Development of a petal protoplast transfection system for Sinningia speciosa.

Premise: Transient gene expression systems are powerful tools for studying gene interactions in plant species without available or stable genetic transformation protocols. We optimized a petal protoplast transformation protocol for Sinningia speciosa, a model plant, to study the development of floral symmetry. Methods and
Results: A high yield of petal protoplasts was obtained using a 6-h enzyme digestion in a solution of 1.5% cellulase and 0.4% macerozyme. Modest transfection efficiency (average 41.4%) was achieved. The viability of the transfected protoplasts remained at more than 90%. A fusion of green fluorescent protein and CYCLOIDEA (SsCYC), the Teosinte branched 1/Cincinnata/Proliferating cell factor transcription factor responsible for floral symmetry, was subcellularly localized inside the nuclei of the protoplasts. Transiently overexpressing SsCYC indicates the success of this system, which resulted in the predicted increased (but nonsignificant) expression of its known target RADIALIS (SsRAD1), consistent with gene network expectations. Conclusions: The transient transfection system presented herein can be effectively used to study gene-regulatory interactions in Gesneriaceae species.

The ornamental plant Sinningia speciosa (Lodd.) Hiern, also known as florist's gloxinia, belongs to the Gesneriaceae family. Different cultivars of S. speciosa have either mostly zygomorphic (bilaterally symmetrical) or actinomorphic (radially symmetrical) flowers, with the developmental transition of zygomorphy to actinomorphy being attributed to the mutation of a single gene, CYCLOIDEA (SsCYC), a Teosinte branched 1/Cincinnata/Proliferating (TCP) cell factor transcription factor that controls floral symmetry (Hsu et al., 2017; Dong et al., 2018). Genes downstream of SsCYC are therefore responsible for the flower shape changes between zygomorphic and actinomorphic S. speciosa accessions. Zygomorphic and actinomorphic S. speciosa can be easily crossed to produce recombinant inbred lines, enabling researchers to use S. speciosa as a model plant to study the floral traits associated with various floral developmental genes (Wang et al., 2015; Hsu et al., 2017). Furthermore, the small nuclear genome (approximately 300 Mbp in the diploid S. speciosa ‘Espírito Santo’), chromosome uniformity (2n = 26), and short generation time (approximately 3–6 months) of S. speciosa make it an excellent model for scientific research (Skog, 1984; Zaitlin and Pierce, 2010). To study the roles of genes downstream of SsCYC in flower development, a stable or transient gene transformation system is required. The Agrobacterium‐mediated stable genetic transformation of S. speciosa has been reported (Xu et al., 2009; Kuo et al., 2018); however, obtaining transgenic plants using this method is a laborious undertaking because of the low transfection efficiency of Agrobacterium in S. speciosa. Developing an Agrobacterium‐free approach in which DNA may be delivered directly into cells is therefore desirable.

Plant protoplasts are cells in which the cell walls have been digested by enzymes. Studying protoplasts can be an effective approach for observing and analyzing cellular processes involved in various biological functions for two reasons (Marx, 2016). First, protoplasts can be isolated from distinct types of plant tissues to explore cell type–specific and tissue‐specific gene expression or function. Second, isolated protoplasts maintain their cellular identity for a certain period of time in vivo (Marx, 2016). Plant protoplast systems therefore facilitate the study of transient gene functions, especially in plant species without well‐established stable genetic transformation systems. Protoplast isolation and transfection systems have been established and optimized for numerous plant species, even non‐model plant species, since their development from the isolation of bacterial and fungal protoplasts and initial use in plant cells in 1960 (Cocking, 1960); however, most efficient protoplast isolation protocols have been developed using leaves or seedlings from model plants and various crops, including Arabidopsis thaliana (L.) Heynh., chickpea (Cicer arietinum L.), Lolium L., Liriodendron L. hybrids, and Phaseolus vulgaris L. (Yoo et al., 2007; Nanjareddy et al., 2016; Huo et al., 2017; Cheng and Nakata, 2020; Davis et al., 2020).

