An evolutionarily diverged CCD4 enzyme negatively regulates mesocotyl elongation in rice

Abstract

Figures

Enzymatic activity of OsCCD4b in vivo and in vitro. (a) Phylogenetic analysis of CCD proteins from different species. Accession numbers were obtained from previous studies (Wang et al., 2019; Zheng et al., 2022). LOC_Os12g24800 encoded protein, referred to as OsCCD4b, clusters within the CCD4 subfamily and is closely related to monocot CCD4 enzymes, which form a distinct subgroup within the CCD4 family. The CCD4 proteins are separated into four groups, consistent with their cleavage site specificity. (b) UHPLC analysis of OsCCD4b in vivo enzymatic activity using β‐carotene and zeaxanthin as substrates. OsCCD4b catalyzed the formation of β‐apo‐8′‐carotenal and 3‐OH‐β‐apo‐8′‐carotenal from β‐carotene and zeaxanthin, respectively. (c) Relative quantification of apocarotenoid products from in vivo assays. The cleavage products were confirmed by UHPLC‐HR‐MS. (d) Relative quantification of apocarotenoids in Nicotiana leaves transiently expressing OsCCD4b or AtCCD4. (e) Relative quantification of apocarotenoids in stable transgenic callus lines overexpressing OsCCD4b (OEOs4b) or AtCCD4 (OEAtC4). (f) UHPLC analysis of in vitro cleavage activity of OsCCD4b, its mutant variant OsCCD4b‐M, and AtNCED2 using 9′‐cis‐neoxanthin as a substrate. Only AtNCED2 cleaved the C11′–C12′ double bond to produce xanthoxin and a C25 aldehyde. (g) Quantification of xanthoxin in in vitro enzymatic reaction products (h). Relative quantification of abscisic acid (ABA) in stable transgenic callus lines overexpressing OsCCD4b (OEOs4b) or AtCCD4 (OEAtC4). Identification and quantification of apocarotenoids and ABA were performed using UHPLC‐HR‐MS. Bars represent SEM. Asterisks indicate statistically significant differences by Student's t‐test: *, P < 0.05; , P < 0.01; *, P < 0.001. ns indicates no significant difference. EV, empty vector; DW, dry weight; mAU, milli‐absorbance units.

OsCCD4b regulates mesocotyl length and cell elongation through its cleavage product 3‐OH‐β‐cyclocitral. (a) Mesocotyls of seedlings grown in darkness for 9 d. The upper and lower arrows indicate the positions of the coleoptile node and the basal part of the seminal root, respectively. The mesocotyl is the tissue between these two arrows. (b) Mesocotyl length of OsCCD4b overexpression lines and CRISPR knockout mutants in the NB (Nipponbare) background. (c) Mesocotyl length of OsCCD4b CRISPR mutants in the ZH11 (Zhonghua 11) background. (d) Effects of different concentrations of HTTC (3‐OH‐β‐cyclocitral) on mesocotyl length. KAR1, a known negative regulator of mesocotyl elongation, was used for comparison. (e) Effects of 20 μM HTTC treatment on mesocotyl elongation in 8‐d‐old dark‐grown Nipponbare WT and Osccd4b CRISPR mutant seedlings. (f) Effects of 20 μM HTTC treatment on mesocotyl elongation in 8‐d‐old dark‐grown d14 and d3 mutants in the SH (Shiokari) background. (g) Representative tissue sections of mesocotyls from the wild‐type (WT), OsCCD4b overexpressing line, Osccd4b CRISPR knockout mutant, and mutant treated with HTTC. Red bars represent 50 μm. (h) Cell length measurements were performed using ImageJ. (i) Quantification of 3‐OH‐β‐apo‐8′‐carotenal and HTTC levels in OsCCD4b overexpression and CRISPR mutant lines. (j) Gene Ontology (GO) term analysis of upregulated genes in mesocotyl of Osccd4b knockout mutants from RNA‐seq. (k) GO term analysis of downregulated genes in mesocotyl of OsCCD4b overexpression lines from RNA‐seq. GO terms with corrected P‐value ≤ 0.05 were considered significantly enriched by up‐ or downregulated genes. Bars represent SEM. Asterisks indicate statistically significant differences by Student's t‐test: *, P < 0.05; , P < 0.01; *, P < 0.001. ns indicates no significant difference. DW, dry weight.

Full text

Disclaimer

The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.

Introduction

Carotenoids constitute a large class of natural isoprenoid compounds synthesized in all photosynthetic organisms and by many nonphotosynthetic microorganisms (Moise et al., 2014; Rodriguez‐Concepcion et al., 2018; Li et al., 2025). In plants, carotenoids contribute to the pigmentation of fruits and flowers, thereby promoting pollination and seed dispersal by attracting animals and insects (Yuan et al., 2015). More critically, they play vital roles by safeguarding the photosynthetic apparatus from photooxidative damage and participate in light harvesting, thereby enabling plant growth and survival (Hashimoto et al., 2016).

