Sucrose interferes with endogenous cytokinin homeostasis and expression of organogenesis-related genes during de novo shoot organogenesis in kohlrabi

Cross-talk between phytohormones and sugars is intensely involved in plant metabolism, growth and regeneration. We documented alterations in cytokinin (CK) homeostasis in four developmental stages during de novo shoot organogenesis (DNSO) of kohlrabi (Brassica oleracea var. gongylodes cv. Vienna Purple) seedlings induced by exogenous CKs, trans-zeatin (transZ) and thidiazuron (TDZ), added together with elevated sucrose concentration (6% and 9%). Significant impact of CK and sucrose treatment and their interaction was recorded in all investigated stages, including plantlet development before calli formation (T1 and T2), calli formation (T3) and shoot regeneration (T4). Results showed remarkable increase in total CK levels for transZ treatment, particularly with 9% sucrose. This trend was observed for all physiological and structural groups of CKs. Application of TDZ contributed to little or no increase in CK levels regardless of sucrose concentration. Analysis of expression profiles of organogenesis-related genes involved in auxin transport, CK response, shoot apical meristem formation and cell division revealed that higher sugar concentration significantly downregulated the analysed genes, particularly in T3. This continued on TDZ, but transZ induced an opposite effect with 9% sucrose in T4, increasing gene activity. Our results demonstrated that phytohormone metabolism might be triggered by sucrose signalling in kohlrabi DNSO.


Results
Effect of sucrose on kohlrabi growth and development during cytokinin-induced de novo shoot organogenesis. Optimal sucrose concentration for growth of in vitro kohlrabi cultures is 3% 6,12,38,39 .
As shown in Fig. 2a, the total CK content was similar in all investigated developmental stages and on all sucrose concentrations applied on CK-free medium. Exception was detected for 3% sucrose in T2 stage with significantly elevated value of approximately 800 pmol g −1 FW compared to others ranging up to around 300 pmol g −1 FW. On the other hand, a remarkable increase in total CK levels was recorded in samples treated with transZ, particularly when 9% sucrose was used in the medium, reaching up to 50,000 pmol g −1 FW in T3 (Fig. 2b). This aspect could be seen in all stages. Addition of a synthetic urea-type CK, TDZ, to the medium showed similar results as when CK was omitted, making no significant difference in total endogenous CK Figure 2. Contents (in pmol g −1 FW) of total cytokinins (a-c) and CK nucleobases (d-f) in four developmental stages (T1-T4) during kohlrabi (cv. Vienna Purple) growth and de novo shoot organogenesis on media containing 3/6/9% sucrose without the presence of cytokinins (a, d), with 2 mg L −1 transZ (b, e) or with 2 mg L −1 TDZ (c, f). Data are shown as the mean ± SE (n = 3 independent biological replicates). Means marked with the same letter within each developmental stage of each treatment were not significantly different according to the LSD post-hoc test, P ≤ 0. 05 www.nature.com/scientificreports/ contents depending on the sucrose concentration used (Fig. 2c). Still, the measured values were noticeably lower compared to transZ treatment. According to ANOVA, values recorded for CK nucleobases differed with regard to the concentration of sucrose applied (Fig. 2d-f). Addition of 6 or 9% sucrose led to a significant increase in the portion of bioactive CK forms in total CK pool in all studied stages. Again, the effect of highest applied sugar concentration was particularly manifested when transZ was added to the growth medium (Fig. 2e). The amendment of transZ generally elicited remarkable rise in contents of CK nucleobases contents compared to CK-free medium, while application of TDZ contributed only to little or no increase (Fig. 2f).
Application of 9% of sucrose resulted in a significant rise in the level of CK ribosides, in all developmental stages except T2, during CK-free cultivation (Fig. 3a). After applying transZ, substantially increased concentrations of CK ribosides were recorded in all T1-T4, reaching the highest values on medium with 9% of sucrose (Fig. 3b). Presence of TDZ in nutritive medium followed the previous trend of statistically significant increase when sucrose was applied in concentration of 9% but to a much lesser extent (Fig. 3c).
