Introduction

The development of de novo roots, or new roots (Davies et al. 2018), on above-ground plant parts like stems is referred to as adventitious rooting (AR) or adventitious root formation (Roussos 2023). This process can occur naturally, sometimes serving as a survival response to abiotic stresses like flooding or salt exposure (Roussos 2023) or it can be induced artificially (Janick 1986) as a tool for producing new plants, including those grown in the horticulture industry. New plants regenerated or reproduced by nonsexual means, derived from plant tissues or plant parts, and not involving sexual recombination, have been propagated vegetatively (Janick 1986). With few exceptions, an explant generated using vegetative or clonal propagation techniques will maintain specific, desired characteristics and remain true-to-type (Dirr and Heuser 2006). The horticulture industry has thereby employed this phenomenon to overcome sexual reproduction barriers and commercialize the production of many desirable horticultural commodities and specialties ranging from ornamental crops to fruit, nuts, and vegetable crops (Davies et al. 1994). The ease of evoking AR on horticultural commodities can vary by species (Stokes et al. 2023) and for this reason, improving AR of desirable clones of plants has become a central theme of horticultural crop development and advancement (Davies et al. 2018), as well as profitability in the horticulture industry (Konjoian 2017). Advancing technologies or techniques that improve AR can yield increase in the diversity of plant selections cultivated, which subsequently become available for application by humans (Preece 2003). In a modern horticultural context, this trend can equate to economic opportunities for producers and broader crop and plant selection availability for consumers. There are a variety of factors that play a role in the success of AR formation on plants; however, this review explores the potential for using new opportunities for chemical biology to improve the AR of plants valuable for horticultural applications.

Early history

AR of fruit trees has a long history in the evolution of human behavior because under domestication the maintenance of desired genotypes becomes practical only by vegetative propagation (Zohary and Spiegel-Roy 1975; Weiss 2015). In fact, in very ancient writings, such as the Tanakh (Tanakh 1985), there are notations about “degenerate plants” that allude to the inherent problems of seed-propagated fruit crops. Propagation by cuttings was discussed in Book 2 of Historia Plantarum (Theophrastus; Einarson 1976) as well as in Natural History (Pliny the Elder; Anthony et al. 2010). The ability of people to select and maintain unique phenotypes likely advanced very early in human history and supported the change from a nomadic lifestyle to resident agriculture because fruit tree agriculture requires site-specific long-term residence. Fruit crops that can be easily propagated vegetatively could be considered preadapted for domestication (Zohary and Spiegel-Roy 1975; Zohary et al. 2012) and formed the basis for an early understanding of innate AR ability. However, it was a rich history that early agriculturists did not rely only on spontaneous rooting and treatments that accelerated root formation (Marston 1955). For example, Weaver (1972) discussed how, for centuries, farmers in Afghanistan and gardeners in Dutch used seeds of grains, such as barley, to induce rooting on cuttings. In addition, before the use of auxins specifically was established for rooting, treatments that likely resulted in a wound-auxin response (LaRue 1941; Xuan et al. 2008; Guo et al. 2008; Canher et al. 2020) were employed. These included manipulations such as simply cutting the stem itself or root removal (Steffens and Rasmussen 2016), as well as treatments with permanganate (Curtis 1918) or carbon monoxide (Zimmerman et al. 1933). In L.H. Bailey’s classic “The Nursery-Manual” (Bailey 1920), he notes three important conditions for successful “cuttage”: a moist and uniform atmosphere, porous soil, and sometimes bottom heat. Elevated root temperature was important in the 19th and early 20th century propagation greenhouse, and it could be due to an auxin regulatory mechanism, the increase in auxin levels with elevated temperature (Gray et al. 1998). Elevated temperature effects on auxin levels are mediated by phytochrome interacting factor 4 (PIF4; Franklin et al. 2011), thus it is connected to phytochrome regulated environmental treatments such as light quality and dark exposure periods (Halliday et al. 2009; Tillmann et al. 2022) that alter plant auxin responses. The early 20th century progress on AR was reviewed by Preece (2003).

Auxins and rooting

The first reports that auxins promoted root formation in cuttings were from Thimann and Went (1934), and Thimann and Koepfli (1935). Zimmerman and Wilcoxon (1935), and later Hitchcock and Zimmerman (1939) reported soon after about their studies on a variety of different auxin-like compounds, including indole-3-butyric acid (IBA) (see Table 1 for a list of compound abbreviations and acronyms). Zimmerman and Hitchcock trademarked IBA as Hormodin, and this started the first commercial use of IBA for rooting cuttings for propagation. In the subsequent decade, many compounds were tested in numerous trials for their ability to initiate AR, mainly with woody plant cuttings (Thimann and Behnke-Rogers 1950). Such early characterizations of auxins as “root-forming hormones” established a long-standing link between auxin and auxin-like compounds and root development (Went 1929; Thimann and Went 1934). Commercial plant propagation quickly recognized and adapted the technique of applying auxin for the rooting of stem cuttings of many forest and nursery crops. These techniques were applied to cuttings collected from taxa historically considered more challenging to root and greatly expanded the number of plants that could be propagated commercially. Evaluation of the merits of various auxin application methods continued through the second half of the 20th century (Huckenpahler 1955), with IBA as a basal quick-dip and powder application methods establishing themselves as the most broadly employed techniques in horticulture (Blythe et al. 2007; http://www.rooting-hormones.com/IBAmethd.htm; http://www.getroots.net/search.html; https://npn.rngr.net/npn/propagation).

Table 1 Compounds discussed in the text, including their proper names, the abbreviations and acronyms used in the text, and alternatives that are commonly used or are trade names. Each compound is also identified by either a primary literature reference or the Chemical Abstracts Service Registry Number (https://scifinder-n.cas.org/). The compounds are listed in order of their appearance in the text
Full size table

When IBA was first found to have significant activity in the rooting process, it was widely assumed that it was a synthetic auxin, and in many growth bioassays, it was shown to be a weaker auxin than indole-3-acetic acid (IAA) (Woodward and Bartel 2005; Simon et al. 2013). A decade or more after its first uses for rooting were described, there appeared an early report of its presence in plants based on paper chromatography and bioassay in potato peelings (Blommaert 1954). It was also described as an endogenous auxin based on gas chromatography of extracts from Nicotiana (Bayer 1969). lBA was identified by gas chromatography-mass spectrometry (GC-MS) in pea root nodules (Badenoch-Jones et al. 1984) and reported in pea root and epicotyl (Schneider et al. 1985). Both free IBA and IBA released by hydrolysis from ester-conjugated lBA were clearly identified by GC-MS in the kernels and leaves of Zea mays (Epstein et al. 1989; Fallik et al. 1989; Ludwig-Müller and Epstein 1991) and subsequently in at least seven other plant species (Epstein and Ludwig-Müller 1993). Arabidopsis plants, for example, can accumulate a detectable amount of IBA (Epstein and Ludwig-Müller 1993). The endogenous presence of IBA is, however, somewhat inconsistent. For example, in maize, one variety exhibited detectable IBA and another did not (Epstein et al. 1989); in Arabidopsis, IBA has been found in both greater or lesser amounts, and in one study it was not detected at all (Novak et al. 2012). What controls the levels of endogenous IBA in specific varieties and growth conditions is not fully understood, although for Arabidopsis the pH of the growth medium, light intensity, and the volume of the culture flask can all have an effect (Ludwig-Müller et al. 1993). Similarly, the processes by which IBA is synthesized in plants need further clarification (Ludwig-Müller 2007; Damodaran and Strader 2019).

