Restorer of fertility like 30, encoding a mitochondrion-localized pentatricopeptide repeat protein, regulates wood formation in poplar

Abstract Nuclear–mitochondrial communication is crucial for plant growth, particularly in the context of cytoplasmic male sterility (CMS) repair mechanisms linked to mitochondrial genome mutations. The restorer of fertility-like (RFL) genes, known for their role in CMS restoration, remain largely unexplored in plant development. In this study, we focused on the evolutionary relationship of RFL family genes in poplar specifically within the dioecious Salicaceae plants. PtoRFL30 was identified to be preferentially expressed in stem vasculature, suggesting a distinct correlation with vascular cambium development. Transgenic poplar plants overexpressing PtoRFL30 exhibited a profound inhibition of vascular cambial activity and xylem development. Conversely, RNA interference-mediated knockdown of PtoRFL30 led to increased wood formation. Importantly, we revealed that PtoRFL30 plays a crucial role in maintaining mitochondrial functional homeostasis. Treatment with mitochondrial activity inhibitors delayed wood development in PtoRFL30-RNAi transgenic plants. Further investigations unveiled significant variations in auxin accumulation levels within vascular tissues of PtoRFL30-transgenic plants. Wood development anomalies resulting from PtoRFL30 overexpression and knockdown were rectified by NAA and NPA treatments, respectively. Our findings underscore the essential role of the PtoRFL30-mediated mitochondrion-auxin signaling module in wood formation, shedding light on the intricate nucleus–organelle communication during secondary vascular development.


Introduction
Mitochondria, thought to have evolved from free-living bacterial ancestors absorbed by an archaeal-like host, have a separate genome [1].During endosymbiotic evolution, most mitochondrial genes were lost or transferred to the cell nucleus [2].Conversely, numerous proteins encoded by nuclear genes are essential for mitochondrial biogenesis and function [3].Specifically, proteins destined for mitochondria actively participate in driving mitochondrial gene expression [4], forming the foundation for nucleus-dependent organelle control.Mitochondrial functionality dynamically modulates nuclear gene transcription, providing context-dependent feedback from mitochondria to the nucleus [5].Bidirectional communication, termed anterograde and retrograde signaling, orchestrates interactions between mitochondrial and nuclear genomes [6], playing a crucial function in plant development and stress responses [5].
Cytoplasmic male sterility (CMS) and its restoration serve as an insightful model for probing nuclear-mitochondrial interactions in plant cells.The CMS phenotype is often linked to atypical chimeric open reading frames (ORFs) expressed in mitochondria [7].These ORFs, arising from mitogenomic recombination and cotranscribed with normal mitochondrial genes, encode cytotoxic proteins that disrupt microsporogenesis and pollen viability [8].Nuclear genes known as restorers of fertility (Rf ) promote CMS restoration by normalizing pollen production in plants with sterile cytoplasm [9].Among various plant species, like petunia, rice, rapeseed, and radish, identified Rf genes encode members of a large RNA-binding protein family characterized by tandem pentatricopeptide repeat (PPR) domains [10][11][12].Initially discovered in Arabidopsis thaliana genomic sequences, the PPR protein family comprises 2-27 P-type 35-amino-acid domains (PPR domains) [10,12].Recent studies have identified two PPR domain variants: L-type domains (∼35-36 amino acids) and S-type domains (31 amino acids) [13,14].Functionally, proteins containing P-type PPR domains are involved in 3 and 5 terminal processing, RNA stabilization, cleavage, translation activation, and RNA intron splicing regulation.In contrast, PLS-type PPR proteins are involved in C-to-U RNA editing [4,15].Specifically targeted to mitochondria, Rf-PPR proteins play a crucial role in recognizing and eliminating CMSinducing transcripts, thereby restoring normal pollen production [16][17][18][19].Altogether, Rf-PPR proteins emerge as pivotal mediators in orchestrating nuclear-mitochondrial interactions during CMS restoration.
The genomes of land plants contain a considerable expansion of the PPR protein family, but the Rf-PPR protein selection pattern is different from that of most other PPR gene families (non-Rf ) [9,20].Most PPR proteins (non-Rf) undergo selection against mutations to conserve functional protein sequences [21].Local sequence duplication is a common method used by plants to create new Rf-PPR genes in response to sterilityinducing genes [16,17,22].Besides, Fujii et al. demonstrated that not only do members of the Rf-PPR protein family face diverse selection pressures, but different amino acids within a single Rf-PPR protein also face varying degrees of selection pressure [20].The amino acids directly involved in CMS transcript recognition are 5-15 times more likely to be susceptible to varied selection than other amino acids in this domain [20,23,24].Diversified selection of PPR proteins helps to more quickly and effectively create new Rf genes to silence newly emerging CMS transcripts, which is a successful strategy for dealing with conf licts between nuclear and mitochondrial genomes in angiosperm evolution.
Recently, mitochondrial-targeted members of the Rf-PPR family have emerged as crucial nuclear response factors for modifying mitochondrial homeostasis, playing indispensable roles in processes beyond CMS restoration during growth and development [25].Rice mutants carrying natural blight leaf 3 (nbl3) not only exhibit delayed growth and premature senescence, but also demonstrate increased resistance to bacterial and fungal infections, as well as enhanced tolerance to salt stress [26].Nine mitochondrial-targeted PPR proteins in Arabidopsis have been found to be involved in the plant's responses to different abiotic or biotic stressors [27][28][29][30][31][32][33][34].In maize, endosperm and embryonic development are almost completely halted due to the functional absence of Small kernel (SMK) genes [35][36][37].Additionally, plants have evolved variable reproductive strategies.Rf-PPR genes are also present in the genomes of dioecious plants; however, their functions remain unclear [20].According to our earlier studies, MiR476a specifically targets a number of PPR genes located in the mitochondria that are part of the RFL gene family.By altering the auxin pathway, MiR476a/RFL-mediated dynamic regulation of mitochondrial homeostasis affects adventitious root development [38].At present, there is still no direct evidence that Rf-PPR genes are involved in wood formation, but recent research indicates that the vascular cambium region accumulates a large amount of reactive oxygen species (ROS), and the type (O 2 − and H 2 O 2 ) and level of ROS are key to maintaining cambium activity [39].Mitochondria are one of the main production sites for ROS in plant cells, and mitochondrial homeostasis directly determines the formation levels of different ROS [40][41][42][43].These findings suggested that mitochondrial homeostasis may be necessary for plant vascular development.Therefore, there is a need to prove whether the Rf-PPR gene mediates mitochondrial homeostasis during wood formation in trees.
In this study, we identified 32 RFL genes in Populus tomentosa, which are exclusively conserved in Salicaceae plants.Among these genes, the member named PtoRFL30 was discovered, exhibiting specific expression in the stem.Through genetic and physiological analyses of mitochondrial-targeted PtoRFL30, we elucidated its role in modulating mitochondrial homeostasis, thereby inf luencing wood formation through the regulation of auxin levels in the secondary vasculature.Our findings offer initial insights into the collaborative involvement of mitochondria and phytohormones in the development of secondary vasculature in woody plants.

