Transgenic aspen lines with the introduced recombinant sp-Xeg xyloglucanase gene from the fungus Penicillium canescens under the transcriptional control of the 35S promoter and nopaline synthase terminator were analyzed. The sp-Xeg gene encodes a chimeric xyloglucanase XegA with a white poplar cellulase signal peptide . The in vitro-derived trees were adapted to in vivo conditions in the climatic chambers for a month and were grown in greenhouses for a further month. In total, 25 transgenic lines, two control lines - non-transgenic wild-type line (Pt) and a transgenic line with the inserted gene β-glucuronidase (Gus-1-5a) were studied, respectively.
Each line (genotype) was represented by 50 plants (ramets). The plants were grown in individual plastic containers with a volume of 1 liter (with peat to perlite ratio of 3:1) and after two months of growth in the greenhouse moved to semi-natural conditions in an open air with additional watering and feeding. Semi-natural conditions are the cultivation of potted plants outside of the greenhouse . Such growing conditions are close to the field and allow us to estimate the resistance of plants to various biotic and abiotic factors. The tests in open air conditions are long and time-consuming, therefore, a small sample of lines from the entire panel of the obtained transformants is usually used to study the effect of the introduced recombinant construct during the transgenesis of woody plants [47-49]. Following this common practice used in studies of transgenic woody plants we selected six transgenic lines from 25 previously transformed greenhouse plants for our analysis under semi-natural conditions . The selected lines were renamed, and the old and new names are presented in Table S1 in Supplementary material. The sampled lines included Xeg-1-1a, Xeg-1-1b, Xeg-1-1c, Xeg-2-1b, Xeg-2-3b, and Xeg-2-5a lines. The selection was based on growth rate data (height) obtained for 2-month-old greenhouse plants and presented in Table S2 in Supplemental material. The defective Xeg-2-1c line was excluded, and lines with growth indicators either a significantly higher (Xeg-1-1a, Xeg-2-1b) or similar (Xeg-1-1b, Xeg-1-1c, Xeg-2-3b, Xeg-2-5a) to the control line (Pt) were included.
After four months of vegetation under semi-natural conditions, six transgenic lines and a non-transgenic control line (Pt) were selected for the further analysis. Selected lines were transplanted into plastic containers with a volume of 2 liters. Nontransgenic control aspens (natural variant) were randomized with aspens of transgenic lines and grown together alongside each other under the same conditions. Under these conditions the plants were grown up to the age of one and a half years with wintering in natural conditions.
After two months of growth in the greenhouse and before transfer to semi-natural conditions, the height and number of leaves were measured. After another 4 months of vegetation, growth was measured, samples were taken for molecular analyses and carbohydrate composition measurement, and some of the plants were used to analyze the decomposition rate of the stem wood. The second part wintered and continued to grow for another year, and then, at the age of 18 months, the growth, content of cellulose and pentosans were measured, and samples were taken for carbohydrate composition, microscopy and libriform, and the annual results of the decomposition experiment were evaluated.
The presence of the recombinant gene and protein expression were confirmed in all six transgenic lines. In addition to growth parameters, analyses of specific content of cellulose, pentosans, carbohydrate composition of xylem, measurement of libriform were also performed on all these six transgenic lines and the control line. Electron microscopy was carried out on plants of the Xeg-3 and Xeg-6 lines, which have the greatest differences in the cellulose content compared to the Pt control line. The decomposition rate was measured in the plants of the Xeg-1-1b, Xeg-2-3b and Pt lines. To carry out the decomposition analysis, we selected these two lines that fell into the same group with the control plants according to Duncan's ANOVA-1 rank tests based on measuring biometric indicators obtained for greenhouse plants, and they differed only by the content of pentosans .