The biochemical conversion of IBA to IAA has been demonstrated in a variety of plants using isotope tracer methods (Epstein and Lavee 1984; Ludwig-Müller and Epstein 1991; Nordström et al. 1991; Van der Krieken et al. 1992; Baraldi et al. 1993, 1995; Kreiser et al. 2016). The isolation and characterization of mutants that are resistant to inhibitory concentrations of IBA or other long-side chain auxins but then respond normally to IAA or synthetic auxins has allowed the isolation of mutants defective in IBA responses, peroxisomal β-oxidation and peroxisome biogenesis (Zolman et al. 2000; Woodward and Bartel 2005; Baker et al. 2006; Damodaran and Strader 2019). The in vivo conversion of IBA to IAA involves six steps (Fig. 1), all essentially analogous to the β-oxidation of fatty acids: activation (thioesterification), oxidation, hydration, dehydration, thiolysis, and hydrolysis (Adham et al. 2005; Spiess and Zolman 2013; Rinaldi et al. 2016; Jawahir and Zolman 2021). IBA, or the synthetic auxin precursor 2,4-dichlorophenoxybutryric acid (2,4-DB) as well as chlorinated and dechlorinated IAAs, are converted to IAA or their respective analogues, indicating a somewhat permissive biochemical process. Following β-oxidation, the IAA- or auxin-like product that is generated seems to be exported from the peroxisomes. The genetic screens for Arabidopsis mutants resistant to exogenous IBA (initially named ibr, IBA resistant, or ped, peroxisome defective) has played an important role in our understanding of auxin metabolism and the special role IBA plays in this process. For example, the analysis of ech2 and other ibr mutants demonstrated that IBA-derived IAA plays an important role in root cell expansion (Strader et al. 2010b) as well as root hair and cotyledon cell expansion (Strader et al. 2010a, 2011). Although IBA application appears to have specificity for induction of root growth, it is nevertheless, only active upon its conversion to IAA (Strader et al. 2010b, 2011), indicating that it is an important auxin precursor rather than a weak auxin as originally proposed. This finding was confirmed when it was shown that the four-carbon side chain of IBA renders it unable to stimulate the formation of the TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX PROTEIN-Auxin/INDOLE-3-ACETIC ACID (TIR1/AFB-Aux/IAA) co-receptor complex required for auxin responsiveness (Uzunova et al. 2016). While studies have shown that IBA behaves differently than IAA and some have proposed it acts as a plant hormone itself (Van der Krieken et al. 1992, 1993; Chhun et al. 2004; Wang et al. 2003, reviewed in Ludwig-Müller 2020), differences in transport, metabolism, and production of nitric oxide have, however, been proposed to account for the differences reported (Schlicht et al. 2013; Fattorini et al. 2017; Michniewicz et al. 2019).

Fig. 1
figure 1

Metabolic pathway for peroxisomal β-oxidation of indole-3-butyric acid (IBA) to indole-3-acetic acid (IAA), the active auxin. IBA and 2,4-dichlorophenoxyacetic acid (2,4-D) response mutants in Arabidopsis have been key to the identification of genes encoding the enzymes for each step (as shown). They were named ibr (IBA response) or ped (peroxisome defective) mutants or, in some cases, renamed based on their protein/enzyme function. The first steps, the activation steps, use adenosine triphosphate (ATP) to form first IBA-adenylate and then the transfer of the IBA to coenzyme A to form IBA-CoA in a two-step reaction utilizing one molecule of ATP and releasing one pyrophosphate (PPi) moiety (LACS4, coenzyme A synthetase; Jawahir and Zolman 2021). Mirroring fatty acid metabolism, the IBA-CoA derivative goes through a single β-oxidation cycle, resulting in its sequential oxidation (IBR3, acyl-CoA dehydrogenase/oxidase-like; Zolman et al. 2007), hydration (IBR10, predicted enoyl-CoA hydratase; Zolman et al. 2008), oxidation (IBR1, short-chain dehydrogenase/reductase; Zolman et al. 2007), and thiolysis (PED1, 3-ketoacyl-CoA thiolase; Zolman et al. 2000) to yield IAA-CoA and an acetyl-CoA (Kaur et al. 2009). Following these steps, IAA is released (ECH2, enoyl-CoA hydratase2; Strader et al. 2011)

Full size image

Auxin induction of AR

The induction of AR can be classified as three distinct stages: the root induction period, where molecular and metabolic changes occur prior to cytological changes; root initiation, when root initials become apparent on microscopic examination; and protrusion, characterized by the emergence of the first root primordia (Berthon et al. 1990; Heloir et al. 1996; Arya and Husen 2022). Auxin primarily plays a critical role at the induction stage; later in the process, higher levels of auxin inhibit subsequent root elongation (Fig. 2). The emerging picture of IBA function in the induction of AR is advancing through the analysis of the genes involved in auxin metabolism, understanding aspects of IBA uptake and transport, and studies aimed at understanding how the endogenous presence of IBA relates to plant developmental events. One of many unanswered questions arises from the observation that when the concentrations of applied IAA and IBA were used on apple cuttings to produce a similar increase in internal IAA levels, IBA still gave more roots than IAA (Van der Krieken et al. 1992, 1993). A possible explanation is provided by the results showing that IBA → IAA conversion is accompanied by peroxisomal nitric oxide, which is important for IBA-induced lateral root formation (Schlicht et al. 2013). Thus, although IBA seems to be only active by virtue of the IBA → IAA conversion process, the spatial/temporal relationships are likely important, and how these might impact plant tissues with different physical and metabolic phenotypes is largely unknown.

Fig. 2
figure 2

Indole-3-butyric acid (IBA) response differs as the rooting process advances from induction to root growth. Data for the concentrations of IBA required for the induction of rooting and for later root growth are typical across many plant species (redrawn from data in Khan et al. 2020 and Šípošová et al. 2019). Stages of rooting have been defined differently by many authors, for example: Pre-emergance stage (S0), Early stage of root formation (S1), Massive root formation (S2), Root post-emergence (S3) (Zhang et al 2017); Induction, Initiation, Expression, Root emergence (da Costa et al. 2013); Pre-branch site formation, initiation, patterning, emergence (Atkinson et al. 2014); induction, initiation and extension (Alallaq 2021). This diagram uses a consensus of the assigned stages, where induction is followed by dedifferentiation, nuclear polarization, and branch site formation as parts of root primordia initiation. These early steps are followed by extension growth, clear expression of the root morphology, and terminating in root emergence and subsequent external root growth and development. AR, adventitious rooting

Full size image

In the absence of a rigorous understanding of why IBA and a few other auxins are effective rooting agents, empirical data has sufficed to allow the propagation of enumerable bedding, herbaceous, woody, and forest plant species. The methods are similar for most species and can be divided into three basic techniques, solvent solution dips, water dips/soaks, and dry powder treatment. Basal end solvent immersion can be for seconds to hours, depending on the concentration and plant species, with an auxin solution in a solvent such as ethanol, ethanol/water or sometimes 2-propanol. The most often used auxins include IAA, IBA, and 1-naphthaleneacetic acid (NAA), or mixtures of IBA and NAA. Over the years, some have used indole-3-butyric acid (K-IBA) or other IBA salts to prepare water-based IBA solutions. In some cases, cuttings are treated with IBA as a foliar spray of water-diluted K-IBA (Taylor and Hoover 2018). A very commonly used method is the end-dipping of plant stems into a dry powder formed from IBA, IBA/NAA or NAA and talcum powder. Various formulations are available commercially, including dilutable concentrates, pre-mixed powders, and gels. Some reports have used diluted 2,4-dichlorophenoxyacetic acid (2,4-D) solutions, although this technique is not commonly used in practice (Moura-Costa and Lundoh 1994; Barnes 2011). Over the last two decades, the evolution of chemical biology has opened up novel high-throughput screening tools to explore auxin regulation of plant development (De Rybel et al. 2009), however, the use of these for the routine manipulation of plant materials has not replaced the regular use of IBA and other simple auxins. The goal of this review is to document the potential for both research and potential applications of the knowledge provided by these emerging discoveries. A summary of these chemical effectors and their target biochemical steps are shown in Fig. 3.