PtoRFL30 is preferentially expressed in the vascular cambium zone
To elucidate the evolutionary relationships among RFL family genes in poplar and other angiosperm species, we identified RFL genes from A. thaliana (25), P. tomentosa (32), P. trichocarpa (34), Salix purpurea (61), Linum usitatissimum (22), and Manihot esculenta (134) (Supplementary Data Table S1).Phylogenetic analysis revealed that the evolution of RFL genes in the Salicaceae occurred after Salicaceae divergence (Supplementary Data Fig.S1).Although most RFL genes within the Salicaceae form orthologous clusters, a subset of genes in S. purpurea forms specific paralogous clusters (Supplementary Data Fig.S1).This suggests that the RFL gene family has not yet reached complete evolution and expansion within the Salicaceae.In P. tomentosa, 32 RFL genes were identified and clustered into two clades (Fig. 1A).Clade A is exclusive to the Salicaceae, while the genes in Clade B share similarities in origin with the RFL genes of other species in the Malpighiales (Supplementary Data Fig.S1 and B).
The expression profiles of all RFL genes in various tissues of Populus were characterized to investigate their potential involvement in wood development (Supplementary Data Fig.S2A).Among these, a member of Clade B, named PtoRFL30, exhibited significantly higher expression levels in the stem compared with other tissues (Supplementary Data Fig.S2A and Fig. 1C).Protein structure prediction via the TPRpred online tool (https:// toolkit.tuebingen.mpg.de)indicated that PtoRFL30 encoded a classic PPR protein harboring 14 typical P-type PPR motifs (Supplementary Data Fig.S3).Furthermore, we assessed coexpression patterns of poplar genes in the secondary vasculature by weighted correlation network analysis (WGCNA) using the datasets from the AspWood database [44] (Supplementary Data Fig.S2B).It was found that the co-expression module containing PtoRFL30 displayed a stronger association with the cambium compared with other tissues (Supplementary Data Fig.S2C).Besides, the GUS reporter driven by the PtoRFL30 promoter in transgenic poplar verified its expression in the cambium zone within the stem (Fig. 1D and E).These findings suggest that PtoRFL30 is likely involved in the regulation of vascular cambial activity during wood formation.

