Arabidopsis thaliana ecotype Columbia (Col-0) was used in all experiments. T-DNA insertion mutants iaa8-1 (CS25210) and iaa8-2 (SALK_202296) were obtained from SALK. T-DNA insertion was confirmed by genotyping PCR using IAA8 gene-specific and T-DNA border primers (listed in Supplementary Table 1). The transcript was confirmed by semi-quantitative RT-PCR using IAA8 gene specific forward and reverse primers (Supplementary Table 1).

Seeds were surface sterilized and then stratified at 4°C for 4 days in the dark. All seeds were germinated on plates containing half-strength Murashige and Skoog (½ MS) medium supplemented with 2% sucrose and 0.25% Phytagel. Plates were then transferred to a growth chamber at 22 ± 2°C under long day conditions (16-h-light/8-h-dark photoperiod) with 100 E m⁻² s⁻¹ light intensity.

The Cauliflower mosaic virus (CaMV) 35S::3xflag-IAA8 construct in binary vector pCAMBIA 1300 was introduced into Agrobacterium tumefaciens strain GV3101 and used for transformation of iaa8-1 mutant plants by floral dipping. Transformed lines were selected on ½ MS medium containing hygromycin (40 g/mL). Three independent homozygous lines overexpressing IAA8 were selected from the T₃ generation and used for all experiments.

Seeds were carefully harvested after siliques were fully mature. The germination assay was performed according to the method of Nguyen et al. (2012). After surface sterilization, seeds of all genotypes were stratified at 4°C for 4 days in the dark and allowed to germinate on ½ MS medium or ½ MS supplemented with 5 µM NAA or 1 µM ABA alone or together at 22 ± 2°C in a growth chamber under a 16-h-light/8-h-dark cycle. Seed germination based on radicle protrusion was quantified from day 0 until day 5. Seeds were considered germinated after radicle protrusion at the indicated time. Statistical analysis was performed, and data are presented as percentage germination rate from three independent experiments with three biological replicates.

Immunoblot analysis was performed according to the method of Kim et al. (2017). Seedlings were treated with or without MG132, cycloheximide (CHX), or H₂O₂. Tissues were ground in liquid nitrogen to fine powder, and total proteins were extracted using extraction buffer containing 50 mM HEPES, pH 7.5, 5 mM EDTA, 5 mM EGTA, 2 mM DTT, 25 mM NaF, 1 mM Na₃VO₄, 50 mM β-glycerophosphate, 20% glycerol (v/v), 2 mM PMSF, 1% Triton X-100 (v/v), and protease inhibitor cocktail (Roche diagnostics, Germany). Following two rounds of centrifugation at 12,000 × g for 15 min, supernatants were transferred to pre-chilled micro centrifuge tubes. Protein concentration was determined using a protein assay kit (Bio-Rad Laboratories, USA) with bovine serum albumin (BSA) as a standard. For immunoblot analysis, 80 µg of total protein from each sample was separated by 10% Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad Laboratories, USA). Proteins were probed using mouse anti-flag (1:5000; Sigma, USA) as primary antibody and horseradish peroxidase (HRP) conjugated anti-mouse as secondary antibody (1:5000) and visualized using an ECL kit (Bio-Rad Laboratories, USA).

Total RNA was extracted from seeds by the LiCl/phenol method according to the protocol of Nguyen et al. (2012). RNA (2 µg) was reverse transcribed using SuperScript II RNase-Reverse Transcriptase (Invitrogen, USA). RT-qPCR was performed according to the method of Kim et al. (2017) with some modification. RT product (1 µL) was mixed with gene specific primers (10 pmol) in 10 µL reaction volume. SYBR Green PCR Master Mix kit (Bio-Rad SYBR Green Supermix) was added to the mixture and incubated in a CFX384 real-time PCR detection system (Bio-Rad Laboratories, USA). Gene expression was quantified during the logarithmic phase using expression of the housekeeping gene Tubulin2 as an internal control. Semi-quantitative RT-PCR was carried out as described by Nguyen et al. (2012). Primers used for PCR are listed in Supplementary Table 2.

Ten-day-old iaa8-1 and iaa8-1/IAA8 OX seedlings were treated with cold (4°C) and H₂O₂ for 12 h. Chromatin immunoprecipitation (ChIP) was carried out as described (Gendrel et al., 2002) using mouse polyclonal anti-flag antibody (1:3000; Sigma, USA). PCR amplification was performed quantitatively using the CFX384 Real-Time System (Bio-Rad, USA). The immunoprecipitation was replicated three times. The ChIP primers are listed in Supplementary Table S2.