Fig. 3
figure 3

Hormonal pathways leading to root organogenesis required for adventitious root formation. Shown are the steps and pools of active compounds that are impacted by the synthetic compounds/agonists, inhibitors and activators discussed in the text. Compound abbreviations are consistent with what are defined in Table 1. IAA, indole-3-acetic acid; IBA, indole-3-butyric acid; NAA, 1-naphthaleneacetic acid; 4-CPA, 4-chlorophenoxyacetic acid; MCPA, 2-methyl-4-chlorophenoxyacetic acid; 2-DP, 2-(2,4-dichlorophenoxy) propionic acid; 4-Cl-IAA, 4-chloroindole-3-acetic acid; 5,6-diCl-IAA, 5,6-dichloroindole-3-acetic acid; 4-Cl-IBA, 4-chloroindole-3-butyric acid; 5,6-diCl-IBA, 5,6-dichloroindole-3-butyric acid; ICapA, indole-3-caproic acid; naxillin, (2E)-2-({5-[3-(trifluoromethyl) phenyl]-2-furyl} methylene) hydrazinecarbothioamide; 2,4DP-glyMe, 2-(2,4-dichlorophenoxy) propanoic acid-glycine methyl ester; 4-CPA-TrpMe, 4-chlorophenoxyacetic acid-L-tryptophan-O-methyl ester; JA, jasmonic acid; DHAP, 2,6-dihydroxyacetophenone; AIEP, adenosine-5’-[2-(1H-indol-3-yl) ethyl] phosphate; kakeimide, 4-(1,3-dioxoisoindolin-2-yl)-N-(3-isopropoxyphenyl) butanamide; nalacin, N-[4-[[6-(1H-pyrazol-1-yl)-3-pyridazinyl] amino] phenyl]-3-(trifluoromethyl)benzamide; IMT, indole-3-methyltetrazole; 4-Cl-IMT, 4-chloroindole-3-methyltetrazole; sortin2, 5-[[5-(3-chlorophenyl)-2-furanyl] methylene]-4-oxo-2-thioxo-3-thiazolidineethanesulfonic acid; retinal, (2E,4E,6E,8E)-3,7-dimethyl-9-(2,6,6-trimethylcyclohex-1-en-1-yl) nona-2,4,6,8-tetraenal; AVG, L-alpha-(2-aminoethoxyvinyl) glycine; AOA, 2-aminooxyacetic acid; rhizobitoxine, 2-amino-4-(2-amino-3-hydropropoxy)-trans-but-3-enoic acid; AIB, 2-aminoisobutyric acid; NBD, 2,5-norbornadiene; TCO, trans-cyclooctene, 1-MCP, 1-methylcyclopropene; triplin, 1-(1-morpholino-1-(thiophen-2-yl) propan-2-yl)-3-(2-(trifluoromethoxy) phenyl) thiourea; SHAM, salicylhydroxamic acid; DIECA, diethyldithiocarbamic acid; jarin-1, biphenyl-4-carboxylic acid [3-(3-methoxy-propionyl)-8-oxo-1,3,4,5,6,8-hexahydro-2H-1,5-methano-pyrido[1,2-a] [1,5] diazocin-9-yl]-amide; COMO, coronatine O-methyloxime; J4, 5-[3-(trifluoromethyl) benzylidene]-1,3-thiazolidine-2,4-dione; Y11, 2-ethoxy-4-(2-nitrovinyl) phenol; Y20, 4-hydroxy-3-[(4-methylcyclohexyl) carbonyl]-2H-chromen-2-one; lyn3, 3-[2-(Pyridin-4-yl) azepane-1-carbonyl]-1,2-dihydroisoquinolin-1-one; GA, gibberellins; H-acid, (1R,4R,5S,8S)-8-(hydroxymethyl)-1,7-dimethyl-4-propan-2-ylbicyclo [3.2.1] oct-6-ene-6-carboxylic acid; Compound 67D, (2(S)-3-phenyl-(9,10-dihydro-9,10-ethanoanthracene-11,12-dicarboximido) propanoic acid; Compound 6, (2(S)-3-methyl-(9,10-dihydro-9,10-ethanoanthracene-11,12-dicarboximido) penthanoic acid; AC94377, phthalimide 1-(3-chlorophthalimido)-cyclohexanecarboxamide; A1, N-(2-aminoethyl)-naphthalene-1-sulfonamide hydrochloride; TSPC, 3-(2-thienylsulfonyl) pyrazine-2-carbonitrile; SA, salicylic acid; chlormequat, 2-chloroethyl) trimethylammonium chloride; mepiquat, 1,1-dimethylpiperidinium chloride; chlorphonium, tributyl(2,4-dichlorobenzyl) phosphonium chloride; AMO-1618, N,N,N,2-tetramethyl-5-(1-methylethyl)-4-((1-piperidinylcarbonyl)oxy)benzenaminium chloride; ancymidol, cyclopropyl-(4-methoxyphenyl)-pyrimidin-5-ylmethanol; flurprimidol, 2-methyl-1-pyrimidin-5-yl-1-[4-(trifluoromethoxy) phenyl] propan-1-ol; HOE 074 784, 1-(2,6-diethylphenyl)-imidazole-5-carboxamide; tetcyclacis, 1-(4-chlorophenyl)-3a,4,4a,6a,7,7a-hexahydro-4,7-methano-1H-(1,2) diazeto (3,4f) benzotriazole; paclobutrazol, 1-(4-chlorophenyl)-4,4-dimethyl-2-(1,2,4-triazol-1-yl) pentan-3-ol; uniconazole, (S)-E-1-(4-chlorophenyl)-4,4-dimethyl-2-(1,2,4-triazole-1-yl) penten-3-ol; inabenfide, 4’-chloro-2’-(alpha-hydroxybenzyl)-isonicotinanilide; daminozide, 4-(2,2-dimethylhydrazinyl)-4-oxobutanoic acid; prohexadione, calcium 4-(1-oxidopropylidene)-3,5-dioxocyclohexanecarboxylate; trinexapac-ethyl, ethyl 4-[cyclopropyl(hydroxy)methylidene]-3,5-dioxocyclohexane-1-carboxylate; SL, strigolactones; TIS108, 6-phenoxy-1-phenyl-2-(1H-1,2,4-triazol-1-yl) hexan-1-one; abamine, methyl 2-[[(E)-3-(3,4-dimethoxyphenyl) prop-2-enyl]-[(4-fluorophenyl) methyl] amino] acetate; tebuconazole, 1-(4-chlorophenyl)-4,4-dimethyl-3-(1,2,4-triazol-1-ylmethyl) pentan-3-ol; 2-MN, 2-methoxy-1-naphthaldehyde; 4RG, 1-[4-(4-hydroxy-but-1-ynyl)-benzyl]-4-(3-trifluoromethyl-benzyl)-piperidine-4-carboxylic acid ethyl ester; TFQ0010, (3R, 4S)-3-methyl-4-phenethyloxetan-2-one; rhodestrin, (2E,4E,6E,8E,10E,12E,14E,16E,18E)24-hydroxy-2,6,10,14,19 pentamethyltetrecosa-2,4,6,8,10,12,14,16,18nonenyl-2(hydroxymethyl)-1H-indole-3-carboxylate

Full size image

Stronger auxins for rooting

Auxins used to induce rooting are active in high concentration dips, often at 2000 mg/L (10 mM IBA) or higher, so it may be a reasoned consideration to employ “strong” auxins to lower the required concentrations or induce rooting in recalcitrant genotypes. Auxin-like compounds considered “strong” auxins include the auxinic herbicides, which typically cannot be applied at concentrations near the high-inductive doses often required due to phytotoxicity (Bottoms et al. 2011). Some have observed that auxinic herbicides at lower concentrations often induce callus and not AR (Verstraeten et al. 2013), although induction of AR has been noted as an aspect of herbicide damage (Warmund et al. 2021). The discovery of the TIR1/AFB-Aux/IAA auxin receptor (see Morffy and Strader 2022) has, nevertheless, renewed interest in auxins with very high receptor binding and less herbicidal effects. These include synthetic auxins such as NAA, 2,4-D, 4-chlorophenoxyacetic acid (4-CPA), 2-methyl-4-chlorophenoxyacetic acid (MCPA), and 2-(2,4-dichlorophenoxy) propionic acid (2-DP). In addition, 4-chloroindole-3-acetic acid (4-Cl-IAA), a naturally occurring auxin (Magnus et al. 1997), as well as related Cl-IAAs (Antolic et al. 1999) including 5,6-dichloroindole-3-acetic acid (5,6-diCl-IAA), have very high auxin activity, ranging from 10X that of IAA to 20X or more in some growth assays. Tests with many chloro-substituted IAAs found these forms to be quite active auxins (Engvild 1994; Antolic et al. 1999) rather than toxic (Slovin 1997) and this phenomenon has been confirmed in several studies. In fact, many of the monochloro and di-chloro-IAA compounds were later shown to have very high TIR1/AFB-Aux/IAA receptor binding activity (Jayasinghege et al. 2019) and minimal toxicity. Few root-induction or growth studies have further tested these compounds in practice. When tested, halogenated auxins showed promise for root induction. 5,6-diCl-IAA-methyl ester treatment significantly increased root numbers on hypocotyl cuttings of mung bean at a lower concentration than IBA (Pan and Tian 1999). Most reports were tests in bioassays other than difficult-to-root systems, possibly because they are considered rare compounds. Fortunately, facile methods for their synthesis exist (Cohen et al. 2022) and some are currently commercially available. IBA derivatives like 4-Cl-IBA and 5,6-diCl-IBA have received scant attention, although an old patent was issued (Marumo et al. 1991). We have found in preliminary studies that the halogenated IBAs have significant activity in vitro, suggesting that they are processed in the peroxisome like IBA to yield the active halogenated auxin.