Our goal was to study gene regulation in flowers (i.e., genes downstream of SsCYC), and existing leaf protoplast systems may not be suitable for studying tissue‐specific gene interactions in flowers. It was therefore necessary to establish a petal protoplast isolation and transfection system for S. speciosa. Although protoplast isolation protocols for Gesneriaceae species (e.g., Streptocarpus Lindl. and Saintpaulia ionantha H. Wendl.) are available, the protocols were developed based on small shoot primordia regenerated from cultured leaf blade explants (Afkhami‐Sarvestani et al., 2012) or young shootlets regenerated from leaf explants (Hoshino et al., 1995; Winkelmann and Grunewaldt, 1995a, 1995b). Published protoplast isolation protocols for flower tissues remain scarce, and none are available for Gesneriaceae species. A few reports have investigated the use of petal protoplasts in other ornamental plants, such as Rosa rugosa Thunb. (Borochov et al., 1976), Tulipa L. (Wagner, 1979), and Dendrobium Sw. (Hu et al., 1998), and more recently, petal protoplasts have been used as transient systems to study various genetic pathways in flowers, such as Petunia Juss. hybrids (Faraco et al., 2011) and Phalaenopsis Blume and Cymbidium Sw., two species in the orchid (Orchidaceae) family (Lin et al., 2018; Ren et al., 2020). However, directly adapting the petal protoplast isolation protocols from orchids to S. speciosa may be difficult due to the differences in texture and cell wall properties between the species. To account for these differences, distinct enzyme digestion and gene transfection treatments must be employed in studies of distinct species.

Herein, we describe an efficient protocol for isolating Sinningia Nees petal protoplasts and transforming them with exogenous DNA to examine the transient expression levels of the gene CYCLOIDEA (SsCYC) and determine the subcellular location of its protein. Using petals from Sinningia floral buds, protoplasts with consistent quality (viability) and quantity (yield) can be obtained. In addition, we optimized a large‐scale protoplast transfection procedure to allow for the simultaneous examination of the transient upstream and downstream regulation of multiple genes. These contributions will facilitate future studies of gene interactions in floral tissues of S. speciosa.

METHODS AND RESULTS

Plant growth conditions

Wild‐type S. speciosa ‘Espírito Santo’ (‘ES’) plants with zygomorphic flowers were cultivated in a walk‐in greenhouse at National Taiwan University (Taipei City, Taiwan) under long‐day conditions (16 h light/8 h dark) and an LED‐produced light intensity of 200 μmol·m–2·s–1 at 70–80% relative humidity and 22°C to 25°C. Under these conditions, S. speciosa ‘ES’ plants can flower twice a year (flowering periods: February to April and August to October). Fertilizer (2000× diluted water‐soluble 25N–5P–20 K) was applied every 1–2 weeks during the nonflowering period. During the flowering period, we reduced the fertilization frequency to once every 3–4 weeks because regular fertilization can drastically reduce the transfection efficiency (to <10%). Protoplasts were not isolated from individuals near the beginning or end of the flowering period.