Furthermore, marked variations were found for O-glucosides ( Fig. 3d-f) when kohlrabi plantlets were treated with transZ (Fig. 3e). In all four stages (T1-T4) application of high sucrose concentration (9%) induced statistically significant increment in content of O-glucosides. As described earlier, TDZ showed no marked effect regardless of sucrose concentration (Fig. 3f). . Contents (in pmol g −1 FW) of CK ribosides (a-c) and O-glucosides (d-f) in four developmental stages (T1-T4) during kohlrabi (cv. Vienna Purple) growth and de novo shoot organogenesis on media containing 3/6/9% sucrose without the presence of cytokinins (a, d), with 2 mg L −1 transZ (b, e) or with 2 mg L −1 TDZ (c, f). Data are shown as the mean ± SE (n = 3 independent biological replicates). Means marked with the same letter within each developmental stage of each treatment were not significantly different according to the LSD post-hoc test, P ≤ 0. 05 Fig. 4a-c), whose distribution reflected the distribution of total CKs. Accordingly, application of transZ directed to the significant increase in concentrations of N-glucosides, again especially at 9% of sucrose (Fig. 4b).
The situation for CK phosphates (Fig. 4d-f), the smallest fraction of total CKs representing immediate biosynthetic precursors, was different compared to other CK types. When CK-free medium was used, variations were mainly detected in combination with 6% sucrose where the CK phosphate content was reduced in T2 and T4 (Fig. 4d). Presence of transZ led to the marked enhancement of these CK forms level in T1 compared to CKfree medium, mainly when 9% of sucrose was applied (Fig. 4e). On the other hand, TDZ had a diverse effect in combination with different concentrations of sugar in all studied developmental stages (Fig. 4f). Table 1 demonstrates the distribution of endogenous CKs based on their side-chain structure into four types 41 : cisZ-types (cisZ, cisZR, cisZ7G, cisZ9G, cisZOG, cisZROG, cisZRMP); transZ-types (transZ, transZR, transZ7G, transZ9G, transZOG, transZROG, transZRMP); DHZ-types (DHZ, DHZR, DHZ7G, DHZ9G, DHZOG, DHZ-ROG, DHZMP) and iP-types (iP, iPR, iP7G, iP9G, iPRMP). Statistical analysis showed that the levels of listed endogenous CK types in all developmental stages were generally dependent on the type of CK treatment and the applied concentration of sucrose. For all CK structural groups, the same trend was observed as previously  www.nature.com/scientificreports/ Effect of sucrose on the expression of organogenesis-related genes during cytokinin-induced de novo shoot organogenesis. The activity of selected genes involved in auxin transport (PIN3), CK response (ARR5), de novo shoot apical meristem formation (RGD3) and cell division (CDKB2;1 and CYCB2;4) was determined in kohlrabi plantlets using quantitative PCR analysis showing variation in expression levels during callus formation (T3) and de novo shoot regeneration (T4) (Fig. 5a-c). Absolute quantification determined for kohlrabi plants grown on 3/6/9% sucrose without CKs was performed to study the effect of sugar alone on the activity of chosen genes. The analysis revealed that a higher concentration of sucrose had a statistically significant impact on the expression of all investigated genes, diminishing their activity, particularly during callus formation. The same trend was observed in transZ-treated plants in T3. On the other hand, the addition of transZ induced contrary effect in combination with high sugar concentration in T4 www.nature.com/scientificreports/ stage leading to increased activity of investigated genes, except for PIN3 (Fig. 5b). Application of TDZ induced a statistically significant decrease in the gene expression when higher sugar concentration was used, particularly 9% in T3 (Fig. 5c). The RGD3, CDKB2;1, and CYCB2;4 genes showed slightly different expression patterns in T4, with 6% sucrose being the treatment with the highest values for absolute expression.
