Changes in the Activity of the CLE41/PXY/WOX Signaling Pathway in the Birch Cambial Zone under Different Xylogenesis Patterns

The balance between cell proliferation and differentiation into other cell types is crucial for meristem indeterminacy, and both growth aspects are under genetic control. The peptide-receptor signaling module regulates the activity of the cambial stem cells and the differentiation of their derivatives, along with cytokinins and auxin. We identified the genes encoding the signaling module CLE41-PXY and the regulator of vascular cambium division WOX4 and studied their expression during the period of cambial growth in the radial row: the conducting phloem/cambial zone and the differentiating xylem in two forms of Betula pendula, silver birch and Karelian birch. We have shown that the expression maximum of the BpCLE41/44a gene precedes the expression maximum of the BpPXY gene. Non-figured Karelian birch plants with straight-grained wood are characterized by a more intensive growth and the high expression of CLE41/44-PXY-WOX4. Figured Karelian birch plants, where the disturbed ratio and spatial orientation of structural elements characterizes the wood, have high levels of BpWOX4 expression and a decrease in xylem growth as well as the formation of xylem with a lower vessel density. The mutual influences of CLE41-PXY signaling and auxin signaling on WOX4 gene activity and the proliferation of cambium stem cells are discussed.


Introduction
Radial or secondary growth promotes the formation of vascular and mechanical stem tissue. The vascular tissues ensure the transport of water, minerals, photoassimilates, and signal molecules over long distances, while the mechanical tissues ensure the maintenance of the tree in an upright state, its resistance to wind, storms, etc. Radial growth is essential for the growth of woody plants and also plays a significant role in the long-term conservation of carbon in terrestrial biomass. The whole variety of structural elements of stem tissues comprises the lateral meristem-the cambium. Vascular cambium generates the initial vascular cells and their immediate derivatives, called mother cells [1]. Although the mother amplifying cells are anatomically indistinguishable from cambial stem cells, only stem cells retain the ability to generate both secondary phloem and secondary xylem elements. These mother amplifying cells have stably gained the cell fate of one of these tissues; they may differentiate directly into different tissues or may divide several times before terminal differentiation [1][2][3][4].
Data on the role of TDIF and their receptors (TDR) in regulating the secondary growth in woody plants are thus far scant. Such studies are not only of purely academic but also of applied value. The co-overexpression of CLE41 and PXY genes in hybrid aspen under tissue-specific promoters corresponding to the original expression domains of both genes increases wood formation [13]. Other experiments have shown that WOX4 plays a major role in controlling the cell identity and division activity in the vascular cambium of hybrid aspen [23]. The downregulation of WOX4 homologs by RNA interference in hybrid aspen causes more dramatic phenotypic changes than are observed in annual species; in the most extreme cases, the resulting reductions in cambial activity and wood formation are severe enough to prevent the trees from remaining upright [2].
Most information on the roles of CLE41/44 and PXY genes comes from studies involving their overexpression in plants or peptide treatment assays [2]. The objects of the present study were two forms of silver birch differing in wood figure: silver birch (B. pendula var. pendula), which forms a typical straight-grained wood, and a form of the silver birch, Karelian birch (B. pendula var. carelica (Merckl.) Hämet-Ahti) with figured wood. When the figured wood of Karelian birch is formed, the development program of vascular cambium cell changes: the program of cell death, leading to the formation of vessels and fibers of the xylem and sieve elements of the phloem, does not start, while the differentiating cambial derivatives preserve the protoplast and turn into storage parenchyma cells, which accumulate large amounts of storage substances [24][25][26]. The unique fact that different scenarios of xylogenesis can be studied within the same trunk makes Karelian birch a unique object for studying the mechanisms of wood formation [25,27,28].
In our previous work, we identified the genes encoding the main enzymes for auxin biosynthesis, transport, and conjugation and studied their expression models in various xylogenesis scenarios in the silver birch. We have shown that the auxin-deficient phenotype in Karelian birch trunk tissues is formed against the background of the overexpression of genes involved in auxin biosynthesis (Yucca), polar auxin transport (PIN), and the conjugation of auxin with amino acids (GH3) and UDP-glucose (UGT84B1) [26,29]. Since the TDIF/TDR-signaling influences the fate of the cambial initials and is closely related to auxin-signaling, in the present study, we focused on investigating the CLE41/44-PXY-WOX4 gene expression in the cambial zone and differentiating xylem in two forms of B. pendula: silver birch and Karelian birch.

