINDING CASSETTE subfamily B19 (ABCB19) by phot1 in phototropic seedlings suggests that phot1 may directly regulate ABCB19 to adjust auxin-dependent leaf responses. Here, abcb19 mutants were analyzed for ﬂ uence and blue light-dependent changes in leaf positioning and morphology. abcb19 displays upright petiole angles that remain unchanged in response to red and blue light. Similarly, abcb19 mutants develop irregularly wavy rosette leaves that are less sensitive to blue light-mediated leaf ﬂ attening. Visualization of auxin distribution, measurement of auxin transport in protoplasts, and direct quanti ﬁ cation of free auxin levels suggest these irregularities are caused by misregulation of ABCB19-mediated auxin distribution in addition to light-dependent auxin biosynthesis.

Leaf positioning and morphology in plants are optimized for photosynthetic light capture and growth. Photoreceptor systems, including members of the phytochrome, phototropin, cryptochrome, zeitlupe, and UV-B/UV-R receptor protein families, monitor light quantity (intensity and duration), quality (wavelength), and direction to regulate downstream growth processes . Light perception by these photoreceptors regulates multiple downstream signaling pathways and developmental processes. These include tropic growth, shade avoidance, and multiple aspects of auxin-regulated shoot and root development (Hangarter, 1997;Christie et al., 2011;Li et al., 2012;Casal, 2013;Michaud et al., 2017).
Several lines of evidence suggest that the ATP-BINDING CASSETTE subfamily B19 (ABCB19) auxin efflux transporter is directly regulated by phototropin during leaf positioning and flattening responses. ABCB19 is a primary regulator of long-distance auxin transport from the shoot apex (Noh et al., 2001;Blakeslee et al., 2007), functions in cotyledon development (Lewis et al., 2009), and has been recently shown to have a similar function in leaf tip to petiole transport in young true leaves (Jenness et al., 2019). During seedling phototropic bending, phot1 directly phosphorylates and inactivates ABCB19 to inhibit rootward auxin fluxes (Christie et al., 2011), and loss of ABCB19 results in accelerated initiation of bending (Noh et al., 2003). Loss of ABCB19 and its paralog ABCB1 results in wrinkled rosette leaves that curled downward (Noh et al., 2001;Blakeslee et al., 2007). Similar rosette leaf morphologies are observed when young Col-0 rosette leaves are treated with the auxin efflux inhibitor 1-naphthylphthalamic acid (NPA; Husbands et al., 2015). As NPA directly binds to a subset ABCB auxin transporters including ABCB19 and inhibits auxin efflux in intact plants, individual cells, and leaves (Noh et al., 2001;Murphy et al., 2002;Petrásek et al., 2006;Blakeslee et al., 2007;Jenness et al., 2019), these results suggest a direct contribution from ABCB19 during regulation of light-dependent leaf positioning and growth.
Here, a consensus set of light response experiments in Arabidopsis establishes a common set of conditions to evaluate leaf responses. The results presented here show that ABCB19 is the primary missing component in blue light-dependent changes in auxin distribution that regulate leaf angle and primary leaf curling/flattening responses.

Leaf Positioning Integrates PHY-and Phot-Mediated Red and Blue Light Perception
To establish the baselines for light-dependent leaf petiole positioning, petiole angles were measured with Col-0 seedlings treated with 50 mmol m 22 s 21 red light plus farred (;4:1 red:far-red), 50 mmol m 22 s 21 red alone, 50 mmol m 22 s 21 red plus 0.1 mmol m 22 s 21 blue, or 50 mmol m 22 s 21 red plus 5 mmol m 22 s 21 blue. Under red plus far-red light, Col-0 display upright petiole angles that are indicative of the shade avoidance response (Supplemental Fig. S1,A and B). When treated with red light alone, petiole angles were nearly horizontal and rose to an intermediary angle with the addition of 0.1 mmol m 22 s 21 (phot1 levels) or 5 mmol m 22 s 21 (phot1 and phot2 levels) blue light (Supplemental Fig. S1,A and B).
Upright petiole angle positioning during the shade avoidance response is primarily controlled by phyB and downstream increases in TAA1-dependent auxin biosynthesis (Tao et al., 2008;Moreno et al., 2009;Keller et al., 2011). Under all conditions tested, phyA phyB double mutants showed constitutive shade avoidance phenotype with upright petiole angles (Supplemental Fig. S1, A and B), which is consistent with previous results (Tao et al., 2008;Keller et al., 2011).
