Protein degradation in auxin response

Abstract

The signaling molecule auxin sits at the nexus of plant biology where it coordinates essentially all growth and developmental processes. Auxin molecules are transported throughout plant tissues and are capable of evoking highly specific physiological responses by inducing various molecular pathways. In many of these pathways, proteolysis plays a crucial role for correct physiological responses. This review provides a chronology of the discovery and characterization of the auxin receptor, which is a fascinating example of separate research trajectories ultimately converging on the discovery of a core auxin signaling hub that relies on degradation of a family of transcriptional inhibitor proteins—the Aux/IAAs. Beyond describing the “classical” proteolysis-driven auxin response system, we explore more recent examples of the interconnection of proteolytic systems, which target a range of other auxin signaling proteins, and auxin response. By highlighting these emerging concepts, we provide potential future directions to further investigate the role of protein degradation within the framework of auxin response.

The authors review the deeply intertwined roles of proteolytic regulation in auxin response.

Introduction

The signaling molecule auxin holds tremendous significance in orchestrating plant growth and development (Mockaitis and Estelle 2008; Vanneste and Friml 2009). Following its synthesis, auxin is transported to cells at both short and long distances, where it elicits a specific response by initiating a signaling cascade (Vanneste and Friml 2009; Leyser 2018). To ensure appropriate auxin responses, signal transduction is subject to multi-layered control mechanisms that rely on a precisely tuned proteome (Mockaitis and Estelle 2008; Leyser 2018). Therefore, control of the abundance, activity, and localization of proteins that participate in auxin response necessitates accurate regulation. This occurs through modulation of protein synthesis by altering transcription rates (Abel and Theologis 1996), by post-transcriptional regulation (Parry et al. 2009) and post-translational modifications (Terrile et al. 2012; Cho et al. 2014), as well as through selective removal of proteins from the proteome via active degradation (Tiwari et al. 2001; Yu et al. 2015). Key switches in molecular signaling events triggered by auxin, and their proper homeostasis, require rapid and well-timed degradation of specific auxin response proteins (Mockaitis and Estelle 2008; Leyser 2018). These degradation events are deeply intertwined with auxin signaling and are facilitated by proteolytic systems that, when disrupted, generally result in aberrant accumulation patterns of auxin signaling effectors, often with severe consequences for plant development (Mockaitis and Estelle 2008).

In this review, we provide an overview of the fundamental role of proteolytic processes within the context of auxin response. Because the stepwise discovery of the nuclear auxin response mechanism from genetic, biochemical, and molecular biological data over several decades is a beautiful example of how biological principles unfold, we start with an overview of the identification of the auxin receptor that mediates gene expression. Much attention in connecting auxin response to protein degradation has focused on the Aux/IAA proteins. We extend the exploration of proteolysis in auxin response by highlighting emerging roles of protein degradation of other effectors within the auxin signaling pathway.

Identification of a core ubiquitin ligase in auxin response

Pioneering scientific work conducted in the late 19th and early 20th century led to the discovery of auxins, with indole-3-acetic acid (IAA) representing the major naturally occurring auxin (Went and Thimman 1937; Thimann and Schneider 1939; Enders and Strader 2015). These signaling molecules elicit specific plant growth responses, and their identification gave rise to the field of auxin research. Decades of efforts toward elucidating the role of auxin in plant biology resulted in the recognition of auxins as crucial coordinators of virtually all aspects of plant development (Mockaitis and Estelle 2008; Vanneste and Friml 2009). In parallel to studies that characterized the physiological roles that auxin plays, a key finding was that auxin can trigger profound changes in gene expression (Kaufman and Thimann 1978; Guilfoyle and Key 1986; Theologis 1986; Key 1989).

