2.1 Plant material and irradiation conditions

Geranium cultivars (Pelargonium × hortorum, Bailey) grown under natural light conditions in a greenhouse were used in this study. The buds on young branches (cut and bound together in groups of three or four), were irradiated by vertical 80 MeV/u ¹²C⁶⁺ beams accelerated by the Heavy Ion Research Facility in Lanzhou (HIRFL) at dosages of 0 Gy, 10 Gy, 15 Gy, 30 Gy, and 40 Gy (Fig. 1). The average linear energy transfer (LET) value within each sample as water-equivalent substance was calculated by using SRIM2010 program to be 34keV/μm [27], and the dose rate was 10 Gy/min. A total of 150 excised young branches were treated in this study, namely, 30 branches for each dosage.

Fig. 1Full-screen View

Geranium plants irradiated by carbon ion beams. Black arrows indicate the position of buds in the branches; yellow circles show the radiation field of carbon ion beams (Ф 50 mm). Scale bar = 10 mm.

Irradiated shoots were transferred into 18-cm flowerpots with regular garden soil and grown under natural temperature and illumination in a greenhouse. Five months later, flower mutants were screened from the treated sample populations of geranium. Since 2010, more than three continuous stem-cutting vegetative propagation culture cycles were performed to maintain the mutagenic phenotype. The stable flower mutant in the third cutting propagation cycle M1V3 was used in the present study.

2.2 Observation of epidermal cells

Isolation of epidermal cells was conducted using the plastic Scotch tape adhesion method. Pieces of 10- plastic Scotch tape (15 cm in length) were laid flat on the laboratory bench with the adhesive side up. Fully expanded petals cut from the plants were carefully stuck to one side of the Scotch tape. The tape on the other side of the petal was pinched and then carefully torn. The tape with samples of epidermal cells was cut into 1 cm × 1 cm square pieces and adhered onto glass slides. The distribution of pigments in the epidermal cells of samples placed on glass slides was examined under a conventional optical microscope at 40× magnification (BX53 Olympus, Tokyo, Japan).

2.3 DNA isolation and polymorphism analysis

Genomic DNA was isolated using a standard procedure [28] with minor modifications. Approximately 3 g of fresh leaves were finely ground in liquid nitrogen, and incubated for 1 h at 65°C in pre-warmed 2× cetyl trimethyl ammonium bromide (CTAB) buffer and then washed twice in chloroform:isoamyl alcohol (24:1). After discarding the supernatant, DNA was precipitated with isopropyl alcohol. DNA was rinsed with 75% ethanol twice, air dried, and dissolved in double distilled water (ddH₂O). Equal quantities of DNA from 10 individual M3 plants were pooled for RAPD analysis.

RAPD analysis was performed in a 20-μL mixture that consisted of 20 ng of a DNA template, 10 μM of primers, and a Golden Easy Polymerase Chain Reaction (PCR) System (Tiangen Biotech, Beijing, China), which was mixed with a 2× reaction mix and 0.25 U of Golden DNA polymerase. A total of 45 cycles of PCR was performed in a programmed temperature control system (Thermal Cycler, BIO RAD, California, USA). The PCR conditions were as follows: First, a 5-min pre-denaturation step at 94°C; followed by 40 cycles of denaturation at 94°C for 1 min, annealing at 36°C for 50 s, and elongation at 72°C for 1.5 min; and a final DNA synthesis step at 72°C for 10 min. A total of 55 oligonucleotide primers were used in screening for DNA variations, of which 49 primers (S1) were selected and used in subsequent analyses. Amplified fragments were detected by 1.5% agarose gel electrophoresis containing Goldview I (Solarbio, Beijing, China) for 1.5 h at 90 V. The amplified DNA fragments were detected and analyzed using a chemiluminescence image analysis system (Tanon, Shanghai, China).

