Trade-off between Plasticity and Velocity in Mycelial Growth

Tip-growing fungal cells maintain cell polarity at the apical regions and elongate by de novo synthesis of the cell wall. Cell polarity and tip growth rate affect mycelial morphology.

or tubulin polymerization. Then, directed trafficking of vesicles and carriers contribute to local membrane and cell wall expansion. In addition, cell growth is controlled by turgor pressure, which drives the expansion of the cell cortex, especially in cell types covered by a cell wall (5,6). Although both polarity and growth are essential for cell morphology, how growth speed and cell polarity cooperatively control cell shape remains unclear.
Filamentous fungi grow as highly polarized tubular cells that elongate through the continuous supply of membrane lipids and de novo synthesis of cell wall at the extending tip (7)(8)(9)(10)(11). The necessary proteins and lipids are delivered to the tip by vesicle trafficking via the actin and microtubule cytoskeletons and their corresponding motor proteins (12)(13)(14)(15)(16). The delivered secretory vesicles accumulate temporarily in an apical vesicle cluster, called the Spitzenkörper (SPK) (17)(18)(19). Vesicle exocytosis at the apical membrane allows release of secretory enzymes and the expansion of apical membrane and cell wall. Recent live-imaging analyses, including superresolution microscopy, have revealed that the multiple steps in polarized growth, such as the assembly of polarity markers, actin polymerization, and exocytosis, are temporally coordinated through pulsed Ca 21 influxes (20)(21)(22).
While the tip growth rate depends on the supply of vesicles, it has been reported that turgor pressure is also one of the driving forces of hyphal tip expansion (6). Turgor pressure in growing hyphae has been directly measured by using microinjection with pressure probes (23). Cytoplasmic bulk flow, which is evident in fast-growing fungi like Neurospora crassa, is also involved in the force to expand the hyphal tip (6,24).
Microfluidic devices-based technology has been used to study the behavior of tipgrowing plant cells (25)(26)(27) and, more recently, of filamentous fungi (28,29). An elastic polydimethylsiloxane (PDMS) microfluidic device enabled measurement of the invasive pressure of tip-growing plant pollen tubes (30). Likewise, scanning probe microscopy (SPM), with a sensor probe that directly indents the cellular surface, is available for measurement of cellular stiffness in a noninvasive manner (31). These methods, in combination with cell biological approaches, are powerful tools to investigate mechanical properties in living cells.
Here, we constructed a microfluidic device with 1 mm-width channels, which are narrower than the diameter of fungal hyphae, and observed growth as hyphae grew into, through, and out of the channels (Fig. 1A). The present study aimed to identify the relationship between cell polarity and growth rate by observing the forced morphological changes of growing hyphae under a microscope. Our results will help to understand fungal invasive growth into substrates or host plant/animal cells, which knowledge can be applied to the fields of fungal biotechnology, ecology, and pathogenicity.

RESULTS
Aspergillus nidulans and A. oryzae, but not N. crassa, hyphae grow through the channels. The PDMS microfluidic device used in this study possesses multiple micro channels, 1 mm wide and 50 or 100 mm long (Fig. 1A). Fungal spores were inoculated to the inlet at the center of the device (IN). The medium solution was continuously supplied to the inlet with the help of a pump (0.8 ml per hour) and flowed out from the four outlet corners (OUT).