Results
The study objects were two forms of silver birch: B. pendula var. pendula, the form of silver birch with a straight-grained wood (hereafter Bp trees), and another silver birch variety, B. pendula var. carelica-Karelian birch. We selected Karelian birch trees possessing a highly figured wood (figured B. pendula var. carelica trees, future Bc FT trees) and nonfigured plants that had a typical straight-grained wood with a weakly expressed figure (non-figured B. pendula var. carelica trees, hereafter Bc NF trees). Previously, we showed that the studied trees differed in the structure of mature xylem. The xylem in Bp trees displayed a typical for the species ratio of structural elements [30]. The xylem in Bc FT was characterized by low vessel density, a high proportion of parenchyma, and fibrous tracheids with thick cell walls. The straight-grained wood of the Bc NF trees was characterized by a high density of vessels [30].

Description of 1 and 2 Fractions in Three Phenotypes at Different Dates of Sampling
On 28 May, fraction 1 (from the bark side) in all three groups of trees included the cambial zone, early conducting phloem (part of the conducting phloem developed in the beginning of the growing season), and a small portion of non-conducting phloem (Figure 1a-c). On 11 June, fraction 1 of the figured Karelian birch species also contained three tissues: the cambial zone, the early conductive phloem, and a small portion of the non-conductive phloem ( Figure 1f). In the silver birch and non-figured Karelian birch trees, on this date, fraction 1 contained several layers of differentiating xylem closest to the cambium, aside from the three tissues above-mentioned (Figure 1d,e). The reason for the different composition of tissues in fraction 1 was the more active division of the cambium toward the xylem in two groups of trees that formed normal wood. On 26 June, fraction 1 included five tissues: several layers of the differentiating xylem closest to the cambium, the cambial zone, early conducting phloem, differentiating late conducting phloem (part of the conducting phloem developed in the second half of the growing season), and a small portion of non-conducting phloem in all three groups of trees (Figure 1g     We failed to count the number of cells in the cambial zone of the samples collected on May 28 because thin-walled and highly hydrated cells of the cambium ruptured during sampling. On 11 June, the number of cells in the cambial zone did not differ in all three studied phenotypes. Two weeks later (25 June), the number of cells of the cambial zone in the Bp and Bc NF trees decreased, especially in the Bc NF trees, while the Bc FT trees remained at the same level (Table 1).  Bp trees 8.6 ± 0.3 7.9 ± 0.5 2.9 ± 0.1 2.9 ± 0.1 3.
As we had already observed a significant increase in the conductive phloem width in the Bp trees on 28 May, by 11 June, its value did not change and increased by 26 µm by the end of June ( Figure 3). The number of differentiating phloem cells at the end of May and the middle of June was the same and increased only at the end of June. The number of sieve tubes increased by 11 June and remained at this constant level until the end of June (Table 1).
From 28 May to 11 June in the Bc NF trees, the width of the conducting phloem increased by 30 µm, and from 11 June to 25 June by 12 µm. The conducting phloem was wider in the Bc NF trees compared to the Bp trees only on 11 June (Figure 3). The number of differentiating phloem cells increased from 28 May to 25 June. Their number in the Bc NF trees was less than that of the Bp trees in May, and did not differ from the Bp trees on 11 June. The number of sieve tubes increased by 11 June and remained constant until 25 June, like in the Bp trees ( Table 1).
The features of the Bc FT trees were the smaller width of the conducting phloem ( Figure 3) and the smaller number of sieve tubes (Table 1)

CLE41/44 Gene Identification in the Silver Birch Genome
Genome-wide analysis followed by manual validation and the removal of false positives allowed us to identify two B. pendula genes encoding putative CLE41/44 proteins (Table 2). One gene was on chromosome 1, and the second on contig 443. Both genes contained one exon and had no introns ( Figure 4).  We found that the number of BpCLE41/44b transcripts was close to zero or not detected in all samples studied. Similar data were obtained by J. Alonso-Serra et al. [31] when studying the molecular fingerprints of eight main tissue types in the stem of B. pendula: phellem, combined phellogen and phelloderm, nonconductive secondary phloem, conductive phloem, cambium, developing xylem, xylem, and last year's xylem tissue. The analysis of the normalized transcript abundances (TPM) for each tissue showed scanty amounts of BpCLE41/44b (Bpev01.c0443.g0012.m0001, Table S2 from [31]) in the conductive phloem and cambium. Only the BpCLE41/44a data will be discussed in this paper.