Leaf positioning also requires blue-light perception and downstream signaling mediated by phototropins (Sakamoto and Briggs, 2002;Inoue et al., 2008;de Carbonnel et al., 2010). phot1, phot2, and phot1 phot2 mutants showed shade avoidance response to red plus far-red light with a slight enhancement observed in phot1 phot2 (Supplemental Fig. S1,C and D). Under red light alone, phot1 exhibited slightly more upright petiole angles, while no significant differences were observed in phot2 or phot1 phot2 (Supplemental Fig. S1, C and D). As phototropins are not active under these conditions, the variations in phot1 and phot1 phot2 under red and far red light, respectively, are likely the result baseline changes in the mutant backgrounds and steady-state interactions with PHY, PIF, and PKS signaling components (Lariguet et al., 2006;Sun et al., 2013;Schumacher et al., 2018). When red light was supplemented with 0.1 or 5 mmol m 22 s 21 blue light, phot1 and phot1 phot2 petiole angles were significantly lower than gl1 controls (Supplemental Fig. S1, C and D). Although phot2 purportedly was not involved in petiole angle responses (Inoue et al., 2008), phot2 petiole angles were slightly but statistically lower than gl1 under 5 mmol m 22 s 21 blue light (Supplemental Fig. S1,C and D). Additionally, the increased petiole angles in phot1 with 5 mmol m 22 s 21 blue light and the severely epinastic petioles in phot1 phot2 mutants suggest phot2 also contributes to petiole angle regulation (Supplemental Fig.  S1, C and D).
NPH3 and RPT2 act downstream of phot1 and phot2 signaling during blue light-mediated responses (Inada et al., 2004;Inoue et al., 2008;de Carbonnel et al., 2010;Harada et al., 2013;Haga et al., 2015). Under red plus far-red light, nph3 and rpt2 mutants exhibited typical upright petiole angles with a slight enhancement observed in nph3 (Supplemental Fig. S1, E to H). nph3 and rpt2 mutants displayed low petiole angles under red light alone and were not different from the controls (Supplemental Fig. S1, E to H). However, when red light was supplemented with blue light, both nph3 and rpt2 petiole angles were significantly lower than the gl1 or Col-0 controls, respectively (Supplemental Fig. S1, E to H).
Overall, these results are consistent with previous reports and validate the experimental system used in this study.

PIN-Directed Transport Is Required for Leaf Petiole Positioning
PIN proteins are upregulated during thermonastic and shade avoidance leaf positioning responses (Devlin et al., 2003;Keuskamp et al., 2010;Park et al., 2019). Mutations in PIN3 have been reported as having defects in direct lateral auxin transport during shade avoidance responses (Pantazopoulou et al., 2017). Other reports have described pin3 mutants as not being different from wild type (Keller et al., 2011), potentially due to compensatory PIN expression and activity as is observed during embryogenesis (Vieten et al., 2005). Under the experimental conditions used here, pin3 exhibited normal shade avoidance response and response to red light alone (Fig. 1). Under red plus blue light conditions, pin3 petiole angles were marginally but statistically more upright than Col-0 ( Fig. 1). A pin3 pin4 pin7 triple mutant did not exhibit upright angles in response to far-red or low/horizontal angles under red light alone (Fig. 1). Instead, pin3 pin4 pin7 petiole angles were equivalent to red plus blue light under all conditions tested. The lack of change in both shade avoidance and red lightinduced petiole lowering suggests that PIN-directed auxin transport is required for both responses. While PIN-directed transport is clearly important for optimal leaf positioning responses, PINs are unlikely to be direct targets for blue light and phototropin-mediated responses (Christie et al., 2011). Instead, their roles appear to be canalization of auxin transport downstream of TAA1-dependent auxin biosynthesis (Michaud et al., 2017).

ABCB19 Regulates Light-Dependent Leaf Positioning
The Arabidopsis homolog of maize BR2/ABCB1, ABCB19, is directly phosphorylated and inactivated by phot1 to prime phototropic bending in response to blue light (Christie et al., 2011). More recently, ABCB19 was shown to mediate leaf tip to petiole auxin transport in true leaves (Jenness et al., 2019). Therefore, ABCB19 and its paralog, ABCB1, were analyzed for contributions to light-dependent changes in leaf petiole angle positioning.