We monitored hyphal growth of Aspergillus nidulans as it grew into, through, and out of the channels. We used a strain whose nuclei were visualized by fluorescence of the nuclear localization signal of the transcription factor StuA tagged with green fluorescent protein (GFP) (32). The hyphal widths were 2 to 3 mm before entering the channel under this condition. All observed hyphae grew into the channels, passed through them, and continued to grow (50 . n) (Fig. 1B, Fig. S1A, Movie S1 in the supplemental material). The kymograph along the growth axis indicated comparable growth rates of 37 6 15 mm/h (n = 20) before, through (Fig. 1B), and after the channels (Fig. S1A). In some cases, two or three hyphae passed through the same channel (Fig. S1B, Movie S1). In the same way, we tested Aspergillus oryzae, which is an important species for traditional food fermentation and modern biotechnology (33). We used a strain in which histone H2B is fused with GFP (34). Again, all observed hyphae went into the channels, passed through, and continued to grow after emerging from the other end without decreased growth rates (84 6 37 mm/h, n = 30) (Fig. 1C, Movie S2). We examined another model filamentous fungus, Neurospora crassa, whose hyphae usually grow faster and have a larger diameter than those of A. nidulans (7, 35) (see below). We used a strain in which histone H1 is fused with GFP (24). Some hyphae penetrated into the channels, but often their growth speed slowed down and stopped before reaching the end of the channel (Fig. 1D arrows, Movie S3). The hyphae that passed through the channels frequently lost tip polarization and started to swell (Fig. 1D asterisks, Movie S3). The depolarized hyphae stopped growing after a while (2 to 10 h), then lost the nuclear GFP signal (Movie S3). The growth arrest inside the channels and the loss of polarity of the hyphae after exiting the channels were characteristic of N. crassa and were never observed for A. nidulans or A. oryzae ( Fig. 1E and F). In N. crassa, 33% of the hyphae grew out of the channel without losing apical polarity (Fig. 1E) (n = 50). In addition, N. crassa spores that were trapped in front of the channels frequently germinated toward the opposite side of channels, i.e., did not enter the channel, (87%, Fig. S1C). This phenomenon was observed less frequently in A. nidulans or A. oryzae, with percentages of 53 and 64%, respectively (Fig. S1D). (D) Images of N. crassa (nuclei labeled with GFP) hyphae that have stopped growing in the channels (arrows) and depolarized hyphae exiting from the channels (asterisks) (from Movie S3). Scale bar 20 mm. (E) Ratio of the hyphae that successfully passed through the channels (pass) or stopped in or exiting from the channels (fail) in A. nidulans, A. oryzae, and N. crassa; n = 50 for each. (F) Ratio of polarized or depolarized hyphae that passed through the channels in A. nidulans, A. oryzae, and N. crassa; n = 50 for each.
Slow Down on Narrow Road ® Cell polarity loss after forced morphological changes in N. crassa. We investigated the cell polarity in N. crassa hyphae growing in the channels by monitoring GFPtagged chitin synthase class III (CHS-1) at the SPK (36). Accumulation of GFP-CHS-1 at the SPK was clearly observed at the tips of growing hyphae before growing into the channels ( Fig. 2A, Fig. S1E, Movie S4). The hyphae entered the channels and then stopped growing, coinciding with a loss of the GFP signal at the SPK and dispersion of the fluorescence signal along the cytoplasm of the tip region, with high intensity level of GFP ( Fig. 2A and B, Fig. S1E and F). Distinct accumulation of GFP-CHS-1 at the SPK was hardly observed in those hyphae inside the channels (Fig. 2C). In the depolarized swollen hyphae exiting a channel, the fluorescence signal was diffused and weak, but became visible again at the apex of the multiple branches that formed when polarized growth resumed after channel exit (Fig. 2D arrows, Movie S5). The kymograph along the growth axis indicated comparable growth rate before and in the channels, 200 and  ® 245 mm/h; however, the hyphae drastically decreased the growth rate after exiting the channel (Fig. 2E).
CHS-1-GFP is also known to localize at septa during their formation (36). We found that the depolarized hyphae possessed two-times more septa within the narrow channels than the hyphae that successfully passed through the channels without presenting polarity defects (Fig. S1G to I), suggesting that the cell cycle progresses and deposition of cross walls continues when tip growth is inhibited.
Relationship between growth rate and polarity maintenance. We tested two plant-pathogenic fungi, Fusarium oxysporum and Colletotrichum orbiculare (37, 38), using the same microfluidic devices. Since plant-pathogenic fungal hyphae have to penetrate the space between tightly connected plant cells, polarity maintenance in spatially confined growth should be important for their pathogenicity. Almost all hyphae of F. oxysporum and C. orbiculare grew into and passed through the channels while maintaining their growth rates (83 6 27 mm/h and 91 6 16 mm/h, n = 53 and 45, respectively) (Fig. 3A, Movie S6).