PtoRFL30 negatively regulates vascular cambial activity and wood formation
To determine the physiological role of PtoRFL30 in wood formation, we conducted both overexpression (OE) and RNA interference (RNAi)-based knockdown experiments targeting PtoRFL30.By detecting the PtoRFL30 expression levels in transgenic plants, two overexpressing lines (L1 and L4) with considerably higher PtoRFL30 gene expression were chosen for further analysis (Supplementary Data Fig.S4A).The PtoRFL30-OE lines displayed reduced plant height, internode number, and stem diameter compared with the wild-type (WT) plants (Supplementary Data Fig.S4A, B, and E-G).Conversely, the opposite phenotype was observed in PtoRFL30-RNAi transgenic poplar (L1 and L6), where PtoRFL30 expression levels were decreased (Supplementary Data Fig.S4C-G).
Vascular development observation in both WT and transgenic plants revealed that the vascular cambium typically consists of four cell layers.However, the PtoRFL30-OE and -RNAi transgenic lines exhibited approximately three and five cell layers, respectively, in most of the vascular cambial files (Fig. 2A-D).The variation in the number of cambial cell layers between WT and transgenic plants indicated that PtoRFL30 inhibits cambial activity.Overexpression of PtoRFL30 led to repression of secondary xylem development, while transgenic plants with reduced PtoRFL30 expression levels showed a significant increase in the number of secondary xylem cell layers (Fig. 2E and F).
In addition, quantitative PCR (qPCR) was employed to assess the expression of key genes during wood development.The results indicated that PtoRFL30 overexpression suppresses the transcription of genes associated with cambial activity (WOX4a, WOX4b, ANT, and CYCD3;4) [45] and secondary xylem differentiation (HB7 and HB8) [46].In contrast, the expression levels of these genes were significantly upregulated in PtoRFL30-RNAi lines (Fig. 2G).These findings suggest that PtoRFL30 represses vascular cambial activity and wood formation.

PtoRFL30 modulates mitochondrial homeostasis
RFL proteins are typically localized in chloroplasts or mitochondria, exerting inf luence over the expression of organellar transcripts [15,47].In order to ascertain its subcellular localization, we introduced PtoRFL30 fused with a green f luorescent protein (GFP) tag into Arabidopsis.The f luorescence signals of PtoRFL30-GFP overlapped with the red f luorescence emitted by mitochondrialspecific dyes (Mito-red) in Arabidopsis root cells and isolated root protoplasts (Fig. 3A-C), indicating that PtoRFL30 targets mitochondria.
Mitochondria play a crucial role in cellular respiration, serving as the main source of ATP and ROS.Therefore, ATP and ROS levels are commonly utilized as key biochemical markers to assess mitochondrial function [48].To confirm the impact of PtoRFL30 proteins on mitochondrial function, we measured the content of ROS and ATP in PtoRFL30 transgenic plants.In comparison with WT, the ATP content decreased by ∼30% in PtoRFL30-OE plants, while it increased by >120% in PtoRFL30-RNAi lines (Fig. 3D).The content of H 2 O 2 in PtoRFL30 transgenic plants also exhibited a significant change of a similar magnitude (Fig. 3D).Additionally, specific dyes for H 2 O 2 (3,30-diaminobenzidine, DAB) and O 2 − (nitrotetrazolium blue chloride, NBT) were used to stain various tissues of poplar.The results demonstrated that the accumulation of ROS (H 2 O 2 and O 2 − ) in leaves, roots, and stems of PtoRFL30-OE lines was lower than that of WT, whereas the accumulation level in PtoRFL30-RNAi plants was higher (Fig. 3E and F).Furthermore, PtoRFL30-OE and -RNAi lines showed variable expression levels of various mitochondrial-encoded genes when compared with WT (Supplementary Data Fig.S5), suggesting that PtoRFL30 functionally modulates mitochondria.These findings indicate that variations in PtoRFL30 expression levels compromise mitochondrial homeostasis.