Multiple reports have analyzed phenotypes of single and double Arabidopsis abcb1, abcb19, and abcb1 abcb19 mutants (Noh et al., 2001;Blakeslee et al., 2007;Wu et al., 2007;Christie et al., 2011;Yang et al., 2013;Zhao et al., 2013). Consistently, abcb19 and abcb1 abcb19 mutants exhibit .50% reductions of rootward auxin flows, enhanced phototropic bending, epinastic cotyledons, increased numbers of secondary inflorescences, shortened anther filaments, and altered leaf morphology. However, differences in ecotypic background utilized in these reports make phenotypic comparisons difficult to interpret since basal auxin levels and auxin transport rates vary among Arabidopsis ecotypes (Peer et al., 2014). For consistency in this study, established abcb1-100 and abcb19-101 single and double mutants in the Col-0 ecotypic background were backcrossed to the wild type and genotyped to assure that no phyD mutations (associated with the Ws ecotype) or erecta mutations (associated with introgressions from Ler) were present.
To assess the contributions of ABCB19 and ABCB1 to leaf petiole positioning, petiole angles were measured in abcb19 and abcb1 mutants under red plus far-red, red, and red plus blue light conditions. abcb1 petiole angles were lower than Col-0 under red plus far-red light, but not different under red or red plus blue light conditions (Fig. 2). This likely points to compensation by ABCB19, which is increased ;4.53 in abcb1 mutants (Blakeslee et al., 2007;Jenness et al., 2019). Under all light conditions tested, abcb19 mutants exhibited more upright petiole angles than Col-0 or abcb1 (Fig. 2). Excessive leaf and petiole curling prevented accurate measurement of abcb1 abcb19 petiole angles. These results are consistent with phot1-mediated inhibition of ABCB19 contributing to increased petiole angles, although the effect is less than the loss of ABCB19 function entirely. Light intensity and red:far-red ratios perceived by phytochromes have been shown to be primary regulators of auxin levels (Aukerman et al., 1997;Keller et al., 2011;Hersch et al., 2014). In young seedlings, free indole-3-acetic acid (IAA) levels were greater under 50 mmol m 22 s 21 white light compared to 130 mmol m 22 s 21 (Supplemental Fig. S2A). These differences in total auxin levels can be attributed at least partially to reported phytochrome regulation of auxin biosynthesis, as fluence-dependent changes did not occur in a phyA phyB mutant (Supplemental Fig. S2A). Similarly, analysis of 10-d light-grown abcb19 mutants also revealed increases in auxin levels (Supplemental Fig.  S3A) that were not accompanied with increases in oxidative catabolism of auxin to the inactive metabolite oxindole-3-acetic acid (Supplemental Fig. S3B). Due to the elevated auxin levels, it was hypothesized that abcb19 mutant petioles would be resistant to lowering with increasing light fluence. As expected, Col-0 and abcb1 petiole angles decreased under 100 and 120 mmol m 2 2 s 2 1 white light compared to 60 mmol m 2 2 s 2 1 (Supplemental Fig. S4, A and B). Petiole angles in abcb19 were more upright than Col-0 and abcb1 under all light conditions tested and were less impacted by light fluence (Supplemental Fig. S4, A and B). These results are consistent with upright petiole angles being caused, at least in part, by elevated auxin levels.
Ethylene enhances auxin biosynthesis (R uzicka et al., 2007;Swarup et al., 2007;Stepanova et al., 2008;Muday et al., 2012). Therefore, it was hypothesized that an eto1 eol1 eol2 triple mutant, which overproduces ethylene (Christians et al., 2009), would also display upright petiole angles similar to phyA phyB and abcb19. As hypothesized, petiole angles in eto1 eol1 eol2 were more upright than wild type under all light conditions tested, although the angle under red light alone was not as severe as phyA phyB or abcb19 (Supplemental Fig. S5, A  and B). These results further support that increased auxin levels lead to upright petiole angles.