PXY Gene Identification in the Silver Birch Genome
Seven genes encoding putative receptor such as kinases homologous to TDR/PXY of other plant species (Arabidopsis thaliana and Populus trichocarpa) were identified in the B. pendula genome ( Table 3). The genes were on four different chromosomes and three contigs. All genes contained two exons, except for Bpev01.c1696.g0001, which contained three exons ( Figure 5).  The phylogenetic analysis revealed that gene Bpev01.c0668.g0016 encoded the protein, which was highly similar to PXY ( Figure 6). The sequence identity (in percentage of the total number of residues) for the A. thaliana PXY and B. pendula PXY protein pair was 60.5%.

WOX Gene Identification in the Silver Birch Genome
Eleven genes encoding putative homologous to WUSCHEL-related homeobox (WOX) transcription factors of other plant species (Arabidopsis thaliana and Populus trichocarpa) were identified in the B. pendula genome ( Table 4). The genes were on seven different chromosomes and contained two to four exons (Figure 7).   Phylogenetic analysis performed on protein sequences revealed that the gene Bpev01. c0640.g0004 encoded the protein, which was highly similar to WOX4 (Figure 8). The sequence identity (in percentage of the total number of residues) for the A. thaliana WOX4 and B. pendula WOX4 protein pair was 55.6%.

Expression of the Genes Encoding CLE-PXY-WOX Signaling Pathway in the Trunk Tissues of Bp Trees during Cambial Growth
We have previously shown that WOX4 is a cambial stem cell pattern, and its activity correlates with the proliferation process [23,32]. Therefore, we used the expression level of BpWOX4 in the studied fractions to estimate cell proliferation. We found that the number of BpWOX4 transcripts in the Bp trees did not differ significantly between Fractions 1 and 2. The expression level of BpWOX4 was higher in the differentiating xylem on 28 May and 25 June, while on 11 June, it was higher in Fraction 1 in contrast (Figure 9a). We assumed that cell proliferation was observed in Fractions 1 and 2 on all sampling dates. The WOX4 homeodomain transcription factors are downstream targets of the TDIF/ CLE41-PXY/TDR signaling pathway [9,32]. We showed that the maximum expression of BpCLE41/44a in the Bp trees was in tissues sampled from the bark side on 28 May, and its value was 10 times higher than that from the xylem side (Figure 9d). Previously, J. Alonso-Serra et al. [31] found the TPM maximum for BpCLE41/44a in non-conductive secondary phloem, conductive phloem, and cambium (Bpev01.c0016.g0065, Table S2 after [31]). At the middle and end of June, the number of BpCLE41/44a transcripts did not differ between Fractions 1 and 2. On 11 June, the expression level of BpCLE41/44a was two times higher than on June 25 (Figure 9d).
We showed that the number of BpPXY transcripts from the side of differentiating xylem compared to the tissue layer including the cambial zone and vascular phloem was higher in all trees studied. We observed the maximum at the middle and end of June (Figure 9g).

The Formation of Straight-Grained Wood of Bc NF Trees Occurs against the Background of Greater Expression of the Genes CLE-PXY-WOX in the Cambial Zone and Differentiating Xylem Compared to the Bp Trees
The maximum expression level of the genes CLE-PXY-WOX in the Bc NF trees was at the same tissues and time as in the Bp trees: for BpCLE41/44a, it was the date of 28 May, Fraction 1 ( Figure 9e); for BpPXY, it was 11 June, Fraction 2 ( Figure 9h); and for BpWOX4, 28 May, Fraction 2 ( Figure 9b). We found that in the Bc NF trees, the expression level of BpCLE41/44a on 28 May on the bark side was five times greater compared to the Bp trees (Figure 9e), and the expression level of BpPXY (11 June) on the differentiating xylem side was three times greater (Figure 9h). The proliferation of cells (28 May) in Fraction 2 in the Bc NF trees was approximately two times higher than that in the Bp trees. The dates of 11 June and 25 June were characterized by the fact that the number of BpWOX4 transcripts decreased and did not differ between the two fractions ( Figure 9b).

