Step-by-step protocol for the isolation and transient transformation of hornwort protoplasts.

Premise: A detailed protocol for the protoplast transformation of hornwort tissue is not yet available, limiting molecular biological investigations of these plants and comparative analyses with other bryophytes, which display a gametophyte-dominant life cycle and are critical to understanding the evolution of key land plant traits. Methods and
Results: We describe a detailed protocol to isolate and transiently transform protoplasts of the model hornwort Anthoceros agrestis. The digestion of liquid cultures with Driselase yields a high number of viable protoplasts suitable for polyethylene glycol (PEG)-mediated transformation. We also report early signs of protoplast regeneration, such as chloroplast division and cell wall reconstitution. Conclusions: This protocol represents a straightforward method for isolating and transforming A. agrestis protoplasts that is less laborious than previously described approaches. In combination with the recently developed stable genome transformation technique, this work further expands the prospects of functional studies in this model hornwort.

The first protoplasts were obtained by digesting the tips of tomato (Solanum lycopersicum L.) seedlings using a fungal cellulase (Cocking, 1960). These protoplasts were unstable and underwent cell lysis, releasing intact vacuoles and other cellular contents. Today, more than 60 years later, plant protoplasts are commonly used to determine the localization of proteins, assess their interaction and function with other cellular components/proteins in vivo, and create somatic hybrids using cell fusion, an important tool of strain improvement in plants, fungi, and prokaryotes (Eeckhaut et al., 2013). In addition, transiently transforming protoplasts enables genome editing using CRISPR/Cas9 without the integration of plasmid DNA into the target genome.

While protocols for protoplast isolation have been established for many plant species, they should be adapted and fine‐tuned for every model organism. Hornworts (Anthocerotophyta) represent one of the three monophyletic groups of bryophytes (liverworts, hornworts, and mosses), and are often resolved as being sister to the clade of mosses and liverworts (Setaphytes) (Puttick et al., 2018; de Sousa et al., 2019; Li et al., 2020). The model organisms Marchantia polymorpha L. and Physcomitrium (Physcomitrella) patens (Hedw.) Mitt. have greatly benefited the study of liverwort (Marchantiophyta) and moss (Bryophyta) biology, respectively, for years; however, model species resources for the hornworts have only become available relatively recently (Szövényi et al., 2015). The model hornwort Anthoceros agrestis Paton can routinely be grown axenically under laboratory conditions, be propagated sexually or vegetatively, and its genome has been sequenced (Li et al., 2020; available from https://www.hornworts.uzh.ch/en/hornwort-genomes.html). A range of molecular techniques have also been adapted for the hornworts (Frangedakis et al., 2021b). This establishment of A. agrestis as a tractable model organism is beginning to fill a crucial phylogenetic gap, enabling comparative analyses across all three bryophyte clades. The introduction of a tractable hornwort experimental system is envisioned to provide more accurate insights into fundamental questions of bryophyte and land plant evolution (Rensing, 2017; Szövényi et al., 2021).

Bryophytes differ from vascular plants by having a gametophyte‐dominant life cycle, and all cells of the haploid gametophyte can potentially regenerate into a fully functional new plant (Frangedakis et al., 2021a; McDaniel, 2021). Due to the exceptional regeneration ability of the gametophyte phase, bryophytes can be manipulated without the use of additional phytohormones (Frangedakis et al., 2021b); thus, protoplasts are an attractive platform for bryophyte genome transformation. For the model moss P. patens, several efficient protocols for the isolation and regeneration of protoplasts have been developed since the 1980s and continuously optimized (Grimsley et al., 1977; Jenkins and Cove, 1983; Schween et al., 2003; Ermert et al., 2019; Sugita, 2021); however, the isolation and regeneration of protoplasts in the model liverwort M. polymorpha have scarcely been studied (Ono et al., 1979; Shibaya and Sugawara, 2007). Even less is known about the isolation, transformation, and regeneration of hornwort protoplasts. To date, only three studies have described a method for hornwort protoplast isolation and regeneration in two different species: A. punctatus L. (Takami et al., 1988; Ono et al., 1992) and A. crispulus (Mont.) Douin (Binding and Mordhorst, 1991). Ono et al. (1992) and Binding and Mordhorst (1991) both reported the first chloroplast divisions two days after protoplast isolation, and all three studies reported callus formation and thalli regeneration after approximately 10–14 days and two months, respectively. Nevertheless, all three protocols lack a detailed description of the conditions used, which makes the replication of the described procedures difficult. Furthermore, none of the above‐mentioned studies attempted a transient transformation of the protoplast. We therefore aimed to establish a protocol for the extraction of a considerable number of protoplasts from gametophyte tissues of the model hornwort A. agrestis and their transient transformation with plasmid DNA. Our method not only provides an easy and time‐efficient approach to test the subcellular localization of proteins, but also opens up the possibility of developing several additional techniques that require the transient or stable transformation of protoplasts.