Due to their conjugated double‐bond structures, carotenoids are highly susceptible to oxidative cleavage, producing a wide array of biologically active metabolites known as apocarotenoids. These include the precursors of the phytohormones such as strigolactones (SLs) and abscisic acid (ABA), and growth regulators such as anchorene, β‐cyclocitral, zaxinone, as well as diverse pigments and volatile compounds (Schwartz et al., 1997; Wang et al., 2019; Moreno et al., 2021; Li et al., 2025). The biosynthesis of apocarotenoids is catalyzed by the carotenoid cleavage dioxygenases (CCD) gene family, which comprises several subfamilies: 9‐cis‐epoxycarotenoid dioxygenases (NCEDs), CCD1, CCD4, CCD7, CCD8, CCD2 (identified in Crocus species), and Zaxinone Synthase (ZAS) (Walter & Strack, 2011; Frusciante et al., 2014; Ahrazem et al., 2016; Jia et al., 2018; Wang et al., 2019; Ablazov et al., 2023a). Each CCD subfamily exhibits distinct substrate specificities and cleavage site preferences, resulting in the formation of diverse apocarotenoids with particular biological functions (Zheng et al., 2020). NCED enzymes catalyze the cleavage of the C9‐C10 double bond in 9′‐cis‐neoxanthin and/or 9‐cis‐violaxanthin to produce the ABA precursor xanthoxin (Schwartz et al., 1997). CCD7 and CCD8 sequentially convert 9‐cis‐β‐carotene into carlactone, a central intermediate in strigolactone biosynthesis (Alder et al., 2012; Al‐Babili & Bouwmeester, 2015; J. Y. Wang et al., 2024). ZAS converts hydroxy‐apocarotenoids, primarily 3‐OH‐β‐apo‐10′‐carotenal, into zaxinone, a growth regulator identified in rice and other plants (Wang et al., 2019; Ablazov et al., 2020). CCD1 enzymes catalyze the cleavage of multiple carotenoids and apocarotenoid substrates at different double bonds, yielding a wide range of volatile compounds and dialdehydes (Simkin et al., 2004; IIg et al., 2014). CCD2L cleaves zeaxanthin to yield crocetin dialdehyde, the precursor of crocin, the pigment responsible for saffron coloration (Frusciante et al., 2014; Ahrazem et al., 2016).

CCD4 enzymes exhibit remarkable diversity in substrate specificity and regio‐selectivity across plant species, reflecting their broad functional divergence (Zheng et al., 2020). For instance, CCD4 enzymes from Arabidopsis thaliana and Solanum tuberosum cleave bicyclic carotenoids at the C9′‐C10′ and/or C9–C10 double bonds, producing C_13_ volatiles and C_27_ apocarotenoids that are supposed to be further degraded to colorless compounds (Gonzalez‐Jorge et al., 2013; Bruno et al., 2016; Mi & Al‐Babili, 2019). By contrast, Citrus CCD4b cleaves carotenoids, particularly zeaxanthin, at the C7–C8 or C7′–C8′ positions, yielding the C_30_ apocarotenoid β‐citraurin (also known as 3‐OH‐β‐apo‐8′‐carotenal), which contributes to the characteristic red coloration of citrus fruit peel (Ma et al., 2013; Rodrigo et al., 2013; Zheng et al., 2019, 2021). Similarly, BdCCD4.1/3 and GjCCD4a from Buddleja davidii and Gardenia jasminoides, respectively, cleave carotenoids at both C7–C8 and C7′–C8′ sites, forming crocetin dialdehyde, the precursor of the saffron pigment crocin (Ahrazem et al., 2017; Xu et al., 2020; Zheng et al., 2022; Lobato‐Gómez et al., 2025). In Crocus, CCD4 enzymes cleave carotenoids at the C9–C10 or C9′–C10′ position to generate C_13_ volatiles (Rubio et al., 2008; Rubio‐Moraga et al., 2014). However, knowledge of CCD4 enzymes in monocot species remains limited. In particular, it is unclear whether CCD4 enzymes from cereals differ in substrate specificity and region‐selectivity from their dicot homologs.

The annotation of LOCOs12g24800, a rice gene encoding a putative carotenoid cleavage enzyme, has varied across studies. While some have classified it as CCD4b (Yang et al., 2017; Ko et al., 2018; Choi et al., 2025), others have referred to it as NCED2 (Oliver et al., 2007; Lyu et al., 2013; Mao et al., 2017; Chen et al., 2021; Gao et al., 2022; Jin et al., 2023a, 2023b; J. D. Wang et al., 2024), suggesting a role in ABA biosynthesis. This divergence in annotation likely arises from sequence similarity within conserved regions and the lack of functional validation at the enzymatic level. Given the distinct cleavage specificities and resulting physiological functions of CCD4 and NCED subfamilies (Schwartz et al., 1997; Zheng et al., 2020), resolving the biochemical identity of LOCOs12g24800 is essential for accurately interpreting its biological function. Several previous studies, guided by the NCED2 annotation, have examined its expression dynamics and promoter regulation in ABA‐related responses (Oliver et al., 2007; Lyu et al., 2013; Mao et al., 2017; Chen et al., 2021; Gao et al., 2022; Jin et al., 2023a, 2023b; Wang et al., 2024b). However, without biochemical validation, it remains unclear whether these observations accurately reflect the true enzymatic function of the encoded protein. To clarify its enzymatic identity and physiological role, we performed a comprehensive analysis integrating phylogenetic analysis, in vitro and in planta enzymatic assays, and phenotypes of corresponding mutants. Our findings demonstrate that LOC_Os12g24800 encodes a CCD4‐type enzyme with cleavage specificity distinct from NCEDs, clarifying that it does not participate directly in ABA biosynthesis and resolving previous ambiguities arising from its NCED2 annotation. Thus, this study provides a biochemical and functional framework that helps reconcile previous interpretations and guides future research on carotenoid cleavage and apocarotenoid signaling in rice.