Decreased Expression of the Genes CLE-PXY When Forming the Wood Figure in Bc FT Trees
In the Bc FT trees, the number of BpCLE41/44a and BpPXY transcripts was lower compared to the other forms. Similar to the Bp and Bc NF trees, the maximum expression of BpCLE41/44a was observed in Fraction 1 ( (Table 1).
Another specific feature of the Bc FT trees is the maximum cell proliferation from the xylem side (11 June), while the expression level of BpWOX4 significantly exceeded those of the Bp and Bc NF trees (Figure 9c).

Discussion
Plant biomass is a huge renewable resource of biofuels and biomaterials, and the major constituent of plant biomass is xylem derived from the vascular meristem [33]. Therefore, much attention has recently focused on the study of the molecular mechanisms of xylogenesis. In studies performed on Arabidopsis as the model plant, it has been shown many times that the maximum expression of CLE41/44 is observed in the phloem tissue and neighboring cells [7,9,16], while PXY and WOX4 are in part of the xylem-facing cambium [3,4]. We have shown that during the period of cambial growth in silver birch, the expression of the gene CLE-PXY-WOX patterns were extended in a radial row «the conductive phloem/the cambial zone (Fraction 1)-the differentiating xylem (Fraction 2)» with the maximum for each gene in that or another faction. The expression of the genes changed during cambial growth (28 May-25 June).
We found that the highest expression of the gene BpCLE41/44a was observed on 28 May, then decreased in all of the studied birch phenotypes (Bp, Bc NF, and Bc FT trees). In contrast, the expression levels of the gene BpPXY were the highest on 11 June. Previously, it has been shown many times in the literature that the spatial separation of the ligand (TDIF) and receptor (TDR) is necessary for the correct spatial orientation of the vascular pattern and the restriction of the stem cell zone [33]. We have shown that, together with spatial separation (Fraction 1 for BpCLE41/44a and Fraction 2 for BpPXY), we also observed temporal separation (BpCLE41/44a and BpPXY maximum at different stages of cambial growth).
In the Bp and Bc NF trees, forming straight-grained wood, the maximum expression of the gene BpWOX4 matched in time to BpCLE41/44a. Unlike BpCLE41/44a, whose peak activity was higher in Fraction 1, the maximum number of BpWOX4 transcripts was observed in Fraction 2. We have shown that the Bc NF trees, compared to Bp trees, are characterized by (1) greater expression of the genes BpCLE41/44a, BpPXY, BpWOX4; and (2) the most active divisions of the cambium toward the xylem. In our previous studies, we characterized the phenotype of the Bc NF trees as faster growing than the Bp trees. The straight-grained wood of the Bc NF trees was characterized by a high density of vessels [30].
We found that in the abnormal areas of the Karelian birch trunk, the maximum expression of the gene BpWOX4 did not coincide in time with the maximum of BpCLE41/44a and was observed on 11 June. The number of BpWOX4 transcripts significantly outnumbered those in the Bp and Bc NF trees. The expression level of the genes CLE41/44-PXY running WOX4 expression [9,32] in the Bc FT trees was significantly lower compared to the Bp and Bc NF trees.
We know that WOX4 is considered as a central regulator of vascular cambium division because it activates a cambium-specific transcriptional network and integrates TDIF/TDR signaling and auxin signaling for cambium division [9,20,[34][35][36]. If TDIF/TDR signaling stimulates the WOX4 transcription and promotes the cambium proliferation in stems [9], then auxin signaling attenuates the activity of the stem cell-promoting WOX4 gene, and cell-autonomously restricts the number of stem cells in stems [34].