PCR analysis using each primer was repeated three times. All amplified bands, which were stably reproducible, were recorded and described as either absent or present (0/1). The degree of polymorphism between the wild-type and mutant plants was estimated by using Jaccard’s similarity coefficient, which was determined by using the formula equation: F = 2Nxy/(Nx + Ny) [29, 30]; wherein Nxy is the number of identical fragments in two samples; Nx is the number of bands detected in the wild-type; and Ny is the number of bands detected in the mutant.

2.4 Extraction and anthocyanin content assay

Anthocyanins were extracted according to the method of Xie [31] with slight modifications. Immediately after collection, fresh petals (0.5 g) of flowers from the wild-type and mutant were ground in liquid nitrogen into a fine powder and suspended in 3 mL of acidic methanol containing 0.1% HCl. After vigorous vortexing, the samples were placed in the dark for 12 h at 4ºC and shaken every 3 h. The homogenized suspension of petal tissues was rotated in a centrifuge at 2,500g at room temperature for 15 min. The supernatant was collected, and the solid was washed with 0.1% HCl in methanol. The extraction steps were repeated twice. The supernatants were pooled and the final volume was adjusted to 10 mL in a volumetric flask. Approximately 2 mL of water and 2 mL of chloroform were mixed with 2 mL of the filtered extract, which was then thoroughly mixed. Next, the mixture was centrifuged at 4,000g for 10 min. The upper methanol/water phase that contained the anthocyanins was separated, and its absorbance was measured using a UV-Vis spectrophotometer (Agilent 8453, Germany) at a wavelength of 524 nm. Anthocyanin content was expressed as A524·g ^–1 FW.

3.1 Mutation induction in geranium

A flower mutant of geranium was obtained from samples irradiated with 30 Gy carbon ions (Fig. 2). The petals of flowers of this mutant were smaller and thinner than those of the wild-type (Fig. 2A) and its color was light pink, whereas that of the wild type was red. In addition, the peduncles of the mutant were red, which significantly differed from that of the wild type. These also had light green pistils and stamens, whereas purple stamens were observed in the mutant (Fig. 2B). The offspring inherited all of these acquired variant phenotypes by cutting vegetative propagation.

Fig. 2.Full-screen View

The floral color mutant obtained by carbon ion irradiation in M1V3. A: Wild-type plant of geranium with red petals and mutant with light pink petals; B: The peduncle and floral organ of the mutant are red and purple, respectively, whereas those of the wild-type are green and white, respectively.

3.2 Assessment of epidermal cells and distribution of anthocyanins

In plant cells, the distribution and quantity of pigments, particularly anthocyanins, affect flower color [32, 33]. To illustrate pigment distribution in epidermal cells, anatomical observation of petals was performed. Fig. 3 shows that the morphology of epidermal cells did not significantly differ between the mutants and the wild-type, which were all nearly hexagonal in shape and anthocyanins were distributed in cell vacuoles and intercellular spaces. Orange and red anthocyanins were distributed in the vacuoles of epidermal cells of the wild-type plant, whereas purple and bluish-purple anthocyanins were rarely observed (data not shown). In addition, the color of the lower epidermal cells was lighter than that of the upper cells (Figs. 3A and 3B). However, the epidermal cells of the mutants were whitish, and colored anthocyanins were hardly observed in the vacuoles (Figs. 3C and 3D).

Fig. 3.Full-screen View

Microscopic assessment of epidermal cells. (A) The upper epidermal cells of a wild-type plant; (B) the lower epidermal cells of a wild-type plant; (C) the upper epidermal cells of the mutant; (D) the lower epidermal cells of the mutant. Scale bars = 50 μm.

No significant differences in the shape of the epidermal cells or the distribution of pigments were observed between the mutant and wild-type plants. Orange and red anthocyanins, which were extensively distributed in the vacuoles of epidermal cells, contributed to the red color of flowers of wild-type plants. In contrast, anthocyanins of lighter color distributed in the vacuoles of the mutant plants rendered a light pink flower color. These results suggested that the difference in petal color between the mutant and wild-type was not associated with the shape of epidermal cells, whereas the absence or decrease in anthocyanins in epidermal cells resulted in the light pink color of the flowers of the mutant plants.

