3.1. Constructs for HvUGPase and HvSPP expression in mesophyll cells
The pB-Chrbc vector was prepared by replacing the 35S CaMV promoter and nopaline synthase terminator in the pBI121 binary plasmid with the promoter and terminator derived from the Chrysanthemum morifolium ribulose-1,5-bisphosphate carboxylase small-subunit gene, GenBank acc. nr. AY163904 [44, 55]. The promoter and terminator corresponding to the AY163904 regions 1-1013 bp and 1719–2662 bp, respectively, were used in the pB-Chrbc vector. The regions used in the genetic constructs also contained UTRs and the first three codons in the case of promoter (Fig. 2a). The proximal and terminal parts of the original promoter and terminator sequences were modified by adding restriction sites allowing the insertion of a gene of interest and named Chrbc-P and Chrbc-T, respectively (Fig. 2b, c) (the sequences of Chrbc-P and Chrbc-T are provided in the Supplementary Materials). The numerous highly conserved regions from rubisco small subunit promoters and cis-acting regulatory elements responsible for efficient expression, enhancement of transcriptional activity and response to light are present in the Chrbc-P promoter, which suggests that the region is functional (Fig. 2b and Supplementary Table S2). The 72–98% conservation between the proximal part of the Chrbc-P promoter and rubisco small subunit promoters from the Solanaceae family indicates that this promoter may be active in N. tabacum (Fig. 2b and Supplementary Fig. S1). In the case of the Chrbc-T terminator, its functional activity is supported by the presence of eight polyadenylation signals organized into two clusters (Fig. 2c and Supplementary Table S3). The synthetic DNA fragments harboring the Chrbc-P and Chrbc-T sequences were digested with restriction enzymes and used to prepare a cassette that was subsequently inserted between the EcoRI and HindIII sites of pBI121, yielding a binary plasmid named pB-Chrbc (Fig. 3a).
The cDNAs encoding UGPase and SPP, which were used herein as transgenes, were cloned from H. vulgare, and their sequences were deposited in the Genbank repository under accession numbers MN584934 and MN584936, respectively. The HvUGPase MN584934 cDNA encodes a protein of 473 amino acids, with a calculated molecular mass of 51.6 kDa and pI of 5.41. The HvUGPase MN584934 protein differs from HvUGPase CAA62689 by one amino acid, Gly-160, which is substituted as Glu-160 and lies outside of the following regions of functional importance: N-glycosylation motif, nucleotide-binding loop, 14-3-3-protein binding site, and motifs involved in phosphorylation [56]. The conserved amino acids involved in maintaining UGPase catalytic activity and substrate binding, such as Trp-191, Trp-302, Lys-260, Lys-326, and Lys-364, are preserved in the MN584934 protein [28, 57–60]. Phylogenetic analysis revealed that HvUGPase MN584934 is found in a cluster of monocotyledonous UGPases along with H. vulgare UGPase CAA62689 and Saccharum hybrid UGPase AIB06871, and the structure and enzymatic activity of these genes have been well characterized (Fig. 4) [24].
The 16-1486 bp region of HvUGPase MN584934 was used as the transgene in the pB-Chrbc-UGPase construct for HvUGPase overexpression in N. tabacum (Fig. 3b). The complete sequence of the construct is submitted in the Supplementary Materials as pB-HvUGPase-construct.txt.
The H. vulgare SPP cDNA MN584936 used in this study encodes a protein of 422 amino acids with a calculated molecular mass of 47.22 kDa and pI of 5.7. The MN584936-encoded protein shares 100% identity with the H. vulgare SPP protein encoded by AK250551 cDNA, 99% identity with another H. vulgare SPP protein, AAO33159, and 82% identity with Zea mays SPP AF283564, which has experimentally confirmed enzymatic activity [13, 14] (Fig. 5).
All the amino acids differentiating the proteins HvSPP MN584936 (used in this study) and ZmSPP AF283564 are located outside of conserved motifs I – III representing active sites in the N-terminal HAD phosphatase domain [37]. The 52-1455 bp region of HvSPP MN584936 cDNA was used to prepare the pB-Chrbc-SPP construct for HvSPP overexpression in N. tabacum (Fig. 3c). The complete sequence of the construct is provided in the Supplementary Materials as pB-HvSPP-construct.txt.