Other unusual substituted IAAs have also been tested for enhanced rooting ability. For example, 4-trifluoromethylindole-3-acetic acid (4-TFM-IAA) was shown to be about 50% better than IBA at root induction in black gram [Vigna mungo (L.) Hepper] cutting, but only half as effective as 4-Cl-IAA (Katayama et al. 2008). α-Alkyl IAA derivatives were described as small-molecule agonists and antagonists of TIR1 receptor function (Hayashi et al. 2008), making them excellent candidate molecules for studies of AR, but at least in the initial reports, this biological activity was not studied.

Longer sidechain indolealkanoic acids

IAA almost never works as well as IBA at inducing rooting, and this observation has perplexed plant propagators who have not yielded a satisfactory explanation of why this happens. In some biological assays of root formation, IAA and IBA act quite differently (Chhun et al. 2004). IBA is predominantly moved as conjugates (Liu et al. 2012) but also has unique uptake mechanisms and is a saturable process (Rashotte et al. 2003), suggesting IBA uptake is carrier mediated and distinct from those involved in IAA uptake (Michniewicz et al. 2014; Frick and Strader 2018). The observation that plants convert IBA to IAA seemed to provide an easy answer to why IBA was effective (Kreiser et al. 2016). However, when concentrations of applied IAA and IBA were used on apple cuttings to produce a similar increase in internal IAA levels, IBA still gave more roots than IAA (Van der Krieken et al. 1992, 1993), so this simple idea cannot be the whole story. Although longer-side chain indolealkanoic acids have shown biological activity in growth studies comparable to that of IAA or IBA (Fawcett et al. 1960), their use in root induction appears to not have been done in difficult-to-root plants, and there is not much information on their utility (Van der Krieken et al. 1997). As with IBA, the metabolic fate of indole-3-caproic acid (ICapA; C6) is not certain, although it would be expected to mimic IBA in some regards but requires two rounds of β-oxidation to reduce the side chain from six carbons to two (Song et al. 2021). Understanding longer-chain compounds could help understand the differences between IAA and IBA in terms of the developmental signaling leading to AR initials. Also, as one increases the chain length, the compounds become more lipophilic, which could change uptake properties and require more β-oxidation activity to derive IAA. Compounds synthesized with up to an 11-carbon side chain, indole-3-undecanoic and (IUndecA; C11), have been reported. The longest even-numbered side chain for which a published synthesis is available appears to be indole-3-octanoic acid (IOctA; C8) (Avramenko et al. 1970), although the decanoic acid compound (C10) should be possible by the same procedures.

Increase rates of β-oxidation

The conversion of IBA → IAA and perhaps ICapA → IBA → IAA and longer chain conversions can also be studied in the presence of the non-auxin probe naxillin that appears to increase IBA → IAA conversion by β-oxidation in the root cap (De Rybel et al. 2012). Mutant analyses suggest that naxillin requires the endogenous IBA conversion pathway and thus acts through IBA-derived IAA to promote the development of lateral roots. Its function in tissues other than the root cap has not been extensively investigated, nor has its effect in AR or on auxin metabolism beyond the targeted reaction IBA → IAA been reported. However, other than attempting to increase the rate of uptake of IBA by methylation (Avery et al. 1937; Zimmerman and Hitchcock 1939; Rayle et al. 1970; Schenck et al. 2010) and optimization of methods of application (Blythe et al. 2007), there are few alternative chemical approaches for altering the efficacy of IBA itself.

A so-far unexplored method for up-regulating the targeted reaction IBA → IAA might be to change the carbon source for explants to favor enhanced peroxisome function. Peroxisome biology is complex, but it is clear that both stress and the need for fatty acid β-oxidation results in significant changes in metabolic capacity (Pan et al. 2020) and rates of pexophagy (Reumann and Bartel 2016). Application of mild stress or growth on fatty acids enriched media could potentially change the capacity of plants in tissue culture for higher peroxisome functions (Poirier et al. 1999) and thus potentially improve IBA → IAA activity.

Auxin conjugated forms

Auxin conjugation appears to play an important and complex role in the efficacy of AR (Haissig 1974). Landmark studies by Haissig (1989) described several esters of IAA and IBA for use in cutting propagation (Boyles et al. 1983). These included the aryl ester and aryl amide forms of IAA and IBA referred to as phenyl-IAA, phenyl-IBA, phenyl thioester-IBA, and phenyl amide-IBA. Their study showed these compounds to be more effective alternatives to the use of IAA, IBA, and NAA. Such modified derivatives have not received wide attention but have been reported to improve rooting and growth in oak and maple (Struve and Arnold 1986a, b; Struve and Rhodus 1988). While the role of conjugation remains complicated with many unresolved issues regarding its role in AR regulation, progress with the development of methods to study this has been observed in recent years. In Arabidopsis, the argonaute1 (ago1) mutants rarely form AR, and AUXIN RESPONSE FACTOR 17 (ARF17), which represses Gretchen Hagen 3 (GH3) gene expression and, thus, auxin conjugation, negatively regulates AR formation in ago1 mutants (Pacurar et al. 2014). Earlier studies of the metabolic fate of applied IBA showed that indole-3-butyryl-L-aspartate (IBAsp) levels reached a maximum 1 day after IBA treatment of cuttings. The conjugates thus formed were active in inducing the rooting of cuttings, with IBAsp being superior to free IBA. It was suggested that IBAsp might serve as an important source of auxin during later steps in AR (Wiesman et al. 1989; Riov 1993). Other IBA conjugates, such as IBAla, have also shown higher activity than IBA (Epstein and Wiesman 1987; Mihaljević and Salopek-Sondi 2012), while under specific conditions, IAA conjugates have also shown activity (Zelená and Fuksová 1991). Several different “slow release” forms have been tested for rooting, including different linkages to bovine serum albumin as well as IBA-anhydride, IBA-amino acids, IBA-polyamine-IBA, and IAA-polyamine-IAA, some with significantly positive results (Van der Krieken et al. 1997). Some synthetic auxin conjugates have shown rooting activity even when their “parent compound” was less effective. For example, auxin conjugate 2-(2,4-dichlorophenoxy) propanoic acid-glycine methyl ester (2,4DP-glyMe) was effective for vegetative propagation of mature pine tree cuttings (Riov et al. 2020). A recent report using a focused chemical screen of conjugated forms of four synthetic auxins (NAA, 2-DP, MCPA, and 4-CPA) identified 4-chlorophenoxyacetic acid-L-tryptophan-O-methyl ester (4-CPA-TrpMe) as enhancing the effect of K-IBA on AR in several recalcitrant woody plants (Roth et al. 2024). 4-CPA-TrpMe was shown not to interact directly with the TIR1-Aux/IAA7 auxin-perception complex, thus the activity seems to be related to the slow release of 4-CPA. However, it should be noted that tryptophan conjugates of IAA or jasmonic acid (JA) are endogenous auxin inhibitors (Staswick 2009). Indole-3-acetyl-L-tryptophan (IATrp) inhibited root gravitropic growth in seedlings, greatly reduced root inhibition from applied IAA, and inhibited the stimulation of lateral roots by IAA (Staswick 2009). Finally, and possibly related, various conjugates of 3-phenyllactic acid and tryptophan and esters of the conjugates showed good rooting activity in an Adzuki bean bioassay (Maki et al. 2022). While in combination, these studies would suggest that the ability to form and hydrolyze conjugates is important for AR, the issue is far from resolved (Salope-Sondi et al. 2015).