To further assess the contribution of auxin levels to petiole angle, free IAA was quantified in the upper sections (shoot apex, petioles, and leaves) of Col-0, phyA phyB, phot1, phot1 phot12, and abcb19 under red and red plus blue light. cry1 cry2 was also included, although cry has been reported not to contribute to the response (Inoue et al., 2008). Under red light alone, phyA phyB had slightly elevated free IAA levels, while cry1 cry2, phot1, phot1 phot2, and abcb19 were not different from Col-0 (Fig. 3A). In Col-0, free IAA was increased ;15% to 20% with the addition of blue light to phyA phyB levels (Fig. 3A). This response did not occur in phot1 or phot1 phot2 (Fig. 3A). Although differences in free IAA could be detected, the overall change in auxin levels were not substantial. However, increases in IAA have been shown to be accompanied by increases in IAA-Asp levels as a mechanism to attenuate localized auxin accumulations (Christie et al., 2011;Zhang et al., 2016). While small differences in IAA-Asp could be detected in cry1 cry2, phot1, and phot1 phot2, phyA phyB, and abcb19 had greater than five times the IAA-Asp of Col-0 under all light conditions (Fig. 3B). These results are consistent with a correlation between upright petiole angles, as observed in phyA phyB and abcb19, and increased IAA and/or IAA-Asp levels.
DII-VENUS is a fluorescent auxin reporter consisting of a degradation domain of an AUX/IAA transcriptional repressor fused to the fluorescent protein VENUS (Brunoud et al., 2012). Therefore, degradation of the VENUS signal reflects areas of auxin accumulation and subsequent activation of auxin signaling. Although DII-VENUS signals are generally low in petioles, differences that correlated with the reported auxin levels were observed (Fig. 4). Under red plus far-red light DII-VENUS signal was absent in Col-0 and abcb19, indicating elevated auxin levels (Fig. 4). Under red light alone, DII-VENUS was clearly detected in Col-0, but the signal was significantly less in abcb19 (Fig. 4). When red light was supplemented with blue light, DII-VENUS signals were reduced in both Col-0 and abcb19, although signals in abcb19 remained significantly less (Fig. 4). These results further support that lower petiole angles in response to red light are caused by a decreased petiole auxin accumulation and supports phot1 inhibition of ABCB19 increasing petiole auxin levels to increase petiole angle in response to blue light.
The recently characterized auxin transporters ABCB6 and ABCB20  have also been shown to be phosphorylated in guard cells under blue light conditions (Hiyama et al., 2017). Although a direct role in blue light-mediated guard cell conductance regulation was not described in these studies, abcb6 and abcb20 were examined for leaf positioning defects. No statistical differences in response to red plus far-red, red, or white light were detected between Col-0, abcb6, abcb20, and abcb6 abcb20 (Supplemental Fig. S6, A and B; Supplemental Table S1). Like abcb1 abcb19, abcb6 abcb20 exhibited twisted and skewed leaves and petioles  pointing to their function in auxin distribution during vegetative growth. However, the lack of defects under varying light conditions suggest ABCB6 and ABCB20 do not play important roles in light-mediated leaf positioning. Additionally, Col-0 exhibited similar angles in response to red plus blue and white light. Since no change was observed in abcb6 and abcb20 mutants with white light, they are unlikely to show differences in response to red plus blue light as well.
abcb19 Rosette Phenotypes Reflect Fluence Dependence of Auxin Levels In mature wild-type leaves, free IAA levels were ;40% greater under 130 mmol m 22 s 21 white light compared to 50 mmol m 22 s 21 (Supplemental Fig. S2B), and previous analysis showed mature abcb19 rosette leaves had reduced free IAA levels compared to Col-0 (Jenness et al., 2019). Therefore, midrosette leaf area was examined as a measure for auxin distribution in actively expanding leaves. As expected, total midrosette leaf area increased with light fluence in Col-0, but also in abcb1 and abcb19 (Supplemental Fig. S4, C and D). However, no growth differences were observed between abcb1 and Col-0 under any conditions (Supplemental Fig. S4, C and D). This, again, is likely due to compensatory ABCB19 expression described previously (Blakeslee et al., 2007;Jenness et al., 2019). However, under both 60 and 100 mmol m 22 s 21 light, abcb19 rosette leaf areas were smaller than both Col-0 and abcb1 (Supplemental Fig. S4, C and D). abcb19 rosette area was not different from abcb1 and Col-0 under 120 mmol m 22 s 21 light, but petioles and leaves remained shorter and more curled (Supplemental Fig.  S4, C and D). Although total leaf area in abcb1 abcb19 increased with light fluence, the impact was minimal due to the leaves remaining tightly curled downward under all conditions (Supplemental Fig. S4, C and D). These results suggest ABCB19, and to some extent ABCB1, contribute to auxin distribution during rosette leaf development, particularly under low light when auxin biosynthesis is limiting. Although phot1 has a more predominant function in leaf petiole angle positioning, both phot1 and phot2 have been reported to contribute to leaf morphology (Sakamoto and Briggs, 2002;Lariguet et al., 2006;Inoue et al., 2008;Harada et al., 2013). Phototropins regulate ABCB19 activity (Christie et al., 2011) (40R 1 4B). Seedlings were grown as in Figure 1 with the alterations in light fluence indicated. Hypocotyls were removed 8 mm above the rootshoot transition zone prior to IAA extraction and quantitation by liquid chromatography mass spectrometry. Data shown are means 6 SD (n 5 3 replicates of 10 pooled seedlings). Asterisks indicate statistical difference by two-way ANOVA (P , 0.0001), followed by Dunnett's post-hoc (*P , 0.05), versus Col-0 for each condition.