IAA8 negatively regulates flower development (Wang et al., 2013). To investigate the biological function of IAA8, we obtained the T-DNA insertion mutants iaa8-1 (CS25210) and iaa8-2 (SALK_202296) from SALK (Supplementary Figure 1A) and confirmed the T-DNA insertion by diagnostic PCR and semi-quantitative RT-PCR assays (Supplementary Figure 1B). The full length transcript of IAA8 was disrupted in the iaa8-1 line, whereas the transcript in iaa8-2 was similar to the Col-0 plants, suggesting that only iaa8-1 is a loss-of-function mutant (Supplementary Figure 1C).

To identify the physiological function of IAA8, we intensively screened the iaa8-1 phenotype. In a time course experiment, seeds of the selected genotypes were germinated on ½ MS medium for 5 days under long day conditions. Germination rate based on radicle protrusion was lower in iaa8-1 mutant seeds than in Col-0 and iaa8-2 seeds (Figure 1A), indicating that IAA8 positively regulates seed germination.

To further confirm the iaa8-1 mutant phenotype, we generated complementation plants expressing p35S::3xflag-IAA8 in the iaa8-1 background (iaa8-1/IAA8 OX). Three independent homozygous lines were selected from the T₃ generation. To verify ectopic expression, semi-quantitative RT-PCR and immunoblot analyses were performed (Figures 1B, D). Aux/IAA repressor proteins are rapidly degraded by the auxin-mediated 26S proteasomal pathway, thereby regulating auxin-dependent transcriptional regulation (Tiwari et al., 2003; Park et al., 2011). We assumed that IAA8 might also be degraded by the 26S proteasomal pathway, similar to other Aux/IAA proteins (Gilkerson et al., 2015). To test this hypothesis, 10-day-old seedlings of iaa8-1/IAA8 OX plants from three independent lines were treated with or without 10 µM MG132 (a proteasome inhibitor) for 3 h. Proteins were extracted and immunoblot analysis was performed. As expected, IAA8 protein was almost undetectable under mock conditions (control); however, high levels of IAA8 protein accumulated in the presence of MG132 (Figure 1D), indicating that IAA8 is degraded by the 26S proteasomal pathway.

We next tested seed germination phenotypes using Col-0, iaa8-1, and three independent iaa8-1/IAA8 OX complemented lines. Seeds of the selected genotypes were allowed to germinate on ½ MS medium under long day conditions. The germination rate of iaa8-1/IAA8 OX complemented seeds was almost the same as that of Col-0 (Figure 1C and Supplementary Figure 2), indicating that the lower germination in the iaa8-1 mutant line resulted from loss-of-function of IAA8. Three independent complemented lines showed similar expression patterns and seed germination phenotype. Therefore, we used the iaa8-1/IAA8 OX (#6-3) for further experiments, and hereafter referred to as iaa8-1/IAA8 OX.

Previous study suggests that auxin inhibits seed germination through ABA-dependent manner (Liu et al., 2013). Therefore, we tested the effect of auxin on germination by using Col-0, iaa8-1, and iaa8-1/IAA8 OX seeds in the presence and absence of ABA (Supplementary Figure 3). Seeds of the selected genotypes were allowed to germinate on ½ MS or ½ MS supplemented with 5 µM NAA and 1 µM ABA alone or together. In our experimental condition, exogenous NAA alone did not inhibit seed germination in Col-0 and iaa8-1/IAA8 OX compared to iaa8-1 mutant (Supplementary Figures 3A, B). These results suggest that even upon NAA treatment, small amount of IAA8 protein can be existed in Col-0 and iaa8-1/IAA8 OX, which inhibits ABI3 expression to promote seed germination compared to iaa8-1 mutant seeds. However, co-treatment of NAA and ABA synergistically inhibited seed germination in Col-0 and iaa8-1/IAA8 OX seeds similar to those of iaa8-1 (Supplementary Figures 3C, D), indicating that ABI3, ABI4, and ABI5 transcription by ABA may cause inhibitory effect on germination, which cannot be rescued by small amount of IAA8 proteins.

