Technological developments
Fungi are characterized by a large diversity of life forms. In order to establish a broadly applicable platform for their manipulation, we selected four model organisms that cover their main distinctive morphological features: the unicellular fungi Saccharomyces cerevisiae and Schizosaccharomyces pombe, which are the two major models for budding and fission yeasts, respectively; the dimorphic yeast Candida albicans, an opportunistic pathogen that can switch from a unicellular budding growth to a multicellular mycelial growth under specific environmental conditions; and Coprinopsis cinerea, a multicellular filamentous fungus and model organism for studying fungal development and the evolution of multicellular fungi. S. cerevisiae, S. pombe and C. albicans belong to the phylum Ascomycota whereas C. cinerea is a representative of the phylum Basidiomycota of the kingdom Fungi19.
The general workflow for injecting or extracting fluid into and from living cells using FluidFM consists of the following steps: (1) positioning the pyramidal FluidFM tip above a selected cell; (2) driving the tip through the cell wall and membrane using force spectroscopy; (3) maintaining the tip in the cell at a constant force while applying over- or underpressure for injection or extraction, respectively; (4) withdrawing the tip from the cell (Supplementary Fig. 1a). This approach works reliably with mammalian cells14. However, it is not readily applicable to fungal cells due to their unique physical properties, particularly the rigid cell wall.
The insertion of the FluidFM tip into a cell requires that the cell is immobilized on a hard substrate. While most animal cells spread and adhere to solid substrates, the adhesion of fungal cells is comparatively weak20,21,22. Therefore, we first evaluated whether fungal cells maintained a fixed position upon indentation as a prerequisite for injection. On flat substrates, the spherical yeast S. cerevisiae and the rod-shaped yeast S. pombe were laterally displaced upon contact with the FluidFM tip. We therefore designed microstructured substrates and could show that the yeast cells were spatially constrained (Supplementary Figs. 2, 3, “Methods”). In contrast, the hyphae of the dimorphic yeast C. albicans and of the filamentous fungi C. cinerea adhered sufficiently on glass and plastic polystyrene, respectively, to prevent cell displacement upon tip insertion.
Furthermore, the cell wall surrounding fungal cells results in a drastically increased cell rigidity, causing an increase in the forces required to insert a probe into the cells compared to animal cells (Supplementary Fig. 1b). The high stiffnesses of fungal cells were reflected in the force-distance curves obtained by indentation and eventual insertion of the probe inside fungal and animal cells (Fig. 1b). For effective probe insertion into a typical adherent mammalian cell (e.g., HeLa), the maximal force was set such that the tip crossed the entire cell to the underlying substrate, which typically required forces up to 200 nN14. In comparison, the force-distance curves on fungal cells showed that, while the indentation distances were similar or shorter than for HeLa cells, the forces required for maximal indentation were 10 times larger, i.e., up to 2 μN. FluidFM probes were thus re-designed, optimizing their length and thickness to increase their stiffness to ~2.3 ± 0.5 N/m, so that forces up to several micro-Newtons could be exerted.
Finally, cell size also represents a crucial aspect for the development of an injection protocol for fungal cells. It determines the upper injectable volumes, especially due to the cell wall, which provides a physical barrier to cell expansion. We estimated the cell volumes of the four selected fungi, in comparison to human HeLa cells (Fig. 1c). The mean (SD) cell volumes of the yeasts C. albicans, S. pombe and S. cerevisiae were 75 (50), 105 (25) and 80 (55) fL, respectively. The average volume of hyphal compartments of the mushroom C. cinerea was 360 (290) fL. In comparison, HeLa cells had a mean volume of 2500 (950) fL, approximately ten- to thirtyfold larger. Because the volumes injected into animal cells were typically on the order of hundreds of femtoliters14, we anticipated that volumes injectable into fungal cells will be on the order of tens of femtoliters. Cell size is also relevant in the context of proper probe insertion, as the probe aperture must be sufficiently narrow and close to the pyramid apex to be fully within the cell just upon puncturing. Yet, at the same time, the aperture area should be large enough for sufficient flow. We thus shaped probe apertures with focused ion beam embracing two of the four sides of the tip, with aperture sizes down to 200 nm in height and width (Fig. 1d).