To investigate the reason why only N. crassa but not the other fungi showed Slow Down on Narrow Road ® growth defects during spatially confined growth, we compared the widths of hyphae and hyphal elongation rates of all fungi growing in the device (Fig. 3B). The results corresponding to before entering and after exiting the channels are shown in dark and bright colors, respectively. The hyphal widths of A. nidulans were 2 to 3 mm, whereas those of N. crassa, F. oxysporum, and C. orbiculare were 3 to 4 mm, and those of A. oryzae were slightly wider. These results suggest that the hyphal widths are not correlated with the growth defect shown in the channels. There was no significant difference in the hyphal widths between before entering and after exiting the channels except with A. oryzae. It is known that A. oryzae increases hyphal width as cultivation time passes (34). Since the widths in mature hyphae of N. crassa are known to be over 10 mm, the hyphae we observed under this condition were considered young hyphae. In contrast, the average hyphal elongation rate measured in A. nidulans was less than 50 mm/h, whereas those in A. oryzae, F. oxysporum, and C. orbiculare were 50 to 100 mm/h (Fig. 3B, lower graph). Notably, the average hyphal elongation rate measured in N. crassa was 150 to 250 mm/h, considerably higher than that of the other fungi.
To examine the relationship between growth rate and growth defect inside the channels, we further examined Rhizopus oryzae and Coprinopsis cinerea dikaryon, whose hyphal elongation rates are known to be relatively high (39,40). The hyphal elongation rates of R. oryzae and C. cinerea in the device were 100 to 250 mm/h (Fig. 3B), whereas the hyphal widths were 3 to 5 mm, indicating that these two fungi grow faster than the other fungi, and similar to N. crassa. At least one hypha entered one channel, however, hyphae of R. oryzae and C. cinerea often stopped growing in or shortly after exiting the channels ( Fig. 3C and D, Movie S7). Depolarized hyphae were sometimes observed after exiting the channels in R. oryzae, similarly to what was observed for N. crassa (Fig. 3C and E). We tested various fungal species of different phylogenetic lineages (41). However, the observed output did not correlate with the phylogenetic distance (Fig. S1K). Altogether, these results indicated that neither phylogenetic relevance, nor the width of hyphae, are correlated with the growth defects seen inside the channels. In contrast, the hyphal elongation rate displayed a strong correlation with observed growth defects inside the channels (Fig. 3F, correlation: 0.86, Fig. S1L and M).
Contribution of turgor pressure for polarity maintenance. Why do hyphae of N. crassa, R. oryzae, and C. cinerea generally grow faster than those of A. nidulans and other species? One possibility points to the fact that N. crassa, R. oryzae, and C. cinerea hyphae have higher turgor pressure. This is supported by the results showing they were sensitive to the high osmotic condition generated by addition of 0.6 M KCl, resulting in decrease of turgor pressure ( Fig. 4A and Fig. S1N and O). In contrast, A. nidulans, A. oryzae, and F. oxysporum were not sensitive to the high osmotic condition.
Indeed, we measured the elastic modulus, which represents forces balanced in the opposite direction of turgor pressure, by using a scanning probe microscope (SPM). The SPM scans sample surfaces with an extremely sharp sensor probe and measures the physical property of fungal cells in a noninvasive manner at high magnifications ( Fig. S1P and Q). The elastic moduli of N. crassa hyphae, 236 6 115 MPa, were significantly higher than those of A. nidulans, A. oryzae, and F. oxysporum (Fig. 4B), reflecting that the turgor pressure in N. crassa is higher. The elastic moduli of N. crassa decreased to 58 6 30 MPa in the hyphae grown in minimal medium (MM) with 0.6 M KCl (Fig. 4B), suggesting that the turgor pressure decreased under high osmotic condition. After the hyphae grown in MM 1 0.6 M KCl were shifted to MM for 30 min, the elastic moduli recovered to 141 6 82 MPa.