Reverse transcription PCR (RT-PCR) analysis
Total plant RNA was extracted from the leaves of all 6-month-old transgenic and control lines growing in the semi-natural conditions by addition of the TRIzol® reagent (Invitrogen, USA) following the manufacturer’s protocol (http://www.invitrogen.com). Genomic DNA was removed from the samples using DNAse I (Fermentas, USA) according to the manufacturer’s protocol. For the cDNA synthesis, 500 ng of total RNA were used. RNA concentration was measured using a microvolume spectrophotometer Nanodrop 2000 (Thermo Scientific, USA). cDNA synthesis was performed using M-MuLV reverse transcriptase, RNaseH- (SibEnzyme, Russia), and oligo-d(T)18 primer (Synthol, Russia) at a temperature of 37°C for 75 minutes. For the revertase inactivation, the mixture was heated to 70°C. We took 2 µl of the cDNA mixture as a template for PCR amplification using Xeg-up (GAAATGGCTAATGCCACTACATT) and Xeg-low (GATTTAGGCAACATCGGCAG) primers (Evrogen, Russia) under conditions similar to the PCR of total DNA . To control RNA contamination by residues of the genomic DNA, PCR was performed with RNA preparations of each of the clones without revertase treatment.
Real-time quantitative PCR (RT-qPCR) analysis
The recombinant gene expression level was assessed by using cDNA preparations obtained with the RT-PCR. An intercalating dye SYBR Green was used as a fluorophore for detecting the reaction products. Amplification was carried out in the ANK 32 apparatus (Syntol, Russia) with the RT-qPCR reaction mixture M-427 (Syntol, Russia). Primers used in the RT-qPCR experiments are listed in Table 7. Amplification conditions included: denaturation at 95°C for 5 min, then 45 cycles of primer alnnealing and elongation at 60°C for 40 sec and denaturation at 95°C for 15 sec. Specificity of amplification was estimated using melting curves. Actin and ubiquitin were used as reference genes. Gene expression levels were calculated according to the 2-ΔΔСt method .
Qualitative Western blot analysis
Total protein extracts were obtained from the leaves of 6-month-old transgenic and control plants ex vitro by the addition of an extraction buffer [0.175 M Tris / HCl (pH 8.8), 5% SDS (w/v), 15% glycerol (v/v), 0.3 M mercaptoethanol] . Protein concentration was determined using a PierceTM BSA protein assay kit (Thermo Scientific, USA): 120 μg of total protein was added to each well, and 2 μg of xyloglucanase was added to the control well. Electrophoretic separation of proteins was performed according to Laemmli  in a 12% polyacrylamide gel. Proteins with molecular weights of 20, 26, 34, 50, 90, and 120 kDa were used as molecular weight markers (Thermo Scientific, Lithuania), and 5 μl of a mixture containg these markers were added to the gel. As a positive control, a xyloglucanase preparation from P. canescens was used after three stages of purification (U / mg tamarind seed xyloglucan = 18) and was kindly provided by the IBPM RAS (Pushchino, Russia). Electrotransfer of the gel-separated polypeptides was carried out on a nitrocellulose membrane (Bio-Rad, Germany) using semi-dry transfer (at 40 mA, 2 hours) on a TE70PWR translotter (Amersham, USA). To amplify the signal, PierceTM Western Blot signal enhancer (Thermo Scientific, USA) was used according to the manufacturer’s protocol. Primary polyclonal rabbit antibodies were obtained at the Institute of Theoretical and Experimental Biophysics of the Russian Academy of Sciences (Pushchino, Russia). Hybridization with primary rabbit (1:200) antibodies was carried out for 18 hours at 8° C. Hybridization with goat anti-rabbit antibodies conjugated with alkaline phosphatase (Sigma, USA) was carried out for 2 hours at room temperature, with dilution according to the protocol. Immunocomplexes were detected using a BCIP/NBT ready-to-use substrate (Serva, Germany).