In addition, a reference construct, pB-Ref, containing a cassette with an unexpressed GUS gene was prepared. The pB-Ref construct is based on the pBI121 vector in which the 35S CaMV promoter was replaced with a 586 bp fragment of the Sorghum bicolor intergenic region (see pB-Ref-construct in the Supplementary Materials). The plants transformed with this construct did not show GUS expression after GUS staining and RT–PCR.
3.2. Transgenic N. tabacum plants expressing HvUGPase and HvSPP
Transgenic N. tabacum plants with pB-Chrbc-UGPase and pB-Chrbc-SPP expression constructs as well as with the reference construct pB-Ref were generated by Agrobacterium-mediated transformation of leaf explants. When the plants grown in vitro on selective media had 4–6 leaves, the activity of the expression constructs was analyzed using standard RT–PCR with the transgene-specific primers and RNA from the leaf as the template. Next, the HvUGPase and HvSPP F0 plants that had transgene expression were rooted and transferred to the soil where they were grown for seeds without selection. At 69 days after transferring the plants to pots, the presence of the transgene transcripts was confirmed using standard RT–PCR, and RNA from leaves that formed after potting the plant was used as a template.
In the case of the F0 reference plants, RT–PCR showed no transgene expression despite the detection of the GUS gene in genomic DNA by PCR. The reference F0 plants grown in soil were tested using PCR with genomic DNA from leaves that formed after potting as a template.
The seeds of self-pollinated F0 transgenic plants were used for growing F1 plants, which were subsequently analyzed with PCR and RT–PCR for the presence of the cassette with the transgene in genomic DNA and its expression activity. The plant material, young leaves and whole roots, was collected 7 weeks after sowing when the plants had 4–6 leaves and the leaves were 6–10 cm in length and still under development.
The analyses revealed that the F1 progeny of plant #24, transformed with the pB-Chrbc-UGPase construct, showed the presence of the transgene in its genomic DNA and transgene expression (Figs. 6 and 7). The same procedure allowed to identify two lineages of plants transformed with the pBChrbc-SPP construct whose F1 progeny possessed and expressed the HvSPP transgene: #22 and #23 (Figs. 6 and 7). Additionally, two lineages of the reference plants, #34 and #35, were identified and applied in further experiments. The F1 progeny of reference plants (in the figures marked as Ref plants) #34 and #35 possess the transgene for GUS integrated into their genome, but its expression was not observed (Figs. 6 and 7).
Two independent sets of F1 plants were analyzed: 1 - young plants grown for 7 weeks after sowing from which young, still developing leaves and whole roots were collected to analyze the promoter specificity, including plants 22 − 11, 22 − 12, 23 − 11, 23 − 12, and 24 − 11, 24 − 12; and 2 - mature plants grown up to 69 days after potting used for analysis of the energy value and cell wall composition, including plants: 22 − 1, 22 − 2, 23 − 1, 23 − 2, and 24 − 1, 24 − 2.
3.3 F1 transgenic plants of the same lineage show the same or a similar number of transgene copies in the genome
The number of transgene copies integrated into the genome of F1 transgenic plants was analyzed with digital droplet PCR (ddPCR). To make this analysis independent of the measurement of absorbance or fluorescence of DNA, the NtL25 gene fragment flanked by the primers Mz116 and Mz117 was used as the standard for genomic DNA. The results showed that F1 transgenic plants from the same lineage had the same or a very similar number of transgene copies in the genome (Table 1 and Supplementary Fig. S2). However, plants from different lineages showed large differences, up to one order of magnitude, in the number of copies of the transgene per genome (Table 1 and Supplementary Fig. S2).
Table 1
Number of HvUGPase and HvSPP transgene copies in the genomes of transgenic plants, average values from two ddPCR experiments. F1 plants of line 22 show a great variability in the number of transgene copies per genome, presumable due to the multiple copies of the transgene integrated into the genome of the F0 plant. The ddPCR with genomic DNA from reference plants as template and primers targeting HvUGPase or HvSPP did not produce any product (positive droplets).