Petal protoplast isolation and protoplast yield

All the chemicals, solutions, and equipment used in this study are listed in Appendix 1. The proposed small‐scale procedures were mainly adopted from optimized protocols for the model plants Arabidopsis and poplar (Populus L.) (Yoo et al., 2007; Lin et al., 2014). We developed the large‐scale procedures to reduce the processing time and labor required in situations where large quantities of RNA must be extracted from transfected protoplasts (e.g., for the analysis of the expression of numerous candidate genes or in transcriptome sequencing). For the small‐scale procedure, 1.5 g of fresh petals from 5–6 pre‐bloom flowers at floral bud stages 10–12 (FB10–12, at which the length of the corolla tube is 21–30 mm; Pan et al., 2022) were collected (Figure 1A). For the large‐scale procedure, 4.5 g of fresh petals from 15 FB10–12 floral buds were used. All the sepals, stamen/staminodes, and carpels were carefully removed and discarded (see Note 1 in Appendix 1). Each petal was cut into 0.5–1‐mm strips using a new sharp scalpel blade, without crushing the petal tissue at the cutting site (Figure 1B, Note 2 in Appendix 1). The petal strips were immediately fully submerged in a freshly prepared enzyme solution (Figure 1C, Note 3 in Appendix 1), before being vacuum‐infiltrated (30–40 mm Hg) at room temperature (23°C to 26°C) in the dark for 30 min. The enzyme digestion was performed for 5–6 h at room temperature in the dark without shaking (Notes 4 and 5 in Appendix 1). After digestion, the enzyme solution gradually turned purple due to the release of purple petal protoplasts (Figure 1D). The shapes of the released protoplasts in the enzyme solution were examined using a microscope, and were confirmed to be intact and round, and approximately 30–40 μm in diameter. To remove the petal debris, the protoplast solution was filtered through a double layer of 75‐μm nylon mesh, which had been cleaned with ddH2O to remove excess water and wetted with MMG solution before filtration. The filtered protoplasts were collected in a 30‐mL round‐bottomed tube (Figure 1E, total volume 20–30 mL, Note 6 in Appendix 1). The tube was centrifuged for 1–2 min with a swinging‐bucket rotor at 200 × g to pellet the protoplasts; the protoplast pellet was visible at the bottom of the tube (Figure 1F, arrow). To harvest the protoplasts, the supernatant was removed as fully as possible, without disturbing the pellet, using a pipette (Note 7 in Appendix 1). After resuspending the protoplast pellet in a small volume of MMG solution by gentle swirling, the cells were counted using a hemacytometer, and then the final concentration of protoplasts was adjusted with MMG solution (small‐scale procedure: 2 × 105/mL, large‐scale procedure: 4.16 × 106/mL) (Figure 1G, Note 8 in Appendix 1). The tube with the resuspended protoplasts was placed on ice for at least 30 min. The isolated petal protoplasts remained intact and round, and some of them contained large purple vacuoles at their centers (Figure 1H), indicating that the protoplasts were viable and healthy.

Figure 1

Overview of procedures for isolating petal protoplasts of Sinningia speciosa ‘Espírito Santo’ (‘ES’). (A) Floral buds harvested at floral development stages FB10–FB12. (B) Petals were cut into strips with widths of 0.5–1.0 mm using a scalpel blade. (C) Petal strips submerged in enzyme solution (after 0 h of digestion). (D) Release of protoplasts from petal strips after 6 h of enzyme digestion. (E–G) Steps of the large‐scale protoplast isolation protocol, performed in a 30‐mL tube. (E) Protoplasts washed with MMG solution. (F) Protoplast pellet (arrow) visualized after centrifugation. (G) Protoplast pellet resuspended in MMG solution. (H) Isolated petal protoplasts. Scale bars = 1 cm for A–D; 50 μm for H.

Enzyme concentration and digestion time are key to obtaining a high yield of petal protoplasts (Huang et al., 2013; Wu et al., 2017; Li et al., 2018). We thus compared the yield of protoplasts obtained using an enzyme concentration of 1× (1.5% cellulase and 0.4% macerozyme) and 2×, with digestion times ranging from 3 h to 6 h at room temperature in the dark without shaking. The use of the small‐scale procedure with a 6‐h digestion time and 1× enzyme concentration produced a yield of 2.84 × 106 protoplasts/g fresh weight (FW; Appendix 2A) or more than 1 × 106 protoplasts/g FW from each isolation procedure. When the large‐scale procedure was conducted using the same enzyme treatment, the average yield was approximately 2.95 × 106 protoplasts/g FW. Studies have reported that the protoplast yield from petals (105–106/g FW) is usually lower than that from leaves (approximately 3.0 × 107/g FW; Wu et al., 2009). Furthermore, most studies have reported low to moderate yields of petal protoplasts, such as 9.5 × 105/g FW from Phalaenopsis orchids (Lin et al., 2018), 2 × 105/g FW from Phaseolus vulgaris (Nanjareddy et al., 2016), and 1.5–8 × 106/g FW from Petunia hybrids (Oh and Kim, 1994). One exception was a study of Cymbidium orchid petal protoplasts, which reported a yield of 3.5 × 107/g FW (Ren et al., 2020). Because our protocol helped achieve moderate yields of S. speciosa petal protoplasts (1–2.95 × 106/g FW), we determined that the 6‐h digestion time and 1× enzyme concentration were the optimal conditions for protoplast isolation and continued to use them in our subsequent small‐ and large‐scale tests.