To study each line, 40 plants were used. The length of the stem was measured from the root neck to the apical bud. Stem diameter was measured at the base of the root neck. The volume (V, cm3) was measured by the formula: V = SD2 H, where SD - stem diameter (cm), H – plant height (cm) . The number of leaves was calculated from the apical bud and to the root neck. The parameters of the leaf blade were measured using the LAMINA software  in the second year of vegetation. Height was measured at the age of 2, 6 and 18 months and analyzed using the Statistica 7.0 software (https://www.tibco.com/products/tibco-statistica).
Analysis of the specific cellulose content
Median internodes of 20 plants per each 18-month-old line were used to determine the content of cellulose by the Kurschner-Hanak nitrogen-alcohol method . The content was recalculated taking into account the weight of an absolutely dry sample.
Analysis of the specific pentosan content
[Please see the supplementary files to view this section.]
Analysis of hemicelluloses monosaccharide
Monomeric sugars of hemicelluloses (arabinose, fucose, galactose, glucose, mannose, rhamnose, and xylose) were measured by standard alditol acetate method [57, 58]. For the analysis of the composition of monosaccharides, two 10 cm long cuts of stem per plant without bark, representing the 1st (6 months) and the 2nd (18 months) year growing wood, respectively, were taken from 18-month-old plants. Samples of 5 mg of wood sawdust were hydrolyzed with 2M trifluoroacetic acid (TFA) at 100° C for 5 hours. The mixture of neutral monosaccharides was converted to alditol acetates and identified by gas chromatography–mass spectrometry (GC-MS) analysis using the GCMSQP 2010 Plus chromatograph (Shimadzu Corporation, Japan) with the HP-5MS column (60 m × 0.32 mm × 0.25 μm). Myo-inositol was used as an internal standard. Helium was used as the carrier gas. The temperature of the injector was 150° C. The column temperature was increased from 60° C to 250° C at a rate of 2° C/min, and, then, held for 10 minutes .
For the microscopy analysis, samples of 18-month-old plants were taken from the lower part of the stem (wood of the second year of cultivation) in three replicates. A total cut of all tissues of the stem was made. To prepare the cross sections, the samples were embedded in the epoxy-resin mixture containing DER-332, DER-732, DDSA, and DMP-30 . The transverse sections were obtained using the ultra microtome Reichert Om U2 (Austria) with glass knives, stained with methylene blue, azur-II, and basic fuchsin  and photographed with the AxioImager M1 light microscope (CarlZeiss, Germany). The slices were scanned, and the thickness of the cell walls was estimated using the AxioVision 4.8.1 software package (CarlZeiss, Germany).
Measurement of the libriform fibers
For the measurement of the libriform fibers, samples of the 18-month-old plants were taken from the bottom of the stem without bark (wood of the second year of cultivation) in three replicates. Samples were macerated using acetic acid and sodium chlorite, and the length and diameter of their libriform fibers were measured .
Measurement of the aspen decomposition rate
The decomposition rate was estimated by analyzing the emission of carbon dioxide during plant material decomposition . In the experiment, sifted through a 0.5 mm fine sieve, washed and sterilized sand was used as a substrate. The plant tissue (stems and roots) from the 6-month-old plants was ground in a porcelain mortar, and, then, dried at 65º C for three days, and 100 mg of this dried and ground tissue were placed in a glass tube with 2 g of sand and sealed with rubber stoppers. To ensure the decomposition in the test tubes, an aqueous extract of the forest plant litter was added. Distilled water was also added to the tubes in an amount of 50% of the total moisture capacity of the sand (taking also into account the water needed to restore the initial mass of plant tissue). Then, the tubes were placed in a thermostat at 22º C for 48 weeks. Samples of air were sampled from the tubes every 8 weeks, and their carbon dioxide gas was analyzed using a gas chromatograph Crystallux 4000M (Research and Production Company «Meta-chrom», Yoshkar-Ola, Russia). All samples were analyzed also for C and N content by gas chromatography using the Euro EA-CHNSO Elemental Analyser (HEKAtech GmbH, Wegberg, Germany).