3.3 RAPD analysis of mutants

X-rays, gamma rays, neutrons, and charged particle irradiation can cause DNA strand breaks and clustered lesions [34, 35]. DNA alterations induced by electron and carbon ion irradiations have been reported in Arabidopsis [36, 37]. Molecular genetic markers, including those used in RAPD analysis, have been extensively used in detecting DNA alterations [38]. In the present study, differences in DNA fingerprints between the mutant and wild-type geranium plants were characterized using an RAPD assay. A total of 500 recognizable fragments were amplified using 49 different primers. Moreover, the average number of amplified bands per primer set was 10. Among all recognizable fragments, 265 fragments were obtained from the wild-type, whereas 235 fragments were amplified from the mutant. A total of 217 fragments was shared by the two fragment populations, thereby indicating that a total of 66 bands were polymorphic. Compared to the wild-type, 18 polymorphic bands were identified in the mutant, whereas 48 bands were determined to be residual. The ratio of the polymorphic bands was 3.6% and 9.6%, respectively. Meanwhile, the similarity coefficient between the mutant and the wild-type, which was calculated by using Jaccard’s formula, was 0.868 (Table S2). Using genomic DNA of the mutant as template, primer P13 amplified a total of 12 bands, whereas only one band was obtained with primer S155. No amplification was observed with primers P12 and P10 in the mutant, whereas 5 and 9 bands were respectively amplified in the wild-type using the same primers. Fig. 4 shows representative polymorphic bands of the RAPD assay involving the wild-type and the mutant flowers.

Fig. 4.Full-screen View

RAPD analysis of geranium mutants. M is the marker; the odd lanes are the amplified DNA fragments of the mutant; and the even-numbered lanes are those of the wild-type. P1–P8 refer to the primers used in amplification. The fragments indicated by white arrows are polymorphic bands.

Goodhead earlier reported that heavy ion beams transferred high energy to the target DNA and induced DNA breaks and large fragment rearrangements such as deletions, translocations, inversions, insertions, and their combinations [39]. These findings indicate that heavy ions could be used as a valuable mutagen for the plant research studies involving forward and reverse genetics. DNA polymorphisms between the control and the early maturity sweet sorghum as induced by carbon ion beam irradiations were successfully investigated using the RAPD technique; and the percentage of polymorphism between the control and the mutant was 5.3% [40]. Recently, in vitro mutation breeding of Paphiopedilum induced by carbon ion beams and gamma ray radiation were undertaken. No variant line was found in gamma ray-irradiated samples, whereas 24 variant lines were screened after carbon ion irradiation. The range of RAPD polymorphisms in the genomic DNA was from 57.8% to 75.6% between wild-type and variant lines [41]. In the present study, polymorphism fragments of DNA were detected between the wild-type and the light pink flower mutant using the RAPD technique, and the results proved that carbon ion radiation induces genomic variation and DNA polymorphism in geranium. In addition, the present study has shown that RAPD is an effective method for analyzing variations of genome DNA of mutant plants generated by heavy ion beam radiation mutation breeding.

3.4 Anthocyanin content analysis

On account of the obvious absence of pink color in the petals of the mutant, the anthocyanin contents of flowers of the mutant and wild-type were assessed by UV-Vis spectrophotometry. Fig. 5 shows that the amount of anthocyanins that accumulated in the wild-type was much higher than that in the light-colored mutant. The anthocyanin content of wild-type flowers was 10 A524·g ^–1 FW, which was about 22-fold higher than that in the mutant, which was only 0.46 A524·g ^–1 FW, and was determined to be statistically significant

Fig. 5.Full-screen View

Comparison of anthocyanin contents of wild-type and mutant of geranium flowers. The data represent the averages and standard deviation of three repeat experiments. The student’s t-test was used to determine the statistical significance of the measurements (**P<0.01).