Materials and Methods

Plant materials and phenotyping

The transgenic callus of Citrus paradise Macf. overexpressing tpCrtB and OsBCH (Zheng et al., 2022) was grown in the dark at room temperature and subcultured at 20‐d intervals. Nicotiana benthamiana plants were grown in a growth chamber at 24°C, under a 12 h day‐night photoperiod. Rice seedlings were grown in a Biochamber at a day/night temperature of 27/25°C, under a 12‐h day/night photoperiod. The d14‐1 and d3‐1 mutants (cv. Shiokari) were described previously (Ishikawa et al., 2005). For mesocotyle phenotyping, we have either used vermiculite or ½‐strength Murashige & Skoog medium (1/2MS) medium consisting of 0.5% (w/v) agar. The vermiculite experiment was performed as described by Patil et al. (2019) with some modifications. Initially, 250 ml of vermiculite was added to the glass vessels (6 cm diameter, 18 cm height). Then, c. 20 (dry) seeds were placed on this layer, and following this, 500 ml of vermiculite was added on top of the seeds. Last, 400 ml of Milli‐Q water (pH 5.8) was added to each bottle, and each was tightly closed and incubated in complete darkness under the above conditions for 10 d. Mesocotyl lengths were scanned and measured from the seminal root emergence point to the coleoptile node using ImageJ software. For agar mesocotyl phenotyping, unshelled rice seeds were sterilized with 1.5% sodium hypochlorite for 30 min and then germinated at 30°C for 1 d. Uniform seeds were selected and placed on solid 1/2MS medium consisting of 0.5% (w/v) agar medium with the indicated concentrations of chemicals (3‐OH‐β‐cyclocitral and acetone as mock) and then grown in the dark at 30°C for 7 d. The mesocotyl length measurement was conducted manually or via ImageJ. For seedling phenotyping, 1‐wk‐old seedlings were grown hydroponically in Hoagland nutrient solution (Wang et al., 2019) for 3 wk, and phenotypic data were recorded. The solution was changed every other day.

Mesocotyl Embedding, Sectioning, and Staining: Mesocotyl samples were first vacuum‐infiltrated in 4% (w/v) paraformaldehyde prepared in 1 × PBS buffer (pH 6.9) at room temperature for 1–2 h. After fixation, samples were embedded in 5% (w/v) low‐melting‐point agarose, and longitudinal sections (80‐μm‐thick) were obtained using a Leica VT1200S vibratome. To visualize cell walls, sections were stained overnight with 0.1% (v/v) SR2200, following the protocol described by Musielak et al. (2016). Confocal images were acquired using a Zeiss LSM 710 inverted microscope with a 405 nm excitation laser, and fluorescence signals were collected in the 430–500 nm emission range. Vertical sections within the coleoptile node to the base of the seminal root were analyzed. Cell lengths within the mesocotyl region were measured using ImageJ software.

In bacterial and vitro assays

The coding sequences of OsCCD4b (LOC_Os12g24800) and OsCCD4a (LOC_Os02g47510), excluding the chloroplast transit peptide region, were cloned into the pThio vector in frame with an N‐terminal thioredoxin tag. Primers used for plasmid construction are listed in Table S1. Constructs pThio‐AtNCED2 and pThio‐CitCCD4b were obtained from previous studies (Zheng et al., 2021; Jia et al., 2022). pThio‐OsCCD4b, pThio‐OsCCD4a, pThio‐CitCCD4b, and the void plasmid were transformed into E. coli competent cells engineered to accumulate either β‐carotene or zeaxanthin. In bacterial assays, UHPLC analysis of enzymatic products was conducted as described previously (Zheng et al., 2022).

For in vitro assay, the pThio‐OsCCD4b, pThio‐OsCCD4b, and pThio‐AtNCED2 were transformed into BL21 E. coli harboring the pGro7 plasmid. E. coli cell induction, crude lysates and substrates preparation, incubation conditions, metabolites extraction, and UHPLC analysis were conducted as previously described (Bruno et al., 2016; Zheng et al., 2022).

Carotenoids and apocarotenoids were analyzed using an UHPLC‐DAD system and a C_30_ YMC Carotenoid column (150 × 3 mm, 5 μm). The mobile phases consisted of (A) Methanol/MTBE in a 1 : 1 ratio and (B) Methanol/MTBE/Water in a 30 : 1 : 10 ratio. Chromatographic separation was carried out at a flow rate of 0.6 ml min^−1^. For in bacteria samples, the elution program began with 100% solvent B, linearly decreasing to 0% over 15 min, and held until 24 min. For in vitro enzymatic products, the gradient transitioned from 100% B to 45% B within 15 min, stepped to 0% B over the next 5 min, and maintained this condition until 24 min. Subsequently, the gradient returned to 100% B within 1 min and held until 33 min.

Plasmid construction and Transient expression in N. benthamiana leaves

The CDS of OsCCD4b (LOC_Os12g24800) was cloned into the pDONR221 vector to generate pDONR221‐OsCCD4b, respectively, by using ‘BP’ reaction kit (Invitrogen). pDONR221‐OsCCD4b was then recombined into the pB2GW7 overexpression vector using an ‘LR’ reaction kit (Invitrogen) to generate pB2GW7‐OsCCD4b. The primers used for plasmid construction are listed in Table S1.

The constructs pB2GW7‐OsCCD4b and pB2GW7‐AtCCD4 (At4g19170) from our previous study (Zheng et al., 2021), along with the empty vector (EV), were individually electroporated into Agrobacterium tumefaciens strain GV3101. The Agrobacterium‐mediated transient expression was performed by using leaves of 5‐wk‐old N. benthamiana plants as previously described (Zheng et al., 2022). The pB2‐OsCCD4b and EV agrobacterium cell pellets were resuspended in infiltration buffer to an OD_600_ of 0.5; the OD_600_ of the p19 agrobacterium cell culture was 0.3. The pB2‐OsCCD4b‐transformed agrobacteria were mixed with p19 agrobacterial culture in a 1 : 1 ratio. The agro‐infiltrated leaves were harvested after 5 d, then frozen in liquid nitrogen for the following apocarotenoid profiling.