The pursuit of the mechanisms underlying auxin action has followed multiple parallel tracks that eventually converged. A prominent track was the genetic analysis of auxin function. Mutagenesis genetic screens based on resistance of Arabidopsis thaliana seedlings to a high concentration of the synthetic auxin 2,4-Dichlorophenoxyacetic acid, or to the auxin transport inhibitor N-1-naphthylphthalamic acid, led to the isolation of loci that are required for auxin response (Fig. 1, A, B and E). The first genes isolated this way were AUXIN-RESISTANT1 (AXR1) and TRANSPORT INHIBITOR RESISTANT1 (TIR1), which directly pointed to the involvement of the ubiquitin-proteasome system (UPS) in auxin response (Fig. 1, B and E; Fig. 2A) (Estelle and Somerville 1987; Leyser et al. 1993; Ruegger et al. 1997). While AXR1 encodes an enzyme involved in the regulation of ubiquitin ligase activity, TIR1 encodes an F-box protein, a substrate-binding subunit of E3 ubiquitin ligase complexes (Fig. 2, A and B) (Leyser et al. 1993; Ruegger et al. 1998; del Pozo and Estelle 1999). Indeed, TIR1 was found to interact with the ubiquitin ligase complex subunits Arabidopsis SKP1-like (AtASK) and AtCULLIN1 (AtCUL1), together forming the SKP1, CULLIN1, F-box (SCF)^TIR1 complex (Fig. 2, D and E) (Gray et al. 1999). The auxin-resistant6 (axr6) mutant carries a mutation in AtCUL1, which affects its interaction with AtASK and results in severe auxin-related phenotypes (Fig. 1G) (Hobbie et al. 2000; Hellmann et al. 2003). SCF complexes are multi-subunit E3 ubiquitin ligase enzymes capable of recognizing and interacting with substrate proteins, conjugating it with ubiquitin proteins by recruiting E2 ligases and thereby targeting ubiquitylated substrates for degradation by the 26S proteasome (Fig. 2A) (Vierstra 2009). The E2 ligase is tethered to SCF^TIR1 by RING BOX1 (RBX1), which associates with the AtCUL1 subunit (Fig. 2E) (Gray et al. 2002). SCF target specificity is conferred by subunits harboring the F-box domain as these often selectively interact with a singular protein or a group of related targets (Gagne et al. 2002). TIR1 was found to be a member of a small, 6-gene clade of F-box proteins, which also included AUXIN SIGNALING F-BOX PROTEIN1-5 (Dharmasiri et al. 2005a, 2005b). As mutations that disrupted SCF^TIR1/AFB assembly, as well as TIR1 overexpression, all interfered with auxin response, it was highly likely that auxin signaling required degradation of a specific protein or group of proteins capable of interacting with TIR1/AFB family members (Gray et al. 1999).

Figure 1.

Phenotypes of auxin-insensitive mutants. A) Schematic illustration of auxin-controlled plant developmental processes, which were used as read-outs in various genetic screens. B–G) Examples of mutants identified in genetic screens. The illustrations are based on (B) Estelle and Somerville (1987), (C) Wilson et al. (1990), (D) Leyser et al. (1996), (E) Ruegger et al. (1997), (F) Tian and Reed (1999), and (G) Hobbie et al. (2000).

Figure 2.

Overview of the UPS, domain architecture of auxin response proteins, and the canonical nuclear auxin pathway. A) Schematic overview of the enzymatic cascade leading to ubiquitylation of a substrate protein, resulting in its recognition and subsequent degradation by the 26S proteasome. Ub, ubiquitin. B) General domain composition of canonical TIR1/AFB, ARF, and Aux/IAA proteins. This figure is not an alignment, and the protein and domain sizes are not to scale. AC, Adenylate Cyclase; DBD, DNA-binding domain; LRR, leucine-rich repeat; MR, middle region; PB1, Phox-Bem1. C–E) Model of auxin signaling through auxin-induced and SCF^TIR1/AFB-mediated degradation of the Aux/IAA repressors. When auxin levels are low (C), ARFs are repressed by Aux/IAAs; however, upon removal of the Aux/IAA repressors by auxin-triggered degradation (D), ARF proteins are able to induce transcription of auxin-responsive genes, which includes Aux/IAA encoding genes. E) Detailed overview of the composition of the SCF^TIR1/AFB complex and hypothesized (dashed lines) as well as known (solid lines) regulatory mechanisms that tune SCF^TIR1/AFB activity.

Identification of auxin-triggered proteolysis substrates

The identification of the elusive substrate proteins for TIR1/AFB also came from genetic screens investigating mutant plants with abnormal development (Fig. 1, A, C, D and F). Several auxin-resistant mutants were identified with aberrant root and hypocotyl elongation, inhibited lateral root formation, and disturbed tropic responses: auxin-resistant2 (axr2), auxin-resistant3 (axr3), and short hypocotyl2 (shy2; Fig. 1, A, C, D and F) (Wilson et al. 1990; Kim et al. 1996; Leyser et al. 1996; Reed et al. 1998). When underlying genetic defects were identified, it was found that these mutants were all mutated in genes encoding proteins belonging to the same protein family—the AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA) proteins (Rouse et al. 1998; Tian and Reed 1999; Nagpal et al. 2000). The Aux/IAA family received its name from the identification of some of its members as transcripts that are rapidly increased upon auxin treatment (Theologis et al. 1985; Abel et al. 1994; 1995). Rapid auxin-induced transcription of these genes did not require the synthesis of new protein, indicating that all molecular components required for this upregulation were already present and could quickly respond (Abel et al. 1995). A plausible explanation for this phenomenon was the existence of a de-repression mechanism whereby short-lived repressor proteins are quickly removed upon auxin stimulation (Ballas et al. 1995; Abel and Theologis 1996).