Therefore, abcb19 was examined for blue light-dependent changes in leaf morphology. When grown under red light plus far-red, both Col-0 and abcb19 leaves exhibited shade avoidance responses, including expanded leaves and elongated and upright petioles (Fig. 5A). Under red light alone, Col-0 and abcb19 developed more compact rosette leaves that curled downward (Fig. 5A). Under blue light supplementation, Col-0 leaves appeared flatter and more expanded, while abcb19 was largely resistant to the added blue light and remained more curled compared to Col-0 (Fig. 5A). Quantification of petiole length showed that abcb19 petioles were shorter than Col-0 under all light conditions tested; however, the relative difference remained the same among the light treatments ( Fig. 5B; Supplemental Table S1). To better assess the leaf blade morphological differences, leaf flattening (as a ratio of overhead curled leaf area to manually flattened leaf area) was quantified. These measurements demonstrated that Col-0 and abcb19 developed flatter leaves under far-red supplementation, and highly curled leaves under red light alone (Fig. 5C). Addition of blue light caused leaf flattening in Col-0, while abcb19 leaves remained more curled (Fig. 5C). While leaves were significantly more expanded and flatter when grown under 100 mmol m 22 s 21 white light ( Fig. 5C; Supplemental Fig.   S4, C and D), under conditions that resulted in elevated auxin levels (Supplemental Fig. S2B), abcb19 leaves remained more curled than Col-0.
DR5:GUS was used to visualize light-dependent changes in auxin distribution at a leaf stage before full expansion had occurred (Supplemental Fig. S7). No observable differences in DR5:GUS signal were found between Col-0 and abcb19 under red light supplemented with far-red or red light alone. Under these conditions, GUS signals were observed at hydathodes, with little accumulation along the leaf margins. Blue light supplementation increased DR5:GUS signals in Col-0 leaf tips, hydathodes, and leaf margins. Similar, but weaker, signals were observed in abcb19, although differential chlorophyll accumulation under the various light conditions and consequent impacts on destaining made quantification unreliable. The pooling of auxin at the leaf tip was consistent with the decreased leaf tip to petiole auxin transport observed with Col-0 treated with NPA and in abcb19 mutants (Jenness et al., 2019). These results suggest that blue light treatments alters auxin accumulations, presumably via phot1 inhibition of ABCB19. However, the alterations in auxin accumulation during blue light-dependent leaf flattening are not to the extent seen with complete loss of ABCB19 function.