PtoRFL30-mediated mitochondrial homeostasis is necessary for wood formation
To investigate whether mitochondrial homeostasis is involved in wood growth, we treated WT plants with two mitochondrial function inhibitors, rotenone and antimycin A, which inhibit complexes I and III of the electron transfer chain in mitochondria, respectively.Both inhibitor treatments resulted in a significant reduction in the cell layer number of vascular cambium and secondary xylem (Supplementary Data Fig.S6A-E).Additionally, the expression levels of key genes related to vascular cambial cell division and secondary xylem differentiation were inhibited, especially under antimycin A treatment (Supplementary Data Fig.S6F).This suggests that maintaining mitochondrial homeostasis is essential for wood formation.
Furthermore, we applied rotenone and antimycin A to PtoRFL30-RNAi transgenic plants.It was found that the expression level of mitochondrial-encoded genes in PtoRFL30-RNAi transgenic plants was restored under treatment with mitochondrial function inhibitors (Supplementary Data Fig.S7).Observation of the wood developmental phenotype showed that the number of vascular cambial cell layers and secondary xylem cell layers in lines treated with rotenone was significantly reduced compared with untreated PtoRFL30-RNAi transgenic plants (Fig. 4A-E), with little variation in the expression level of genes related to wood development (except for WOX4b and CYCD3;4, which did not return to the expression level in WT) (Fig. 4F).Antimycin A treatment restored excess wood development in PtoRFL30-RNAi plants to a state similar to WT, including the expression level of these key genes (Fig. 4).This indicates that PtoRFL30-mediated mitochondrial homeostasis is crucial for maintaining normal wood development.

PtoRFL30 inhibits accumulation of auxin in the secondary vasculature
Previous studies have established that auxin transport and homeostasis are essential for microRNA476a-RFL modulemediated mitochondrial homeostasis, leading to adventitious root formation [38].Therefore, we examined the expression levels of genes associated with auxin homeostasis in response to mitochondrial function inhibitor treatment and PtoRFL30 transgenic plants to investigate whether PtoRFL30 inf luences auxin accumulation during wood formation.Under mitochondrial function inhibitor treatment, the expression levels of the auxin biosynthesis gene (YUC1), polar transport genes (PIN1a, PIN1b), and primary auxin-responsive gene (GH3.5)were significantly reduced (Fig. 5A).Simultaneously, PtoRFL30 overexpression markedly suppressed the expression of auxin biosynthesis and transport genes (Fig. 5B).A decrease in auxin signaling in vivo was indicated by the downregulation of the primary auxin-responsive gene GH3.5, as further supported by testing the auxin content of the poplar stem (Fig. 5B and C).In contrast to PtoRFL30-OE plants, PtoRFL30-RNAi plants exhibited upregulation of auxin homeostasis-related genes and increased auxin accumulation (Fig. 5B and C).Immunof luorescence imaging using the antibody against IAA revealed that PtoRFL30 altered auxin content in the stem but did not affect the auxin distribution pattern in vasculature (Fig. 5D and E).

PtoRFL30 regulates wood formation by modulating auxin homeostasis
Excessive PtoRFL30 expression and mitochondrial function inhibitors (antimycin A and rotenone) disrupt mitochondrial homeostasis, impeding vascular cambial activity and secondary xylem differentiation, contrary to the known inf luence of auxin on wood formation in poplar [45,46].Conversely, exogenous auxin supplementation mitigates the effects of mitochondrial function inhibitors, delaying wood development (Supplementary Data Fig.S8).To further validate the necessity of auxin homeostasis for PtoRFL30-mediated wood development,PtoRFL30-OE transgenic plants were treated with NAA.NAA treatment partially restored the defects in secondary vascular development caused by PtoRFL30 overexpression compared with the mock control (Fig. 6).Auxin treatment restored the number of vascular cambial cell layers in PtoRFL30-OE transgenic plants to WT levels (Fig. 6A-D), while the number of xylem cell layers increased but did not reach the WT level (Fig. 6A and E).In PtoRFL30-RNAi plants, blocking polar auxin transport with NPA significantly reduced the number of cell layers in the vascular cambium and secondary xylem, leading to defects in wood development (Fig. 6).Together, these findings indicate the dependence of PtoRFL30-mediated wood formation on auxin homeostasis in secondary vasculature.

