**3. Molecular and morphological characterisation of transgenic** *AtCKX* **centaury plants**

The possibility of direct regeneration of centaury shoots and/or somatic embryos makes root cultures a good model for studying *in vitro* morphogenesis of this plant species. Root cultures, in addition to easy manipulation, are characterised by other advantages such as small differences in physiological responses as well as high growth potential and metabolic activity. During the time, it has been shown that root cultures are suitable for production of genetically, biochemically and phenotypically stable plant material without somaclonal variation. The plant cultures, including a callus phase, theoretically promote a higher mutation rate [108]. The first literature data describing centaury root culture appeared almost twenty years ago, when the adventitious buds regeneration was shown to be induced with callus formation on initial root explants [109]. Spontaneous regeneration of adventitious buds in solid root culture was described, for the first time, by Subotić et al. [94]. These results motivated us to use root tips growing on½MS hormone-free medium as initial explants for further spontaneous regeneration and propagation of centaury plants. In our investigations, a solid root culture has been originally proved to be also a suitable and grateful model for genetic transformation of this medicinal plant species using *A. tumefaciens*.

Because of prosperous selection efficiency and adequate regenerative capacity of the transformants, the selection of suitable explants for initial infection with *A. tumefaciens* represents the most crucial step for a successful process of genetic transformation [110]. The potentially transformed centaury shoots formed directly on the root explants, without callus induction and thus minimized possible somaclonal variation [84]. Genomic PCR with specific primers confirmed the successful integration of the *AtCKX1* and *AtCKX2* transgenes into centaury genome in 30% and 28.2% of the analysed hygromycin-resistant lines, respectively. The quantification of transgene expression was confirmed by highly sensitive and precise method, qRT-PCR. As transgenic *AtCKX1* lines showed low expression and unspecific amplification of cDNA, optimization of PCR and qPCR protocols was necessary. The optimized protocol used for real-time quantification of the transgene expression included combining a gradient of annealing temperatures with the application of seven different PCR enhancers: formamide, DMSO, glycerol, ethylene glycol, trehalose, BSA and Tween-20 [111]. The expression of *AtCKX1* and *AtCKX2* transgenes was detected and quantified in 50% and 63.6% transgenic lines, respectively. The *AtCKX1* roots showed slightly higher expression than shoots whereas the *AtCKX2* transgene was better expressed in shoots compared to roots. Evidently, quantitative differences in transgene expression after incorporation into genome of different centaury lines depended on the transgene copy number and positional variations of transgenes.

All of the *AtCKX* transgenic centaury lines showed decreased regeneration capacity in root culture on the solid hormone-free ½MS medium in comparison to control. However, regenerated shoots spontaneously rooted on the same medium. On the other hand, fresh weight of shoots in most of analysed transgenic centaury lines showed no significant difference in comparison to control. In analogy, the root culture of *Dendrobium* orchid overexpressing *DSCKX1* transgene proved reduced shoot regeneration with decreased biomass compared to the non-transformed controls [112]. Similarly, the increased CKX expression affected on shortened internodes and reduced leaf surface of transgenic *Arabidopsis* and tobacco shoots [44].