DNA–polyethylene glycol–calcium transfection

We transfected the pUC19 plasmid with green fluorescent protein (GFP) driven by the 35S promoter (pUC19‐35S:GFP) to evaluate the transfection efficiency of the protoplasts. We calculated the transfection efficiency by dividing the number of protoplast cells displaying GFP fluorescence under a dark‐field UV microscope by the total number of protoplast cells observed under a bright‐field microscope. To conduct small‐scale transfections, 100 μL of protoplasts (2 × 104 protoplasts) were aliquoted into a 2‐mL round‐bottomed tube and gently mixed with 10 μL (1 μg/μL) of plasmid DNA. To conduct large‐scale transfections, 1.6 mL of protoplasts (approximately 6.65 × 106 protoplasts) were added to a 30‐mL tube and mixed gently with 200 μL (1 μg/μL) of plasmid DNA. Freshly prepared polyethylene glycol (PEG)‐CaCl2 transfection solution (110 μL and 1.76 mL in the small‐ and large‐scale procedures, respectively) was added to the protoplasts and immediately mixed by gently tapping the tube. The transfection mixture was incubated at room temperature for 10–15 min; transfection was stopped by diluting the mixture with W5 solution (440 μL and 7.04 mL in the small‐ and large‐scale procedures, respectively) at room temperature and gently inverting the tube to mix. Equal volumes of the solution were dispensed into 16 new 2‐mL round‐bottomed tubes. In both the small‐ and large‐scale protoplast transfection procedures, the tubes were centrifuged at 200 × g for 2 min at room temperature, after which the supernatant was removed. To produce a plate‐coating buffer, 5 mL of 5% BSA was dispensed on the plastic surface of a 12‐well culture plate for a short time (1–2 s). This buffer prevented the protoplasts from sticking to the plate. The transfected protoplasts were resuspended in 0.5 mL of WI solution, transferred into the wells of the coated culture plate, and incubated at room temperature under light (daylight fluorescent lightbulb) for 16–18 h. The transfection efficiency was then calculated.

Optimizing the PEG concentration and amount of exogenous DNA is crucial for successful transfection (Huang et al., 2013; Nanjareddy et al., 2016; Wu et al., 2017; Li et al., 2018; Page et al., 2019); PEG facilitates the introduction of exogenous DNA (i.e., plasmids) into the targeted cell by adjusting the osmotic pressure of the cell membrane, thereby increasing its permeability. To determine the effect of PEG 4000 concentrations on transfection efficiency, the protoplasts were transfected with 40% or 50% PEG 4000 in combination with digestive enzymes (1× or 2× concentration; Appendix 2B). After a 16‐h incubation, the transfection efficiencies of the four combined conditions were between 44.3% and 56.1%. Under the combination of 50% PEG 4000 and 2× concentration conditions, the highest transfection efficiency (56.1%) was achieved. However, the digestive enzymes with 1× and 2× concentration under 40% PEG 4000 resulted in a similar high transfection efficiency (49.9% and 51.2%, respectively). Hence, 40% PEG 4000 and 1× digestive enzyme concentration was therefore used in subsequent tests.

To assess the effect of the amount of plasmid DNA on transfection, the protoplasts were transfected with different amounts of plasmid DNA. The transfection efficiencies obtained using 10, 15, and 25 μg were 49.9%, 35.9%, and 51.1%, respectively (Appendix 2C). A transfection efficiency of nearly 50% was achieved using 10 μg of plasmid DNA; therefore, we used 10 μg of plasmid DNA in our subsequent tests for cost‐efficiency. Accordingly, the key factors required to achieve optimal transfection in S. speciosa petal protoplasts can be summarized as follows: a 1× concentration of digestive enzymes, 10 μg of exogenous DNA, and 40% PEG 4000.