Blue Light Regulates ABCB19 Auxin Efflux in Leaf Mesophyll Protoplasts
Due to the integration of multiple signaling pathways involved in leaf flattening, identifying specific phot1/ABCB19-dependent contributions to whole-leaf growth is difficult. However, decreased PIN1 function at the plasma membrane and well-defined protocols established for assaying ABCB19 transport function in protoplasts Feraru et al., 2011) suggested a better avenue for functional evaluation. In leaves, both phot1 and ABCB19 are expressed in epidermal and mesophyll cells (Sakamoto and Briggs, 2002;Lewis et al., 2009). To see how light alters rates of auxin transport directly, Radiolabeled indole-3-acetic acid ( 3 H-IAA) efflux from Col-0, phot1 phot2, phyB, abcb1, abcb19, and abcb1 abcb19 protoplasts was analyzed (Fig. 6). Assays were conducted as described in the "Materials and Methods" section, with protoplasts handled under green safe light and kept at 2°C to 4°C between light treatments to reduce red and blue lightdependent metabolic and transcriptional effects. Auxin efflux from Col-0 protoplasts was reduced when exposed to red plus blue light compared to red light alone (Fig. 6A). This reduction appeared to be phototropin dependent, as no reduction was observed in phot1 phot2 protoplasts (Fig. 6B). Reductions in net auxin efflux under red light with no change after addition of blue light in abcb19 protoplasts (Fig. 6C) is consistent with direct phot regulation of ABCB19. In contrast, auxin efflux was slightly increased in abcb1 with blue light treatment (Fig. 6D), presumably an effect of compensatory ABCB19 expression (Blakeslee et al., 2007;Jenness et al., 2019) and incomplete ABCB19 inhibition. As expected, basal auxin efflux in abcb1abcb19 protoplasts was even more reduced than in abcb19 but was not changed by blue light treatment (Fig. 6E). Net efflux under both red and red plus blue light was also slightly reduced in phyB protoplasts (Fig. 6F).
These results suggest that blue light excitation of phot in leaf cells reduces but does not eliminate the auxin efflux activity of ABCB19. The impacts on PIN-directed auxin efflux are also likely to be less visible in protoplasts, as digestion of cell walls delocalizes PIN proteins on the plasma membrane (Feraru et al., 2011), and PIN expression in protoplasts is low . Another factor to consider in the interpretation of these results is epidermal cell contamination of mesophyll protoplast preparations (estimated 2% to 3%), as epidermal cells are enriched in phot1 compared to mesophyll cells (Sakamoto and Briggs, 2002).

DISCUSSION
The results presented here implicate ABCB19 as a modulator of blue light-/auxin-dependent leaf positioning and morphology (Fig. 7A). Direct phosphorylation of ABCB19 by phot1 has been shown to inhibit auxin transport in single-cell systems and intact plants (Christie et al., 2011), and loss of ABCB19 results in enhanced blue light-mediated phototropic bending (Noh et al., 2003;Nagashima et al., 2008b;Christie et al., 2011). During leaf-positioning responses, far-red light inactivation of phyB (Pfr conversion to Pr) results in increase auxin biosynthesis and promotes upright petiole angles ( Fig. 7B; Supplemental Fig. S1, A and B). Similarly, blue light activation of phot1 promotes auxin accumulations that result in upright petiole angles (Fig. 7B, Supplemental Fig. S1, C and D; Inoue et al., 2008;de Carbonnel et al., 2010;Harada et al., 2013), although not to the degree associated with shade avoidance responses. These changes in overall auxin levels require directional PIN-directed auxin transport (Fig. 7B), as pin3/4/7 mutants fail to respond to red lightinduced petiole lowering or far-red-induced petiole raising (Fig. 1). Like phyB mutants, abcb19 mutants display constitutively upright petiole angles that correlate with elevated auxin and auxin conjugate levels (Figs. 2 to 4). This is consistent with the downregulation of ABCB19 observed in leaves treated with high red:farred light (Liu et al., 2019). Together, these results suggest that ABCB19 is required for petiole positioning in response to red light and direct ABCB19 inactivation by phot1 is required for proper auxin distribution during blue light-mediated leaf positioning responses (Fig. 7, A  and B). Further, the lack of petiole angle lowering in abcb19 under red light suggests that both ABCB19 and PIN-directed transport are needed for proper response to occur. ABCB19 and PIN3/4/7 function coordinately in maintaining seedling auxin flows that are redirected in phot-mediated phototropism (Friml et al., 2002;Christie et al., 2011;Ding et al., 2011;Willige et al., 2013). These transporters appear to exhibit analogous function during leaf-positioning responses.