Inhibiting the induction of amide conjugation

While studies with applied auxin conjugates, mutants in conjugation, and the process of conjugation during IBA application logically all seem to suggest that conjugate formation and hydrolysis are both important aspects of AR, the results with small molecule inhibitors of auxin conjugation point to a specific role for conjugate formation. The first reported inhibitor of auxin amino acid conjugation, 2,6-dihydroxyacetophenone (DHAP) (Lee and Starratt 1986), doubled the number of roots in IBA treated cuttings (Epstein et al. 1993). Better characterized inhibitors of IAA-amido synthetase activity such as adenosine-5’-[2-(1H-indol-3-yl) ethyl] phosphate (AIEP), which mimics the adenylated intermediate of the GH3 auxin-amido synthetase reaction (Böttcher et al. 2012), bring specificity to studies of the enzyme activity resulting in indole auxin amino acid conjugation. Cano et al. (2018) found that in difficult-to-root carnation stem cuttings enhanced conjugation of auxin by GH3 enzymes leads to poor AR, and rooting ability could be restored with AIEP, the inhibitor of conjugation. After the description of AIEP, kakeimide (Hayashi et al. 2021; Fukui et al. 2022) and nalacin (Xie et al. 2022) were also found to be potent inhibitors specific to the GH3 enzyme that catalyzed the conjugation of IAA with an amino acid. Kakeimide and nalacin were shown to inhibit GH3 enzymatic activity in competition with IAA and alter root growth or AR development. A working model is that auxin homeostasis regulated by auxin-amido synthetases accounts for cultivar dependency of AR formation, and if this is true for other species, where stem cuttings show cultivar-dependent variation, this could be an important advancement (Fig. 4).

Fig. 4
figure 4

Metabolic pathways that have been described for indole-3-butyric acid (IBA) and endogenous auxin relationships that impact auxin availability for adventitious rooting. Pathways for the amino acid conjugation of indole-3-acetic acid (IAA) and, possibly IBA, by Gretchen Hagen 3 (GH3) encoded auxin-amido synthetases can be inhibited by adenosine-5’-[2-(1H-indol-3-yl) ethyl] phosphate (AIEP), kakeimide and nalacin (Böttcher et al. 2012; Fukui et al. 2022; Xie et al. 2022). The hydrolysis of specific amino acid conjugates can be inhibited by the appropriate N-carbobenzyloxy-amino acid (Chou et al. 2002). The conversion of IBA to bioactive IAA via β-oxidation can be accelerated using naxillin (De Rybel et al. 2012) and IAA biosynthesis through the two step indole-3-pyruvate (IPyA) pathway can be inhibited using either the pyruvamine PVM2153, an inhibitor of tryptophan aminotransferase, or yucasin DF, a persistent monooxygenase inhibitor or by using both together (Tillmann et al. 2022). DAO, auxin dioxygenase; IAGLU, maize IAA glucosyltransferase; ILR1-like, IAA amino acid hydrolases; TaIAR3, wheat IAA amino acid hydrolase; TGW6, rice IAA-glucose hydrolase; oxIAA, 2-oxoindole-3-acetic acid; oxIAasp, 2-oxindole-3-acetyl-aspartic acid; oxIAglu, 2-oxindole-3-acetyl-glutamic acid; UGT, uridine diphosphate (UDP)-glycosyltransferases

Full size image

A chemical probe for the involvement of the conjugation process in AR will be to employ an auxin-like compound that cannot itself be conjugated. A possible chemical form for such studies would be auxins with the tetrazole functional group that is a bioisostere (a chemical mimetic that sustains biological activity) of carboxylic acids (Hamilton et al. 1960; Quareshy et al. 2018). 3-((1H-Tetrazol-5-yl) methyl)-1H-indole (IMT) and its 4-Cl derivative, 4-Cl-IMT, have strong affinity for the TIR1/AFB-Aux/IAA receptor but not for the homologous AFB5 protein (Quareshy et al. 2018). The tetrazole analogs to IAA and 2,4-D are both weak auxins (Hamilton 1960), but in a preliminary study in our labs, we found IMT was very poor at inducing roots in a hazelnut micropropagation system (growth conditions as described in Pincelli-Souza et al. 2022). Thus, it is not clear if the observed low rooting was due to the ‘weak auxin’ character or the possible involvement of AFB5-type receptors in the induction itself. This might be something worthy of explorations using Arabidopsis auxin receptor mutants and with 4-Cl-IMT (Quareshy et al. 2018).

IAA and IBA ester conjugates

The formation and utilization of specific esters with IAA and IBA have fewer tools available to alter the process. Transgenic tomato plants expressing the maize IAA-glucose synthetase (IAGLU) gene in either sense or antisense orientation have, however, shown some potential for understanding the role of this process in AR. The maize IAGLU probe hybridized to two transcripts (1.3 kb and 2.5 kb) in wild-type tomato vegetative tissue and green fruit. Sense transformants showed an almost complete lack of root initiation and development. Antisense transgenic plants, on the other hand, had unusually well-developed root systems at early stages of development, and the amount of the endogenous 75 kDa IAGLU protein was reduced. IAGLU antisense plants also had reduced levels of IAA-glucose and lower free and esterified IAA (Iyer et al. 2005). Expression of the IAGLU gene in Arabidopsis did not have the profound effect on root growth as seen with tomato, but the plant roots were reduced in size relative to the controls (Ludwig-Müller et al. 2005). Many of the IAA glucosyltransferase enzymes, including the one originally isolated from maize (Szerszen et al. 1994), have significant activity with IBA as an alternative substrate, and the activity is inhibited by 2,4-D (Ciarkowska et al. 2021). An IAA-LEU RESISTANT1 (ILR1)-like amidohyrolase was described from wheat with a preference for longer side chain indoles (Campanella et al. 2004), and some GH3 acyl acid amidosynthetases, such as in Arabidopsis, AtGH3.15, prefer longer side chain auxins like IBA (Sherp et al. 2018; Jez 2022). The lack of good chemical biology approaches to study ester conjugation and GH3 enzymes specific for longer side chain auxins makes these aspects of auxin metabolism difficult to study in woody plant systems by interference methods using chemical inhibitors and currently must be approached by analysis of metabolic pathways. Inferences drawn from changes noted in lines with different developmental responses or studies at different time points in the developmental process are of only limited utility. A critical point for understanding auxin ester biochemistry is that, in contrast to the formation of amino acid conjugates, which reduce free auxin levels, ester formation often increases the free hormone levels (Iyer et al. 2005). This suggests that esterification is a possible mechanism for homeostatic regulation of the active hormone, and the free energy changes of the associated reactions support this regulatory role (Slovin et al. 1999), where auxin levels are tied to the levels of the ester precursor uridine diphosphate (UDP)-glucose.

Other compounds that alter tryptophan metabolism or developmental response

As noted by Bellini et al. (2014), “lateral and adventitious roots share key elements of the genetic and hormonal regulatory networks but are subject to different regulatory mechanisms.” As such, chemical modifiers of lateral root formation may impact AR. Pérez-Henríquez et al. (2012) studied how auxin responses and components of auxin signaling pathways initiate lateral root development. Using a chemical biology compound discovery approach, they found that Arabidopsis lateral root formation is at least partially independent of the auxin receptor SCFTIR. The bioactive compound found by their screen, sortin2, increased lateral root occurrence by increasing mitotic activity. The compound itself did not display auxin activity and appeared to act upstream of auxin signaling. Sortin2 accelerated endosomal trafficking and affected late endosome/PVC/MVB trafficking and morphology. While there are significant differences in the processes of lateral and AR formation, sortin2 is one of several compounds that alter cellular activity, which is important for the early stages of organogenesis.