Correlations between endogenous cytokinins and activity of the analysed genes during later stages of de novo shoot formation. In order to investigate if altered endogenous CK homeostasis affected the activity of given organogenesis-related genes, correlations between content of CK nucleobases, representing bioactive CKs, and absolute expression of analysed genes were evaluated in response to each particular CK/sucrose treatment, at the T3 and T4 stages of organogenesis. As demonstrated in Fig. 6, at the T3 stage of organogenesis, there is positive correlation between CK nucleobases and expression of most genes at 3% and 9% sucrose treatments. When it comes to transZ treatments, correlation between bioactive CKs and the expression of most genes is negative at 3% sucrose, but this negative correlation becomes less pronounced with increasing sucrose concentration in the media. At all TDZ treatments, there is strong positive correlation between bioactive CKs and the expression of RGD3 and the cell cycle genes CDKB2;1 and CYCB2;4, with the exception of CYCB2;4 at 9% sucrose, where the correlation shifts from strongly positive to strongly negative. At the T4 stage of organogenesis, the situation is somewhat different. On media without added CKs, the correlation between the plant content of bioactive CKs and the expression of the investigated genes is dependent on sucrose in the media, and, with the exception of ARR5, shifts from strongly positive at 3% sucrose, to strongly negative at 9%. It is noteworthy that unlike with the other genes, the expression of ARR5 is remarkably negatively correlated with the plant bioactive CK content at almost all the CK/sucrose treatments during the T4 phase. Besides, at the transZ treatment, the expression of RGD3 and CDKB2;1 is strongly negatively correlated with the plant bioactive CKs regardless of sucrose concentration. On the contrary, at the TDZ treatment, the expression of RGD3 and both examined cell cycle genes is strongly positively correlated with plant CK content regardless of sucrose concentration, whereas the expression of PIN3 shows negative correlation with plant bioactive CKs at moderate sucrose concentrations, but this correlation becomes strongly positive at 9% sucrose (Fig. 6.).

Discussion
Sugar levels influence plant development, including various processes from embryogenesis to senescence 42 . In emerging plant embryos, glucose and sucrose are involved in regulation of cell division, cell expansion as well as accumulation of reserve carbohydrate 43 . Furthermore, sucrose is implicated in cell division and starch synthesis in mature plant cells 44 . This impact is accomplished due to the fact that sugars act as signalling molecules and can regulate the expression of various genes 24 .
Plant shoots grown in vitro are characterised by a reduced ability of photosynthesis, so as compensation carbon source must be included in the nutritive media 45 . The most exploited sugar for this purpose is sucrose, which besides being an energy source also plays a significant role as osmotic regulator.
The optimal concentration of sucrose in shoot cultures is dependent not only on the plant species but also on different genotypes and cultivars 46 . Our previous work demonstrated that kohlrabi in vitro cultures grow www.nature.com/scientificreports/ well on an optimal concentration of 3% sucrose 6,12,38,39 , which enables an efficient regeneration process. In the present study we investigated the application of 6% and 9% sucrose on kohlrabi germination, seedling growth and regeneration. The results were obtained showing a proportional decrease in germination rate as a response to the increased amount of sugar in the medium after six days, regardless of exogenous CK addition. Other studies previously demonstrated the sugar-induced delay of seed germination, indicating involvement of abscisic acid (ABA); namely it was suggested that elimination of endogenous ABA is slower in seeds germinating under the influence of sugar, hence the delay 47 .
Similarly, further growth and development of kohlrabi seedlings were affected by the presence of a higher amount of sucrose-the development of shorter and thicker plantlet stem, larger root system, as well as a lower number of leaves were recorded in a dose-dependent manner. Described growth changes are in accordance with our former work where growth and development of treated kohlrabi seedlings were significantly affected due to sugar treatment in a distinctive manner, with a variety of morphological traits being altered in comparison to matching controls 39 .
Seedlings germinated on media with higher concentrations of sugars, such as glucose or sucrose, usually have shorter hypocotyls whose length is inversely proportional to applied sugar concentration, less developed root system and leaves 48,49 . On the other hand, in Brassica napus seedlings application of exogenous sucrose led to inhibition of hypocotyl growth while it stimulated the root growth 31 , as in our study with kohlrabi plantlets grown on CK-free media. Improved root growth under a high sucrose level (6%, w/v) was also reported in Arabidopsis seedlings overexpressing melon tonoplast sugar transporter CmTST1 50 . Authors proposed that CmTST1 plays a key role in importing sugar into the vacuoles of roots in response to metabolic demands to maintain cytosolic sugar homeostasis.