Fluorescent tracer injection
To assess the ability to cross the cell wall and inject solutions into fungal cells and to adjust the injection parameters, we first used the membrane-impermeable fluorescent tracer Lucifer yellow (LY). We performed force spectroscopy on the fungal cells and subsequently monitored LY fluorescence in real time during and after applying a pressure difference of +1000 mbar for several minutes. Using a fluorescent probe allowed detection of both cell staining resulting from a successful injection and extracellular leakage in the event of a compromised tip insertion (Supplementary Fig. 4). Consistent with the force-distance curves (Fig. 1b), the application of a force of 2.0 ± 0.1 μN resulted in 100% successful injection of the dye into fungal cells without apparent leakage (N = 9), indicating full insertion of the probe aperture into the cell (Fig. 2). Of note, when we tested the injection with lower forces, i.e., 1.0 ± 0.1 μN and 1.5 ± 0.1 μN, this resulted in intracellular staining, but with concomitant leakage of the fluorophore into the extracellular environment in 100% of the cases (N = 11), and 83% (N = 18), respectively. The simultaneous release of LY both inside the cell and in the extracellular environment indicated effective cell wall pervasion but incomplete insertion of the probe aperture. With applied forces lower than 1.0 ± 0.1 μN, the cell wall was not pierced, and the fluorophore was released entirely in the extracellular medium. These results highlight the need to apply sufficient force for robust injection into fungal cells, which we achieve with 2 µN and represents an important step from earlier studies18.
a Bright field and fluorescence images of the injection of lucifer yellow (LY) into S. cerevisiae, before, during, and after the injection. The images show a single oval microwell (in the center) containing two yeast cells; the right cell in the microwell was injected and was labeled, while the adjacent cell to its left was not injected and remained unlabeled. b S. pombe injected with LY. c Germ tube (left), hyphae (center left, center right), and pseudo-hyphae (right) of C.albicans injected with LY. d LY injections in C. cinerea hyphal compartments. The white arrows in (b), (c), and (d) indicate the injection sites. The asterisks in (c) and (d) indicate the vacuoles. The orange arrows in (d) indicate the septa of the injected hyphal compartment. Scale bars are 20 µm.
Following injection, the unicellular yeasts S. pombe and S. cerevisiae were uniformly labeled, whereas neighboring yeasts remained unlabeled (Fig. 2a, b). Although homogenous staining may also result from extracellular binding of the fluorophore to the cell wall, the selective labeling of the targeted but not the neighboring yeasts indicated successful delivery of the fluorophore inside the cell.
Unequivocal intracellular targeting of the fluorophore was achieved by injection of hyphal and pseudohyphal cells of C. albicans (Fig. 2c). While single cells of the dimorphic yeast are similar in size to S. pombe and S. cerevisiae (see Fig. 1c), C. albicans forms multicellular structures under certain environmental conditions. Upon injection of hyphal germ tubes, a homogenous staining of the entire structure was observed, as expected. In fact, germ tubes are hyphal projections that develop in the first cell cycle, before septation. When injecting hyphal compartments, the released fluorophore selectively stained the targeted compartments and did not cross the mature septa. In addition, injection of pseudo-hyphae resulted in labeling of some but not all compartments, in agreement with the dye passing through primary but not mature septa. In addition, vacuoles in the injected compartments remained unlabeled, validating the cytoplasmic location of the injected fluorescent tracer.
Similarly, in C. cinerea mycelia, the injected fluorophore dispersed throughout the targeted hyphal compartments, included their branches, but did not pass through the septa delimiting the cells, indicating effective intracellular injection (Fig. 2d). The intracellular delivery of LY was further validated by injecting vacuolated hyphal compartments, whereby the injected dye clearly located in the cytoplasmic fluid and was excluded from the vacuoles (Fig. 2d, right panels).