Discussion
RFL proteins, recognized as sequence-specific RNA-binding proteins crucial for CMS restoration [7,9], have been implicated in diverse functions throughout the plant life cycle, yet their involvement in secondary vascular development remains unexplored [25,49].This study unveils PtoRFL30 as a mitochondrial-targeted PPR protein in poplar, demonstrating specific expression in the vascular cambium and a negative regulatory role in wood development by inf luencing mitochondrial homeostasis.Importantly, we elucidate the PtoRFL30-mediated mitochondrial regulation of wood formation, emphasizing its dependency on auxin accumulation within the secondary vasculature (Fig. 7).
The RFL gene family, well known for its rapid evolution, engages in a molecular 'arms race' between the mitochondrial and nuclear genomes [9,20,21].This swift evolution is attributed to the emergence of new PPR genes in response to evolving infertilityinducing genes.The intricate interplay between the mitochondrial and nuclear genomes has been recognized as a dynamic process [20,50,51].Notably, RFL genes exhibit species specificity, adapting to random mitochondrial genome mutations, with each Rf gene recognizing the transcript sequence of a specific mitochondrial gene [8,16,19].Our study reveals that rapid evolution persisted after Salicaceae divergence, leading to the emergence of distinct RFL genes in Salicaceae plant genomes (Fig. 1B, Supplementary Data Fig.S1).However, in dioecious Salicaceae plants, where cross-pollination circumvents the negative effects of CMS on reproduction, the presence of RFL genes may not be as pronounced.Previous studies have demonstrated the regulatory role of poplar RFL genes in adventitious root formation through a mitochondria-dependent pathway [38].Building on this, we demonstrate that PtoRFL30 regulates wood formation in Populus through mitochondrial regulation, intertwined with auxin homeostasis.Understanding the specific functions of RFL genes in dioecious species unveils novel insights into how mitochondrial signaling orchestrates plant growth and development.
Mitochondrial function is tightly integrated into metabolism and biosynthetic processes and is also crucial for signaling and responding to stresses [52][53][54][55].In plants, the levels of signals and energy metabolism are used to determine the mitochondrial functional state [48].The levels of mitochondrial energy metabolites ATP (or the ATP/ADP ratio) and NAD/NADH are known to directly or indirectly affect plant development [56].While ROS serve as mitochondrial retrograde signals, their excessive production poses a dual role, regulating diverse biological processes yet inducing oxidative stress [40,42,43,57,58].In poplar, mitochondrial signaling through ROS has been implicated in responses to biotic and abiotic stress [59][60][61].However, a delicate balance must be maintained, as heightened ROS levels can lead to cellular damage.The importance of sustaining mitochondrial functional homeostasis is underscored by its role in secondary vascular development (Supplementary Data Fig.S6).We further identified that the PtoRFL30 protein is localized in mitochondria and regulates the levels of ATP and ROS in poplar (Fig. 3).The secondary vascular developmental phenotype caused by PtoRFL30 deficiency was restored by altering the mitochondrial state (Fig. 4).All of these findings point to a role for PtoRFL30 in preserving mitochondrial homeostasis throughout the growth of wood.
It is generally believed that ROS and Ca 2+ are related to the mitochondrial retrograde signaling cascade [62][63][64].Given that ROS are generated through the mitochondrial respiratory electron transport chain (mETC), they often serve as a messenger for mitochondrial metabolic status, interlinked with key mitochondrial metabolites like ATP, acetyl-CoA, TCA cycle intermediates, and NAD/NADH [48,65,66].Despite their crucial roles, ROS presently lack well-defined receptors [38,48].While mitochondrial retrograde signaling has undergone extensive exploration in yeast and mammalian cells, the regulatory network governing mitochondrial and nuclear communication during plant development remains an area in need of further refinement.In the course of PtoRFL30-mediated wood development, the perturbation of mitochondrial homeostasis exerts inf luence on the levels of auxin transport and homeostasis-related gene expression, consequently impacting auxin accumulation within the secondary vasculature (Fig. 5).This suggests a regulatory mechanism wherein nuclear genes undergo transcription under the governance of signaling pathways originating in the mitochondria and translocating to the nucleus.
As RFL30 encodes a typical P-type PPR protein (Supplementary Data Fig.S3), the possibility cannot be excluded that it still acts during reproductive processes like CMS in gynodioecious horticultural plants.RFL genes encode P-type PPR proteins that generally participate in 3 and 5 terminal processing, RNA stabilization, cleavage, translation activation, and RNA intron splicing regulation of mitochondrial genes [4,15].In order to restore CMS, Rf-PPRs mainly target genes that are toxic to reproductive development in the mitochondrial genome, and restore the fertility of plants by inhibiting their target protein production [16][17][18][19].However, overexpression of PtoRFL30 exhibits an inhibitory effect on secondary vascular development (Fig. 2), suggesting that its target genes may not be negative regulatory factors for normal plant development.In addition, the regulatory mechanism of PtoRFL30 in wood formation is more likely to contribute to maintaining mito-chondrial homeostasis by limiting mitochondrial function and affecting mitochondrial energy metabolism and electron transfer (Fig. 3 and Supplementary Data Fig.S5).Meanwhile, PtoRFL30 utilizes the mitochondrial retrograde signaling pathway to inf luence endogenous signals (such as auxin) to regulate secondary vascular development (Figs 5 and 7).This is an unrecognized role of RFLs to regulate plant development, which relies on a different mechanism than RFLs to restore CMS.Considering the large number of RFL genes in the plant genome and their function in de novo root organogenesis and vascular cambium activity, it is likely that RFL proteins represent a group of key regulators in multiple developmental programs of plants, including dioecious horticultural plants.According to this conclusion, RFL genes may possess potential roles in regulating important traits of gardening plants, which affect the vegetative growth and reproductive development of plants through diverse mechanisms.
In conclusion, a comprehensive examination of the evolutionary relationships among RFL family genes in poplar and other angiosperm species reveals that, despite poplar being dioecious, its genome retains a significant number of RFL genes.Within the poplar RFL family, PtoRFL30 exhibits preferential expression in the vascular tissue of the stem, playing a crucial role in upholding mitochondrial functional homeostasis.Moreover, PtoRFL30, through its inf luence on auxin accumulation in the secondary vasculature, orchestrates mitochondrial states that regulate vascular cambial activity and xylem development.These findings provide insights into the novel role of PtoRFL30 in poplar, highlighting its involvement in coordinating mitochondrial homeostasis and auxin levels during wood formation.