Lineage
|
Plant
|
Transgene/1 copy of NtL25 gene
|
Copies of transgene per genome assuming 3 or 4 copies of NtL25 gene
|
|
Plant
|
Transgene/1 copy of NtL25 gene
|
Copies of transgene per genome assuming 3 or 4 copies of NtL25 gene
|
a
|
b
|
c
|
d
|
|
e
|
f
|
g
|
|
Promoter specificity analysis / young plants
|
|
Energy value analysis / mature plants
|
Transgene - HvSPP
|
|
Transgene - HvSPP
|
22
|
22 − 11
|
5.3
|
15.9–21.20
|
|
22 − 1
|
3.9
|
11.7–15.6
|
22 − 12
|
4.2
|
12.6–16.80
|
|
22 − 2
|
2.24
|
6.72–8.96
|
23
|
23 − 11
|
0.41
|
1.23–1.64
|
|
23 − 1
|
0.73
|
2.19–2.92
|
23 − 12
|
0.83
|
2.49–3.32
|
|
23 − 2
|
0.34
|
1.02–1.36
|
Transgene - HvUGPase
|
|
Transgene - HvUGPase
|
24
|
24 − 11
|
0.36
|
1.08–1.44
|
|
24 − 1
|
0.32
|
0.96–1.28
|
24 − 12
|
0.39
|
1.17–1.56
|
|
24 − 2
|
0.34
|
1.02–1.36
|
The genome of N. tabacum has not yet been sequenced, and therefore, there are no data on the copy number of the reference sequence. Thus, the number of copies of the reference sequence in the genomes of closely related plants, such as Solanum tuberosum and Solanum lycopersicum, was estimated using the blastn program, and the L18908 cDNA fragment from N. tabacum flanked by primers Mz116 and Mz117 as a query [61]. The genome of S. lycopersicum contains four regions with high similarity (77–89% identity) to this sequence along its entire length, while the S. tuberosum genome has three such regions with 84–90% identity and a fourth with 85% identity to part of this DNA fragment. Thus, it is further assumed that the N. tabacum genome contains 3 to 4 copies of the reference sequence (hereafter referred to as the NtL25 gene), similar to the genomes of the closely related plants S. tuberosum and S. lycopersicum. Based on this assumption, it was estimated that HvUGPase transgenic plants 24 − 1 and 24 − 2 or 2411 and 24 − 12 possess one copy of the transgene per genome. HvSPP F1 plants from lineage #23 possess from one to three copies of the transgene per genome and the F1 HvSPP plants in lineage #22 have from 6 to 21 copies of the transgene per genome. This variability is most likely due to the presence of multiple copies of the transgene in the ancestor of lineage #22, (Table 1).
3.4. Chrbc-P promoter specificity and the expression level of transgenes
The specificity of the Chrbc-P promoter and its efficiency were assessed by comparing the expression levels of the transgenes in leaves and roots. The preliminary results obtained from standard RT–PCR showed a much higher promoter efficiency in leaves than in roots (Fig. 6), so the ddPCR technique was used for accurate, quantitative analysis. Relative transgene expression in leaves and roots was used in these analyses to make the results independent of RNA concentration estimation by measuring absorbance or fluorescence. The NtL25 gene was used as a reference gene together with the primers Mz116 and Mz117, the same as in the genomic DNA analysis [47]. The plant material used for Chrbc-P promoter analysis was collected from young F1 plants grown for 7 weeks after sowing and included young, still developing leaves (6–10 cm in length) and whole roots.
The expression level of the transgenes expressed from the Chrbc-P promoter in leaves was 7 to 48 times higher than that of the reference gene NtL25 (Table 2, Supplementary Figs. S3 and S4). Conversely, in the roots, the activity of the Chrbc-P promoter was 10 to 50 times lower than that of the reference gene. The comparison of the Chrbc-P promoter performance in leaves and roots revealed that the frequency of the transcripts generated from this promoter in leaf cells was 67 to 2489 times higher than that in root cells (Table 2). The leaf-to-root ratio of relative expression of the transgene was similar in transgenic plants of the same lineage but varied between different lineages by as much as 30-fold.