Supertransformation of engineered callus

The abovementioned pB2GW7‐OsCCD4b and pB2GW7‐AtCCD4 were electroporated into A. tumefaciens EHA105, respectively. The 20‐d‐old transgenic callus was incubated with suspension (OD = 0.3) EHA105 harboring the above‐constructed pB2GW7‐OsCCD4b or pB2GW7‐AtCCD4 in liquid MS B5 medium for 10 min. After 3 d of co‐cultivation, the transformed callus was selected in MS B5 solid medium with 50 mg l^−1^ glufosinate‐ammonium (Sigma‐Aldrich) for 5 wks. The solid and liquid MS B5 medium used in this study was according to Zheng et al. (2023). The small piece of each recovered transformed callus was then transferred onto selective solid medium and sub‐cultured for at least six cycles. Twenty‐day‐old transgenic lines grown in solid medium without antibiotics were harvested for later apocarotenoid profiling and molecular analysis.

Generation of

OsCCD4b overexpression and CRISPR mutant lines in rice

The aforementioned pDONR221‐OsCCD4b was recombined into the pH7WG2D overexpression vector using an ‘LR’ reaction kit (Invitrogen) to generate pH 7‐OsCCD4b. The plasmid was electroporated into Agrobacterium tumefaciens (EHA105) and was transformed into Nipponbare wild‐type Japonica rice cultivar as previously described (Hiei & Komari, 2008). The transgenic rice seedlings were selected by 50 mg l^−1^ hygromycin in a 1/2 MS medium. Twelve seedlings from each T1 transgenic line were grown to identify homozygote lines at the T2 generation. Transgenic lines' seeds that showed a 90–100% germination rate on a 1/2 MS medium containing 50 mg l^−1^ hygromycin were considered homozygote lines and then were propagated to obtain T3 generation seeds. The OsCCD4b mutant in the Nipponbare background using the clustered regularly interspaced short palindromic repeat (CRISPR)‐Cas9 technology was previously generated (Yang et al., 2017). The OsCCD4b mutant in Zhonghua 11 (O. sativa L. japonica, ZH11) background was generated by Biogle (Lu et al., 2017) using the CRISPR–Cas9‐mediated editing method. The transgenicity and mutagenicity of the CRISPR mutant lines were confirmed as described by Ablazov et al. (2023b).

Quantitative Real‐time PCR

Total RNA was isolated using the Trizol method as previously described (Zheng et al., 2023). First‐strand cDNA synthesis was performed with the iScript™ cDNA synthesis Kit (Bio‐Rad) following the manufacturer's protocol. Quantitative real‐time polymerase chain reaction was conducted on a CFX384 Touch RT‐PCR System (Bio‐RAD) using SsoAdvanced™ Universal SYBR® Green Supermix (Bio‐Rad) Kit. Relative gene expression levels were calculated using the E^−ΔΔCt^ method. The qRT‐P2CR primers were provided in Table S1.

UHPLC‐HR‐MS analysis of apocarotenoids

Apocarotenoids were extracted from freeze‐dried tissues according to a previous protocol (Mi et al., 2018). Dried extracts were re‐dissolved in 90 : 10 (v/v) acetonitrile/water, followed by filtration through a 0.22 μm filter, and subjected to LC‐MS analysis. The separation and detection of all apocarotenoids except 3‐OH‐β‐cyclocitral were conducted using UHPLC coupled with a Q‐Orbitrap high‐resolution mass spectrometer (HR‐MS), as previously described (Mi et al., 2018; Zheng et al., 2022). Detection of 3‐OH‐β‐cyclocitral followed a separate UHPLC‐HR‐MS method optimized for glycosylated apocarotenoids (Mi et al., 2019). Apocarotenoids were identified and quantified by comparing retention times and MS spectra with authentic and isotope‐labeled internal standards (Buchem B.V., Apeldoorn, the Netherlands) as previously described (Mi et al., 2018).

ABA quantification using UHPLC–MS

For ABA analysis, c. 5 mg freeze‐dried powder was extracted twice with 600 μl of 10% methanol containing 1% acetic acid and internal standard (D6‐ABA, 1 ng each sample) in an ultrasound bath for 5 min, followed by incubation on ice for 30 min. After centrifugation, the two extracts were combined and filtered with a 0.22 μm filter before UHPLC–MS analysis. Identification and quantitative analysis of ABA were performed as previously described (Jia et al., 2022).

UHPLC analysis of carotenoids and chlorophylls

Approximately 10–20 mg of freeze‐dried tissue powder was extracted on ice for 20 min using a mixture of chloroform and methanol (2 : 1, v/v), as described by Fraser et al. (2000). After adding 1 volume of water, the mixture was centrifuged to allow phase separation. The lower organic phase was collected, vacuum‐dried, and used for UHPLC analysis of carotenoids and chlorophylls. UHPLC separation and detection were performed as previously described in Zheng et al. (2023), using mobile phase A (Methanol/Methyl tert‐butyl ether, 50 : 50, v/v) and mobile phase B (Methanol/water/methyl tert‐butyl ether, 6 : 3 : 1, v/v/v) at a flow rate of 0.8 ml/min and a column temperature of 35°C. The gradient started at 30% A/70% B, linearly increased to 100% A over 19 min, held until 34 min, then returned to initial conditions by 36 min, and maintained until 38 min. Compounds were identified by comparing retention times and absorption spectra with authentic standards (CaroteNature, Switzerland) and published data (Fraser et al., 2007). Quantification was performed using standard calibration curves.