Phylogenetic analysis
All the RFL amino acid sequences used to construct the phylogenetic tree are listed in Supplementary Data Table S1.The JTT + G model and 1000 bootstrap replicates based on amino acid sequences were used to build the phylogenetic tree.

Weighted correlation network analysis
WGCNA (1.72) was used to identify co-expressed genes according to the protocol [67].The list and expression data of differentially expressed genes (DEGs) were downloaded from Asp-Wood [44].WGCNA network construction and module detection were performed with an unsigned type of topological overlap matrix (TOM), a power β of 6, a minimal module size of 30, and a branch merge cut height of 0.15.The sample classification information used for trait association analysis came from AspWood [10].

Gene cloning and plasmid construction
Both PtoRFL30-OE and PtoRFL30-GFP plant expression vectors were constructed by Xu et al. [38].The promoter sequence upstream of the PtoRFL30 was amplified from genomic DNA of P. tomentosa and inserted into XcmI-digested pCXGUS-P to drive GUS expression [68].The 208-bp specific sequence of PtoRFL30 and its reverse complementary sequence from cDNA of P. tomentosa were constructed into the PtoRFL30-RNAi plant expression vector as described previously [69].

Genetic transformation and growth conditions
The above plant expression vectors were stably transformed into leaf disks from WT P. tomentosa by Agrobacterium-mediated infiltration as described previously [70].For phenotypic analysis, seedlings were cultivated in the greenhouse [23-25 • C, 16 h light (10 000 lux):8 h dark] for 3 months.

RT-qPCR
Total RNA was isolated with the Biospin Plant Total RNA Extraction Kit (Biof lux, China).Then, the PrimeScript RT Reagent Kit with gDNA Eraser (0047A; TaKaRa, China) was used to transcribe RNA into cDNA.RT-qPCR was performed in a Real-Time PCR machine (qTOWER3G; Analytik Jena, Germany) with Hieff qPCR SYBR Green Master Mix (11201ES08; Yeasen, China).The UBIQUITIN (UBQ) gene of poplar was used as the reference gene when calculating the expression data using the ΔΔCt method.The sequences of primers used are listed in Supplementary Data Table S2.

Quantification of ATP and H 2 O 2 contents
The ATP and H 2 O 2 contents of the eighth internodes from 3month-old P. tomentosa WT and transgenic plants were measured using a hydrogen peroxide assay kit (S0038; Beyotime Biotechnology, China) and an ATP assay kit (S0026; Beyotime Biotechnology, China).

IAA content quantification
As previously described [71], 0.5 g poplar stem samples of WT and transgenic plants were gathered for IAA purification.Quantification was carried out using an LC-ESI-MS/MS system (4000 Q-Trap; Sciex, USA).As an internal standard, 50 ng of [ 13 C 6 ] IAA was added to each extraction buffer.