Canonical Aux/IAA proteins share 4 conserved domains (numbered I to IV), and most of these proteins are markedly short-lived, with half-lives ranging from minutes to tens of minutes (Fig. 2B) (Oeller et al. 1993; Abel et al. 1994; Worley et al. 2000; Ramos et al. 2001; Zenser et al. 2001; Dreher et al. 2006). The short-lived nature of canonical Aux/IAAs relies on a conserved 13 amino acid “degron” embedded in domain II (Ramos et al. 2001). This degron is composed of a conserved core GWPP[VIL] motif and is required for proteasome-dependent degradation (Fig. 2B) (Ramos et al. 2001). Mutations within the degron dramatically increase Aux/IAA protein stability (Ramos et al. 2001; Tiwari et al. 2001). Interestingly, iaa7/axr2, iaa17/axr3, and iaa3/shy2 all had hypermorphic gain-of-function mutations, which are located within the degron, causing protein stabilization (Rouse et al. 1998; Tian and Reed 1999; Nagpal et al. 2000). While degron integrity is essential for Aux/IAA degradation, later work revealed that a KR motif between domain I and II and other regions outside of the degron also strongly influence Aux/IAA stability (Dreher et al. 2006; Calderon Villalobos et al. 2012; Havens et al. 2012; Moss et al. 2015; Niemeyer et al. 2020).

Molecular biological assays had meanwhile shown that Aux/IAAs are negative regulators of auxin-dependent gene activation (Ulmasov et al. 1997b). The mechanism by which they act was revealed by the use of a minimal auxin-regulated promoter and screening for transcription factor binding (Ulmasov et al. 1997a). This led to the identification of AUXIN RESPONSE FACTOR1 (ARF1), which directly binds to Auxin Responsive Elements (AuxREs) in auxin-induced promoters (Ulmasov et al. 1997a). ARF proteins are encoded by a multigene family and are discussed in more detail in other reviews (Roosjen et al. 2018; Cance et al. 2022). Canonical ARFs have a general topology consisting of an amino-terminal DNA-binding domain (DBD), a middle region (MR), and a carboxy-terminal Phox and Bem 1 domain (Fig. 2B) (Finet et al. 2013). The latter domain mediates direct interaction with Aux/IAA proteins (Kim et al. 1997; Ulmasov et al. 1997a). Thus, ARF activity is inhibited by the interaction with Aux/IAAs, whose degradation would relieve inhibition (Weijers and Wagner 2016; Roosjen et al. 2018).

These 2 strands of research were connected when IAA7/AXR2 and IAA17/AXR3 were found to be degraded in response to external application of auxin, which depended on an intact degron, TIR1, and the proteasome (Gray et al. 2001). In fact, TIR1 directly interacted with domain II of IAA7/AXR2 and IAA17/AXR3, and this interaction was stimulated by auxin (Fig. 2, B–E) (Gray et al. 2001).

Identification of TIR1 as an auxin receptor

Most pieces of the puzzle were now identified, but a crucial piece, the auxin receptor, was still missing. Further characterization of the Aux/IAA-SCF^TIR1 interaction revealed that this interaction remained intact in cell-free environments and that the site of auxin perception must therefore be soluble (Dharmasiri et al. 2003a). A real breakthrough occurred when 2 seminal studies demonstrated that TIR1 itself was capable of binding auxin (Dharmasiri et al. 2005a; Kepinski and Leyser 2005). The Aux/IAA-SCF^TIR1 interaction relied on the presence of auxin and remained auxin-inducible in heterologous systems without additional plant proteins (Fig. 2, C–E) (Dharmasiri et al. 2005a; Kepinski and Leyser 2005). It was clear that the TIR1 subunit of SCF^TIR1 was the auxin receptor, thereby bypassing the need for additional signaling components, which grants SCF^TIR1 its ability to rapidly respond to auxin. The mechanism of auxin signaling was cemented by solving the structure of the AtASK1-TIR1 complex bound to auxin and an IAA7/AXR2 peptide (Tan et al. 2007). Auxins extend the TIR1-Aux/IAA protein-interaction interface and thereby stabilize IAA degron binding; this phenomenon was famously described as auxin acting as molecular glue (Fig. 2, D and E) (Tan et al. 2007).