Table 2
Specificity of the ChmrbcS promoter – relative expression (RE) of the transgenes in young leaves and roots, average of the two ddPCR experiments.
Plant ID
|
Aver. RE in young leaves
(+/- standard dev.)
|
Aver. root RE
(+/- standard dev.)
|
Aver young leaves RE / Aver. root RE
|
HvSPP transgene; HvSPP plants, lineage #22
|
HvSPP 22 − 11
|
48.8 (+/- 16.0)
|
0.020 (+/- 0.003)
|
2440.0
|
HvSPP 22 − 12
|
35.8 (+/- 9.2)
|
0.016 (+/- 0.004)
|
2237.5
|
HvSPP transgene; HvSPP plants, lineage #23
|
HvSPP 23 − 11
|
19.3 (+/- 2.6)
|
0.036 (+/- 0.004)
|
536.1
|
HvSPP 23 − 12
|
15.5 (+/- 4.1 )
|
0.040 (+/- 0.007)
|
387.5
|
HvUGPase transgene; HvUGPase plants, lineage #24
|
HvUGP 24 − 11
|
7.3 (+/- 1.5)
|
0.109 (+/- 0.019)
|
67.0
|
HvUGP 24 − 12
|
7.6 (+/- 0.3)
|
0.084 (+/- 0.016)
|
90.5
|
3.5. Expression of the transgenes and native genes for UGPase and SPP in mature leaves
The level of transgene expression in mature HvUGPase- and HvSPP-overexpressing F1 plants was analyzed by RT ddPCR. The plant material, two leaves at least 20 cm long, from the third and fourth lower internodes, was collected from mature, glasshouse-grown F1 plants at 69 DAP.
The results showed a higher average relative expression level of HvUGPase compared to HvSPP in transgenic lines, with values of 2.93 and 0.91–1.21, respectively (Table 3). The difference in the expression level of HvSPP varied between transgenic plants by as much as 2.7-fold. Due to the small representation of plants overexpressing HvUGPase, less variation was observed in the relative expression level of HvUGPase between individuals.
Table 3
The relative expression levels of the transgenes - HvUGPase and HvSPP and native genes - NtUGPase and NtSPP - in mature leaves of transgenic and reference N. tabacum plants, the average values from three individual ddPCR experiments. Expression of transgenes – HvUGPase and HvSPP (column d) is 10–14 times higher than the corresponding native genes (column f).
Lineage
|
Plant
|
Aver. RE of given transgene
|
Aver.. RE of the transg. in given lineage
|
|
Aver. RE of given native gene
|
Aver. RE of nat. gene in given lineage
|
Aver. RE of a native gene as a % of the aver. RE of a given gene from all reference plants
|
a
|
b
|
c
|
d
|
|
e
|
f
|
g
|
Transgene - HvSPP
|
|
Native gene - NtSPP
|
HvSPP plants,
22
|
22 − 1
|
0.717
|
0.912
|
0.079
|
0.090
|
93.7%*
|
22 − 2
|
1.107
|
0.100
|
HvSPP plants,
23
|
23 − 1
|
1.953
|
1.504
|
0.127
|
0.106
|
110.4%*
|
23 − 2
|
1.055
|
0.085
|
Ref. plants,
34
|
34 − 1
|
-
|
-
|
0.094
|
0.09
|
93.7%*
|
34 − 4
|
-
|
0.086
|
Ref. plants,
35
|
35 − 1
|
-
|
-
|
0.083
|
0.101
|
105.2%*
|
35 − 2
|
-
|
0.115
|
35 − 3
|
-
|
0.104
|
|
|
All Ref. plants
|
0.096
|
100%*
|
Transgene - HvUGPase
|
|
Native gene - NtUGPase
|
HvUGPase plants,
24
|
24 − 1
|
3.440
|
2.927
|
|
0.267
|
0.243
|
91.7%**
|
24 − 2
|
2.413
|
0.219
|
Ref. plants,
34
|
34 − 1
|
-
|
-
|
0.246
|
0.283
|
106.8%**
|
34 − 4
|
-
|
0.320
|
Ref. plants,
35
|
35 − 1
|
-
|
-
|
0.243
|
0.254
|
95.8%**
|
35 − 2
|
-
|
0.250
|
35 − 3
|
-
|
0.268
|
|
|
All Ref. plants
|
0.265
|
100%**
|
The results showed high expression of both transgenes, similar to or 2.9 times higher than that of the native NtL25 gene. However, compared to young leaves, the expression of the transgenes in mature leaves decreased 11–46 times in plants with HvSPS overexpression and 2.5 times in plants with HvUGPase overexpression, Table 2 and Table 3. The expression of the native genes for NtUGPase and NtSPS was not disturbed by the activity of the transgenes and remained in the transgenic plants similar to that of the reference plants. The expression of the transgenes in mature leaves significantly exceeded the expression of the corresponding native genes: 9.0–14 times for SPP and 10.9–12.7 times for UGPase, see Table 3 column d and f.