METHODS AND RESULTS

Protoplastation and transformation

Here, we provide a step‐by‐step protocol for the extraction and transformation of hornwort protoplasts. In this publication, we use the word protoplastation, which is sometimes also referred to as enzymolysis. Aiming to develop a simple method, we used a protoplastation protocol available for P. patens as a starting point (Hohe et al., 2004). We used a liquid culture of A. agrestis (Bonn strain) thallus fragments as the starting material, which is the most promising procedure for most plants (Eeckhaut et al., 2013), and achieved a yield of 35,000–65,000 protoplasts·mg–1 of tissue (dry weight; see Appendix 1). By changing several parameters (e.g., using more tissue, increasing the fragment size, and prolonging the duration of the digestion with Driselase from 45 min [described for P. patens] to 12–13 h), we were able to obtain three to six times more viable protoplasts than was previously reported for the hornwort A. crispulus (10,000 protoplasts·mg–1) (Binding and Mordhorst, 1991). A detailed protocol for the protoplastation and transient transformation can be found in Appendix 1.

Protoplast regeneration

A major challenge is the regeneration of a whole plant from totipotent protoplasts. Even though gametophytic cells of the hornwort A. agrestis typically feature a single chloroplast, protoplasts containing two chloroplasts were present throughout our experiments, from right after the digestion to more than 50 d post‐digestion (dpd) (Figure 1). This is an indication of active cell division because chloroplast divisions in hornworts occur right before nuclear, and thus cell, division (Vaughn et al., 1992). Generally, we would expect symmetric cell divisions in the undifferentiated protoplasts, as compared with the regeneration of other plant protoplasts or callus tissue. Although we could only observe evidence of the lead up to symmetric cell division, asymmetric cell division was detected as early as 4 dpd. Asymmetric cell divisions have been observed during regeneration in other plant protoplasts (Cove et al., 1996; Wiszniewska and Piwowarczyk, 2014). We could also observe two chloroplasts per cell, as well as budding, in transiently transformed protoplasts (Figure 1). The regeneration of a functional cell wall could be observed starting at 48 h post‐digestion (Figure 2). We tested for the presence of a cell wall by staining protoplasts that were 5, 24, and 48 h old with Calcofluor white (CFW) which was only successful for the 48‐h‐old protoplasts. The observed regeneration is slower than the 12 h stated for A. punctatus (Ono et al., 1992), and might be the result of the 3 dpd incubation in the dark used in this protocol. The functionality of the cell wall was tested by staining for several components of the cell wall. Furthermore, we observed that protoplasts regenerating cellulose microfibrils (visualized using CFW) did not let the membrane stain FM 1‐43 pass through. This was not the case for shriveled and potentially dead protoplasts (data not shown).

Figure 1

Protoplast isolation and regeneration. (A) Thallus fragment after 13 h of digestion with Driselase. (B) Protoplasts in a counting chamber at a concentration of 2.6 × 106·mL–1. (C) Anthoceros agrestis protoplast with a typical single chloroplast. (D–F) A protoplast at 4 d post‐digestion (dpd) that contains two chloroplasts, potentially entering symmetric cell division. (D) Differential interference contrast (DIC) microscopy image. (E) Red autofluorescence of the two chloroplasts. (F) Merged image of D and E. (G–I) Protoplast containing two chloroplasts at 23 dpd. (G) DIC image. (H) Red autofluorescence of the two chloroplasts. (I) Merged image of G and H. (J–L) A chain of protoplasts, chloroplasts, and potential buds or vesicles at 4 dpd. From the top‐right to the bottom‐left object, these appear to be an extracellular chloroplast, an intact protoplast, an extracellular or potentially shared chloroplast, an intact protoplast, and a putative vesicle or bud. (J) DIC image. (K) Red autofluorescence of the chloroplasts, with the extracellular chloroplast displaying a reduced autofluorescence. (L) Merged image of J and K. (M–O) Budding protoplast potentially undergoing asymmetric cell division at 23 dpd. The arrow indicates the transition of the protoplast to the bud. (M) DIC image. (N) Red autofluorescence of the shared chloroplast. (O) Merged image of M and N. (P–S) Budding of a protoplast transformed with the L2‐MW‐AA42‐CsA plasmid at 10 d post‐transformation. The budding cells share a cell membrane. The arrows indicate the transition to the first bud. (P) DIC image. (Q) Red autofluorescence of the chloroplast located close to the bud. (R) Enhanced green fluorescent protein (eGFP) membrane‐localized signal (AaTip1;1 promoter–driven eGFP fused to the membrane‐localization signal Lti6b). (S) Merged image of P, Q, and R. (T–W) Transformed protoplast (plasmid L2‐MW‐AA42‐CsA) and budding cell sharing a chloroplast at 10 d post‐transformation. The transition zone is indicated with an arrow. (T) DIC image. (U) Red autofluorescence of the potentially shared chloroplast. (V) eGFP membrane‐localized signal. (W) Merged image of T, U, and V. (X–ZII) Protoplast with two chloroplasts (arrows) at 23 d post‐transformation with the L2‐MW‐AA42‐CsA plasmid. (X) DIC image. (Y) Red autofluorescence of the two chloroplasts. (ZI) eGFP membrane‐localized signal. (ZII) Merged image of X, Y, and ZI. The red autofluorescence was detected using a Leica DM6000B Tx2 filter (excitation 520–600 nm, emission 570–720 nm), and the green fluorescence was detected using an L5 filter (excitation 440–520 nm, emission 505 nm). Black scale bar = 100 µm, red scale bar = 10 µm