Phylogenetic tree and conserved motif analysis

The phylogenetic tree was constructed using the Neighbor‐Joining method in MEGA5 software (Tamura et al., 2011), with 1000 bootstrap replicates. Protein accession numbers used for phylogenetic analysis are listed in the legend of Fig. S1. The conserved motifs and specific residues within the CCD sequences of both rice and Arabidopsis were examined through the MEME web server (http://meme‐suite.org/tools/meme) (Bailey et al., 2009). The server settings were configured with a minimum conserved motif width of 15 and a maximum of 40, allowing for up to 10 motifs per theme. All other parameters were set to their default values.

RNA‐seq experiment and bioinformatic analysis

Total RNA was extracted from rice mesocotyls of the OsCCD4b CRISPR mutant and the overexpression lines using TRI‐Reagent with Direct‐zol RNA MiniPrep Kit according to the manufacturer's instructions (Zymo Research, Irvine, CA, USA). Seedlings that were 9 d old, with 4–5 seedlings grouped together for each sample, were collected in a dark room illuminated by green lights. The quality and quantity of RNA were assessed using the NanoDrop 6000. Library preparation, sequencing, and subsequent data analysis were conducted by Novogene Technology. Sequencing took place on the Illumina Novaseq platform, yielding 150 bp paired‐end reads. The clean paired‐end reads were aligned to the Nipponbare reference genome (http://rice.uga.edu/pub/data/Eukaryotic_Projects/o_sativa/annotation_dbs/pseudomolecules/version_7.0/) utilizing Hisat2 v.2.0.5. To identify differentially expressed genes, the DESeq. 2R package (v.1.20.0) was employed under the criteria of an adjusted P‐value ≤ 0.05 and a fold change (FC) ≥ 0.58. PCA was conducted to illustrate consistency among biological replicates. The clusterProfiler R package was used for GO enrichment analysis of the DEGs through the NovoMagic online platform (https://ap‐magic.novogene.com/), considering GO terms with a corrected P‐value ≤0.05 as significantly enriched by the differentially expressed genes.

Statistical analysis

All the data are presented as mean ± SEM from at least three biological replicates unless otherwise noted. Statistical significance was assessed using GraphPad Prism v8 and Microsoft Excel 2010. Significant differences are denoted as *P‐value <0.05, **P < 0.01, and ***P < 0.001.

Results and Discussion

We first constructed a phylogenetic tree of CCD proteins from diverse plant species, revealing that LOC_Os12g24800 encoded protein clusters within the CCD4 subfamily clade (Fig. 1a). In parallel, motif analysis of rice and Arabidopsis CCD proteins revealed notable differences in motif composition among subfamilies: NCEDs, CCD4s, CCD1s, ZASs, CCD8s, and CCD7s contained 15, 13, 11, 11, 7, and 7 conserved motifs, respectively (Supporting Information Fig. S1). We found that motifs 14 and 15, characteristic of NCED family members, such as AtNCED2/3/5/9 and OsNCED3/4/5, are absent in LOC_Os12g24800 and CCD4 proteins (Fig. S1). These findings support the classification of LOC_Os12g24800, which we call in the following CCD4b, as a CCD4‐type enzyme rather than a member of the NCED subfamily. Furthermore, its phylogenetic separation from Arabidopsis CCD4 indicated that it may possess distinct enzymatic activity.

*Enzymatic activity of OsCCD4b in vivo and in vitro. (a) Phylogenetic analysis of CCD proteins from different species. Accession numbers were obtained from previous studies (Wang et al., 2019; Zheng et al., 2022). LOC_Os12g24800 encoded protein, referred to as OsCCD4b, clusters within the CCD4 subfamily and is closely related to monocot CCD4 enzymes, which form a distinct subgroup within the CCD4 family. The CCD4 proteins are separated into four groups, consistent with their cleavage site specificity. (b) UHPLC analysis of OsCCD4b in vivo enzymatic activity using β‐carotene and zeaxanthin as substrates. OsCCD4b catalyzed the formation of β‐apo‐8′‐carotenal and 3‐OH‐β‐apo‐8′‐carotenal from β‐carotene and zeaxanthin, respectively. (c) Relative quantification of apocarotenoid products from in vivo assays. The cleavage products were confirmed by UHPLC‐HR‐MS. (d) Relative quantification of apocarotenoids in Nicotiana leaves transiently expressing OsCCD4b or AtCCD4. (e) Relative quantification of apocarotenoids in stable transgenic callus lines overexpressing OsCCD4b (OEOs4b) or AtCCD4 (OEAtC4). (f) UHPLC analysis of in vitro cleavage activity of OsCCD4b, its mutant variant OsCCD4b‐M, and AtNCED2 using 9′‐cis‐neoxanthin as a substrate. Only AtNCED2 cleaved the C11′–C12′ double bond to produce xanthoxin and a C25 aldehyde. (g) Quantification of xanthoxin in in vitro enzymatic reaction products (h). Relative quantification of abscisic acid (ABA) in stable transgenic callus lines overexpressing OsCCD4b (OEOs4b) or AtCCD4 (OEAtC4). Identification and quantification of apocarotenoids and ABA were performed using UHPLC‐HR‐MS. Bars represent SEM. Asterisks indicate statistically significant differences by Student's t‐test: *, P < 0.05; **, P < 0.01; **, P < 0.001. ns indicates no significant difference. EV, empty vector; DW, dry weight; mAU, milli‐absorbance units.