The TDIF/TDR signaling, aside from stimulating WOX4 transcription and promoting the cambium proliferation in stems [9], leads to the inactivation of the BES1 [37,38] and BIL1 [37,39] transcription factors. The BES1 (BRI1 (BRASSINOSTEROID INSENSITIVE)-EMSSUPPRESSOR) 1) transcription factor promotes xylem identity and also suppresses the expression of WOX4. The interaction between TDR and GSK3 (Glycogen Synthase Kinase 3) proteins (BIN2) is thought to result in the negative regulation of the BES1 by phosphorylation [37]. The BIL1 (BIN2 (BRASSINOSTEROID INSENSITIVE 2)-LIKE 1) refers to GSK3-proteins. The BIL1 links the TDIF/TDR module with auxin-cytokinin signaling during secondary growth. The BIL1 functions as a negative regulator of cambial activity downstream of the TDR by targeting the ARF5 (auxin response factor) protein [37]. Phosphorylation of ARF5 by BIL1 weakens its binding to the repressor Aux/IAA (Aux/INDOLE-3-ACETIC ACID) and enhances the negative effect ARF5 on the activity of vascular cambial, which upregulates ARR7 (ARABIDOPSIS RESPONSE REGULATOR 7) and ARR15 (two negative regulators of cytokine signaling). The BIL1 activity is inhibited by TDR, attenuating the effect of ARF5 on the expressions of ARR7 and ARR15 and increasing the vascular cambial activities [36,39,40]. High concentrations of auxin promote ubiquitin-mediated proteolysis of Aux/IAA transcriptional repressors, releasing ARF5 to regulate xylem differentiation. ARF5 autonomously limits the number of stem cells through the weakening of WOX4 activity and increases the expression of the HB8 gene, positively affecting xylem differentiation [2,41] (Figure 10).
From this, it follows that (1) the presence of a high expression of CLE41/44-PXY and/or (2) reduced auxin signaling is necessary to maintain a high expression of the gene BpWOX4. We have previously investigated the expression level of BpARF5 and its direct target, BpHB8, in the trunk tissues of different forms of silver birch [30]. We found that the level of BpARF5 expression in the differentiating xylem of Karelian birch plants with both figured and non-figured wood was lower than in the Bp trees. The formation of the figured wood of the Bc FT trees was associated with the low expression of BpHB8, whereas in the Bc NF trees with a straight-grained wood, the expression of BpHB8 was indistinguishable from that of the Bp trees [30]. Collectively, the data obtained through this study and previous ones allow us to propose that the maximum expression of the gene BpWOX4 in Bc FT trees on 11 June was due to (1) continuing TDIF/TDR signaling and (2) reduced auxin signaling. This suggestion is consistent with the anatomical evidence of lower xylem increases in the Bc FT trees compared to the Bp and Bc NF trees. Figure 10. The schematic representation of the hormonal and molecular genetic regulation of cambium activity in woody plants. The TDIF signal peptide is produced in the phloem and migrates to the cambium, where it binds to its TDR receptor. The TDIF/TDR signaling activates WOX4 by regulating cambium cell proliferation [4,12,23]. TDIF binding to TDR activates GSK3 (BIN2) and leads to the phosphorylation and degradation of the BES1 protein. BES1 promotes xylem differentiation and also suppresses WOX4 expression, so its degradation preserves cambium pluripotency [37]. The TRD acts as a suppressor of another GSK3-BIN2-LIKE 1 (BIL1). BIL1 in the absence of TDR phosphorylates the auxin-dependent transcription factor of ARF5, releasing it by weakening the interaction of ARF5 with the AUX/IAA repressor [40]. Phosphorylation of ARF5 enhances its activity in suppressing the cytokine response through the expression of ARR7/15 and leads to a decrease in cambium proliferation [39]. Auxin signal transduction in cambial stem cells is required to maintain cambial activity [20]. The level of auxin signal transmission increases in the differentiating xylem. High concentrations of auxin promote the ubiquitin-mediated proteolysis of Aux/IAA transcriptional repressors, releasing ARF5 to regulate xylem differentiation. ARF5 autonomously limits the number of stem cells by downregulating the WOX4 activity and increasing the HB8 gene expression, positively influencing xylem differentiation [2].

Study Objects
The study objects were two forms of silver birch: Betula pendula Roth var. pendula-the form of silver birch with straight-grained wood, and B. pendula var. carelica-Karelian birch. All plants grew in the same soil and climatic conditions on the experimental plot of the KarRC RAS near Petrozavodsk city. The trees were 14-15 years old. The Karelian birch plants were grown from seeds obtained from the controlled pollination of Karelian birch plus trees. The number of trees in each group was nine. The samples were collected during the period of cambial growth (28 May 2020, 11 June 2020, and 26 June 2020).