Different concentrations of constituents of nutritive media affect morphogenesis and growth of the cultured plant cells, tissues and organs partially due to their nutritional value but can also alter its osmotic potential 51 . Stress-shocked cells showed cell growth retardation and the induction of plasmolysis 52 . Biahoua and Bonneau 53 reported in their study of the spindle tree somatic embryogenesis that sucrose played an osmotic role stimulating plant regeneration rather than acting exclusively as a carbohydrate source. It was estimated that larger part of the applied sucrose that was needed for inducing an optimal rate of adventitious shoot formation in hypocotyl explants of Digitalis obscura were required as energy source, while the rest osmotically regulated morphogenesis 54 .
Observed developmental alterations and growth retardation were enhanced by the presence of CKs, so they might not represent a primary consequence of osmotic stress but could be instead associated to sucrose crosstalk with other response pathways such as e.g. phytohormone response pathways. Sugar sensing processes and their effects on plant development arrest are already shown to be separated from their nutritive role 55 .
We can assume that supra-optimal sucrose combined with exogenous CKs largely contributed to the stunted and deformed kohlrabi phenotypes, by promoting excessive stimulation. TDZ manifested even stronger effect than transZ, and also induced the formation of a larger number of newly formed buds when sugar was applied at 9%. It is widely considered that synthetic substances with cytokinin-like biological effects (such as TDZ) have additional ways of influencing plant morphogenesis, that are not typically attributed to isoprenoid cytokinins (such as transZ) 56,57 . In some cases higher concentrations or prolonged treatment of TDZ could obstruct further growth and development of the regenerants, which might represent the consequences of TDZ effect on metabolism of endogenous auxin and cytokinin 58,59 . On the other hand, these morphological changes could be the result of alterations in kinetics of plant enzymes, as many of the TDZ-stimulated enzymes were shown to be associated with cell walls, membranes and membrane fluidity 60 . Thus, TDZ might influence morphological alterations that are qualitatively distinct from those induced by zeatin, and not necessarily related to the dynamics of DNSO as described in the standard two-step shoot regeneration protocol for Arabidopsis.
Levels of endogenous phytohormones are influenced not only by plant species and genotypes but also by the type of explant and the developmental stage 6,61 . As previously shown, the presence of PGRs in nutritive media affects the content of endogenous plant hormones through alteration of their homeostasis 5,6 .
Although the total CK content in in vitro grown kohlrabi plantlets was not affected by higher sugar application alone, a substantial rise in the CK level was documented after transZ application together with 9% sucrose regardless the developmental/DNSO stage. Our earlier report on kohlrabi DNSO demonstrated that after six weeks of seedling cultivation on different CKs, exogenously applied transZ into the medium with 3% sucrose contributed not only to the notable increase of endogenous levels of zeatin-type CKs but also to the enhancement of the total CK pool 6 . The stadium of analysed kohlrabi samples in that work approximately corresponded to the T4 stage in the present study (emergence of de novo shoots), with comparable total CK level on 3% sucrose (441 pmol g −1 FW). We assume that the CKs from regeneration media affect endogenous levels of particular CK derivatives upon uptake from the media, ultimately contributing to the rise in their content, as previously reported 4,41,62 .