To investigate the enzymatic activity of OsCCD4b, we expressed the gene in Escherichia coli strains engineered to accumulate either β‐carotene or zeaxanthin. UHPLC (Ultra‐High‐Performance Liquid Chromatography) analysis of cell extracts revealed the production of β‐apo‐8′‐carotenal and 3‐OH‐β‐apo‐8′‐carotenal, respectively, but not β‐apo‐10′‐carotenal or 3‐OH‐β‐apo‐10′‐carotenal (Fig. 1b,c). LC‐MS‐based apocarotenoid profiling confirmed the UHPLC results (Fig. S2), indicating that OsCCD4b preferentially cleaves the C7′–C8′ double bond to generate C_30_ and C_10_ apocarotenoids. This cleavage specificity differs from that of Arabidopsis CCD4, which targets the C9–C10 or C9′–C10 double bond, but resembles the activity of the CitCCD4b, consistent with their phylogenetic clustering (Fig. 1a). We also investigated the second rice CCD4 enzyme encoded by LOC_Os02g47510, referred to here as OsCCD4a, using the same in vivo assay system. However, we did not detect cleavage of β‐carotene under our experimental conditions (Fig. S2).

Given the relatively low enzymatic activity of OsCCD4b in the bacterial system, we further evaluated its function in planta. Transient expression of OsCCD4b in Nicotiana benthamiana leaves significantly increased the levels of β‐apo‐8′‐carotenal and 3‐OH‐β‐apo‐8′‐carotenal, as well as their corresponding C_10_ co‐products, β‐cyclocitral and 3‐OH‐β‐cyclocitral (Fig. 1d), but not of C9′–C10′ cleavage products (Fig. S3), further supporting the cleavage specificity inferred from bacterial assays.

To corroborate these findings, we generated stably transformed citrus callus lines overexpressing OsCCD4b in a background preengineered to accumulate carotenoids via the introduction of bacterial PSY (phytoene synthase) and BCH (β‐carotene hydroxylase) genes (Fig. S4). UHPLC‐Q‐Orbitrap‐MS analysis revealed a significant increase in β‐apo‐8'‐carotenal and 3‐OH‐β‐apo‐8'‐carotenal, along with their respective co‐products, β‐cyclocitral and 3‐OH‐β‐cyclocitral (Fig. 1e). By contrast, callus overexpressing AtCCD4 did not show an increase in C_30_ apocarotenoid levels. Likewise, no increase in C9′–C10′ cleavage products was detected in OsCCD4b‐overexpressing citrus callus (Fig. S5), consistent with the results obtained from both Nicotiana leaves and in bacterio assays. We also observed that overexpression of OsCCD4b resulted in a significant reduction in total carotenoid content, including β‐carotene and zeaxanthin (Fig. S6). In light of the previous annotation of LOCOs12g24800 encoded protein as NCED2, we conducted in vitro enzymatic assays with 9'‐cis‐neoxanthin as a substrate and the Arabidopsis NCED2 (AtNCED2) as a positive control. As expected, AtNCED2 efficiently converted 9'‐cis‐neoxanthin into the ABA precursor xanthoxin (Figs 1f,g, S7). By contrast, we did not detect xanthoxin in the OsCCD4b incubation, which supports its distinct enzymatic activity. Confirming the results of the in vitro assay, ABA levels remained unchanged in OsCCD4b‐overexpression citrus callus lines (Fig. 1h). We also tested OsCCD4bM, a mutant variant previously suggested to enhance ABA content in planta (Huang et al., 2024); however, we did not observe any formation of xanthoxin (Fig. 1f,g). Confirming the stereo‐specificity and excluding the direct contribution to ABA biosynthesis, OsCCD4b did not convert 9‐cis‐β‐carotene (Fig. S8). Overall, our enzymatic and transgenic analyses demonstrate that OsCCD4b preferentially cleaves at the C7'–C8' double bond to produce C_30 and C_10_ apocarotenoids, highlighting an evolutionary divergence in cleavage‐site specificity among CCD4 enzymes. While CCD4s in dicots such as A. thaliana and S. tuberosum predominantly catalyze the cleavage of the C9–C10 or C9′–C10 double bonds to generate C_13_ volatiles, monocot CCD4s show distinct regio‐selectivity. Notably, while its cleavage activity closely resembles that of Citrus CCD4, the two genes differ in expression patterns, suggesting functional divergence. While Citrus CCD4 is predominantly expressed in the fruit peel (Rodrigo et al., 2013; Zheng et al., 2019), OsCCD4b shows only weak expression in rice vegetative tissues. Moreover, the catalytic activity of OsCCD4b is markedly lower than that of its citrus counterpart (Fig. 1c), suggesting a limited role in bulk carotenoid degradation in rice. Given that certain apocarotenoids function as signaling molecules at low concentrations, OsCCD4b may instead contribute to the regulation of plant growth and development through the production of specific bioactive cleavage products.