Plant Sampling
The trunk tissue samples were taken at breast height (1.3 m above ground level). The sections of the trunk with the highest degree of manifestation of structural anomalies were selected for the sampling of plant material from the Karelian birch trunks. For microscopic analysis blocks including the phloem, the cambial zone and the last 2-3 annual increments of wood were cut out (5 × 5 × 3 mm, length x width x height). For molecular genetic analysis, «windows» of 6 × 8 cm were cut out of the trunk and the bark was separated from the wood (Figure 11). During the period of cambial growth, the bark moves away from the wood along the expanding xylem zone. Tissue complexes (hereafter Fraction 1) were prepared from the inner surface of the bark. The layers of tissue (hereafter Fraction 2) were scraped off the exposed wood surface with a blade. The sampling of stem tissues was monitored under a light microscope (Figure 1). The material was frozen in liquid nitrogen and stored at −80 • C.

Microscopy
Three samples of phloem and xylem tissues from each tree were analyzed. Sample preparation for microscopy was conducted as described previously [26]. The samples were fixed in glutaraldehyde, dehydrated in a series of alcohols of rising concentrations, and embedded in an Araldite-Embed-812 mixture following the published technique [42]. Cross sections 2 µm thick were cut with an Ultrotome IV (LKB, Bromma, Sweden) and stained with a 1% aqueous solution of safranin. Permanent slides were made using synthetic mounting medium Vitrogel (BioVitrum, St. Petersburg, Russia). Microscopic analysis was carried out under an AxioImager A1 light microscope (Carl Zeiss, Jena, Germany) equipped with an ADF PRO03 camera. Images were processed with ADF Image Capture software (ADF Optics Ningbo, China). Anatomical measurements were made following the available guidelines using panoramic cross sections with an area of 7-10 mm 2 [43][44][45].

Gene Retrieval from the Silver Birch Genome by Bioinformatics Methods
The search for the CLE41/44, PXY, and WOX4 genes was carried out using the published genome of Betula pendula Roth [46]. To this end, the CDS of Arabidopsis thaliana and Populus trichocarpa CLE41/44, the PXY and WOX4 genes, and the amino acid sequences of the corresponding proteins were obtained from The Arabidopsis Information Resource (TAIR) database (release 13, https://www.arabidopsis.org, accessed on 20 April 2020) and the Phytozome database (http://www.phytozome.net/poplar, release v3.0, accessed on 20 April 2020), respectively. The resulting sequences were then used as a BLAST search query across the genome of B. pendula var. pendula (release 1.2, https://genomevolution.org/coge, accessed on 20 April 2020) to identify homologous sequences.
The structures of candidate proteins were predicted using the National Center for Biotechnology Information (NCBI) resource (http://www.ncbi.nlm.nih.gov/Structure/ cdd/cdd.shtml, accessed on 20 April 2020) [47]. The prediction of protein subcellular localization was performed using DeepLoc 2.0 [48]. Phylogenetic analysis was carried out using MEGA X software [49]. Multiple sequence alignment of the protein sequences was performed using ClustalW. Phylogenetic trees were constructed using the neighbor-joining method based on the Poisson correction model with 1000 bootstrap replicates [50][51][52]. The different genes in B. pendula were named according to the Phylogenetic analysis. The percent identity of B. pendula and A. thaliana proteins was determined using the EMBOSS Needle online tool (https://www.ebi.ac.uk/Tools/psa/emboss_needle/, accessed on 20 April 2020). The gene structures were analyzed using the Gene Structure Display Server (//gsds.gao-lab.org/, accessed on 20 April 2020) [53].