To further investigate the specific role of IAA8 in seed germination, we searched the public Arabidopsis microarray database (http://bar.utoronto.ca/eplant/) and found that the IAA8 transcript is induced by seed imbibition. We therefore proposed that IAA8 transcription might be induced during seed germination. To determine the expression pattern of IAA8 during seed germination, we performed RT-qPCR using freshly harvested Col-0 seeds. The IAA8 transcript level peaked in germinating seeds at day 2 (Figure 2A), indicating that IAA8 transcription is induced during seed germination.

Because IAA8 transcription was induced during seed germination, we assumed that the IAA8 protein might also accumulate following initiation of germination signals. We therefore germinated seeds of iaa8-1/IAA8 OX plants and performed immunoblot analysis. As expected, IAA8 protein accumulated to high levels during seed germination. Interestingly, IAA8 protein was detectable at day 2 and accumulated rapidly until day 5 (Figure 2B). By day 10, however, IAA8 protein was almost undetectable (Figure 1D), suggesting that IAA8 proteins accumulate only during germination.

Cold treatment is considered one of the most efficient stimuli of seed dormancy release (Penfield et al., 2005b). Germinating seeds produce ROS upon imbibition, which promotes radicle protrusion by rupturing the testa (Leymarie et al., 2012). Cold and imbibition therefore work in parallel during seed germination. As IAA8 protein accumulated to high levels during seed germination after cold imbibition (Figure 2B), we investigated whether cold and ROS affect IAA8 protein level. Seedlings of iaa8-1/IAA8 OX plants were exposed to cold (4°C) and 5 mM H₂O₂ alone or together for 12 h. Total proteins were then extracted and immunoblot analysis was performed (Figure 3). Under mock conditions, IAA8 protein was detected at low levels, indicating that iaa8-1/IAA8 OX plants were under mild stress during experimental handling. However, treatment with cold or H₂O₂ induced a rapid accumulation of IAA8 (Figure 3A). Interestingly, more IAA8 protein accumulated when iaa8-1/IAA8 OX seedlings were incubated under cold and H₂O₂ together (Figure 3A). These results demonstrate that cold and ROS act together to promote IAA8 protein accumulation during germination.

We next asked whether accumulation of IAA8 protein under cold and H₂O₂ conditions was caused by de novo protein synthesis or increased protein stability. We exposed iaa8-1/IAA8 OX seedlings to MG132, cold or H₂O₂ for 12 h, and then incubated them in medium containing 500 µM cycloheximide (CHX; an inhibitor of translation) with or without 20 µM 1-naphthaleneacetic acid (NAA) for 1, 3, and 6 h. As expected, IAA8 protein accumulated rapidly in the presence of MG132; however, preventing de novo protein synthesis by treatment with CHX caused significant depletion of IAA8 (Figure 3B). Next, we treated iaa8-1/IAA8 OX seedlings with cold or H₂O₂, followed by incubation with CHX. Surprisingly, preventing de novo protein synthesis by treatment with CHX did not result in complete loss of IAA8 protein in cold and H₂O₂ treated seedlings (Figures 3C, D), indicating that the increased abundance of IAA8 protein caused by cold and H₂O₂ is independent of de novo protein synthesis. Furthermore, treatment with exogenous NAA potentiated the effect of CHX, leading to faster turnover of IAA8 in MG132-treated seedlings (Figure 3B). Surprisingly, however, IAA8 protein was still detectible after 6 h of NAA treatment in both cold- and H₂O₂-treated seedlings (Figures 3C, D). Together, these results suggest that cold and H₂O₂ delay auxin-mediated degradation of IAA8 protein.

ABI3 functions in ABA-mediated seed dormancy and inhibition of seed germination. ABI3 transcript levels are high in dormant seeds and decrease rapidly after germination (Liu et al., 2013). Likewise, ABI3 protein levels decrease in germinating seeds in a light dependent manner (Lopez-Molina et al., 2002). We therefore proposed that IAA8 accumulation might regulate ABI3 expression during or after seed germination. To test this hypothesis, we germinated seeds of Col-0, iaa8-1, and iaa8-1/IAA8 OX under long day conditions and collected samples from day 0 to day 5. Total RNA was then extracted, and RT-qPCR analysis was performed. Interestingly, the transcript levels of ABI3 were comparatively high in non-germinating Col-0 seeds but declined during germination. The ABI3 transcript levels were very high in iaa8-1 seeds compared to those in Col-0 seeds (Figure 4A). Furthermore, the transcript levels of other ABA-responsive genes such as ABI4, ABI5, Em1, Em6, and RAV1 were also analyzed. The results showed that transcript levels of ABI4, ABI5, Em1, and Em6 were significantly elevated in iaa8-1 (Figures 4B–E). In contrast, the transcript level of RAV1 was very low in iaa8-1 compared to Col-0 plants (Figure 4F). Consistent with the germination results (Figures 1A, C, Supplementary Figure 2), IAA8 complementation down-regulated ABI3, ABI4, ABI5, Em1, and Em6 expression to levels almost comparable to those in Col-0 (Figure 4), indicating that IAA8 negatively regulates ABI3, ABI4, ABI5, Em1, and Em6 expression during seed germination.

