Skip to main content
NCBI home page
RSC Chemical Biology logoLink to RSC Chemical Biology
. 2026 Sep 30. Online ahead of print. doi: 10.1039/d6cb00197a

High-throughput droplet screening of filamentous fungi on solid substrates

Chiara Leal Alves a,b,†, Jeremy Z Wu b,c,†, Elham Khanjani b,c,†, Adiphol Dilokpimol d, Zhiyang Deng a,c, Biplu Prajapati d, Niloufar Rahbari b,c, Antti P Aalto d, Mari Valkonen d,✉, Steve CC Shih a,c,✉
PMCID: PMC13625276  PMID: 42819911

Abstract

Droplet-based microfluidics has been frequently used for high-throughput single-cell microbial screening, where individual cells are encapsulated into nanoliter or picoliter-sized droplets and sorted based on measurable phenotypes such as fluorescent proteins or enzyme activity using fluorescence or absorbance-based substrates. However, using droplet microfluidics with filamentous fungi poses a unique challenge: fungal strains tend to escape from droplets as their hyphae grow, causing individual droplets to merge and eventually burst which prevents measuring individual phenotypes. In our work, we demonstrate a strategy to robustly encapsulate and retain spores of filamentous fungi within water-in-oil droplets by incorporating solid substrates into the droplet microenvironment. Using this approach, we show for the first time the encapsulation and growth of two model filamentous fungi species in droplets – Aspergillus oryzae and Trichoderma reesei – with solid substrates supporting spore retention, germination, and early hyphal development without compromising droplet stability. Our goal was to screen for enhanced enzyme production (e.g., α-amylase and cellulase) using fluorescence-based activity assays. To this end, we further characterized our top hits through scale-up fungal fermentation to evaluate improvements in both enzyme production and total protein secretion. Our work overcomes a key technical barrier in applying droplet microfluidics to filamentous fungi that allows for high-throughput screening and discovery of new fungal strains that produce active enzymes and heterologous proteins relevant to industrial applications.

A droplet-based microfluidic and solid substrate approach towards screening fungal industrial enzymes.graphic file with name d6cb00197a-ga.webp

Introduction

Filamentous fungi are among the most widely used microorganisms for industrial-scale production of proteins and metabolites.1–4 Their adaptability to diverse environments enables them to degrade a variety of materials by secreting a broad spectrum of hydrolytic enzymes and bioactive compounds, including phenolic compounds,4 lipases for oil-based substrates,5 amylases for starch degradation,6 cellulases for lignocellulosic biomass conversion,7 and antibiotics (e.g., penicillin).8 This versatility makes them indispensable for the biofuels and pharmaceuticals industries.

Filamentous fungi, such as Aspergillus and Trichoderma species, exhibit remarkable morphological plasticity,9 growing in various formats, from submerged fermentation (SmF)10 in liquid media to solid-state fermentation (SSF)11 on solid substrates. They can metabolize a wide range of carbohydrates as carbon sources and form complex multicellular structures, where cells are organized into an interconnected hyphal network, which is essential for fungal growth, protein secretion, and environmental adaptation.12 These fungi are known producers of industrially relevant enzymes, and efforts to enhance protein and metabolite secretion in filamentous fungi have gained significant traction across multiple industries. However, screening fungal libraries is challenging due to the unique growth characteristics of fungi.

Traditionally, improvements in protein production have been made through iterative rounds of random mutagenesis followed by phenotypic screening to identify rare high-producing variants. Although random mutagenesis can generate productive strains without detailed prior knowledge of the underlying molecular mechanisms, identifying improved mutants typically requires extensive hands-on screening, making the overall workflow time-consuming and labor-intensive. Conventional well-plate screening of filamentous fungi offers excellent throughput, but challenging in isolating single spores into individual wells and fungal growth is heterogeneous, with differences in spore germination, mycelial branching, and biomass accumulation which often masks subtle phenotypic differences between strains.13,14 Another option is fluorescence-activated cell sorting (FACS), which can significantly boost throughput, however, a key limitation is that secreted proteins are not typically produced until after germination. After germination, the fungus forms filamentous, branched, multicellular hyphae that cannot readily be sorted by FACS. Hence, single-spore containment and incubation are essential for effective fungal screening. Techniques such as droplet-based microfluidics can encapsulate and analyze individual fungal spores in pico- or nanoliter-scale droplets, and facilitate the screening of vast mutant libraries while minimizing cross-contamination and handling difficulties.15–21

Integrating filamentous fungi into high-throughput microfluidic pipelines remains challenging. Fungal hyphae tend to grow heterogeneously in droplets, making it difficult to confine them for the extended incubation times necessary for target molecule production.15,16 In previous studies, hyphae remained confined within picolitre droplets for approximately 12 h, which generally is insufficient for efficient production and secretion of target proteins into the culture medium. Various microfluidic strategies have been explored, with most efforts focusing on Aspergillus niger, a species known for its high protein secretion titers. One effort is to increase droplet volumes from pL to 20 nL, which has extended incubation times to 36 h.15 Alternatively, using a core–shell encapsulation technique with a 2 nL outer layer of gelatin methacrylate and an aqueous inner core containing the spore has enabled incubation for up to 48 h.18 These advancements highlight the potential of microfluidics for fungal screening but show the need for further optimization to accommodate longer incubation times and to accommodate species having highly branched and aggressive hyphal characteristics, such as Aspergillus and Trichoderma.22,23

