Salicylic acid (SA) is closely linked to the PPP and known to be an important signaling compound in plant defense responses against both biotic and abiotic stresses (Fragnière et al., 2011). This combination makes SA a likely candidate to explain some of the growth defects observed in lignin mutants. Production of SA is facilitated via two different routes (Lefevere et al., 2020) of which one branches directly from the PPP, more specifically from the intermediate t-CA (Richmond and Bleecker, 1999; Figure 1). A second route, which produces most of the SA in Arabidopsis, goes via the production of isochorismate through the shikimate pathway (Wildermuth et al., 2001; Torrens-Spence et al., 2019) and is positioned upstream of the PPP (Figure 1). Possibly as a consequence of the position of these pathways, either upstream or at the entry point of the PPP, SA typically accumulates when the PPP is blocked downstream of PAL (Schoch et al., 2002). For example, an increase in SA was observed in a dwarfed HCT-RNAi line (Gallego-Giraldo et al., 2011a). Blocking SA accumulation in these mutants by crossing them with either an isochorismate synthase 1 mutant (sid2; Figure 1) deficient in SA biosynthesis or the SA-conjugating NahG line partially restored plant growth, indicating an involvement of SA in the induced dwarfism of lignin mutants. Evidence was provided that a mediation of gibberellic acid signaling might be involved in the SA-induced dwarfism (Gallego-Giraldo et al., 2011a), although the exact mechanism is still unclear.

In concordance with the HCT-RNAi line, SA levels were increased in a c3′h mutant (ref8-1) (Kim et al., 2014) and restoration of the pathway by reintroducing C3′H expression restored growth and brought SA content back to WT levels. However, in contrast to the HCT-RNAi line, preventing SA accumulation in the ref8-1 mutant by crossing it with the NahG line did not result in growth restoration. A parallel study wherein the ref8-1 mutant was crossed with a sid2 mutant also showed SA accumulation not to be at the basis of the growth phenotypes (Bonawitz et al., 2014). A ref8-1 mediator (med)5a/5b double mutant, however, showed a full growth restoration while still having increased levels of SA. It is striking that SA would cause growth reduction of HCT-downregulated plants but not that of c3′h mutants, given their proximal position in the PPP and their comparable mutant phenotypes. Therefore, whereas SA accumulation seems to be at the basis of dwarfism of HCT-downregulated plants, the results obtained for the c3′h mutant question the role of SA accumulation as a general mechanism underlying growth reduction in PPP mutants.

Cinnamic acid (CA) is produced upon deamination of phenylalanine by PAL (Figure 1). Although the enzymatic reaction results in the formation of t-CA, which is further metabolized by the PPP, also its bioactive cis-isomer (c-CA) is present in plants (Yin et al., 2003; Wong et al., 2005; El Houari et al., 2021). Both isomers are readily interconvertible under the influence of UV-light, and this mechanism has long been considered as a source for c-CA production (Yin et al., 2003; Wong et al., 2005). Recently, evidence was also provided for a UV-independent and possibly dedicated enzymatic biosynthesis of c-CA in plants (El Houari et al., 2021). The possibility of accumulating CA-esters or their derivatives having a role in the observed growth perturbation of PPP mutants was already coined by Franke et al. (2002) and CA has indeed been implicated in the growth defects found in several c4h mutants (Schilmiller et al., 2009; Kurepa et al., 2018; Kurepa and Smalle, 2019; El Houari et al., 2021). Genetic studies toward C4H were mainly performed on weak c4h mutants (ref3-1; ref3-2; ref3-3; Schilmiller et al., 2009) nonetheless showing large changes in lignin content and composition. The depletion in lignin went paired with dwarfism, sterility, and increased branching (Schilmiller et al., 2009). However, it was later suggested that the increased branching of the ref3-1 mutant, as well as a then-observed increase in lateral rooting, were the result of a altered auxin sensitivity caused by the accumulation of a PPP intermediate upstream of C4H (Kurepa et al., 2018). Subsequent investigation indicated the responsible bioactive compound to be downstream of PAL. Specifically, t-CA or a t-CA derivative was put forward as responsible for the increased branching and lateral rooting. Another allelic mutant, ref3-3, showed an increase in rosette size and biomass (Kurepa and Smalle, 2019). This growth increase was again suggested to be caused by an upstream accumulation of t-CA, as exogenous application of t-CA to wild-type plants facilitated a similar growth-promotion (Kurepa and Smalle, 2019).