Moderately high transfection efficiency for petal protoplasts

The high petal protoplast yield of this protocol means that a moderate transfection efficiency (maximum 50% and 58% using the small‐ and large‐scale procedures, respectively) can be achieved (Table 1, Figure 2A, B). The average efficiency (41.40% ± 7%, range: 30–50%, 35S:GFP) obtained in the small‐scale transfections was higher than that obtained through large‐scale transfections, the efficiencies of which were similar to the small‐scale protocol but considerably more variable. In the large‐scale procedures, the transfection efficiencies achieved using 35S:GFP and 35S:SsCYC‐GFP were 20–58% and 19–42%, respectively (Table 1). Fluorescein diacetate staining revealed that the average viabilities of the petal protoplasts transfected using the small‐ and large‐scale procedures were 83.15% and 75.13%, respectively (Table 1, Figure 2C, D). The transfection efficiency for S. speciosa petal protoplasts in our study is almost equivalent to that reported for Arabidopsis leaf protoplasts (50%) in previous studies (Yoo et al., 2007; Wu et al., 2009); however, recent studies have reported exceptionally high transfection efficiencies (approximately 80%) for petal protoplasts of Cymbidium (Ren et al., 2020) and Phalaenopsis (Lin et al., 2018). One possible explanation for this discrepancy may be differences in the extracellular matrices or the cell wall residues in the petal protoplasts of orchids and S. speciosa.

Table 1

Yield, cell viability, and transfection efficiency of the petal protoplasts isolated using our small‐scale and large‐scale procedures.

Transfection scale	Protoplast yielda	Cell viabilityb	Construct	Transfection efficiency
Small scale	1.21 × 106 (n = 2)	83.15 ± 7% (n = 2, 197/246; 256/298)	35S:GFP	41.40 ± 7% (30–50%, n = 6)
Large scale	2.95 × 106 (n = 22)	75.13 ± 7.2% (n = 6, 124/168; 187/232; 113/147; 119/162; 114/166; 159/205)	35S:GFP	30.55 ± 9% (20–58%, n = 18)
			35S:SsCYC‐GFP	30.30 ± 6% (19–42%, n = 17)

Protoplasts/g fresh weight.

Cell viability = the number of cells stained with fluorescein diacetate divided by the total number of cells.

Figure 2

Transfection efficiency, viability, and gene regulation analysis of transfected Sinningia speciosa ‘ES’ petal protoplasts. (A, B) Images of GFP‐transfected petal protoplasts taken under (A) bright light and (B) ultraviolet light to quantify transfection efficiency. (C, D) Images of transfected protoplasts stained with fluorescein diacetate under (C) bright light and (D) ultraviolet light to determine cell viability. (E–G) Subcellular localization of SsCYC–GFP protein in a transfected petal protoplast. (F) Fluorescence of SsCYC‐GFP protein under dark‐field microscopy. (F) Fluorescence of 4′,6‐diamidino‐2‐phenylindole dihydrochloride (DAPI) staining under dark‐field microscopy. (G) Transfected petal protoplast under bright‐field microscopy. Scale bars = 10 μm. (H) Validation of SsCYC overexpression in three paired petal protoplast replicates transfected with pUC19‐35S:SsCYC‐35S:GFP (OX1–OX3) compared with SsCYC expression in protoplasts transformed with vector‐only pUC19‐35S:GFP constructs (Vector1–Vector3) and untransfected protoplasts (Untransfected). (I–J) Expression of SsRAD1 (I) and SsRAD2 (J), putative targets of SsCYC, in petal protoplast lines overexpressing SsCYC (OX1–OX3). Normalized mRNA expression levels were calculated as 2−ΔCt. The mean values ± SD are from technical repeats. Statistical analyses were conducted using Student's t tests. *P ≤ 0.05.