Figure 2

Transformed and untransformed protoplasts regenerating their cell walls. (A–D) Non‐transformed protoplast 4 d post‐transformation. (A) Differential interference contrast (DIC) microscopy image. (B) Weak autofluorescence of the chloroplast detected in the filter for eGFP fluorescence. (C) Red autofluorescence of the chloroplast. (D) Merged image of A, B, and C. (E–H) Protoplast at 4 d after transformation with the L1‐AA026‐Ck2 plasmid. (E) DIC image. (F) AaTip1;1 promoter–driven eGFP fused to the membrane localization signal Lti6b. (G) Red autofluorescence of the chloroplast. (H) Merged image of E, F, and G. (I–L) Protoplast at 5 d after transformation with the L1‐AA016‐Ck3 plasmid. (I) DIC image. (J) AaEf1α promoter–driven mTurquoise2 fluorescent protein fused to the nuclear localization signal N7. (K) Red autofluorescence of the chloroplasts. (L) Merged image of I, J, and K. (M–P) Regeneration of the hornwort protoplast cell wall at 2 d post‐digestion (dpd). The protoplast on the right shows a cellulose layer stained by Calcofluor white (CFW), while the protoplast on the left does not have a cell wall. (M) DIC image. (N) Cellulose in the cell wall stained with CFW. The protoplast to the left is degrading and its chloroplast is emitting a signal indicating the presence of chlorophyll products. (O) Red autofluorescence of the chloroplasts. (P) Merged image of M, N, and O. (Q–T) Different stages of the regeneration of the cell wall component cellulose at 2 dpd. The large protoplast in the center has two chloroplasts. (Q) DIC image. (R) Cellulose in the cell wall stained with CFW (shown in turquoise). (S) Red autofluorescence of the chloroplasts (arrows). (T) Merged image of Q, R, and S. (U–X) Cell wall regeneration at 52 dpd. In this image, at least four protoplasts share a chloroplast (indicated by the arrows). (U) DIC image. (V) Cellulose in the cell wall stained with CFW. (W) Red autofluorescence of the chloroplasts with the arrows marking the shared chloroplast. (X) Merged image of U, V, and W. The red fluorescence in C and G was detected using a Leica DM6000B Tx2 filter (excitation 520–600 nm, emission 570–720 nm) and the green fluorescence in B and F was detected using an L5 filter (excitation 440–520 nm, emission 505 nm). The red fluorescence in K was detected using a DSR ET filter (excitation 530–560 nm, emission 590–650 nm) and the green fluorescence in J was detected using an ET GFP filter (excitation 450–490 nm, emission 500–550 nm). The images of the cell wall regeneration were taken using a Leica TCS SPE. Blue scale bar = 50 µm, red scale bar = 10 µm

When the protoplasts were left undisturbed in glass tubes for ~50 d, clusters of protoplasts with cell walls could be observed (Figure 2). We tested several conditions and treatments, such as different light conditions after digestion, several protoplast densities in both liquid culture and solid media (BCD medium [Cove et al., 2009] or regeneration medium, solidified with 1% Gelrite [ref. G1101.0500; Duchefa Biochemie, Haarlem, the Netherlands]), as well as supplementing the regeneration medium with the auxin analog 2,4‐dichlorophenoxyacetic acid to enhance protoplast regeneration, with the results summarized in Table 1. Although slight improvement could be achieved, none of these treatments considerably increased the protoplast regeneration rates.