Aux/IAA proteins are transcriptional repressors that regulate the expression of auxin responsive genes by inactivating ARFs activity (Tiwari et al., 2003). Recent studies suggest that ABI3 functions as a major downstream regulatory component of auxin-mediated seed dormancy (Belin et al., 2009; Liu et al., 2013). We therefore asked how IAA8 might affect the expression of ABI3. For this purpose, chromatin immunoprecipitation (ChIP) assay was performed to confirm the binding ability of IAA8 with ABI3 promoter using anti-flag antibody. We treated the iaa8-1 and iaa8-1/IAA8 OX plants with mock, cold and H₂O₂ for 12 h, and then ChIP-qPCR analysis was performed. The results showed that IAA8 associates to AuxRE motif on ABI3 promoter under mock condition in iaa8-1/IAA8 OX plants (Figure 5). Interestingly, the ABI3 promoter fragment were highly enriched in iaa8-1/IAA8 OX plants in response to cold and H₂O₂ signals compared to ChIP samples in mock and without anti-flag treated plants (Figure 5B), suggesting that binding of IAA8 to ABI3 promoter presumably through ARFs could down-regulate ABI3 transcription during seed germination.

The biological functions of most Aux/IAA genes have been identified using gain-of-function mutants, with axr5-1 (iaa1), shy2 (iaa3), iaa6 (shy1), axr2-1(iaa7), iaa8, iaa12 (bdl), iaa14 (slr), iaa16, axr3-1(iaa17), iaa18, iaa19 (msg2), and iaa28 reported to display auxin related phenotypes (Timpte et al., 1994; Kim et al., 1996; Leyser et al., 1996; Rouse et al., 1998; Hamann et al., 1999; Nagpal et al., 2000; Rogg et al., 2001; Fukaki et al., 2002; Tian et al., 2003; Tatematsu et al., 2004; Yang et al., 2004; Uehara et al., 2008; Rinaldi et al., 2012; Wang et al., 2013). Among these, the IAA8 gain-of-function mutant negatively regulates flower development (Wang et al., 2013); however, the IAA8 loss-of-function mutant shows no developmental phenotype in aerial plants parts (Overvoorde et al., 2005). In this study, we found that the IAA8 loss-of-function mutant (iaa8-1) showed delayed seed germination compared to the wild type (Figures 1A, C). This result correlates well with previous observations. First, external auxin application negatively regulates seed germination (Liu et al., 2013). Second, the gain-of-function mutants axr2-1 (IAA7) and axr3-1 (IAA17) show enhanced seed germination and exhibit stronger ABA insensitivity than the wild type (Belin et al., 2009; Liu et al., 2013). Third, tir1, tir1afb2, tir1/afb3, and tir1afb1afb2afb3 mutants with impaired ability to degrade Aux/IAA proteins show enhanced seed germination (Liu et al., 2013). These observations suggest that Aux/IAA proteins act as positive regulators of seed germination. The biological roles of other Aux/IAA proteins during seed germination should be investigated to better understand which and how many Aux/IAA proteins are involved.

The amount of different proteins in cells is controlled by translation of mRNA, which is dependent on gene expression. In addition, protein accumulation can also be achieved by increasing protein stability in response to external or internal signals (Finch-Savage et al., 2007; Carrera et al., 2008; Holdsworth et al., 2008). We found that IAA8 protein rapidly accumulated to high levels during seed germination in response to germination signals (Figure 2B). The increase in IAA8 protein was achieved not only by de novo protein synthesis but also by protein stabilization (Figure 3). Similarly, the expression of many germination related genes is highly up-regulated in response to germination-inducing signals (Wang et al., 2018). Several Aux/IAA genes, such as IAA1, IAA2, IAA3, IAA16, IAA20, IAA26, IAA28, and IAA29, are highly expressed during germination (Fujii et al., 2000; Winter et al., 2007; Carranco et al., 2010). Furthermore, transcription of IAA30, IAA11, and IAA19 is enhanced under stress conditions in germinating seeds (Park et al., 2011), suggesting that many Aux/IAA genes may play important roles during seed germination in a stress-dependent or -independent manner.