In order to decrease the turgor pressure in hyphae of N. crassa grown inside the device, N. crassa was grown under high osmotic condition attained with 0.6 M KCl. The hyphal elongation rate just before the channels decreased in high osmotic condition from 239 6 50 to 151 6 21 mm/h (Fig. 4C), whereas the hyphal widths increased from 3.8 6 0.7 to 5.2 6 1.0 mm (Fig. 4D). Notably, the ratio of hyphae that passed through the channels increased from 33% to 75% in the high osmotic condition (Fig. 4E), which is correlated with the decreased hyphal elongation rate (Fig. 3F, Nc 1 KCl). Although 25% of the hyphae still stopped growing in the channels, depolarized hyphae were not observed (Fig. S1L). We compared the hyphal elongation rate just before the channels in the hyphae that passed or failed to pass, and found that hyphal elongation rate was lower in the hyphae that passed the channels than seen in hyphae which failed to pass in normal and high osmotic conditions (Fig. 4F).
Under high osmotic conditions, the SPK labeled by GFP-CHS-1 was clearly observed  Fig. S1J). (G, right) Kymograph of GFP signal along the growth axis (from Movie S8). Total 8 h; scale bar 100 mm. The hyphal elongation rates before entering and after exiting the channel are shown by arrows.
Slow Down on Narrow Road ® at the tips of growing hyphae even in the channels (Fig. 4G, Fig. S1J, Movie S8). Although hyphae swelled slightly when exiting the channel (Fig. 4G right, asterisk, Movie S8), they grew into and passed through the channels while maintaining their growth rates (Fig. 4G right, arrows). These results indicate that the growth rate is important for the maintenance of cell polarity in spatially confined growth derived from passing through the channels (Fig. 5).

DISCUSSION
This study showed that hyphae from several fungal species of different phylogenetic lineages were able to grow into microchannels narrower than their width, as described before for plant tip-growing cells (27). Microfluidic devices have also been used recently to study the hyphal growth of filamentous fungi (28,29,42,43). It was first found that hyphae of N. crassa, R. oryzae, and C. cinerea either ceased growing when passing through the channels or lost polarity upon exiting the channels. The observed effects did not correlate with their taxonomic classification or with the width of hyphae, but correlated with the hyphal elongation rate. Fast-growing fungi possess the advantage of quickly covering new nutrient-rich substrates or free open spaces. However, at the same time, they may lack the ability to regulate cell shape properly when growing in spatially confined environments. As far as we know, this is the first report indicating a trade-off between growth rate and morphological plasticity, which suggests the significance of slow growth for the cooperative control of cell polarity and cell growth. This characteristic is considered a case of convergent evolution, given that each fungus possesses a similar morphology and physiology adapted to different environmental factors despite that they are phylogenetically distant. It will be fascinating in the near future to study whether a similar relationship is observed in other tipgrowing cells, such as pollen tubes and root hairs of different plant species.
Our results indicate that hyphal tip growth requires a very delicate balance of ordered exocytosis to maintain polarity under spatially constrained circumstances. In fast growing hyphae, such as N. crassa, a large number of secretory vesicles per time unit FIG 5 Relationship between extension rate and the ability to adapt to spatially confined growth. Cartoon representation of trade-off between cell plasticity and growth rate in spatially confined growth, and how they are correlated and act cooperatively to determine cell shape. When fast-growing hyphae pass through a narrow channel, a massive number of vesicles are forced to be congregated and mislocalized at sites others than the tip region, resulting in the depolarized growth and tip swelling when exiting the channels. In addition, a high turgor pressure makes all the machinery collect isotropically at the tip in an uncontrolled manner, thus generating a swollen tip. In contrast, a lower growth rate allows hyphae to maintain positioning of the cell polarity machinery, thereby permitting growth in highly confined spaces.