3.6. The phenotypes of plants overexpressing HvUGPase and HvSPP
The F1 plants overexpressing HvUGPase and HvSPP showed faster growth and earlier flowering than the control reference plants. The F1 HvUGPase and HvSPP transgenic plants showed a greater increase in height that was most evident between 30–69 DAP, and they started the generative phase 11 days earlier than the reference plants (Figs. 8 and 9).
The F1 plants overexpressing HvUGPase were 35.7% taller than the reference plants, and plants overexpressing SPP were 41.5% taller. Additionally, the dry mass (d.m.) of F1 HvUGPase and HvSPP plants, measured at 69 DAP (when the cultivation was completed), was greater than that of the reference plants (Fig. 9b). The differences in d.m. were smaller, amounting to 9% for HvUGPase plants and 6% for HvSPP plants, but still evident, Fig. 9b. No differences in leaf length and width were noted between HvUGPase or HvSPP plants and reference plants, and no disturbances in the flower morphology of the transgenic plants were observed.
3.7. The composition of carbohydrates in plants overexpressing HvUGPase and HvSPP
The carbohydrate composition of the aerial part of mature plants was analyzed to determine the effect of HvUGPase or HvSPP overexpression on plant metabolism and physiology. The aboveground part of mature plants was harvested at 69 DAP, and the plant material, consisting of leaves and stems, was dried and ground. The sum of all carbohydrates, including mono- and disaccharides, hemicellulose, cellulose, and lignin, in HvUGPase and HvSPP plants was higher than that in reference plants, but this difference was subtle, with values of 1.71 [% d.m.] for HvUGPase plants and 3.16 [% d.m.] for HvSPP plants (Fig. 10). Monosaccharides are one of the major constituents of the analyzed plant material, and their content in HvUGPase and HvSPP plants was slightly lower than that in reference plants. Disaccharides, which constitute only approximately 1–2% of the dry mass of the mature N. tabacum plant, have a much smaller impact on the energy value of the analyzed plant material. In the case of HvUGPase plants, their content in d.m. was slightly lower than that for reference plants, and for HvSPP plants, this value was higher but still below 2% of d.m. (Fig. 10b). For all insoluble carbohydrates, such as hemicellulose, cellulose and lignin, their share in d.m. was greater in plants with HvUGPase and HvSPP overexpression than that of reference plants (Fig. 10c-e).
3.8. Heat of combustion and sequestered energy
For all F1 plants, the energy value (heat of combustion) was determined to compare the upper limit of the available thermal energy obtained from the complete combustion of one gram of given plant material (Fig. 11a). We observed that the energy value was slightly higher in plants overexpressing HvUGPase and HvSPP than in reference plants, probably due to the higher content of total carbohydrates in these plants, but the difference was in the range of 1.26–1.65%. Data on the energy value and dry mass of a given plant allowed calculation of the total energy sequestered by the aboveground part of the plant (Fig. 11b). A comparison of the total energy stored by the plant shows that the plants overexpressing HvUGPase and HvSPP sequestered nearly 20% more energy than the reference plants during the same growing period. This indicates the beneficial effects of the overexpression of the genes for enzymes in the photosynthetic sucrose synthesis pathway on the energy value of plants, even if the plants do not have an organ dedicated to the sequestration of newly bound carbon.