To further investigate its potential biological function, we examined the OsCCD4 (LOC_Os12g24800) locus and observed that it colocalizes with quantitative trait loci and genome‐wide association study signals associated with mesocotyl length (Liu et al., 2019; Zhan et al., 2020). However, its direct functional relevance to this trait has not been experimentally validated. Therefore, we first screened previously reported Osccd4b CRISPR lines (Yang et al., 2017) and identified a Cas9‐free mutant (Figs S9, S10), which exhibited significantly elongated mesocotyls (Fig. 2a,b). To confirm this phenotype, we generated an independent OsCCD4b knockout line in the ZH11 background (Fig. S11), which similarly displayed longer mesocotyls (Fig. 2c). Confirming the role of the enzyme as a negative regulator of the mesocotyl length, overexpression of OsCCD4b in the Nipponbare background (Fig. S12) led to shorter mesocotyls (Fig. 2a,b). Apart from the mesocotyl phenotype, we did not detect significant differences in other traits, including root growth on solid media (Fig. S13) as well as root length, shoot length, shoot biomass, or root biomass in hydroponically grown seedlings (Fig. S14).

*OsCCD4b regulates mesocotyl length and cell elongation through its cleavage product 3‐OH‐β‐cyclocitral. (a) Mesocotyls of seedlings grown in darkness for 9 d. The upper and lower arrows indicate the positions of the coleoptile node and the basal part of the seminal root, respectively. The mesocotyl is the tissue between these two arrows. (b) Mesocotyl length of OsCCD4b overexpression lines and CRISPR knockout mutants in the NB (Nipponbare) background. (c) Mesocotyl length of OsCCD4b CRISPR mutants in the ZH11 (Zhonghua 11) background. (d) Effects of different concentrations of HTTC (3‐OH‐β‐cyclocitral) on mesocotyl length. KAR1, a known negative regulator of mesocotyl elongation, was used for comparison. (e) Effects of 20 μM HTTC treatment on mesocotyl elongation in 8‐d‐old dark‐grown Nipponbare WT and Osccd4b CRISPR mutant seedlings. (f) Effects of 20 μM HTTC treatment on mesocotyl elongation in 8‐d‐old dark‐grown d14 and d3 mutants in the SH (Shiokari) background. (g) Representative tissue sections of mesocotyls from the wild‐type (WT), OsCCD4b overexpressing line, Osccd4b CRISPR knockout mutant, and mutant treated with HTTC. Red bars represent 50 μm. (h) Cell length measurements were performed using ImageJ. (i) Quantification of 3‐OH‐β‐apo‐8′‐carotenal and HTTC levels in OsCCD4b overexpression and CRISPR mutant lines. (j) Gene Ontology (GO) term analysis of upregulated genes in mesocotyl of Osccd4b knockout mutants from RNA‐seq. (k) GO term analysis of downregulated genes in mesocotyl of OsCCD4b overexpression lines from RNA‐seq. GO terms with corrected P‐value ≤ 0.05 were considered significantly enriched by up‐ or downregulated genes. Bars represent SEM. Asterisks indicate statistically significant differences by Student's t‐test: *, P < 0.05; **, P < 0.01; **, P < 0.001. ns indicates no significant difference. DW, dry weight.

Since CCD enzymes exert their function through the production of apocarotenoids, we hypothesized that a downstream metabolite of OsCCD4b may mediate its effect on mesocotyl elongation. We focused on 3‐OH‐β‐cyclocitral, one of the main products of OsCCD4b cleavage, as β‐cyclocitral is highly volatile and difficult to apply consistently. Exogenous application of 3‐OH‐β‐cyclocitral significantly suppressed mesocotyl elongation in a dose‐dependent manner, with concentrations above 20 μM causing significant inhibitory effects (Fig. 2d). However, its activity was weaker than that of the karrikin KAR1 (3‐methyl‐2H‐furo[2,3‐c]pyran‐2‐one), a known growth‐regulating compound that negatively affects mesocotyl length. Notably, 3‐OH‐β‐cyclocitral treatment partially rescued the elongated mesocotyl phenotype of the OsCCD4b CRISPR knock‐out mutant (Fig. 2e), suggesting that this apocarotenoid may function downstream of OsCCD4b. Given that carotenoids are also precursors for SLs, we tested whether 3‐OH‐β‐cyclocitral acts through known SL or karrikin (KAR) signaling pathways. However, SL‐ and KAR‐insensitive mutant responded similarly to wild‐type plants upon 3‐OH‐β‐cyclocitral treatment (Fig. 2f), indicating that its mode of action is likely independent of these hormone signaling pathways.

To understand the underlying cellular mechanism, we measured the cell length of the mesocotyl in the dark‐grown seedlings. The OsCCD4b CRISPR knockout line exhibited increased cell length, whereas overexpression lines showed shorter cells. Moreover, exogenous application of 3‐OH‐β‐cyclocitral partially rescued the elongated cell phenotype in the mutant line (Fig. 2g,h). Consistent with these observations, apocarotenoid profiling revealed increased levels of 3‐OH‐β‐cyclocitral and its co‐product of 3‐OH‐β‐apo‐8′‐carotenal in both shoots and roots of OsCCD4b‐overexpressing plants, and reduced levels in the CRISPR knock‐out mutants (Figs 2i, S15, S16).

To determine whether 3‐OH‐β‐cyclocitral is derived from zeaxanthin, we quantified carotenoids in both OsCCD4b knockout and overexpression lines. Overexpression of OsCCD4b led to a significant decrease in β‐carotene and zeaxanthin, whereas other carotenoids remained largely unchanged. Interestingly, the CRISPR knockout mutant did not show a significant difference in total carotenoid levels (Fig. S17), suggesting that OsCCD4b is not a major regulator of overall carotenoid content or tissue pigmentation in rice, in contrast to CCD4 enzymes in some other species (Gonzalez‐Jorge et al., 2013; Ma et al., 2013; Zheng et al., 2021). Chl content and ABA level were also unaffected (Figs S18, S19), further supporting the conclusion that OsCCD4b is not directly involved in ABA biosynthesis. This conclusion is consistent with the findings mentioned above, which show the enzyme's inability to generate the ABA precursor xanthoxin (Fig. 1f) and its functional independence from SL and KAR signaling pathways (Fig. 2f), suggesting that 3‐OH‐β‐cyclocitral may act as an independent apocarotenoid signal.