During early seedling growth, lateral roots branch early from the primary root. This developmental process is queued by the plant circadian clock that times seedling development throughout the day and night (Davis et al. 2022). Lateral root development is driven by a temporal series of oscillating changes in gene expression referred to as the “root clock” (Dickinson et al. 2021). The function of the root clock is rephased during lateral root development and controls the levels of auxin and auxin-related genes (Voß et al. 2015) as well as a set of largely unidentified metabolic events. Blocking carotenoid metabolism either genetically or with inhibitors has profound consequences for the function of the root clock and disrupts lateral root formation. Treatment with retinal, which rescues similar clock defects in animals, significantly increased the amplitude of root clock oscillations that regulate lateral root initiation (Dickinson et al. 2021), opening the possibility that AR is also a clock-related activity that might be similarly regulated.

Auxin biosynthesis inhibitors

While some studies have found that easier-to-root varieties contain higher levels of free auxin and more recalcitrant plants have lower endogenous auxin, there are notable problems with such correlational studies (Blakesley 1994). In addition, high levels of auxin conjugates correlate with lower rooting potential (Blakesley et al. 1991; Blakesley 1994). While it might seem at odds with the general theme of application of auxins and auxin activity modifiers to consider the inhibition of native auxin biosynthesis to induce rooting, one theory is that plants generating higher levels of IAA cannot form effective auxin gradients (Benková et al. 2003) required for organogenesis or that increased production of auxin activates conjugation pathways (Li 2021). For our work on understanding basic IAA metabolism, we produced a working set of more than a dozen compounds that interfere with IAA biosynthesis, from inhibition of aspects of tryptophan synthase to the amino transferase TAA1 and the monooxygenase YUCCA (Tillmann et al. 2021). Few studies have employed such a strategy to reduce the levels of endogenous IAA biosynthesis while also providing an extrinsic auxin signal to induce AR. However, without applied auxin, root meristems will not regenerate if IAA biosynthesis is inhibited; but this can be restored by exogenous auxin (Matosevich et al. 2020).

Other plant hormones and effectors

While it is clear from the studies discussed already that auxin is a central signal in AR, it also interacts with other signaling systems (Mazzoni-Putman et al. 2021). Interactions occur via a very complex crosstalk network (Fig. 5), where signaling pathways modulate each other’s biosynthesis, catabolism, transport, and signaling response processes (Pacurar et al. 2014). AR is a complex and plastic development that originates from various cell types and requires coordination to result in a precise organogenic result regardless of cell origins (Lakehal and Bellini 2019). Because of the need for complex coordination, it is not surprising that other phytohormones and compounds that interact with hormonal signaling have effects on the process.

Fig. 5
figure 5

Circle of plant hormone interactions showing only some of the many pathways of crosstalk between auxin and other hormonal signals in a plant

Full size image

Auxin and ethylene interaction

Auxin and ethylene biosynthesis are biochemically linked via VAS1, an aminotransferase enzyme of unique specificity that catalyzes the transfer of amino functionality to indole-3-pyruvate (IPyA) to form tryptophan, thus decreasing the levels of the IAA biosynthetic precursor. Similarly, it does this at the expense of the ethylene biosynthetic precursor, methionine, to produce 2-oxo-4-methylthiobutyric acid. Thus, mutants in vas1 have higher levels of both IAA and the ethylene precursor 1-Aminocyclopropane-1-carboxylic acid (ACC) (Zheng et al. 2013). The tight linkage between the biosynthesis of these two signal regulators suggests a need for coordination. An ethylene-mediated increase in auxin sensitivity was early noted for AR formation under hypoxia (Visser et al. 1996), ethylene regulation of auxin biosynthesis and transport is critical for the control of root growth (Růzicka et al. 2007; Qin et al. 2019), and auxin-ethylene crosstalk has now been shown to regulate a plethora of root and seedling developmental events (Zemlyanskaya et al. 2018).

The biosynthetic linkage through VAS1 suggests potential problems with the use of classical ethylene biosynthesis inhibitors of ACC synthase, such as L-alpha-(2-aminoethoxyvinyl) glycine (AVG), 2-aminooxyacetic acid (AOA) (Schaller and Binder 2017), or rhizobitoxine (Owens et al. 1971). Indeed, these have a secondary effect on auxin biosynthesis because of the sensitivity of the pyridoxal-phosphate-dependent aminotransferases that generate the intermediate IPyA (Schaller and Binder 2017; Le Deunff et al. 2019; Tillmann et al. 2021).

Other ethylene biosynthesis inhibitors target ACC oxidase, including AIB (Satoh and Esashi 1980) and Co2+ (Lau and Yang 1976). Compounds that target the ethylene receptor directly can also potentially be used in studies of the ethylene effect. These might include silver ions; commonly silver thiosulfate or silver nitrate are used (Rodríguez et al. 1999; McDaniel and Binder 2012), but silver has a secondary effect on IAA efflux that complicates its use in ethylene/auxin studies (Strader et al. 2009). Perhaps more specific are the strained alkenes, including bicyclo [2.2.1] hepta-2,5-diene (NBD), trans-cyclooctene (TCO), and 1-methylcyclopropene (1-MCP) (Pirrung et al. 2008; Binder 2020), that have been used to inhibit ethylene signaling. New approaches have evolved out of chemical genetic screens, including triplin which is a mimic for ethylene (Li et al. 2017), as well as other compounds that induce the triple response (Oh et al. 2017). Many potential compounds affecting the ethylene-induced triple response are now being reported that will enhance experimental approaches (Hu et al. 2014; Schaller and Binder 2017).

The crosstalk between ethylene and auxin affects both hormones at many levels, including synthesis, response and movement (Park et al. 2017). Auxin increases ACC synthase transcription, and similarly, ethylene increases the expression of genes involved in IAA biosynthesis (Strader et al. 2010b). Some auxin responses seem to require an ethylene response, based on response mutant analysis, and conversely, an ethylene response is dependent on auxin response signaling (Stepanova et al. 2007). The interaction between ethylene and auxin on root formation, however, is different between lateral root formation and AR. Treatment of Arabidopsis and tomato seedlings with the ethylene precursor ACC reduced both the number of lateral roots and their elongation (Ivanchenko et al. 2008; Negi et al. 2008, 2010). However, both mature tomatoes and seedlings respond positively to ACC, increasing in root numbers with increasing treatment levels (Clark et al. 1999; Kim et al. 2008; Negi et al. 2010). ACC reduced auxin transport in the hypocotyl, while treatment with the ethylene signaling antagonist AgNO3 increased auxin transport (Negi et al. 2010). However, these complementary findings are complicated due to the effect of silver ions on auxin efflux (Strader et al. 2009). While these results are consistent with the primary effect of ethylene being modulation of auxin transport, and although little changes in free auxin levels was found in hypocotyls (Negi et al. 2010), auxin biosynthesis can be increased by ethylene (Qin et al. 2019). In dark-grown Arabidopsis, ACC enhanced IBA → IAA conversion and increased AR were dependent on IAA influx via AUX1/LAX3. In contrast and apparently different from tomato, ACC decreased rooting in dark-grown Arabidopsis when applied alone (Veloccia et al. 2016). Ethylene is not always a major regulator of rooting; nevertheless, as Fogaca and Fett-Neto (2005) noted in microcuttings of Eucalyptus, ethylene appears to have only a minor role in the development of AR and that the observed response results from a more direct effect of auxins. Ethylene-auxin crosstalk also regulates the initiation of AR near cut sites where the levels of auxin and ethylene both increase (Guan et al. 2019). However, auxin inhibiting AR elongation later in the process was shown to be a consequence of stimulated ethylene production (Bai et al. 2020).

There are few successful uses of ethylene and ethylene biosynthesis or response modifications for AR, likely because it acts in both a narrow dose-dependent and developmental-specific manner, functioning as either a promotive signal or an inhibitor of AR formation (De Klerk and Hanecakova 2008; Druege et al. 2019). For example, when applied together with IBA, ethylene at times promotes the conversion of IBA to IAA and thus improves the development of AR (Veloccia et al. 2016). Although the relationships are complex, it is clear that ethylene-auxin interactions play key roles in AR, and future efforts to bring regulation of both processes into practice have potential for practical uses as well as further research into these important signaling processes and crosstalk.