Schilmiller et al. (2009) reported local swellings in the branch junctions of the stems of the ref3-1 mutant and coined accumulating c-CA as being responsible. In agreement with these findings, a later study showed c-CA rather than t-CA to be the bioactive isoform of CA (Steenackers et al., 2017). Supplementing plants with c-CA inhibits auxin transport in the root tip of the plant, causing a local build-up in auxin concentrations, which results in strong proliferation of lateral rooting. In addition, a follow-up study showed c-CA and not t-CA to facilitate growth-promotion upon exogenous application (Steenackers et al., 2019). By preventing isomerization using light conditions devoid of UV or using conformationally constrained phenylcyclopropanoid analogs of both CA isomers, c-CA was shown to be the growth-promoting isomer. Finally, increased levels of c-CA were recently indicated to block auxin transport in the hypocotyl of Arabidopsis seedlings upon inhibition of C4H. This lead to a local proliferation of adventitious roots in the upper part of the hypocotyl (El Houari et al., 2021). These observations on the bioactive properties of c-CA are in line with the observed increased lateral rooting and branching of the ref3-1 mutant and the increased rosette size of the ref3-3 mutant. This thus suggests that the accumulation of c-CA and not t-CA lies at the basis of these phenotypes.

The flavonoid biosynthetic pathway branches from p-coumaroyl-CoA and further steps result in the production of a wide range of metabolites, including flavones, flavonols, anthocyanins, chalcones, and flavan-3-ols (Lepiniec et al., 2006). These all share a common backbone consisting of two phenyl rings and one heterocyclic ring (Figure 1) and are involved in a range of processes in the plant, including plant defense (Treutter, 2005), oxidative stress responses (Nakabayashi et al., 2014), nodulation (Kobayashi et al., 2004), and pigmentation (De Jong et al., 2004; Eichhorn and Winterhalter, 2005). Flavonoids have also been coined to steer plant development via two mechanisms. Firstly, flavonoids have antioxidative properties (Agati et al., 2012) and impaired biosynthesis of flavonoids results in higher reactive oxygen species (ROS) levels in the plant (Watkins et al., 2014; Gayomba and Muday, 2020). For example, the flavonoid kaempferol was described as a negative regulator of lateral root growth, most likely by regulating ROS levels in the lateral root primordia (Chapman and Muday, 2021). Secondly, the flavonoids naringenin, quercetin, and kaempferol have also been described to be auxin transport inhibitors (Jacobs and Rubery, 1988; Brunn et al., 1992; Faulkner and Rubery, 1992) and endogenous over- or underproduction of flavonoids was shown to influence auxin transport using mutants in the flavonoid biosynthetic pathway (Murphy et al., 2000; Brown et al., 2001; Peer et al., 2004; Buer et al., 2013). An hct mutant with severe growth reduction and reduced leaf size showed an apparent increase in flavonoid levels, as the leaves of this mutant showed a purple coloration (Hoffmann et al., 2004). Such an accumulation of flavonoids, anthocyanins, and other flavonoid derivatives upon blocking HCT was later confirmed in HCT-RNAi plants (HCT^–) (Besseau et al., 2007) and reducing flavonoid levels by growing the HCT^– plants under low-light conditions was correlated to a restoration in growth. Additionally, reducing flavonoid content in HCT^– plants by crossing them with flavonoid-deficient CHS-RNAi plants (CHS^–; Figure 1) also correlated with a growth restoration. Both the increased flavonoid content and growth inhibition coincided with an inhibition in auxin transport in these plants. A later study, however, indicated that silencing of HCT in a chs knockout (tt4) background deficient for flavonoids did inhibit plant growth (Li et al., 2010). In fact, the degree of growth inhibition upon silencing HCT in the tt4 background was similar to that upon silencing HCT in WT plants. This indicated that the observed growth inhibition of the HCT^– plants was most likely not due to the accumulation of flavonoids. Notably, the growth restoration in the double silenced HCT^–/CHS^– plants did go paired with a slight restoration in lignification. This is most likely due to promoter silencing of the HCT-RNAi construct, as both HCT and CHS RNAi constructs were driven by a 35S promoter (Li et al., 2010). The partial growth restoration of the HCT^–/CHS^– plants could thus find its origin in the slight restoration in lignification of these plants rather than the reduction in flavonoid content. Moreover, repetition of flavonoid quantification indicated that the total amount of flavonoids per rosette is the same for WT and HCT^– plants. Similarly, a CRISPR-generated hct mutant (hct^D7; Kriegshauser et al., 2021) showed an 80-fold increase in levels of p-coumaroyl-glucose but not in kaempferol 3-O-rhamnoside 7-O-rhamnoside, indicating that also here flavonoid levels are not significantly altered when blocking HCT. In summary, flavonoid accumulation does not seem to be responsible for the dwarfism of HCT^– plants. Therefore, whereas altered flavonoid content does modulate auxin transport in flavonoid biosynthesis mutants and possibly also in the HCT^– plants, there is no conclusive evidence that their differential accumulation in lignin mutants causes dwarfism.