In this regard, the mechanism by which high sucrose concentration enhances the plant CK levels specifically in combination with transZ is particularly intriguing. This increase is presumably a consequence of the uptake from nutrition media, upon which CKs may either stay in the form of free bases, or undergo conjugation into other CK types, as reflected in the increase of all CK types (organised based on the CK conjugation status) on the transZ/9% sucrose treatment. The accumulation of these CK classes, particularly at 9% of sucrose, could be not only an implication of the increased uptake, but also of the disturbed hormonal homeostasis and, consequently, metabolic glucosylation 4,62,63 . Overaccumulation of sugars due to higher carbohydrate supply might have altered the metabolic processes in the plant cells, which subsequently altered the CK levels and ultimately growth and development of plantlets. It was shown that plants display improved acclimation to osmotic stresses when endogenous CK content is high. Elevated CK levels under drought stress neutralize drought-induced signalling, which enables plants to preserve normal levels of leaf water content, photosynthetic rate and stomatal conductance under stress conditions 64  www.nature.com/scientificreports/ transport, or signalling of particular hormones, including auxin and cytokinins 23,65,66 . It was suggested that sugars promoted cytokinin accumulation through upregulation of some of the IPT genes 23,37 . The elevation of CK levels in plantlets grown on media with 9% sucrose was identified in all four investigated developmental stages during DNSO, with the maximum increment observed for T3 stage characteristic by callus formation at the base of the kohlrabi plantlets. The detected values were ca. 40-fold to over 100-fold higher for total and bioactive CKs, respectively, compared to control (3% sucrose). Earlier literature data suggested that plant phytohormone signalling depended exclusively on the levels of endogenous hormones occurring in plant tissues, which may not relate to the PGRs present in the regeneration media 67 . However, it was shown that both exogenous and endogenous plant hormones could initiate the callus formation and subsequent cell differentiation in plant tissue culture, interacting with each other to induce shoot organogenesis 68 .
Similar situation was found when endogenous CKs were classified based on their N 6 side-chain structure ( Table 1). Considering this categorisation, higher concentration of sucrose synergistically acted with transZ mostly affecting transZ-type CKs, upraising their levels up to nearly 40,000 pmol g −1 FW in T3.
The lowest contribution of exogenously applied transZ to the increase of endogenous levels was noticed for iP-type CKs considered primary precursors for biosynthesis of zeatin-types 69 . The decrease of iP-type CKs in transZ-treated plants suggests that enhanced uptake of transZ might have triggered homeostatic mechanisms that lead to the decrease in the levels of its biosynthetic precursor and related conjugates, consistently with our previous findings 6 . However, it should be mentioned that application of 9% of sucrose, without any PGRs also induced elevation of iP-types compared to control on 3% in all stages, suggesting the interference of sugar with CK homeostasis.
Accordingly with our previous work 6 , the presence of TDZ in the shoot regeneration media did not significantly affect the endogenous levels of any of CK groups, classified based on their N 6 side-chain structure, as compared to the control regardless of sucrose concentration. Earlier literature data showed that treating plant tissues with synthetic CK homologues initiated at least a transitory elevation in levels of natural zeatin, and zeatin riboside 70 . It has been known for a long time that TDZ has an inhibitory effect on CK catabolism through the inhibition of CK oxidase/dehydrogenase (CKX) 71 . Our results show that although it did not generally contribute to an increase in plant CK levels, TDZ affected the signalling of organogenesis-related genes.
On the other hand, the level of applied sugar seems to be important cofactor with TDZ affecting some CK groups, such as CK nucleobases and ribosides, as well as phosphates. Thidiazuron is a synthetic derivate of phenylurea with proven effects similar to CKs. Although it structurally differs from both isoprenoid and aromatic CKs, it possesses the ability to efficiently bind CK receptors 72 . However, along with indisputable CK-like activities, TDZ can also manifest some other biological effects on plant tissues, such as auxin-like ones 56 . Furthermore, the capacity of TDZ to provoke a defensive response in plant tissues can also initiate down-or up regulation of other phytohormones and secondary metabolites, and at the same time moderate the influx/efflux of specific cations, such as calcium, through biological membranes 56,73,74 . TDZ, as a competitive inhibitor of CKX 71 can affect pathways of purines and cytokinin metabolites by preventing their breakdown 56 . Despite the range of effects associated to TDZ and intensive research, its precise mode of action in induction of in vitro plant morphogenesis has not been fully clarified to date 56,57 .
Kiba et al. 75 suggested that de novo biosynthesis of CKs is triggered by photosynthetically generated sugars in Arabidopsis grown in elevated CO 2 conditions. Their research has indicated the existence of a signalling pathway in which the photosynthesis-derived sugars regulate CK levels to control plant growth and development. Earlier transcriptome analysis revealed that sugar treatments upregulate AtIPT3 and CYP735A2 genes 36,76 suggesting that sugar-induced expression of genes is implicated in de novo CK biosynthesis.