This study showed that IAA8 protein is stabilized by cold and ROS signals during seed germination (Figure 3). Similarly, IAA7 and IAA17 are highly stabilized by salt and salicylic acid (Wang et al., 2007; Liu et al., 2015). Thus, we suspect that the stabilization of IAA8 is achieved by unidentified post-translational modification (PTM). Various PTMs, such as phosphorylation, ubiquitination, and sumoylation, can alter protein stability, subcellular localization, and protein interactions in response to external signals (Yang et al., 2003). Interestingly, various Aux/IAA proteins, such as IAA1, IAA5, IAA8, IAA11, IAA13, IAA15, and IAA31, are putative substrates of kinases (Popescu et al., 2009). Therefore, there is a strong possibility that IAA8 is stabilized by cold and ROS through phosphorylation during seed germination.

ABI3 is a well-known key negative regulator in seed germination (Liu et al., 2013). ABI3 is transcriptionally induced by ABA and involved in ABA-mediated inhibition of seed germination (Bentsink and Koornneef, 2008; Liu et al., 2013). Auxin also negatively regulates seed germination through auxin response factor 10 (ARF10)- and ARF16-mediated increase of ABI3 transcription (Nemhauser et al., 2006; Santner and Estelle, 2009; Liu et al., 2013). We showed that the iaa8-1 mutant has delayed seed germination and higher levels of ABI3, ABI4, ABI5, Em1, and Em6 transcript (Figures 4A–E), while the RAV1 transcript level was markedly repressed in iaa8-1 mutant compared to wild type (Figure 4F). Similar observations in seeds of various auxin biosynthesis and signaling mutants (Liu et al., 2013) demonstrate that auxin signaling can negatively control seed germination through ARF-mediated regulation of ABI3 transcription. In addition, exogenous auxin application up-regulates the transcription of ABI3, resulting in inhibition of seed germination (Liu et al., 2007; Liu et al., 2013). By contrast, the arf10arf16 double mutant shows enhanced germination with reduced transcription levels of ABI3 (Liu et al., 2013), suggesting that ARF10 and ARF16 are required for maintenance of ABI3 transcription. Moreover, ABI3 transcription factor positively regulates ABI4 and ABI5 expression, which in turn promotes Em1 and Em6 transcription, thereby, negatively regulating seed germination (Soderman et al., 2000; Skubacz et al., 2016). In contrast, the higher transcript level of RAV1 positively regulates seed germination (Feng et al., 2014). Taken together, these results suggest that IAA8 acts as a key upstream regulator during seed germination. Additionally, ChIP assay confirmed that IAA8 associates to ABI3 promoter, specifically to AuxRE motif via unidentified interacting ARFs (Figure 5). These observations indicate that ABI3 transcription can be down-regulated through IAA8-mediated transcriptional inactivation of ARFs proteins. Interestingly, various ARFs, such as ARF4, ARF5, ARF6, ARF7, ARF8, ARF11, ARF14, ARF15, ARF16, ARF19, ARF20, and ARF22 have been previously investigated as IAA8 interacting partners (Vernoux et al., 2011; Arase et al., 2012; Wang et al., 2013; Piya et al., 2014). Therefore, we suspect that stabilized IAA8 down-regulates transcription of ABI3 by inactivating ARF proteins, which directly bind to the promoter of ABI3. A better understanding of auxin signaling in seed germination will be achieved by investigating which specific ARF proteins bind to IAA8 and regulate the transcription of ABI3.

We propose a model (Figure 6) explaining how Aux/IAA proteins regulate seed germination. Germination-inducing signals promote seed germination by increasing ROS, which triggers ABA catabolism and GA biosynthesis (Liu et al., 2010). In response to these signals, IAA8 accumulates not only by induction of protein synthesis, but also by protein stabilization. The accumulated IAA8 increases seed germination through transcriptional inhibition of ABI3 by inactivating ARFs activity.