Immunohistochemical localization of auxin
The procedures previously outlined for sample preparation and immunof luorescence detection were followed [72].In short, stem cross-sections of 3-month-old WT and transgenic plants were treated with a polyclonal antibody against IAA (Agrisera, Sweden) as a primary antibody.DyLight550-labeled secondary antibody (Abcam, UK) was used to detect the signal on a confocal microscope (TCS SP8; Leica, Germany) with excitation set at 550 nm and emission ranging from 560 to 600 nm.

Accession number
The GenBank accession number of the P. tomentosa gene PtoRFL30 is MN242837.

Figure 1 .
Figure 1.Phylogenetic analysis and expression pattern of PtoRFL30 in P. tomentosa.A Phylogenetic relationship of all the RFL genes in P. tomentosa subgenome A. All the RFL sequences used to construct the tree are listed in Supplementary Data Table S1.The JTT + G model and 1000 bootstrap replicates based on amino acid sequences were used to build the phylogenetic tree.Every branch displays support above 500 as a circle dot.B Schematic representation of the evolutionary history of the RFL genes in Supplementary Data Fig.S1.C PtoRFL30 relative expression level in various P. tomentosa plant tissues as assessed by RT-qPCR.The dots represent the values of each biological replicate.Error bars represent ± standard deviation; n = 3. D, E Histological staining of the stem from GUS reporter lines driven by the promoter of PtoRFL30.The seventh internodes of 3-month-old P. tomentosa plants were cross-sectioned for GUS staining.Ca, cambium; Ph, phloem; Xy, xylem.Scale bars: 500 μm (D); 200 μm (E).

Figure 2 .
Figure 2. PtoRFL30 regulates vascular cambial activity and xylem development-related phenotypes in P. tomentosa.A, B Phenotype associated with vascular cambial cell proliferation in PtoRFL30-OE, PtoRFL30-RNAi, and WT transgenic lines.The images were captured on toluidine blue-stained cross-sections of the eighth internode of 3-month-old plants.A The areas between lines represent cambial zones.B Cambial cells in the same cell file are represented by dots.Scale bars: 50 μm (A); 20 μm (B).C Cambial cell layer number in each file of WT, PtoRFL30-OE, and PtoRFL30-RNAi transgenic lines.D Frequency distributions of cell numbers in each cambial cell file in the stems of WT, PtoRFL30-OE, and PtoRFL30-RNAi transgenic lines.The WT frequency curve's highest value is shown by the line.E Detailed observation of secondary xylem development of WT, PtoRFL30-OE and PtoRFL30-RNAi transgenic lines.Lines represent xylem zones.Scale bar: 50 μm.F Number of xylem cell layers in stems of WT, PtoRFL30-OE, and PtoRFL30-RNAi corresponding to E. G Expression levels of wood formation regulators WOX4a, WOX4b, ANT, CYCD3;4, HB7, and HB8 of WT, PtoRFL30-OE, and PtoRFL30-RNAi transgenic lines.Dots represent the values of biological replicates.Error bars show ± standard deviation.Significant differences from WT are indicated by asterisks (one-way ANOVA, pairwise comparisons, Dunnett's test): * P < 0.05; * * P < 0.01; * * * P < 0.001; n = 24 (C), 40 (F), and 3 (G).

Figure 3 .
Figure 3. PtoRFL30-mediated changes in mitochondrial state.A-C Mitochondrial location of PtoRFL30 protein was found in the roots (A), root cells (B), and root protoplasts (C) of transgenic Arabidopsis that were carrying ectopically expressed PtoRFL30-GFP.The mitochondrial dye Mito-tracker Red (Mito-red) released f luorescence, which confocal microscopy was used to detect.D Quantitative measurements of ATP and H 2 O 2 in the stems of 3-month-old PtoRFL30-RNAi, PtoRFL30-OE, and WT.Error bars show ± standard deviation.Significant differences from the WT are indicated by asterisks (one-way ANOVA, pairwise comparisons, Dunnett's test): * * * P < 0.001; n = 3. (E, F) 3,30-Diaminobenzidine (DAB) (E) and nitrotetrazolium blue chloride (NBT) (F) staining of leaves, roots, and stems of WT, PtoRFL30-RNAi, and PtoRFL30-OE plants.