qRT-PCR
The isolation of total RNA was performed using an extraction CTAB buffer (pH 4.8-5.0): 100 mM Tris-HCl (pH 8.0), 25 mM EDTA, 2M NaCl, 2% CTAB, 2% PVP, and 2% mercaptethanol was added to the mixture before use. Plant tissue lysates were additionally treated with DNase and RNase inhibitors (Syntol, Moscow, Russia). Separation of the aqueous and organic phases was conducted using a mixture of chloroform-isoamyl alcohol (24:1). RNA was precipitated using 25 mM LiCl, then re-precipitation was carried out using an extraction SDS buffer: 1 M NaCl, 0.5% SDS, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA [54]. RNA was re-precipitated with absolute isopropanol. The quality and quantity of the isolated RNA and synthesized cDNA were checked spectrophotometrically and by electrophoresis in 1% agarose gel.
Specific primers (Sintol, Moscow, Russia) for amplification of the studied genes were designed using the software Beacon Designer 8.21 (PREMIER Biosoft, San Francisco, CA, USA) ( Table 5). As a reference gene for the normalization of quantitative PCR data, we used genes of the Ef1a and GAPDH family [55]. In previous studies, we evaluated the suitability of five genes-GAPDH1, Actin1, Ef1a(1), Ef1a(2), and 18SrRNA-in two forms of silver birch for use as a reference when staging qRT-PCR based on the stability of their expression. The results of studying the level of expression of the potential reference genes were analyzed using the NormFinder and BestKeeper programs. We showed that among the studied genes, the most stably expressed in the B. pendula samples (leaves and trunk tissues) were Actin1, GAPDH1, and Ef1a(1), while the GAPDH1 and Ef1a(1) genes were suggested by NormFinder as the best combination of two reference genes and had the lowest stability index [55]. Families of genes coding for (1) peptide ligand TDIF/CLE41 (BpCLE41/44a and BpCLE41/44b); (2) the TDR/PXY receptor (BpPXY); and (3) WUSCHEL-RELATED HOMEOBOX4 (BpWOX4) involved in the regulation of cambium cell proliferation as the identified targets of the TDIF/TDR signaling pathway were studied.
Amplification of the samples was performed using an iCycler instrument with an iQ5 optical module (Bio-Rad, Hercules CA, USA) and an amplification kit with an intercalating dye SYBR Green (Evrogen, Moscow, Russia). The specificity of the amplification products was checked by melting the PCR fragments. The relative quantity of gene transcripts (RQ) was calculated from the formula: where ∆Ct is the difference in the threshold cycle values for the reference and target genes, and E is the effectiveness of PCR. To determine the efficiency (E), PCR amplification was performed with each pair of primers in a series of 10-fold dilutions (10 −1 , 10 −2 , 10 −3 , 10 −4 , and 10 −5 ) of cDNA. Using Excel software, a plot of Ct versus Lg (conc. cDNA) was plotted, and using the values of the slope of the curve (slope, k), the efficiency was calculated using the formula [56]: E = (10 −1/K − 1) × 100.
The level of expression of specific genes was expressed in relative units (arbitrary units).

Statistical Data Processing
The results were statistically processed with PAST (version 4.0). Before starting the statistical analysis, raw data were initially tested for normality using the Shapiro-Wilk test. The significance of differences between variants was estimated by the Mann-Whitney U-test. Different letters indicate a significant difference at p < 0.05. All data in the diagrams appear as mean ± SE, where SE is the standard error. Sample sizes are denoted as n.
The research was carried out using the equipment of the Core Facility of the Karelian Research Center of the Russian Academy of Sciences.

Conclusions
The distribution of the expression of the genes CLE-PXY-WOX was studied during the period of cambial growth in the radial row: the conducting phloem/cambial zone and the differentiating xylem under different scenarios of xylogenesis using two forms of silver birch as an example. We showed that the expression maximum of the CLE41/44 and PXY genes are separated not only in space (conductive phloem/cambial zone (Fraction 1)-differentiating xylem (Fraction 2)), but also in time. The expression maximum of the BpCLE41/44a gene precedes the expression maximum of the BpPXY gene. TDIF/TDR signaling induces BpWOX4 gene expression and stem cell proliferation in the cambium. Thus, the phenotype of the Bc NF trees with high expression of CLE41/44-PXY-WOX4 is characterized by more intensive growth. Auxin signaling against the background of a decrease in the expression of the CLE41/44 genes leads to a decrease in the expression of the BpWOX4 genes, and, accordingly, the proliferation of cambial initials. The auxin-deficient phenotypes of Bc FT trees showed high levels of BpWOX4 expression and a decrease in xylem growth as well as the formation of xylem with a lower vessel density. A significant decrease in free (physiologically active) auxin because of its conjugation with UDP-glucose and amino acids described previously [26,29] may lead to more dramatic scenarios in the Bc FT trees in the zones of development of structural abnormalities; the differentiating cambial derivatives preserve the protoplast and turn into the storage parenchyma cells.

Conflicts of Interest:
The authors declare no conflict of interest.