The role of ABCB19 during leaf-flattening responses is not as clear due the highly overlapping light signaling pathways provided by phytochromes, crys, and Plants grown under red plus far-red light leaned due to enhanced hypocotyl elongation, as is observed with Col-0 and abcb19 leaning away from and toward the camera, respectively. Scale bar 5 1 cm. B, Petiole lengths of Col-0 and abcb19 fifth rosette leaves. Data shown are means 6 SD (n 5 10). Asterisks indicate statistical difference by Student's t test (*P , 0.05). No differential response to light was detected by two-way ANOVA effect test between petiole length and the interaction between light and genotype. C, Leaf flattening in Col-0 and abcb19. Leaf flattening index was calculated by dividing the intact leaf area of the fifth rosette leaf by the manually uncurled leaf area. Data shown are means 6 SD (n 5 10). Asterisks indicate statistical difference by two-way ANOVA (P , 0.0001), followed by Dunnett's post-hoc (*P , 0.05), versus Col-0 for each condition.
phototropins. However, differences in leaf morphology and pavement cell size associated with loss of ABCB19 and ABCB1 that are not accompanied by changes in vascularization (Jenness et al., 2019;Verna et al., 2019) suggests a direct role in regulating leaf expansion. In wildtype leaves, blue light activation of phot1 enhances auxin accumulations in the leaf tip and along the leaf margins, resulting in leaf flattening (Fig. 7C). These accumulations are consistent with the DR5:GUS distributions and inhibition of auxin efflux in protoplasts presented here ( Fig. 6;  Supplemental Fig. S7). Based on these results, abcb19 mutants would be expected to display constitutively flat leaves. This was not the case, as abcb19 leaves were always more curled than wild-type leaves under blue or white light conditions (Fig. 5). The observed alterations in leaf flattening in abcb19 are more pronounced under low-light conditions supporting a primary role for ABCB19 in mediating auxin distribution particularly when auxin levels are low (Supplemental Figs. S2B and S4,C and D). However, the lack of major defects in abcb19 points to the complex regulation of leaf flattening that likely involves direct phot1 regulation of ABCB19 among several other auxin biosynthesis and transport processes.
Additional ABC transporters may make indirect contributions to blue light responses. For example, abcg37 mutants exhibited reduced indole-3-butyric acid transport in leaf mesophyll protoplasts, and abcg36 mutants exhibit enhanced cotyledon expansion (R u zička Figure 6. Auxin efflux in protoplasts is ABCB19 and blue light-dependent. 3 H-IAA efflux from leaf mesophyll protoplasts prepared from Col-0 (A), phot1 phot2 (B), abcb19 (C), abcb1 (D), abcb1 abcb19 (E), and phyb (F). Protoplasts were prepared as in Geisler et al. (2003) then incubated under 50 mmol m 22 s 21 red light at 20°C for 3 h. Protoplasts were then transferred to 4°C and incubated with 3 H-IAA (28 Ci mmol 21 ), and net efflux was assayed as in Geisler et al. (2005)  et al., 2010). ABCB14 functions in guard cell organic acid uptake and stomatal closure (Lee et al., 2008) and could also contribute indirectly to the observed changes in leaf positioning. However, it appears that manipulation of the activity of orthologs of ABCB19 in dicots and combinations of ABCB1 and ABCB19 can be sufficient to effect substantive changes in leaf absorption of photosynthetically active radiation in crop species.

Leaf-Positioning Measurements
Leaf-positioning measurements were made based on protocols described in Inoue et al. (2008) andde Carbonelle et al. (2010). In brief, soil was placed in 90 3 15-mm plastic petri dishes with holes punched in the bottom, then the dishes were placed in trays and spoil allowed to saturate by watering from below. Seeds were spread evenly across the leveled soil surface and then stratified for 2 d. Seeds were placed in a growth chamber set to 100 mmol m 22 s 2 1 light, 16 h photoperiod. Upon reaching stage 1.02 (Boyes et al., 2001), plants were transferred to continuous 50 mmol m 22 s 21 white light, 50 mmol m 22 s 2 1 red, red plus far-red, red plus 0.1 m 22 s 21 blue light, or red plus 5 m 22 s 21 blue light and allowed to grow an additional 3 or 5 d. First, true leaf petiole angles were then imaged from the side and measured using ImageJ. Angle was determined by measuring the angle formed between the hypocotyl and petiole minus 90°. Assays were conducted in triplicate.

Leaf-Curling Quantifications
Leaf curling was quantified using fully expanded fifth rosette leaves from plants grown under 100 mmol m 22 s 21 white light, 16 h photoperiod. Upon reaching stage 1.02, plants were transferred to continuous 50 mmol m 22 s 21 red light, red plus 0.1 mmol m 22 s 21 blue light, or red plus 5 mmol m 22 s 21 blue light until they were 25 d old. Following removal, the adaxial side was scanned to image the visible overhead leaf area. Leaves were then manually uncurled by cutting them along the margins and placing them flat between two small glass plates before scanning once again. Leaf areas were measured using ImageJ (Schneider et al., 2012). Data presented is the ratio of overhead leaf area to flattened leaf area. Assays were repeated twice with similar results.