Next, we estimated the volumes of solution that we injected based on the fluorescence intensities of the injected cells. The volumes were similar for the four organisms and ranged between 1 and 100 femtoliters (Supplementary Fig. 5). We note here that eventual extracellular staining of the cell wall may result in additional fluorescence and the injected volumes may thus be slightly overestimated. For comparison, volumes up to 900 fL were injected in HeLa cells14. The estimated injected volumes were thus consistent with the respective sizes of fungal and animal cells described above.
Regarding the pressure pulse applied to release the fluorophore solution in the cell, we applied a pressure difference up to more than 1000 mbar, during several minutes. In mammalian cells, such pressure pulses resulted in injection volumes larger than 1 pL, which led to inflation, membrane rupture and cell death. In the fungal cells, no cell deformation was observed, consistent with the rigid cell wall maintaining cell shape and the high internal pressure of the cells.
Post-injection cell viability
The results described above showed that injection into fungal cells is indeed feasible. The next set of experiments was aimed to assess whether the method preserves cell viability. We therefore monitored the growth of the fungi following injection. In addition, we used a fluorescently-labeled histone H1 (H1-AF488) injected into S. cerevisiae and C. cinerea. When the exogenous nuclear protein is delivered into the cell, it is expected to be actively transported to the nucleus in fully functional cells.
In the unicellular S. cerevisiae, the injected histone protein effectively accumulated in the nuclei, while a diffuse fluorescence signal was observed in the cytoplasm and cell wall; moreover, the fluorescent protein was excluded from vacuoles (Fig. 3a). After injection of a budding mother cell, the labeled histone was transported into the nucleus of the mother and the daughter cells, with a lower fluorescence intensity in the daughter cell (Fig. 3b). Monitoring of the injected yeasts by time-lapse microscopy showed that the injected cells were moving, and continued to produce a daughter cell in the hours following injection (Fig. 3c). After the budding event, the labeled histone was observed in both nuclei of the mother and daughter cells (Fig. 3d).
a H1-AF488 injected in S. cerevisiae accumulated in the cell nucleus (orange arrow), and was excluded from the vacuole (darker). The white arrow shows the injection site. b Following injection of a budding S. cerevisiae (lower cell), the labeled histone accumulated in both the lower and the upper cell nuclei (orange arrows). The white arrows show the injection site. c Time-lapse brightfield images of S. cerevisiae after injection. The first raw shows the growth post-injection of a single yeast. Budding was observed 1 h after the injection (blue arrow), and 2 cells were then observed on the following time frames. The second raw shows the growth of a yeast that was budding at the time of injection (same cell as in b). The injected yeast continued its growth, budding ~2.5 h after injection (blue arrows), and resulting in 4 cells at later time frames. d After division of an injected single cell, the labeled histone was found in the nuclei of both the mother and the daughter cells. The white arrows show the injection site, the orange arrows show the labeled nuclei. All the images were created by summing the slices of a Z-stack. Scale bar: 5 µm.
Similarly, in the filamentous fungi C. cinerea, the injected histone proteins were effectively transported into the nuclei, while slightly labeling the walls of the cells (Fig. 4a). The homokaryon strain of C. cinerea most frequently contains two nuclei per hyphal segment23. Consistent with this, we mostly observed the staining of two nuclei per injected hyphal compartments, with a weaker stain for the nucleus most distant from the injection site (Fig. 4a). Vegetative mycelia of C. cinerea grow by apical extension of hyphae and the formation of subapical branches, each of which becomes an apically elongating hypha. To evaluate post-injection growth, we injected hyphal tip cells and subsequently monitored them by time-lapse microscopy (Fig. 4b). Apical growth was observed for all injected hyphal tip cells, with a similar elongation rate as for non-injected hyphae. Upon mycelial growth, the injected H1-AF488 remained in its location and observable for at least 2 days.
a H1-AF488 injected in C. cinerea accumulated in the cell nuclei (orange arrows). Two representative injections are shown. The white arrow shows the injection sites. b Following injection of a hyphal tip, time-lapse monitoring showed the growth of the injected cell, similar to the surrounding hyphae. The injected H1-AF488 remained at its location along the whole time-course. The fluorescence images in (a), and the brightfield and fluorescence images at 0, 24, and 48 h in (b) are summed slices of Z-stacks. Scale bar: 20 µm.