Regulation of SCF^TIR1/AFB activity

TIR1/AFBs and Aux/IAAs, as members of protein families, engage in diverse combinatorial interactions influenced by coexpression, binding affinities, and relative concentrations in cells (Del Bianco and Kepinski 2011; Das et al. 2021). Precise regulation of auxin-responsive gene expression requires degradation of other response proteins in addition to Aux/IAAs. Mutations that untether TIR1 from the SCF complex lead to increased TIR1 protein accumulation (Yu et al. 2015). This is caused by increased stability due to reduced TIR1 turnover by the proteasome. Structural and biochemical analysis indicates that these mutations (E12K and E15K) impede interaction with the SCF backbone AtCUL1 (Zheng et al. 2002; Yu et al. 2015). This implies that TIR1, when assembled into SCF^TIR1, is targeted for autocatalytic degradation (Fig. 2E). Alternatively, TIR1 might be targeted for degradation by an unknown E3 ligase, although this is less likely as experimental data suggest that such an E3 ligase would have to be present in yeast (Fig. 2E) (Yu et al. 2015). It is noteworthy that the amino acid residue in AFB1 that corresponds with the E12 position in TIR1 is a lysine (K8), which coincides with the finding that AFB1 does not efficiently assemble into an SCF complex and is concomitantly highly stable (Parry et al. 2009; Yu et al. 2015). This implies that AFB1 is a naturally occurring auxin receptor mutant with a distinct function from the rest of the TIR1/AFB family members (Yu et al. 2015).

Aside from being a part of an enzymatically active E3-ligase, TIR1/AFBs were recently shown to possess adenylate cyclase (AC) activity (Qi et al. 2022). The AC motif is located at the C terminus of the protein and neighbors the auxin binding pocket in the TIR1/AFB structure (Fig. 2B). In silico docking of ATP onto the TIR1/AFB structure revealed that ATP may be positioned near the Aux/IAA degron, which potentially immobilizes ATP. TIR1 AC activity increased when bound to Aux/IAA7 and 17, and it was dependent on the presence of auxin. Mutant tir1 proteins that were AC deficient were incapable of complementing the tir1-1 afb2-3 double mutant and were unable to induce transcription of auxin-responsive genes, suggesting that this AC activity has an important physiological role. However, it is not yet clear if these mutations in the AC motif also affect Aux/IAA degradation. To understand the kinetics of TIR1/AFB Aux/IAA binding in a broader context, TIR1/AFB AC activity and its effects on Aux/IAA recruitment and degradation should be examined. These recent findings, however, reinvigorated the exploration of a potential role of cAMP as a second messenger in plants, which remains enigmatic and requires further study (Qi and Friml 2023).

Plasticity of the cellular SCF repertoire is ensured through the adaptive exchange model, wherein SCF complexes undergo dynamic dis- and reassembly as well as (de)activation (Pierce et al. 2013; Liu et al. 2018). This occurs through RELATED TO UBIQUITIN1 (RUB1) modification of CUL1, which activates the SCF. But this is blocked when CUL1 is associated with CULLIN ASSOCIATED AND NEDDYLATION DISSOCIATED1 (CAND1), which dissociates the F-box protein from the SCF complex, thereby deactivating it (Pierce et al. 2013; Wu et al. 2013; Zemla et al. 2013). This system is also implicated in auxin response. First of all, the aforementioned AXR1 and E1 C-TERMINAL-RELATED1 (ECR1), together with RUB-CONJUGATING ENZYME1 (RCE1), modify AtCUL1 with RUB1 (Fig. 2E) (del Pozo et al. 1998; del Pozo and Estelle 1999; del Pozo et al. 2002b). The COP9 signalosome (CSN) promotes RUB1 cleavage and associates with SCF^TIR1 components in vivo (Lyapina et al. 2001; Schwechheimer et al. 2001). Aberrant RUB1 (de)conjugation of AtCUL1 decreased SCF^TIR1 activity and auxin response (Schwechheimer et al. 2001; del Pozo et al. 2002b; Gray et al. 2002; Dharmasiri et al. 2003b; Woodward et al. 2007). This was also observed for mutants deficient in CAND1 modification of AtCUL1 (Gray et al. 2003; Cheng et al. 2004; Chuang et al. 2004; Zhang et al. 2008). It is important to note that the RUB1 and CAND1 (de)conjugation machinery is a general mechanism for SCF regulation and is not SCF^TIR1/AFB or auxin specific. Nevertheless, the data above show that these mechanisms are important for appropriate auxin responses.

An auxin-inducible degron in nonplant systems

Throughout the characterization of the auxin response system, many of its key interactions have been explored in heterologous assays, including in yeast (Calderon Villalobos et al. 2012; Pierre-Jerome et al. 2014; Shimizu-Mitao and Kakimoto 2014; Pierre-Jerome et al. 2016). In recent years, the entire auxin response system has been ported to yeast, allowing for systematic analysis of quantitative system properties and specificity (Pierre-Jerome et al. 2014). The ability to trigger auxin-dependent degradation of proteins in heterologous systems did not go unnoticed outside of the plant research community and inspired the development of the auxin-inducible degron (AID) system, which uses the central components in the auxin response system to target any protein for degradation and thus allows for controlled and fast degradation of proteins of interest in nonplant eukaryotes (Nishimura et al. 2009, 2020; Holland et al. 2012; Li et al. 2019; Yesbolatova et al. 2020). AID involves the fusion of a target protein to an Aux/IAA degron coupled with heterologous expression of a TIR1/AFB (Nishimura et al. 2009, 2020; Holland et al. 2012; Li et al. 2019; Yesbolatova et al. 2020). The target protein is rapidly degraded upon treatment with (synthetic) auxins. AID can be used to investigate protein function during specific developmental events, such as cell cycle stages. For example, precisely timed depletion of condensins and cohesins through AID revealed distinct roles for both complexes during specific mitotic phases (Rao et al. 2017; Wutz et al. 2017; Gibcus et al. 2018; Rhodes et al. 2020; Thiecke et al. 2020). AID is a versatile customizable tool, and further innovative utilization could shed light on various other complicated biological processes in nonplant eukaryotic systems.