Flower color is generally influenced by the components and quantities of anthocyanins. A previous study has shown that a decrease in anthocyanin content leads to a lighter color [42]. For instance, in gentian, the blue flower of the wild-type plant is associated with high levels of anthocyanins, whereas no detectable pigments were observed in two white flowered cultivars [43]. Saito showed that in white flower tissues of Petunia, a decrease in anthocyanin levels was accompanied by a significant rise in caffeoyl and coumaroyl glucoside concentrations [44, 45]. In the present study, anthocyanin levels in the wild-type were significantly higher than that in the light pink-colored mutant, indicating that the marked decrease in anthocyanin content in the mutant might be responsible for the light pink color of the mutant flowers.

Changes in anthocyanin levels are associated with changes in the transcription of key genes that are responsible for anthocyanidin biosynthesis [24]. During anthocyanin biosynthesis, the CHS gene is downregulated. Nonetheless, downregulation of the CHS gene is not considered as the ideal strategy in developing white-colored varieties. A more feasible preference to induce white-flowered cultivars is to downregulate the F3H or DFR gene [19]. In spinach and pokeweed, deficiencies in anthocyanin accumulation were attributed to the absence in the expression of the DFR and ANS genes [46, 47]. Yamamizo determined that the mutations in flower color mutants of I. quamoclit and I. hederifolia might involve defects in enzymes involved in flavonoid biosynthesis, but not genes that encode transcriptional activation factors. The white flower petals of I. hederifolia and I. quamoclit were also shown to be deficient in DFR and ANS expression [48].Two different white-flowered gentian plants were compared to establish the molecular basis of white flower coloration, and the results showed that the white color of Homoi petals was due to a mutation in the structural gene, ANS, and the absence of one of two ANS loci was confirmed by southern blotting [43]. Taken together, these results suggest that the light pink petal of geranium mutants lack the ability to synthesize anthocyanins, which in turn might be related to the suppression or downregulation of one or more anthocyanin biosynthesis genes. Future studies should thus focus on the transcriptional patterns of anthocyanin biosynthesis genes in this flower color mutant.

In the present study, the color of the peduncles of mutant geranium plants was determined to be red, whereas that of the was green and colorless. The significance of the red coloration in the stem, peduncle, stamen pistil, and other tissues has been examined. Anthocyanins in the peduncle and stamen provide visual information to pollinators and frugivores [49, 50]. Red pigment in stems indicates a response to abiotic stressors such as drought, low temperature, ultraviolet radiation, or a high ratio of red:far-red light. Cooney earlier showed that peduncle reddening reduces the propensity for photoinhibition, resistance to drought, salinity, heavy metals, or oxidative stress [51]. Previous studies have shown that oxidative stress is one of the major damages induced by heavy ion irradiation; therefore, the red peduncle in mutants might be an adaptive response to heavy ion irradiation.

The spatiotemporal expression of structural genes involved in the biosynthesis of anthocyanins is thought to be determined by regulatory genes and their interactions, which include R2R3-MYB, bHLH, and WD40 factors in Ipomoea purpurea, Caryophyllales, and maize [52-54]. In addition, regulatory genes regulate the anthocyanin biosynthesis in various tissues. For instance, in Japanese morning glory, frameshift mutations involving the MYB transcriptional regulator alter the accumulation of pigments in different tissues. The expression of the InMYB1 gene is only detected in flower tubes and limbs, and the InMYB2 gene is expressed in stems, petioles, and sepals. On the other hand, the mutant with an InMYB1 deficiency exhibits white flowers and red stems [55]. Therefore, variations in transcriptional regulators contribute to color variations in mutants that were generated using heavy ion irradiation. Additional researches on transcriptional regulators in the anthocyanidin biosynthesis pathway and their tissue-specific expression in geranium will be conducted by our group in the near future.