The expression of IPT, and consequently CK levels are influenced by the accessibility of macronutrients like nitrogen, phosphate and sulphate [77][78][79] . Considering that macronutrients and carbon metabolism are closely linked, one can hypothesise that sugars regulate expression of genes involved in the de novo biosynthesis of CKs indirectly via signalling pathways of macronutrients 75 .
Some reports indicate that sugars and CKs can function both agonistically and antagonistically on gene expression 23,36 . The antagonistic effect of CKs was demonstrated through direct interaction between sugar and CKs using hxk1/gin2 mutant 32 . However, other studies from the same research group demonstrated synergistic effects as reported for Arabidopsis seedlings, where CKs and glucose positively regulated root elongation through an HXK1-dependent pathway 80 . This response implicated cytokinin receptor AHK4 (ARABIDOPSIS HISTIDINE KINASE 4) as well as three type B response regulators (ARR1, ARR10, and ARR11) that are positive regulators of CK signalling pathway, with glucose enhancing CK-dependent root elongation upstream of the auxin effect on root growth 23,80 .
It was found that glucose influences the expression of a majority of genes involved in CK metabolism, perception and signalling. Upregulation of AtIPT3 by glucose indicates the possibility of glucose upregulating CK biosynthesis. Additionally, glucose upregulates specific components of CK signalling, namely A-type response regulators, which function as negative regulators of CK signalling pathway 81 , and CK response factors (CRFs) 36 . This kind of glucose-modulated expression of certain components of CK signalling suggests that sugar and CK signalling can interact with each other, with sugars presumably affecting CK homeostasis from biosynthesis to signalling.
Currently it remains unclear whether transporter activity, membrane damage, osmotic stress or specific signals underlie observed remarkable increase in CK content when exposed to high sucrose/transZ combination. The phenotypic flexibility of plant growth and development is a result of signalling network that merges signals, both from inside and outside, and comprises nutrients, phytohormones, as well as physical factors. Besides, identical signals can stimulate diverse responses at different concentrations, making the phenomenon of sugar sensing even more complicated. To provide clear evidences supporting sucrose effect on the CK homeostasis detailed expression analyses of genes regulating CK levels in planta need to be performed.

Scientific Reports
| (2021) 11:6494 | https://doi.org/10.1038/s41598-021-85932-w www.nature.com/scientificreports/ Investigation of sugar effects and endogenous CK levels during DNSO in kohlrabi seedlings in vitro was complemented with determinations of absolute expression profiles of organogenesis-related genes involved in auxin transport, CK response, de novo shoot apical meristem formation and cell division. The selected genes were previously shown to be essential elements of CK signalling cascade in the course of de novo formation of shoots in established efficient kohlrabi regenerating system 12 . Application of sucrose in high concentration without the presence of exogenous CKs resulted in a significantly decreased expression of all investigated genes during callus and de novo shoot development. This decrease could be associated with the observed retardation in kohlrabi development compared to standard 3% sucrose treatment.
It has been shown in our previous work 12 that these genes are upregulated during callus formation (T3) on media with 3% sucrose; their upregulation is most likely closely linked to the molecular processes that accompany callus formation 9,12 . Here we demonstrate that increased concentration of sucrose in regeneration media causes downregulation of these genes compared to 3% sucrose, likely accounting for the delay in morphogenic events that accompanied DNSO in culture.
The same effect of sucrose is present for the genes PIN3 and ARR5 during the T4 regeneration phase (shoot formation), whereas it is not evident, or in some cases even reversed, for RGD3 and the cell cycle genes CDKB2;1 and CYCB2;4. These genes are predominantly downregulated during the shoot formation phase on media with 3% sucrose 12 , consistently with a different expression pattern and different response to sucrose compared to their expression profiles during T3.
Pearson's correlation analysis revealed interesting features of the correlation between plant bioactive CK levels and the expression of organogenesis-related genes in various exogenous CK/sucrose treatments. An interesting phenomenon is negative correlation between bioactive CK levels and expression levels of most investigated genes, in most of the transZ treatments, in both T3 and T4 organogenesis stages. As all these genes were shown to be CK-responsive 12 , a highly positive correlation between plant CK levels and their expression would be expected; yet, we demonstrate that the opposite is mostly the case. A possible explanation lies in very high plant CK levels that accompany the transZ treatment; as a response to excessive presence of the hormone, signalling may be silenced by homeostatic mechanisms, resulting in negative instead of positive correlation between the plant CK levels and gene expression. Similarly, a strong negative correlation is present during the T3 phase of organogenesis in hormone-free media with 6% sucrose, where endogenous CK levels were particularly high during the previous T2 phase.