Beyond TIR1: auxin control of protein degradation

After the discovery of TIR1 as an F-box protein mediating auxin response, a second, unrelated F-box protein was reported to be controlled by auxin. S-PHASE KINASE-ASSOCIATED PROTEIN2 (AtSKP2A) is a constituent of a functional SCF^SKP2A ubiquitin ligase that targets cell cycle–regulating transcription factors for degradation (del Pozo et al. 2002a; del Pozo et al. 2006; Jurado et al. 2008). Notably, AtSKP2A itself is unstable and subject to degradation by the UPS (Jurado et al. 2008). In the mutant axr1, AtSKP2A accumulated while being less stable in the mutants axr2 and axr3 (del Pozo et al. 2002a; Jurado et al. 2008). Chemical inhibition of auxin signaling increased AtSKP2A levels, whereas treatment with various auxins in a cell-free system destabilized AtSKP2A, suggesting a direct effect of auxin on SKP2A stability (Jurado et al. 2008, 2010). Molecular docking simulations revealed a potential auxin binding site; when key residues within this site were mutated, AtSKP2A retained less auxin and was more stable (Jurado et al. 2010). These findings, combined with the requirement of AtSKP2A to bind auxin in order to interact with one of its substrates, suggest that AtSKP2A regulates cell proliferation as a noncanonical auxin receptor, but this awaits substantiation. It should be noted that, following a series of studies that initially reported the auxin regulation of SKP2A, there has been a paucity of further reports, and it is unclear if and how SKP2A connects to other auxin response components or to development.

In addition to being an end station for ubiquitylated Aux/IAA proteins, the 26S proteasome regulates the abundance of virtually all plant proteins, either directly or indirectly (Vierstra 2009). Beyond its role in promoting specific E3 ligases, auxin was shown to also inhibit the proteolytic activity of the 26S proteasome, leading to a profound global effect on the proteome (Yang et al. 2016). Repression of 26S proteasome activity by auxin requires PROTEASOME REGULATOR1 (PTRE1), which, counterintuitively, acts as a stimulant of 26S proteasome activity (Yang et al. 2016). Auxin depletes nuclear PTRE1 and causes PTRE1 enrichment at the plasma membrane (Fig. 3) (Yang et al. 2016). This sequence of events occurs comparatively slower than the rapid SCF^TIR1/AFB-mediated Aux/IAA degradation and potentially counteracts the rapid degradation process (Yang et al. 2016). It is proposed that PTRE1 and its homologs are components of a feedback loop, ensuring proper homeostasis of the auxin signaling cascade and on the proteome in a more general sense. Furthermore, it is hypothesized that PTRE1 has a direct role in regulating Aux/IAA degradation, as PTRE1 co-immunoprecipitated with an Aux/IAA degron peptide, although this avenue remains to be explored (Yang et al. 2016).

Figure 3.

Regulation of ARF activity and accumulation via proteolysis and PTRE1-mediated tuning of proteasomal functioning. Factors influencing degradation are color coded as indicated in the legend. Abbreviations preceding protein names indicate that these proteins are from the following plant species: At, Arabidopsis thaliana; Md, Malus domesticus; Mp, Marchantia polymorpha; Pp, Physcomitrium patens; Zm, Zea mays.

Degradation of auxin response factors

ARFs are final executors of the canonical auxin signaling pathway, as they modulate the expression of auxin-responsive genes (Weijers and Wagner 2016; Cance et al. 2022; Caumon and Vernoux 2023). It is becoming clear that the spatiotemporal components of ARF abundance, stoichiometry, and interactions are essential parameters driving auxin response (Vanneste and Friml 2009; Del Bianco and Kepinski 2011; Das et al. 2021; Caumon and Vernoux 2023). In recent years, evidence is accumulating that tuning the auxin response machinery involves ARF proteostasis. Interestingly, the abundance of multiple ARF proteins has been reported to be regulated by proteolysis. AtARF1 has a half-life of 3 to 4 hours (Salmon et al. 2008). Although AtARF1 degradation was shown to be proteasome dependent, neither knockout of AtCUL1 nor exogenous auxin treatment influenced its degradation. Therefore, AtARF1 is likely routed to the UPS through a path different from the Aux/IAAs (Salmon et al. 2008). AtARF1 contains a degron in its MR, and components required for its degradation likely reside in the nucleus; however, these components have not been identified (Fig. 3) (Salmon et al. 2008).