Auxins and JA crosstalk

JA functions as an integral part of a stress-related signaling network in coordination with other phytohormone processes. They induce responses during wounding and biological stress, and the elicitation or enhancement of secondary metabolite production is often seen with plants under in vitro culture. JAs can, in some cases, stimulate the proliferation of shoots, roots, and calluses and induce storage organ formation. However, negative effects of JAs, such as the induction of leaf senescence, reduced growth and inhibited somatic embryogenesis, are also possible (Kamińska 2021). AR initiation involves JA and auxin regulatory controls that intersect via ARF proteins and the GH3 gene family. GH3 genes encode enzymes responsible for the conjugation of various amino acids to either auxin or JA, leading to their activation, inactivation, or degradation. Together, ARF and GH3/JAR1 regulate the level of JA-Ile, which is the active form of JA that interacts with the CORONATINE INSENSITIVE1 (COI1)-JAZ co-receptor (Monte et al. 2022). JA then negatively regulates AR through the activation of the COI1 signaling pathway (Gutierrez et al. 2012). However, it seems that the effect of JA on AR depends on the conditions (Pan et al. 2021). At < 10–6 M, methyl-JA has been shown to promote AR development when applied together with IBA (Fattorini et al. 2009), and this points to the apparent synergy between JA and auxin signaling, where together they promote root regeneration (Fattorini et al. 2018; Zhou et al. 2019). An important aspect of that relationship is the regulation of auxin levels by JA where COI1-dependent JA signaling controls the expression of the IAA oxidation process genes DIOXYGENASE FOR AUXIN OXIDATION 1 (DAO1) and its closely related paralog DAO2 (Lakehal et al. 2019).

Research in the 1980s established that JA is formed from α-linolenic acid (LA) esterified in chloroplast membranes. Lipoxygenase (LOX) is the first enzyme that converts LA to 13(S)-hydroperoxyoctadecatrienoic acid (13-HPOT), the first precursor of JA (Wasternack and Song 2017). Salicylhydroxamic acid (SHAM) is an efficient JA biosynthesis inhibitor that blocks LOX activity (Yoshihara and Greulich 1999). Also, the common reagent diethyldithiocarbamic acid (DIECA) has been used to inhibit JA biosynthesis, as it is a strong reducing agent and inhibits the octadecanoid pathway. DIECA apparently can efficiently convert 13-HPOT to 13-hydroxylinolenic acid, thereby diverting the pathway into a dead-end product (Farmer et al. 1994; Li et al. 2020).

Neomycin has been described as a non-specific inhibitor of phospholipase C that affects the formation of inositol trisphosphate (IP3). Because a role for inositol phosphates is emerging in JA perception (Peiter 2011), primarily from the observation that inositol pentakisphosphate (IP5) binding is crucial for the COI1-JAZ co-receptor complex to perceive JA-Ile with the required sensitivity (Mosblech et al. 2011). Extending this, Vadassery et al. (2019) showed that neomycin inhibited JA-mediated responses in several diverse plant species.

Chemical biology approaches have also been employed to identify several compounds that target COI1 or the COI1/JAZ co-receptor and prevent JA signal transduction (Monte et al. 2022). Two inhibitors of JA signaling were developed about a decade ago, including jarin-1, which is a cytidine alkaloid derivative (Meesters et al. 2014) identified from screening a library of natural and semisynthetic compounds. A second compound, a derivative of coronatine (which was isolated from Pseudomonas syringae, a plant pathogen that produces this JA-Ile mimic), leads to coronatine O-methyloxime (COMO) by rational design (Monte et al. 2014). These two inhibitors are fundamentally different in that COMO prevents the COI1-JAZ interaction, whereas jarin-1 inhibits the formation of JA-Ile at the level of the enzymatic activity of JAR1 (Monte et al. 2014; Meesters et al. 2014). More recently, new compounds have been added to target the co-receptor. These include J4, Y11 and Y20, which appear to be direct antagonists of JA-Ile perception by COI1-JAZ complexes (Chini et al. 2021). Both Y11 and Y20, although active in vitro, failed to prevent JA-mediated responses in planta. J4 was active in plants at inhibiting JA responses but was also shown to be an antagonist for aspects of auxin signaling via SCFTIR1. However, J4 does not seem to affect other hormonal pathways. Finally, based on a virtual screening procedure informed by the crystal structure of the COI1-JAZ1 co-receptor, Lin et al. (2022) described a compound from the ZINC database of commercially available compounds (https://zinc.docking.org/) they called lyn3. Lyn3 suppresses the expression levels of JA-related genes in Arabidopsis and decreases herbivore resistance in tea plants, but possible interactions with auxin signaling were not examined. Clearly, new chemical biology approaches have now been developed and continue to appear, and these will allow JA and auxin interactions in AR to be more fully explored.

Auxin and GA interactions

Auxin and GA have similar or overlapping functions in their signal pathways that regulate aspects of plant development including root growth and tissue expansion. Both auxin and GA signaling have several points of convergence that allow crosstalk for the regulation of developmental events (Franklin et al. 2011; Richter et al. 2013), but they are not redundant in their functions. However, GA could apparently alleviate aspects of Aux/IAA gain-of-function phenotypic expression (Frigerio et al. 2006). Such studies suggest that changes in GA levels can, in part, also mediate aspects of auxin developmental activities (Hanson 1976). Other studies suggest that GA inhibits the formation of AR in plants by disrupting endogenous hormonal processes, including hormone levels and auxin transport (Willige et al. 2011; Mauriat et al. 2014; Li et al. 2015; Zhang et al. 2021). However, this conclusion needs context since pretreatment of cuttings with GA can result in enhanced AR in cuttings and in vitro (Ford et al. 2002; Zhao et al. 2022), and in some cases, at least GA can substitute for auxin in root induction (Kim and Cha 2015). Pizarro and Díaz-Sala (2020) suggested, based on their studies in which GA inhibited AR and GA biosynthesis inhibitors had little effect, that changes in AR may be more likely related to the maturation-related decline of AR formation in contrast to a direct change in hormonal processes per se. These studies are complex, and the relationship between auxin and GA still demands additional studies to decipher how auxin controls the growth of roots by modulating cellular responses to GA (Fu and Harberd 2003). Or, conversely, how auxin response may be modified by GA, or if these are simply coincidences because of other aspects of plant development (Brumos et al. 2018). There are, however, many chemical and genetic approaches, and these need to be applied to sort out such relationships in high detail.

There are several approaches to GA chemical biology. First, GA can be supplied as the active compound, typically either GA4 or GA1 (Hedden 2016; Miller and Bassuk 2022), but sometimes GA5 or GA6 (King et al. 2001), the biosynthesis of which, via GA20ox and GA3ox gene products, is regulated by auxin (Hedden and Phillips 2000). Often, treatment with bioactive GA3 is used in place of the more prevalent bioactive GA1 and GA4. However, although GA catabolism is complex and still an active area of study, GA3 is not a substrate for the GA 2-oxidase enzyme that deactivates both GA1 and GA4 (He et al. 2020) and thus may be a more persistent form (Shechter et al. 1989). An alternative to studies using active forms of GA is to employ a receptor agonist and several are known. The GA agonist helminthosporol from Helminthosporium sativum showed GA-like bioactivity in a rice bioassay (Briggs 1966). The analog H-acid was shown to have higher GA-like activity and chemical stability than helminthosporol (Miyazaki et al. 2017). Indeed, H-acid is active in both rice and Arabidopsis, where it can regulate the expression of GA-related genes. The H-acid also induced DELLA degradation by forming a GIBBERELLIN INSENSITIVE DWARF1 (GID1)-(H-acid)-DELLA complex, showing it to be a true chemical agonist. Later, an H-acid analog, in which the hydroxymethyl group at the C-8 position of H-acid was converted to a keto group, acts as a selective GA receptor agonist. While the keto analog showed higher activity for Arabidopsis hypocotyl elongation than H-acid, additional studies of the complex with GID1-DELLLA, strongly suggested that the selectivity of such analogs depended on the specificity of their GA receptor binding activity (Miyazaki et al. 2018). The succinimide “Compound 67D” and a related “Compound 6” (a more active derivative) were obtained from a chemical library screen and shown to promote DELLA degradation, thus functioning as additional agonists (Jiang et al. 2017b). This group also studied a different GA agonist, a substituted phthalimide AC94377 that also mimics the growth-regulating activity of GAs in various plants, despite its structural differences (Jiang et al. 2017a). AC94377 was shown to be selective for a specific subtype among three Arabidopsis GID1s, and the selectivity of AC94377 was related to a single residue in the hydrophobic pocket of GID1. The GID1 family member specificity of keto H-acid and AC94377 could be valuable to understanding exactly which receptor family members interact with auxin signaling in AR (Jiang et al. 2022). A related compound, “A1” acts synergistically with GA to enhance degradation of DELLA proteins but requires the presence and GID1 perception of endogenous GAs (Sukiran et al. 2022).