Ferulic acid is an intermediate of the PPP, being produced from caffeic acid by COMT (Figure 1). Like other phenylpropanoids it has known antioxidant properties and ROS scavenging potential (Graf, 1992; Kanski et al., 2002). Whereas many studies have investigated the bioactive properties of ferulic acid (Lippincott and Lippincott, 1971; Li et al., 1993; Locher et al., 1994), convincing evidence for a bioactive role in plants is scarce. In ccr1 mutants, ferulate conjugate levels accumulate and higher levels of ferulic acid are incorporated in the lignin (Derikvand et al., 2008; Vanholme et al., 2012; De Meester et al., 2018). The accumulation of ferulic acid was found to be correlated to growth phenotypes in ccr1-4 plants (Xue et al., 2015), and its antioxidant potential was proposed to bring about a reduced leaf size, as the ccr1-4 cells remained longer in a mitotic state. This delay in cell proliferation exit resulted in plants with a higher cell count but smaller cell size, causing a reduction in overall leaf size. ROS are known to be required for the shift of cells toward cell proliferation exit (Boonstra and Post, 2004; Tsukagoshi et al., 2010). The accumulation of ferulic acid was therefore proposed to bring about the scavenging of ROS through its antioxidant action, hereby being at the basis of the delay in cell proliferation exit. A later study, however, found dwarfism in a ccr1-6 mutant to be fully restored upon restoration of lignin specifically in the xylem vessels, despite increased levels of ferulic acid coupling products in the leaves (De Meester et al., 2018). In addition, the reduced cell proliferation exit in the vessel-complemented ccr1-6 plants was mitigated, indicating (1) that ferulic acid is not at the basis of the dwarfism, and (2) that ferulic acid is not at the basis of the observed delay in cell proliferation exit.

Dehydrodiconiferyl alcohol glucosides or DCGs form a specific class of glucosylated phenylpropanoid coupling products with proposed hormone-like activity. The DCG aglycon is a coniferyl alcohol dimer, coupled via a coumaran linkage (Figure 1). Coupling of the two coniferyl alcohol radicals leads to two chiral centers in the molecule, resulting in different stereoisomers. Interestingly, the bioactivity of DCGs is restricted to particular diastereoisomers that were initially isolated from tumor cells of Vinca rosea (Lynn et al., 1987) and several follow up studies described the accumulation of DCGs in rapidly dividing tissues or cell cultures (Binns et al., 1987; Attoumbré et al., 2006; Hano et al., 2006). In tobacco tissue culture, DCGs were also able to replace cytokinin in cell division assays (Binns et al., 1987) and cytokinin treatment effectively stimulated DCG accumulation (Teutonico et al., 1991), suggesting that DCG biosynthesis in the plant is controlled by cytokinin. Based on these observations, these molecules have been described as having cell division-promoting activities (Binns et al., 1987; Lynn et al., 1987; Teutonico et al., 1991). As they are formed from coniferyl alcohol, positioned at the final step of the PPP (Figure 1), DCG concentrations are often lowered in PPP mutants (Vanholme et al., 2012; Dima et al., 2015). Together with their possible role in the stimulation of plant cell division, this depletion is frequently used to explain the dwarfism of Arabidopsis PPP mutants (Franke et al., 2002; Abdulrazzak et al., 2006; Do et al., 2007; Li et al., 2010). However, a causal role for DCG depletion in PPP mutant dwarfism has not yet been shown. Moreover, feeding the probable DCG precursor coniferyl alcohol to Nicotiana benthamiana seedlings did severely impair instead of stimulate growth, although this could also likely be the result of an induced lignification (Väisänen et al., 2015). In addition, to our knowledge no further physiological support for the cell division promoting activity by DCGs has been provided since the initial reports over 30 years ago (Binns et al., 1987; Lynn et al., 1987; Teutonico et al., 1991). Together, the involvement of DCGs in PPP mutant dwarfism thus remains purely speculative.