Another ARF, AtARF6, was previously reported to be unstable, with a shorter half-life than AtARF1 (Lakehal et al. 2019). Abscisic acid (ABA) and cold treatments promoted proteasome-dependent degradation of AtARF6, showing that hormonal crosstalk and environmental stimuli can affect ARF degradation (Fig. 3) (Li et al. 2020). Proteasome-dependent degradation of AtARF2 was also reportedly enhanced in response to ethylene treatment but was inhibited when etiolated seedlings were exposed to light (Fig. 3) (Li et al. 2004). Likewise, AtARF8 and 17 were observed to be degraded through the proteasome (Fig. 3). In none of these cases is there any insight into the protein domains or motifs (degrons) required for this degradation or the identity of the ubiquitin ligases responsible for ARF targeting. Nonetheless, ARF degradation appears to be a prevalent and potentially relevant component in auxin response.

Very recently, a first component in ARF degradation was identified. A mutation in the AUXIN RESPONSE FACTOR F-BOX1 (AFF1) was recovered in a genetic screen for mutants with excessive AtARF7 protein accumulation. AFF1 is a component of an SCF^AFF1 E3 ubiquitin ligase that recruits AtARF7 and its close paralog AtARF19, causing their proteasomal degradation (Jing et al. 2022). Counterintuitively, AtARF7 and 19 displayed lower nuclear but increased cytoplasmic abundance in root meristem cells of aff1 mutants, suggesting a regulatory role for AFF1 in driving ARF nucleo-cytoplasmic partitioning (Fig. 3). Cytoplasmic AtARF7 and 19 form biomolecular condensates, a process that is dependent on protein concentrations. Impaired SCF^AFF1-mediated degradation might be indirectly responsible for increased AtARF7 and 19 concentrations and condensate formation, although underlying mechanisms have not yet been identified (Fig. 3). The aff1 mutants have reduced auxin response, signifying that non-nuclear cytoplasmic AtARF7 and 19 are unable to contribute to auxin response even when accumulated to abnormally high levels. AFF1-directed intracellular localization of AtARF7 and 19 occurs in a cell type–specific manner, most prominent in root meristematic cells. Such ARF degradation may serve to eliminate damaged and unwanted protein yet might also provide an avenue to influence auxin response to react to biological stimuli. In either scenario, there are implied roles for regulatory modules to ensure AFF1 activity and therefore AtARF7 and 19 nuclear accumulations, but these have yet to be identified.

ARF degradation is not exclusive to Arabidopsis thaliana, but it has also been reported in other plant species. In apple (Malus domestica), the BROAD-COMPLEX, TRAMTRACK, AND BRIC A BRAC and TRANSCRIPTION ADAPTOR PUTATIVE ZINC FINGER (BTB-TAZ) PROTEIN2 (MdBT2) promotes MdARF8 degradation (Fig. 3) (Ji et al. 2022). BTB-TAZ proteins form E3 ubiquitin ligases together with CULLIN3 (CUL3) proteins (Gingerich et al. 2005). Interestingly, the MdARF8 interaction partner MdIAA3 also appeared to associate with MdBT2 (Ji et al. 2022). Overexpression of MdBT2 increased MdIAA3 abundance, and although a transcriptional effect is not ruled out, it is possible that MdBT2 stabilizes MdIAA3 post-transcriptionally since these 2 proteins were shown to interact (Ji et al. 2022). Thus, MdBT2 might affect auxin signaling by regulating MdARF8 and MdIAA3 accumulation, but this requires further experimental verification (Ji et al. 2022). Another apple ARF, MdARF5, is part of an ethylene-auxin–mediated transcriptional module that regulates anthocyanin production (Li et al. 2023). MdARF5 ubiquitylation and degradation is regulated by the E3 ligase SEVEN IN ABSENTIA11 (MdSINA11) (Li et al. 2023). A second E3 ligase, SEVEN IN ABSENTIA4 (MdSINA4), acts antagonistically to MdSINA11 and ubiquitylates and destabilizes the latter (Li et al. 2023). NAA treatment induced and reduced MdSINA4 and MdSINA11 transcription, respectively, revealing a complex degradation feedback loop to manage MdARF5 turnover (Fig. 3) (Li et al. 2023).