The observation that the injected histone protein accumulated in nuclei not only validated the successful intracellular injections, in both uni- and multicellular fungi, but also demonstrated the short-term preservation of cellular functions. Further monitoring of fungal growth after injection confirmed the absence of adverse effect on cell viability.
Extraction
Next, we evaluated the ability to extract cellular fluid from fungal cells. We used an oil-prefilled FluidFM probe to allow confinement and visualization of cytoplasmic fluid extracted into the probe. Tip insertion into C. cinerea hyphal compartments was performed as for injection, but instead of applying overpressure, we applied underpressure for the suction of the cell content. When exerting underpressure, we effectively observed the collection of fluid in the cantilever probe (Fig. 5a, inserts). When targeting a hyphal tip cell, continued growth of the hypha was observed after tip insertion, but stopped after aspiration of the cytoplasmic fluid into the probe (Fig. 5a, blue arrows). The extraction was accompanied by the appearance of vacuoles in the targeted hyphal compartment. No morphological changes were observed in the adjacent hyphal compartments (Fig. 5a). These observations indicated successful extraction of cytoplasmic fluid.
a Brightfield images showing the extraction of cytoplasmic fluid from C. cinerea. Upon application of suction, the cellular fluid is collected in the front of the FluidFM probe (inserts). The targeted hyphae was growing until suction was applied for extraction (blue arrows). Vacuoles are visible after the extraction, exclusively within the targeted compartment. b Extraction (0 h) and subsequent monitoring of the dTomato fluorescent reporter. c Quantified dTomato fluorescence intensity of individual hyphal compartments over time after extraction. Red curves profiles were obtained from extracted hyphal compartments (N = 5). Gray curves are dTomato expression profile from control, non-perturbed hyphal compartments within the same mycelia (N = 5). Scale bars: 20 µm.
Examining single cells within a multicellular fungus
After demonstrating extraction from individual hyphal cells, we applied this method to address an open question in C. cinerea research. The filamentous fungus represents a prey for fungivorous nematodes, such as Aphelenchus avenae24. It has been shown that A. avenae feeds on C. cinerea by perforation of the cell wall with its stylet and suction of the cellular fluid25. This attack in turn triggers a defense response of C. cinerea that involves the expression of the defense gene cgl225,26. Previous results suggest that the physical cue of penetration and suction is necessary to trigger cgl2 expression but it remained unclear whether it is sufficient or whether chemical cues, such as the concomitant release of molecules by the fungivorous nematode, are needed27. We hypothesized that extraction with the FluidFM could mimic nematode feeding and thus, allow us to test whether hyphal content removal is sufficient to induce the fungal defense response. To address this question, we used a reporter strain of C. cinerea carrying a dTomato expression cassette driven by the promoter of the cgl2 gene25. Volumes ranging from 210 fL up to 1750 fL were extracted from different hyphal compartments, and the mycelia was then monitored overnight for the expression of the fluorescent reporter. Time-lapse fluorescence microscopy after extraction showed an increasing dTomato signal, selectively in the hyphal compartments that had been extracted (Fig. 5b). Quantitative measurements of the fluorescent reporter confirmed a steady increase of the reporter fluorescence over time within hyphal compartments that had been extracted, whereas neighboring cells, which did not undergo extraction, did not express the fluorescent reporter (Fig. 5c). This suggests that the suction of cytoplasm was sufficient to induce a defense response in C. cinerea. As pointed out above, we had observed that single hyphal cell extraction resulted in growth arrest. It is thus interesting to note that the cells remained viable and maintained the capacity to express the fluorescent reporter protein, despite the important loss of cytoplasmic fluid. A previous study revealed that upon localized challenge with nematodes, the fungal defense response was generally confined to the site of predation, but it propagated over several millimeters in a distinct subset of specialized hyphae, characterized by a large diameter25. While such “trunk” hyphae were not targeted in the present study, the FluidFM extraction approach will provide an attractive tool to investigate the propagation of the fungal defense response at the single-cell resolution.