The correlation analysis also showed that during the T4 phase, the correlation between plant bioactive CK levels and the expression of most organogenesis-related genes depends on the concentration of sucrose in the media, being, except for ARR5, strongly positive at 3% sucrose and shifting towards strongly negative at 9%. This is another indication that sucrose from the nutrient media interacts-or at least interferes-with the CKmediated regulation of these genes. The only exception is ARR5, whose expression levels are negatively correlated with plant CK levels during the T4 phase on almost all CK/sucrose media combinations. This is unsurprising, as ARR5 is known to undergo a strong inhibition by CK-induced WUSCHEL around the time of shoot development in Arabidopsis 82,83 and the downregulation of ARR5 upon shoot development was also shown in kohlrabi in our previous work 12 .
Finally, correlation analysis also revealed interesting relations between plant CK content and organogenesisrelated gene expression in TDZ treatments during the T4 stage. While the expression of ARR5 was strongly negatively correlated with plant CK content regardless of sucrose concentration in the media, the opposite was true for RGD3 and the two cell cycle genes CDKB2;1 and CYCB2;4. Interestingly, the expression of the auxin transporter gene PIN3 was negatively correlated with plant CK content at moderate sucrose concentrations (3% and 6%), whereas this correlation was positive on media with high sucrose, suggesting that sucrose might interfere with either TDZ or its impact on endogenous CKs affecting the expression of this auxin transporter gene.
In conclusion, we used our previously established efficient one-step DNSO system as a useful tool to investigate the effect of sugar on plant growth and organ development, and ultimately intricacy of the sugar and phytohormone signalling networks. Application of moderate sucrose concentration did not alter the CK homeostasis significantly, with results similar to control for both types of CK treatment. Contrary to that, a higher concentration of sucrose in nutritive media containing transZ contributed to a substantial increase in endogenous CK levels throughout kohlrabi development and DNSO. Absolute quantification of the selected genes revealed that higher concentration of sucrose significantly declined their expression, particularly during callus formation. This trend was demonstrated for TDZ as well as transZ treatment in T3. On the other hand, higher sugar concentration induced contrary effect in combination with transZ in T4 stage, leading to the increase of gene expression. Correlation analysis suggests that sucrose might interfere with how plant CK levels affect the expression of organogenesis-related genes during callus and shoot formation in kohlrabi. Our results indicate that phytohormone metabolism may be affected by sucrose in kohlrabi shoot organogenesis. It should be, however, yet elucidated whether sugars act as signalling molecules, energy sources or building units to impact the expression of the particular genes. Undoubtedly, the regulatory network is more complex and must comprise a role for sugars in the regulation of hormone metabolism as well as hormone signalling.

Methods
Plant material and growth conditions. Establishment of kohlrabi (Brassica oleracea var. gongylodes cv. Vienna Purple) in vitro culture was performed as previously described in Ćosić et al. 6,12 . Accordingly, seeds were primarily surface sterilized by a 5 min submersion in 70% ethanol. This step was followed by 30 min submersion in 30% commercial bleach (4-6% NaOCl) containing a drop of detergent (Fairy; Procter and Gamble, London, UK) and rigorously rinsing in autoclaved distilled water. Sterilized seeds were aseptically transferred to particular growth media. www.nature.com/scientificreports/ A basal, PGRs-free growth medium comprised of Murashige and Skoog 84 mineral salts, Linsmaier and Skoog 85 vitamins, 100 mg L −1 inositol and 0.6% agar. For the investigation of sucrose concentration effects, media containing 3%, 6% or 9% sucrose were used. At the same time, different regeneration media have been additionally complemented with trans-zeatin or thidiazuron at 2 mg L −1 . The respective CK concentrations have been previously established through an optimization of the regeneration protocol for various explant types of kohlrabi cultivars Vienna Purple and Vienna White, as described earlier 6 . All media used in this study were adjusted to pH 5.8 with 1 N NaOH before autoclaving at 114 °C and 80 kPa for 25 min. Cultures were maintained in a growth chamber at 25 ± 2 °C under cool white fluorescent light (16 h light photoperiod; 47 μmol m −2 s −1 irradiance).