Recent preprints provide evidence for ARF degradation by the UPS in monocots and bryophytes as well (Das et al. 2022; Prigge et al. 2023). Comparative analysis of maize (Zea mays) and Physcomitrium patens mutants identified in genetic screens revealed a deeply conserved degron in the DBD of ZmARF28 and PpARF1-4, which was also conserved in AtARF2 (Fig. 3) (Prigge et al. 2023). Mutations in residues that are part of the degron resulted in ARF stabilization, suppression of auxin response, and strong developmental defects (Prigge et al. 2023). The unveiling of these destabilizing residues as components of a degradation signal raises questions regarding its identity and function and thus provides a stepping stone for further exploration. For example, it can be hypothesized that this degron forms an interaction site for an E3 ligase or is modified through PTMs which affect protein stability. To address these and other possibilities, subsequent studies are required. ZmARF28, PpARF1-4, and AtARF2 are class-B ARFs, which are considered repressors of auxin-responsive genes, whereas ARFs that fall into class A are generally viewed as activators of these genes (Finet et al. 2013; Mutte et al. 2018). Class-B ARF-mediated repression of auxin-responsive genes revolves around a competition model, whereby class A and B ARFs compete for the same DNA-binding sites (Lavy et al. 2016; Roosjen et al. 2018; Kato et al. 2020). It is therefore interesting that another preprint shows that Marchantia polymorpha MpARF1 and MpARF2 protein levels decline during breakage of dormancy of asexually produced clonal propagules named gemmae (Fig. 3) (Das et al. 2022). Chemical inhibition of the UPS resulted in stable MpARF1 and accumulated MpARF2 (Das et al. 2022). MpARF1 and MpARF2 fall into classes A and B, respectively. After the MpARF degradation event, MpARF1 and MpARF2 stoichiometry seemingly changed (Das et al. 2022). Experiments and mathematical modeling showed that MpARF degradation essentially reprograms the auxin response pathway and its transcriptional output (Das et al. 2022). These studies suggest that the stoichiometry of class A and B ARFs, and consequentially their competition for DNA-binding sites, can be tuned through controlled degradation, but this theory requires further substantiation. Apart from PTM through ubiquitylation, MdARF8 is also subject to SUMOylation, which is the covalent attachment of a SMALL UBIQUITIN-LIKE MODIFIER (SUMO) protein to a substrate protein, thereby altering its functioning (Johnson 2004; Zhang et al. 2021; Ji et al. 2022). MdARF8 SUMOylation occurs through the E3 ligase SAP AND MIZ1 (MdSIZ1) (Zhang et al. 2021). This required SUMO-CONJUGATING ENZYME1 (MdSCE1), which directly interacted with MdARF8, whereas MdSIZ1 did not (Zhang et al. 2021). MdSIZ1 did enhance SUMO-conjugating activity of MdSCE1, suggesting that both are required for MdARF8 SUMOylation (Zhang et al. 2021). MdARF8 SUMOylation by MdSIZ1 reduces its susceptibility to degradation by the 26S proteasome; this could be because the SUMO moiety occupies lysine residues that would otherwise be ubiquitylated (Fig. 3). MdARF8-mediated induction of transcription of auxin-responsive genes requires SUMOylation (Zhang et al. 2021). The necessity of SUMO modification of MdARF8 might rely on the fact that it confers increased protein stability, but it could also regulate activity in an alternative, thus far unknown manner (Zhang et al. 2021). Alternative consequences of SUMOylation on ARF activity are observed for AtARF7, a class A ARF that positively regulates auxin-induced lateral root formation (Okushima et al. 2007; Orosa-Puente et al. 2018). Lateral root emergence is strongly and asymmetrically promoted on the side of the root exposed to high water availability compared with the side facing dry conditions. This process is known as hydropatterning (Robbins and Dinneny 2018). AtARF7 is expressed throughout the whole root system, however, it is SUMOylated on the dry side of the root when the latter is unilaterally exposed to moisture (Fig. 3) (Orosa-Puente et al. 2018). SUMOylation of AtARF7 reduces its DNA-binding activity and is required for recruitment of the repressor IAA3, thus impairing AtARF7 function as an auxin-induced transcriptional activator (Fig. 3) (Orosa-Puente et al. 2018). Preventing AtARF7 SUMOylation abolished hydropatterning, highlighting the pivotal role of this PTM in modulating ARF activity in response to environmental cues (Orosa-Puente et al. 2018).

Lastly, a recent preprint reports on another degradation mechanism for AtARF7 and 19, which involves proteolytic turnover through macroautophagy (Fig. 3) (Ebstrup et al. 2023). Macroautophagy describes engulfment of a portion of cytoplasm by a double-membraned autophagosome, which then fuses with the vacuole where its cargo is degraded (Wang et al. 2021). Interference with the autophagy machinery led to increased accumulation of AtARF7 in the cytoplasm and reduced auxin response (Ebstrup et al. 2023). These findings are similar to observed AtARF7 behavior in aff1 mutants, although nuclear AtARF7 levels were not assessed (Wang et al. 2021; Jing et al. 2022). Autophagy entails a new mechanism through which nuclear auxin response can be modulated. Future studies should address this possibility.