Use of the petal protoplast transient expression system to examine gene regulation and protein subcellular localization

To test the usability of this system, we transfected the petal protoplasts with a gene encoding a floral symmetry transcription factor (SsCYC) fused with GFP and observed the subcellular localization of the fused protein. If the SsCYC–GFP signals can be observed entering the protoplast nuclei, the petal protoplast system is viable and can be used to study transient gene expression. We therefore introduced a pUC19‐35S:SsCYC‐GFP construct into petal protoplasts of S. speciosa using our protocol. After an overnight incubation for 16 h, GFP fluorescence was detected in the nucleus of the petal protoplasts, which were stained using 4′,6‐diamidino‐2‐phenylindole dihydrochloride, indicating their successful transformation with the ectopic SsCYC gene (Figure 2E, F). Our protocol is therefore useful for the detection of the expression of transiently transfected genes.

The CYC genes are master regulators of floral symmetry and have been reported to both autoregulate and to activate RADIALIS expression in the dorsal petals of zygomorphic flowers in S. speciosa and other Gesneriaceae species, as well as in Antirrhinum L. (Costa et al., 2005; Yang et al., 2012; Dong et al., 2018). To further verify that this system can be used to study well‐characterized upstream–downstream genetic regulatory relationships, such as the ones between CYC and RAD genes, we investigated whether the overexpression of SsCYC in petal protoplasts could alter the expression of SsRADs. To this end, we transfected the effector construct, pUC19‐35S:SsCYC‐35S:GFP (SsCYC OX), into the S. speciosa petal protoplasts and quantified the transcriptional levels of the SsRAD genes SsRAD1 and SsRAD2, both of which have TCP binding sites on their 5′ regulatory regions, using real‐time quantitative PCR (Figure 2H–L). We first performed three independent SsCYC‐overexpression experiments in petal protoplasts (OX1–OX3), in which the SsCYC expression level in the transfected protoplasts was significantly higher than those in the empty vector controls (vectors 1–3) and in the untransfected protoplasts (negative control; Figure 2H). The expression of SsRAD1 was higher, but not significantly upregulated when SsCYC was overexpressed, whereas the expression of SsRAD2 did not differ between the three SsCYC‐overexpression lines and the controls (Figure 2I–J). The significant overexpression of SsCYC after transfection indicated the success of our petal protoplast transient expression system. In addition, the induced upregulation of its known target SsRAD1, while not significant, is consistent with the expectations in the SsCYC‐overexpressing lines, which demonstrates the utility of this system in studying the corresponding gene regulatory pathway.

CONCLUSIONS

We optimized a petal protoplast transformation protocol for S. speciosa ‘ES’ that was developed for both small‐scale and large‐scale procedures. The quality and quantity of the isolated petal protoplasts of S. speciosa are sufficient for DNA transfection and subsequent examination of protein subcellular localization and regulation of SsCYC transcription factor to SsRAD gene expression. Other recently reported methods have used similar modifications of existing protocols from the model plant Arabidopsis to optimize protoplast isolation and transfection from various plants, but our protocol implements attributes of the protoplast isolation of S. speciosa; for example, the 1× concentration of the digestion enzyme is cost‐effective and successfully releases abundant petal protoplasts, and at least 5 to 8 h of incubation time is needed for enzyme digestion to ensure the high yield of protoplasts. The protocol can be easily scaled up to fit the experimental needs of researchers to study a wide range of cellular processes, such as protein–protein and protein–DNA interactions, transcriptional activity, and signal transduction. We expect that our method will be of broad utility in diverse non‐model plant species, particularly in plant species that have previously presented challenges in stable gene transformation.

AUTHOR CONTRIBUTIONS

Z.J.P, Y.L.H., T.Y.C., and Y.A.S. performed the experiments and analyzed the data. Z.J.P, Y.L.H., Y.C.J.L., and C.N.W. conceived the protocol and designed the experiments. Z.J.P, Y.L.H., and C.N.W. wrote the manuscript. All the authors approved the final version of the manuscript.