Free IAA Quantitations
Free IAA levels were quantified as previously described (Jenness et al., 2019). A summary of retention times and mass transitions can be found in Supplemental Table S2.

Protoplast Auxin Transport Assays
Mesophyll protoplasts were prepared as previously described (Geisler et al., 2003) then incubated under 50 mmol m 22 s 21 red light for 3 h. 3 H-IAA loading and elimination of excess radiolabeled auxin were performed at 2 to 4°C under green safe light. Auxin transport assays were then conducted as previously Figure 7. Model for ABCB19 regulation in leaf petiole angle positioning. A, Low-fluence blue light activates phot1, which phosphorylates to inactivate ABCB19, resulting in reduced long-distance auxin transport and increased auxin in the petioles. Additionally, phot1-dependent NPH3 and RTP2 signaling contributes to leaf positioning as well as phot2 under high blue light fluence. Together, these changes result in auxin accumulations and PIN-directed transport to mediate leaf positioning and flattening responses. B, Treatment with red light, red plus blue light, and red plus far-red light result in increasing petiole auxin accumulations and leaf petiole angles, respectively, that correlate with inhibition and down-regulation of ABCB19. These accumulations are PIN dependent, as petiole angles do not change in pin3 pin4 pin7. C, Under red light alone, limited auxin flux through the leaf margins results in downward leaf curling. Addition of blue light results in increases auxin flux along the margins and leaf flattening. Under red plus far-red, increases in overall auxin biosynthesis and transport contribute to leaf flattening.
described  with incubation under 50 mmol m 22 s 21 red light or red light supplemented with 5 mmol m 22 s 21 blue light for the times specified. Collection of protoplasts was conducted under red safe light, as they were not visible under green light. Collected fractions were quickly inspected for intact protoplasts under red safe light prior to quantification by liquid scintillation counting. All assays were repeated with at least three replicates and the overall experiment was piloted once and repeated twice.

DII-VENUS Imaging
DII-VENUS in Col-0 and abcb19-101 were grown according to the conditions described in the "Leaf-Positioning Measurements" section. Before imaging, cotyledons, cotyledon petioles, and true leaves were removed to allow the true leaf petioles and coverslip to lay as flat as possible on the microscope slide. Care was taken to limit the amount of time between exposure to light and dissection to imaging to less than 10 min. DII-VENUS imaging was performed on a Zeiss LSM 710 confocal using 203 objective with an excitation of 514 nm (20%) and emission of 515 to 562 nm. Fluorescence signal was quantified using Zeiss Zen software.

Histochemical Staining
For GUS staining, tissues were incubated in 90% (v/v) acetone for 20 min at 4°C, then immersed in staining solution (50 mM sodium phosphate buffer [pH 7.0], 0.1% Triton X-100, 0.5 mM potassium ferrocyanide, 0.5 mM potassium ferricyanide, and 1 mM X-gluc) and incubated in the dark at 37°C for 5 h. Stained samples were cleared overnight with 70% (v/v) ethanol before imaging. Assays were repeated twice with similar results.

Statistical Analyses
All statistical analyses, as indicated in figure legends, were performed using JMP PRO 14. Full factorial effect tests were performed on data with two-way ANOVA, P , 0.001, to ensure light, genotype, and the interaction between light and genotype had a statistically significant effect on the parameter measured prior to applying post-hoc analysis. ANOVA and effect test P values are shown in Supplemental Table S1.

Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Leaf positioning integrates red and blue light perception and signaling.
Supplemental Figure S2. IAA levels change with light fluence and are in part PHY dependent.
Supplemental Figure S3. Loss of ABCB1 and ABCB19 results in elevated shoot-free IAA levels.
Supplemental Figure S4. Petiole angles and leaf areas are fluence and ABCB19 dependent.
Supplemental Figure S6. abcb6 and abcb20 do not exhibit leaf-positioning defects.
Supplemental Figure S7. Blue light-dependent auxin distribution is altered young abcb19 leaves.
Supplemental Table S1. Statistical analysis by two-way ANOVA and effect tests.
Supplemental Table S2. Mass transitions from liquid chromatography tandem mass spectrometry quantitations.