The alternative to GA and GA agonists, are GA inhibitor antagonists. Using a yeast two-hybrid system for monitoring for two proteins interactions with GID1-DELLA identified a compound, 3-(2-thienylsulfonyl) pyrazine-2-carbonitrile (TSPC), that was an inhibitor for GA perception both in vitro and in planta (Yoon et al. 2013). The phytohormone salicylic acid (SA) also functions as a GA antagonist by promoting the polyubiquitination and degradation of GID1 (Yu et al. 2022), but of course SA also has other regulatory effects making its use for studies of GA responses quite complex.

Inhibitors of GA biosynthesis have both a practical use for plant size regulation and play an important role in research on this hormone class. These inhibitors have been classified into four chemical groups (Rademacher 2000): (1) Onium compounds, such as chlormequat chloride, mepiquat chloride, chlorphonium, and N,N,N,2-tetramethyl-5-(1-methylethyl)-4-((1-piperidinylcarbonyl) oxy) benzenaminium chloride (AMO-1618), which function by blocking early steps of GA metabolism by inhibition of the cyclases copalyl-diphosphate synthase and ent-kaurene synthase; (2) Nitrogen heterocycles such as ancymidol, flurprimidol, 1-(2,6-diethylphenyl)-imidazole-5-carboxamide (HOE 074 784), tetcyclacis, paclobutrazol, uniconazole-P, and inabenfide that inhibit the cytochrome P450-dependent monooxygenases, thereby preventing the oxidation of ent-kaurene into ent-kaurenoic acid. (3) Structural mimics of α-ketoglutarate, which is the co-substrate of dioxygenases, thus blocking the late steps of GA formation. These include the cylcyclohexanediones, such as daminozide, prohexadione-Ca, and trinexapac-ethyl, that block, for example, the 3β-hydroxylation step to prevent the formation of the active GAs from inactive precursor GA compounds; (4) Structural mimics, like 16,17-dihydro-GA5, and similar structures that are competitive inhibitors of these late-step dioxygenases (Zhou et al. 2004).

Although chemical biology approaches for studies of GA response and for control of its biosynthesis are well established in both research and for practical uses, the application of these chemical approaches to explore auxin-GA interactions in AR has so far been quite limited (see Vaičiukynė et al. 2019; Žiauka and Kuusienė 2010). Likely, in part, because of this, our understanding of the interaction of these two phytohormones in organogenesis is fragmentary, and some aspects remain contradictory. Hopefully, as tools continue to expand and as interest increases, these can be studied in detail.

Auxin interactions with other signal messengers

Other signaling messengers and signaling mimics also affect rooting, although the chemical tools for their study may currently be more limited. Strigolactones (SL) is a relatively “new” hormone, as compared to the classical five (Butler 1995), and it is particularly noted for being released as a root exudate. In this way, it was found to be important for the establishment of mycorrhizal and possibly other microbial associations. SL also acts as a facilitator of plant growth by enhancing lateral root formation and root hair elongation (Kapulnik et al. 2011) but inhibiting shoot branching as well as AR formation (Gogna et al. 2022). The complex structures and rapid rates of degradation of the various natural SL have limited their experimental uses, but more stable synthetic analogs are available, such as SL (GR24), and several different SL agonists show promise (Jiang and Asami 2018). Both abscisic acid (ABA) and SL are derived directly from carotenoids, but other plant hormones, such as GA, cytokinins, and brassinosteroids, are derived from the precursors for cytosolic and plastidic isoprenoids, so finding inhibitors is complex. Often, compounds that inhibit the biosynthesis of SL are shown to have non-target effects (Ito et al. 2013a). Also, GA and GA-mimics have been shown to suppress SL biosynthesis, which might provide a secondary strategy to reduce SL levels (Ito et al. 2017; Jiang and Asami 2018). Nevertheless, compounds that target the carotenoid cleavage dioxygenase enzymes max3 and max4 (max = “more axillary growth”) have been developed. For example, 6-phenoxy-1-phenyl-2-(1H-1,2,4-triazol-1-yl) hexan-1-one (TIS108) has been reported as a SL biosynthesis inhibitor (Ito et al. 2011, 2013a) and shows a better level of specificity than several similar structures tested. The SL biosynthesis inhibitors abamine (Sergeant et al. 2009) and tebuconazole (Ito et al. 2013b) have also been described as inhibitors in the later steps of SL biosynthesis from carotenoids. Receptor antagonists for SL include 2-methoxy-1-naphthaldehyde (2-MN) (Mashita et al. 2016) and soporidine (Holbrook-Smith et al. 2016; Yao et al. 2017). Xiang et al. (2017) developed multiple β-lactones, such as (3R, 4S)-3-methyl-4-phenethyloxetan-2-one (TFQ0010), which covalently modifies the SL receptors, leading to an irreversible antagonist response to GR24 application. However, the biological utility of TIS108, abamine, and tebuconazole that inhibit SL biosynthesis, as well as 2-MN, soporidine, and TFQ0010 antagonists, all need to be explored further in relation to their effect on AR.

Other novel metabolites have been shown to alter AR, at least in the specific plants tested. The anthranilate metabolite produced by Rhodobacter sphaeroides via photobiotransformation was termed rhodestrin (Sunayana et al. 2005) and markedly stimulated AR at 50 nM with in vitro cultures of mulberry.

The process of AR development is complex, with many overlapping and crosstalk possibilities. As discussed in a recent review, brassinosteroids, SL, sphingolipids, and nitric oxide all interact to alter AR in some ways (Altamura et al. 2023); however, the precise regulatory basis for “the observed interactions in root formation and plasticity are still to be discovered”.

Conclusions and perspectives

The discovery of the effects of IBA on AR 90 years ago had a profound impact on applied plant propagation methodology, bringing new forests, gardens, and fruit crops into wider use. AR, however, is a complex response, as might be expected for the generation of new organs from differentiated or partially differentiated tissues. The AR response is also sensitive to the plant’s environment, including light or darkness (Monteuuis and Bon 2000; Sorin et al. 2005; Klopotek et al. 2010; Pincelli-Souza et al. 2022), temperature (Corrêa and Fett-Neto 2004), as well as stress factors such as water availability and mineral nutrition (De Almeida et al. 2017). Regardless of the advances in research that have established IBA natural occurrence, the conversion of IBA to IAA as a critical process, and an array of hormonal and environmental needs, most of the commercial plant propagation remains primarily focused on the proper application of IBA, NAA, or combinations of the two (Sharma and Thapa 2022). It should be possible to do better, and as outlined in the review there are several opportunities to improve and many avenues to explore. A quickly evolving library of chemical effectors suggests new opportunities for future investigations that will allow us not only practical avenues for pretreatments but will also enable us to identify the molecular and signal transduction mechanisms underlying adventitious rooting as a key determinant in clonal propagation efficiency. A better understanding of cellular signals and regulatory cascades in development that are involved in adventitious root formation, underpinned by advances in systems and chemical biology, will provide a more complete understanding of rooting recalcitrance. It is the hope of this review that it will encourage a 21st century effort to bring new ideas to both research and the application of AR for plant improvement.