To investigate how OsCCD4b influences mesocotyl and cell development at the transcriptomic level, we performed RNA‐sequencing on mesocotyl tissues from the WT, Osccd4b knockout mutant, and overexpression lines. PCA revealed clear transcriptomic separation among the three genotypes (Fig. S20), suggesting that OsCCD4b exerts a substantial impact on gene expression in rice mesocotyls. GO (Gene Ontology) enrichment analysis of differentially expressed genes indicated significant alterations in sugar‐related metabolic pathways. Specifically, the OsCCD4b knock‐out mutant showed upregulation of genes involved in pentose and arabinose metabolism (Fig. 2j), whereas the overexpression lines displayed downregulation of genes related to polysaccharide and glucan metabolism (Fig. 2k), indicating that OsCCD4b may regulate cell elongation, at least in part, by modulating sugar metabolism.

Conclusion

In conclusion, this study highlights the functional diversification of CCD4 enzymes and the evolutionary emergence of distinct C_30_ and C_10_ apocarotenoids across plant species. Our findings reveal the biological function of the LOC_Os12g24800 gene as OsCCD4b, regulating mesocotyl elongation and sugar metabolism through a previously uncharacterized, hormone‐independent apocarotenoid signaling pathway. This work advances our understanding of apocarotenoid biology – particularly the function of 3‐OH‐β‐cyclocitral – in monocot species and provides a foundation for future investigations into growth‐regulating carotenoid‐derived metabolites.

Competing interests

None declared.

Author contributions

SA‐B conceived the project and supervised the experiments; XZ and SA‐B designed the experiments; YZ, XZ performed callus transformation and q‐RT‐PCR analysis; CR generated rice overexpression lines; YZ and IB performed confocal imaging of mesocotyl cells; YZ, AA, YA, LB, and JYW performed phenotyping; AA, XZ, and YA conducted and analyzed RNA‐seq experiments; XZ and AA carried out bioinformatics analysis; XZ, KXL, and JM performed metabolite analysis. XZ and AB performed transient expression in Nicotiana, in vivo and in vitro assays. YZ, XZ, and AA prepared figures and wrote the manuscript. SA‐B edited and approved the manuscript. YZ and AA contributed equally to this work.

Supporting information

Fig. S1 Analysis of conserved motifs in the CCDs protein sequences of Arabidopsis and rice was conducted using MEME software. Fig. S2 Quantification of β‐apo‐8′‐carotenal, β‐apo‐10′‐carotenal, and 3‐OH‐β‐apo‐10′‐carotenal in in vivo assays of CCD4 using β‐carotene or zeaxanthin as substrates. Fig. S3 Apocarotenoid profiling of Nicotiana leaves transiently overexpressing OsCCD4b or AtCCD4. Fig. S4 Relative expression levels of transgenes in independent OsCCD4b‐ or *AtCCD4‐*overexpression citrus callus lines. Fig. S5 Apocarotenoid profiling in transgenic citrus callus overexpressing OsCCD4b or AtCCD4. Fig. S6 Carotenoid analysis of wild‐type and OsCCD4b or AtCCD4 overexpressing citrus callus lines. Fig. S7 High‐resolution MS and UV/vis spectra of xanthoxin. Fig. S8 UHPLC analysis of the in vitro assays of OsCCD4b and its mutant using 9‐cis‐β‐carotene as substrate. Fig. S9 Mutations in OsCCD4b CRISPR knockout Nipponbare plants result in truncated proteins with loss of enzymatic activity. Fig. S10 Identification of Cas9‐free OsCCD4b CRISPR knockout mutant plants. Fig. S11 Sequencing analysis confirmed a OsCCD4b knockout CRISPR line in the Zhonghua 11 (ZH11) background. Fig. S12 qRT‐PCR confirmation of transgene overexpression in OsCCD4b‐overexpressing Nipponbare plants. Fig. S13 Root phenotypes of wild‐type, OsCCD4b CRISPR knock‐out, and overexpression line seedlings grown on agar medium. Fig. S14 Phenotypic characterization of hydroponically grown Nipponbare wild‐type and OsCCD4b CRISPR knockout mutant seedlings. Fig. S15 Quantification of β‐apo‐8′‐carotenal and 3‐OH‐β‐apo‐8′‐carotenal in different OsCCD4b overexpression lines and CRISPR knockout mutants. Fig. S16 Apocarotenoid profiling in hydroponically grown shoots of Nipponbare wild‐type and OsCCD4b overexpression lines. Fig. S17 Carotenoid analysis in hydroponically grown shoots of wild‐type, OsCCD4b overexpression lines, and CRISPR knockout mutants. Fig. S18 Chlorophyll levels in hydroponically grown shoots of wild‐type, OsCCD4b overexpression lines, and CRISPR knockout mutants. Fig. S19 ABA quantification in shoots and roots of Nipponbare wild‐type and OsCCD4b CRISPR knockout Nipponbare plants. Fig. S20 The PCA based on FPKM value of wild‐type (NB), OsCCD4b CRISPR knockout mutant, and overexpression lines obtained from RNAseq. Table S1 Sequences of primers used in this study.Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.