Figure 4 .
Figure 4. Rescue of wood formation phenotypes resulting from reduced expression of PtoRFL30 by mitochondrial function inhibitors.A, B Detailed observation of toluidine blue-stained cross-sections of the eighth internode of 3-month-old WT and PtoRFL30-RNAi transgenic lines under treatment with mitochondrial function inhibitors (10 μM rotenone and 50 μM antimycin A).Dots in B represent cambial cells in the same cell files.Ca, cambium; Xy, xylem.Scale bars: (A) 50 μm; (B) 20 μm.C Cambial cell layer number in each cell file of WT and PtoRFL30-RNAi transgenic lines treated with mitochondrial function inhibitors.D Cell number frequency distributions in cambial cell files in the stems of WT and PtoRFL30-RNAi transgenic lines treated with mitochondrial function inhibitors.The peak value of the WT frequency curve is indicated by the line.E Number of xylem cell layers in stems of WT and PtoRFL30-RNAi transgenic lines treated with mitochondrial function inhibitors, corresponding to A. F Expression levels of wood formation regulators WOX4a, WOX4b, ANT, CYCD3;4, HB7, and HB8 of WT and PtoRFL30-RNAi transgenic lines treated with mitochondrial function inhibitors.Dots represent the values of each biological replicate.Error bars show ± standard deviation.Significant differences are shown by letters above the error bars (one-way ANOVA followed by pairwise comparisons using Tukey's test); n = 10 (C); n = 49 (E); n = 3 (F).

Figure 5 .
Figure 5. Auxin accumulation is altered by PtoRFL30-mediated mitochondrial homeostasis during wood formation.A, B Expression levels of YUC1, TAA1, PIN1a, PIN1b, and GH3.5, which are genes related to auxin homeostasis, in stems of WT under treatment with mitochondrial function inhibitors (10 μM rotenone or 50 μM antimycin A) (A) and in PtoRFL30-OE and PtoRFL30-RNAi transgenic lines (B).C Content of auxin (IAA) in stems of WT, PtoRFL30-OE, and PtoRFL30-RNAi transgenic lines.Dots represent values of each biological replicate.Error bars show ± standard deviation.Significant differences from WT are indicated by asterisks (one-way ANOVA, pairwise comparisons, Dunnett's test): * P < 0.05; * * P < 0.01; * * * P < 0.001; n = 3. D Immunof luorescence of auxin (IAA) in secondary vasculature of WT, PtoRFL30-OE, and PtoRFL30-RNAi transgenic lines.Ph, phloem; Ca, cambium; Xy, xylem.Scale bar: 100 μm.E Intensity analysis of immunof luorescent signals of auxin (IAA) in secondary vasculature of WT, PtoRFL30-OE, and PtoRFL30-RNAi transgenic lines.Fluorescence intensity in the negative control is indicated by the dashed line.The highest value of f luorescence intensity in WT is shown by the dotted line.

Figure 6 .
Figure 6.Wood formation phenotypes resulting from auxin and auxin transport inhibitor treatment in PtoRFL30-related transgenic plants.A, B Anatomical sections of the eighth internode of 3-month-old wild-type (WT), PtoRFL30-OE transgenic lines treated with auxin (0.1 μM NAA), and PtoRFL30-RNAi transgenic lines treated with auxin transport inhibitor (5 μM NPA), stained with toluidine blue.B Dots represent cambial cells in the same cell files.Ca, cambium; Xy, xylem.Scale bars: (A) 50 μm; (B) 20 μm.C Cambial cell layer number in each cell file of WT and PtoRFL30-OE transgenic lines under 0.1 μM NAA treatment and PtoRFL30-RNAi transgenic lines under 5 μM NPA treatment.D Frequency distributions of cell numbers in cambial cell files in the stems of WT, PtoRFL30-OE transgenic lines under 0.1 μM NAA treatment and PtoRFL30-RNAi transgenic lines under 5 μM NPA treatment.The peak location of the WT frequency curve is indicated by the red line.E Number of xylem cell layers in stems of WT, PtoRFL30-OE transgenic lines under 0.1 μM NAA treatment, and PtoRFL30-RNAi transgenic lines under 5 μM NPA treatment, corresponding to (A).Error bars show ± standard deviation.Significant differences are indicated by letters above the error bars (one-way ANOVA followed by Tukey's test for pairwise comparisons); n = 12 (C); n = 20 (E).

Figure 7 .
Figure 7. Model for mitochondrial homeostasis mediated by PtoRFL30 during wood formation in poplar.During wood formation, nuclear-encoded protein PtoRFL30 targets mitochondria, and affects mitochondrial functional homeostasis.Furthermore, PtoRFL30-mediated mitochondrial signaling regulates vascular cambial activity by altering auxin accumulation in the secondary vasculature.