Future challenges and outlook

It should be clear that auxin response is deeply intertwined with various forms of protein degradation, and these are indispensable for plant survival and development. However, large open questions remain. We outline some of the outstanding questions, and directions for their investigation, below.

1. The position of SCF^TIR1/AFB at the core of nuclear auxin response was thought to be thoroughly understood, but its recently reported AC activity raises questions regarding other, hitherto undiscovered functional roles of this complex. Two questions emerge: what roles does cAMP play in plants, and is there a connection between AC activity of SCF^TIR1/AFB and Aux/IAA degradation rates? An attractive hypothesis is that AC activity influences auxin signaling by modulating ARF activity, either directly through cAMP, or indirectly by, for example, affecting Aux/IAA stability. Evidence for this hypothesis is thus far lacking and would require further testing.

2. SCF^TIR1/AFB can be viewed as a central auxin signaling hub, and a comprehensive understanding of how SCF^TIR1/AFB activity is modulated is therefore essential. While some of these regulatory mechanisms have been observed, such as TIR1 degradation and RUB modification, underlying mechanisms remain elusive and it is unclear if, and in what contexts, these regulatory layers are relevant for the specificity and amplitude of auxin response output.

3. PTRE1 modulates 26S proteasome activity, thus integrating auxin-regulated proteostasis distinct from transcriptional regulation by the typical auxin signaling response. The exact, proteome-wide effects of this remain unclear and will be important to address to determine the relevance of this system. Furthermore, mechanistic data on how PTRE1 governs 26S proteasome activity, and on how PTRE1 itself is regulated, are required. PTRE1 interacts with an Aux/IAA degron peptide, hinting at a more direct role in managing Aux/IAA degradation, which deserves additional attention as PTRE1 might be a component of an auxin signaling feedback loop.

4. Plants are capable of responding to auxin within very short time frames, and these responses largely predate transcriptional feedback, which is relatively slow (Caumon and Vernoux 2023; Fiedler and Friml 2023). Pathways underpinning fast auxin responses have recently been identified and are reviewed by others (Friml et al. 2022; Ang and Ostergaard 2023; Caumon and Vernoux 2023; Fiedler and Friml 2023). Physiological outputs of fast auxin response are, for instance, membrane depolarization and increased velocity of cytoplasmic streaming, which are likely mediated by rapidly occurring proteome-wide (de)phosphorylation (Friml et al. 2022; Kuhn et al. 2024). PTM by (de)phosphorylation can have profound effects on ubiquitin ligase activity, protein stability and turnover, and the UPS (Nelson and Millar 2015; Lee et al. 2023). It is likely that there is crosstalk between rapid auxin response and the proteolytic machinery, which deserves more attention. Identification of the rapid auxin phospho-response revealed the power of large-scale proteomic approaches, which should be harnessed for future investigations into auxin response.

5. It is evident that qualitative and quantitative nuclear accumulation of ARFs are important parameters, defining which genes will be regulated by auxin and in what way. Dynamic signaling requires timely and selective removal of ARFs from the nucleus by degradation and/or translocation. This facet of auxin biology has remained relatively underexplored until recently. There is mounting evidence for the controlled degradation of ARFs through multiple systems; however, the underlying mechanisms remain ambiguous. To disentangle these, motifs constituting ARF degrons, stability-defining PTMs, and external factors impacting ARF degradation, and its kinetics, need to be identified and thoroughly investigated. A handful of E3 ligases that regulate ARF stability are known, but others must exist; once identified, these will provide targets for mutagenesis or interaction screens. Furthermore, some environmental conditions and hormonal treatments have demonstrated the ability to promote or inhibit ARF degradation, but how this occurs remains unknown. Lastly, the recent report of ARFs being degraded through autophagy opens new opportunities to identify novel mechanisms by which the plant cell can dictate nuclear accumulation of ARFs.

Author contributions

M.d.R. conceived of and wrote the article. D.W. and J.W.B. guided the writing process and provided corrections.

Funding

This work was supported by the Netherlands Organization for Scientific Research (OCENW.M20.031 to J.W.B.).

Data availability

No new data were generated or analysed in support of this article.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• ECR1 Gramene: AT5G19180

• ECR1 Araport: AT5G19180

• axr1 Gramene: AT1G05180

• axr1 Araport: AT1G05180

• axr3 Gramene: AT1G04250

• axr3 Araport: AT1G04250

• AFF1 Gramene: AT3G49150

• AFF1 Araport: AT3G49150

• CUL1 Gramene: AT4G02570

• CUL1 Araport: AT4G02570

• shy2 Gramene: AT1G04240

• shy2 Araport: AT1G04240

• AFB1 Gramene: AT4G03190

• AFB1 Araport: AT4G03190

• RCE1 Gramene: AT4G36800

• RCE1 Araport: AT4G36800

• TIR1 Gramene: AT3G62980

• TIR1 Araport: AT3G62980