The experimental setup for evaluating the impact of sucrose on CK homeostasis and organogenesis-related genes in four different stages (T1-T4) of CK-induced DNSO in kohlrabi seedlings implied eight 375 mL jars for each specific combination of sucrose and CK. Each jar contained six surface sterilized seeds. Control seeds were grown on basal growth medium with 3% of sucrose.
Plant material was collected in four developmental stages (T1-T4), and included seedlings bearing two cotyledons (T1), plantlets with emerged two leaves (T2), plantlets forming callus at the base of the stem (T3) and de novo shoots appearing from the calli (T4). Plantlets grown on CK-free medium were collected at the same time as corresponding CK treated plantlets, having in mind that DNSO occurred only on media supplemented with transZ or TDZ. For each stage point, three representative plantlets exposed to the same treatment were collected and pooled into a single biological sample. All samples were immediately frozen in liquid nitrogen and stored at -70 °C until analysis. The entire experimental setup was conducted in three independent biological replicates.

Determination of endogenous cytokinin levels.
Phytohormone analysis was carried out as described in Dobrev and Vankova 86 and Djilianov et al. 87 . Approximately 100 mg of plant material of each biological sample was homogenised in liquid nitrogen and then lyophilised.
Cold extraction buffer consisting of methanol/formic acid/water (15/1/4; v/v/v) was added to the plant homogenates, along with a mixture of stable isotope-labelled internal standards (10 pmol ]iP-9-glucoside (iP9G); [ 2 H 6 ]iPR-5´-monophosphate (iPRMP) (all CK standards were from OlChemIm, Olomouc, Czech Republic). The concentrations of cis-zeatin (cisZ) derivatives were determined based on the retention times and mass spectra of unlabelled standards and the response ratio of their transZ counterparts.
Detection and quantification of CKs were performed using HPLC (Ultimate 3000, Dionex, Sunnyvale, CA, USA) that was coupled to a hybrid triple quadrupole/linear ion trap mass spectrometer (3200 Q TRAP, Applied Biosystems, Foster City, CA, USA) set in selected reaction-monitoring mode. The mass spectrometer was set at electrospray ionisation mode with following ion source parameters: ion source voltage + 4000 V (positive mode); nebuliser gas 50 psi; heater gas 60 psi; curtain gas 20 psi; heater gas temperature 500 °C. The endogenous CKs were quantified using the isotope dilution method with multilevel calibration curves. All data were processed with Analyst 1.5 software (Applied Biosystems). The CK concentrations were calculated as the amount per 1 g of fresh weight plant material.
For evaluation of endogenous CK levels, three biological replicates were used for each of kohlrabi seedlings treated with transZ/TDZ + 3/6/9% sucrose as well as control. Analyses were repeated twice with comparable results.
Statistical analysis. The data were analysed using SAS software (SAS Institute, 2002. SAS/STAT, ver. 9.00. SAS Institute Inc., Cary, NC, USA). To analyse the results for total endogenous CK pool as well as the content of individual CK forms distributed based on their conjugation status (CK nucleobases, CK ribosides, O-glucosides, N-glucosides and CK phosphates), analysis of variance (ANOVA) was used for each developmental stage point of each treatment, sucrose concentration being the factor analysed. The same type of analysis was used for the content analysis of cisZ-, transZ-, DHZ-and iP-types CKs determined based on their side-chain structure, for each applied CK treatment in each developmental stage. To compare the results for mean values of particular absolute gene expression levels for T3 and T4, one-way ANOVA was employed, with sucrose concentration as analysed factor. Statistical correlation was evaluated between content of plant CK nucleobases and absolute expression of the analysed genes and presented using the Pearson's correlation coefficient (r). The differences found among means were determined using the Fisher's least significant difference (LSD) post-hoc test at 95% of confidence level.