Table 1

Survival rate of Anthoceros agrestis (Bonn strain) protoplasts under various regeneration conditions. The relatively low survival rate is partly due to protoplasts that did not survive the isolation procedure. The initial survival rate right after the isolation was not estimated

Treatment	Survival rate after 9 d (SD)a
Control, Ø 6 cm Petri dish	48% (6.1%)
2 mM 2,4‐D, Ø 6 cm Petri dish	65% (6.1%)
6 mM 2,4‐D, Ø 6 cm Petri dish	58% (11.1%)
10 mM 2,4‐D, Ø 6 cm Petri dish	60% (7%)
Micropore tape Ø 2 cm Petri dish	52% (9.2%)
Parafilm, Ø 2 cm Petri dish	56% (7.6%)

Note: 2,4‐D = 2,4‐dichlorophenoxyacetic acid.

Survival rate is reported as the mean (standard deviation) estimated by counting three biological replicates with ≥100 protoplasts, respectively.

It must be noted that the regeneration of protoplasts is difficult not only in this model plant, but is a challenge in most plant species (Eeckhaut et al., 2013). Further assessment of the regeneration parameters will improve protoplast viability and thus lead to the regeneration of fully functional plants. Multiple studies have assessed the conditions for successful protoplast regeneration, e.g., phytohormone supplements, adjusting protoplast densities, or immobilizing the protoplasts in agar beads or layers of agar (Schween et al., 2003; Pati et al., 2005; Wiszniewska and Piwowarczyk, 2014). We note that not all forms of regeneration treatment have been tested or evaluated thoroughly in the present study, and should be explored in future studies.

Transient transformation

As a high number of protoplasts were achieved, a transient transformation can be performed successfully. We tested the DNA delivery into A. agrestis protoplasts using both polyethylene glycol (PEG) with gentle mixing over a longer period (30 min) and PEG combined with a heat‐shock treatment (10 min incubation with PEG at room temperature followed by 3 min at 45°C). Our results suggest that PEG‐mediated DNA delivery alone is preferable over the frequently used heat shock–mediated approach, which drastically decreased the number of viable protoplasts (results of this comparison not shown). Using the 13,407‐bp L2‐MW‐AA42‐CsA plasmid (Appendix 1), we achieved a 10% protoplast‐transformation rate using the PEG‐mediated approach.

CONCLUSIONS

We established an easy‐to‐use method for isolating and transiently transforming A. agrestis hornwort protoplasts that is simpler than earlier laborious approaches and yields a high number of viable protoplasts. In addition to being the first report on hornwort protoplasts in 30 years, this study provides new insight into the cell wall regeneration of hornwort protoplasts, by showing that the cell wall regenerates between 24 and 48 h post‐digestion, and that the cellulose rebuilds as a tight mesh of microfibrils that eventually becomes impenetrable for the cell membrane staining color FM 1‐43 after 48 h. Furthermore, this protocol opens up new avenues for simple assays of protein–protein interaction, protein localization, protein function, and for CRISPR/Cas9 editing in the model organism A. agrestis. We are currently focusing on developing an efficient method for the regeneration of A. agrestis protoplasts, which is crucial for CRISPR/Cas9 editing.

AUTHOR CONTRIBUTIONS

P.S. coordinated the project. A.N. wrote the manuscript. P.S., E.F., A.B., S.I.N., and S.W. revised the manuscript. F.‐W.L. and P.S. carried out preliminary protoplastation trials. A.N., S.R., and M.W. carried out the isolation and transformation of protoplasts. A.N. obtained the microscopy images. A.B. and A.N. produced the cell wall staining images. E.F. and M.W. developed the plasmids. All authors approved the final version of the manuscript.

Appendix S1. L1_AA016‐Ck3 plasmid sequence.

Click here for additional data file.

Appendix S2. L1_AA026‐Ck2 plasmid sequence.

Click here for additional data file.

Appendix S3. L2‐MW_AA42‐Csa plasmid sequence.

Click here for additional data file.

Click here for additional data file.

Table A1

Details of the plasmids used

Plasmid name	Promoter	Localization tag	Fluorescence marker
L2‐MW‐AA42‐CsA	AaTip1;1	Membrane	eGFP
L1‐AA026‐Ck2	AaTip1;1	Membrane	eGFP
L1‐AA016‐Ck3	AaEf1α	Nucleus	mTurquoise2