are presumably supplied to the hyphal tips, resulting in a conspicuous SPK (44). When fast growing hyphae enter into the narrow channels, a massive number of vesicles are forced to congregate in the tip region. The cytoplasmic space in those squeezed hyphae is probably too small for all the secretory vesicles to fit within the tip region. The space constraints therefore cause the excess of secretory vesicles to mislocalize at sites others than the tip region, resulting in loss of apical polarity and tip swelling when exiting the channels (Fig. 5). In fact, the lack of localization of some vesicular markers such as CHS-1 at the tip, as well as the dispersed fluorescence observed instead when N. crassa hyphae grew through the channels, supports the idea than an excess of vesicles accumulates in a nonorganized manner at the subapical region. When hyphae are forced to grow through a very narrow channel, under a high turgor pressure yet maintaining the same growth speed, large portions of the cell wall building machinery accumulate at the subapical region. Upon exiting the channel, all the machinery gets incorporated in an uncontrolled manner at the tip, resulting in isotropic growth and generating a swollen tip. This causes several new polarity axes to become established, such that growth resumes in the form of multiple branches. In the case of A. nidulans, even if hyphae are squeezed when entering the channel, the vesicles, presumably less abundant, manage to maintain their flow rate, spacing, and movement, and hence growth remains unaffected during passage and upon exiting the channel (Fig. 5).
Filamentous fungi play a major role in degradation of biopolymers found in nature for organic material recycling (45,46). Some fungi are useful in biotechnology and traditional food fermentation (33,47), where solid-state cultivation is especially important (48). Hyphal invasive growth into host cells (plant or animal) is essential for pathogenicity and for symbiosis (37,49,50). Our results help understand the mechanisms of fungal invasive growth into substrates or host cells by spatially confined growth and how cell morphology is controlled by cell polarity and cell growth, and are thus highly relevant to fungal biotechnology, ecology, and pathogenicity.

MATERIALS AND METHODS
Fungal strains and media. A list of filamentous fungi strains used in this study is given in Table S1 in the supplemental material. Supplemented minimal medium for A. nidulans and standard strain construction procedures have been described previously (51).
Microfluidic device. The microfluidic devices originally designed for culturing tip-growing plant cells and reported by Yanagisawa et al. (27) were adapted for the current fungal cell studies. Briefly, photoresist (SU-8 3005 and 3010) based microstructures were created on a silicon wafer using a maskless lithography system (DL-1000; Nano System Solutions, Inc.). Then, the polydimethylsiloxane (PDMS, Sylgard 184; Dow Corning) device was prepared through a standard soft-lithography technique. Finally, the PDMS and cover glass (24 Â 60 mm, Matsunami) were both treated with O 2 plasma (CUTE, Femto Science) for permanent bonding.
Growth condition. The minimal medium was placed into a 20-ml plastic syringe (SS-20ESZ, Terumo) and infused into the PDMS devices using a positive displacement syringe pump (YSP-101, YMC) at a rate of 0.8 ml per hour through a polyethylene tube (inner diameter 0.38 mm, outer diameter 1.09 mm; BD intramedic).
Microscopy. Cells were observed using epifluorescent inverted microscopy, including the Axio Observer Z1 (Carl Zeiss) microscope equipped with a Plan-Apochromat 63Â 1.4 oil or 10Â or 20Â lens objective, an AxioCam 506 monochrome camera, and a Colibri.2 LED light (Carl Zeiss). The temperature of the stage was kept at 30°C by a thermo-plate (TOKAI HIT, Japan). Images were collected and analyzed by the Zen system (Carl Zeiss) and ImageJ software.
Scanning probe microscopy. Cells were grown in minimal medium on coverslips at 30°C for 24 h. The medium was removed by pipetting and the cells were analyzed using a scanning probe microscope SPM-9700HT (Shimadzu) with high magnification optical microscope unit, active vibration isolation table, wide area scanner (XY: 125 mm, Z: 5 mm), and fiber light. We used scanning probes (tips), PointProbe-Plus silicon-SPM-Sensor, PPP-NCHAuD (NANOSENSORS), thickness 4.0 6 1 mm; length 125 6 10 mm; width 30 6 7.5 mm; resonance frequency 204-497 kHz; and tip height 10 to 15 mm. Images were collected and analyzed by Nano 3D mapping software (Shimadzu). The nanoindentation results were fitted to the JKR model.
Statistical methods. Correlations were measured by CORREL function in Excel. Hierarchical clustering was performed with complete linkage using hclust function in R.