In this work, we significantly extend our previously reported microfluidic platform (see Table S1 for detailed advances), which introduced a solid substrate within droplets to emulate solid-state fermentation and confine fungal hyphae for screening.16 Our earlier study established the feasibility of this approach for enzyme screening; however, it was limited to one media formulation, a single fungal strain and moderate sorting throughput. Here, we introduce several key advances that collectively transform the platform into a more industrially relevant screening system (Fig. 1). First, we systematically expand the material design space of the droplet microenvironment by evaluating multiple colloidal substrates – including colloidal chitin (CC), colloidal cellulose (CCEL), and hydrogel beads – revealing substrate-dependent effects on hyphal confinement and viability. This enables a substantial extension of incubation times (up to 7 days with colloidal chitin and 4 days with colloidal cellulose), overcoming a major limitation of droplet-based fungal systems. Second, we demonstrate the cross-species applicability of the platform using two industrially relevant yet physiologically distinct filamentous fungi: Aspergillus oryzae, a major producer of amylases,24 and Trichoderma reesei, a leading producer of cellulolytic enzymes.25 This comparison highlights the adaptability of the platform across different enzyme classes, regulatory networks, and strain backgrounds rather than limiting validation to a single model organism. Third, we achieved electrostatic sorting of solid substrate fungal droplets at 10 Hz, representing one of the highest sorting throughputs reported for droplet microfluidic platforms designed for filamentous fungi.16 Our microfluidic pipeline provides a front-end screening strategy for industrial enzyme discovery, enabling large fungal strain libraries to be rapidly screened and sorted according to enzyme activity. High-performing variants can be recovered and transferred to conventional low-volume cultivation for secondary screening and validation, followed by evaluation under fermentation conditions. By enriching promising variants, this workflow can accelerate the identification of strains with improved enzyme-production performance for future strain engineering programs.

Fig. 1. High-throughput screening and sorting of industrial filamentous fungi. Schematic of single-spore co-encapsulation with colloidal material for solid-state fermentation in droplets, enabling enzyme production screening through offline incubation, followed by electrostatic sorting of high performing fungal strains and scale-up validation of top performing mutants.

Fig. 1

Open in a new tab

Results and discussion

Fungal screening and solid-state fermentation in droplets

Droplet-based microfluidics has recently emerged as a powerful tool for the cultivation and screening of filamentous fungi. The most successful platforms use a solid or semi-solid substrates, mirroring the widely used fungal fermentation approach known as solid-state fermentation.15–17,19,26 In such systems, fungal spores are co-encapsulated within droplets containing solid particles, which serve to mitigate common issues such as hyphal penetration through the droplet interface and clogging in the microfluidic channels.16 The inclusion of solid substrates provides an in situ substrate for spore adhesion and hyphal containment, enabling extended incubation periods and promoting sustained fungal growth and enzymatic activity. To encapsulate the solid material within the droplets, our microfluidic device uses a T-junction configuration, which centers the solid substrate in the channel and minimizes contact with channel walls (Fig. S1). Furthermore, to ensure stable suspension of colloidal material during droplet formation, we optimized the loading configuration by orienting the syringe horizontally and elevating it relative to the device inlet, while increasing the flow rate to accelerate droplet generation. These improvements enabled reliable and uniform encapsulation of colloid-containing droplets (Fig. S2 and S3) with coefficient of variation of 8 and 18% for cellulose and chitin respectively.

Two engineered strains were used in this study: T. reesei and A. oryzae. Both strains are compatible with glucose-based cultivation, but through different genetic modifications. The T. reesei strain carries a mutation in xylanase regulator 1 (xyr1) that alleviates carbon catabolite repression and enables cellulase expression on glucose.27 The A. oryzae strain uses a synthetic expression system that drives target protein production constitutively and independently of native glucose repression mechanisms.3 Both strains had mCherry reporter constructs that linked fluorescence to production of target secreted protein (Fig. S4). In T. reesei, mCherry was fused to β-lactoglobulin, and the expression cassette was integrated into a major cellulase (cbh1) locus. In A. oryzae, mCherry was fused to α-amylase, and the expression cassette was integrated to the amyC locus. These fluorescent fusion proteins enabled visualization of target protein production and assessment of colloid association and hyphal position within the droplets. As an initial growth test for A. oryzae, without any colloidal material, we evaluated the glucose concentrations of 0, 0.5, 1, and 2% to determine the condition that will maximize production of α-amylase mCherry fusion protein. The fluorescence of the mCherry is the highest when there is 1% glucose in the media after 48 h (Fig. 2a), indicating optimal α-amylase production.

Fig. 2. Optimization of culture and alpha-amylase production conditions for A. oryzae. (a) Optimization of glucose concentration to maximize alpha-amylase expression via mCherry fluorescence. (b) Fungal growth and droplet stability were assessed in 2 nL droplets with 1% glucose with and without colloidal chitin (No CC and CC, respectively) incubated at 21 °C. (c) Comparison of droplets with colloidal chitin (SSF) versus without colloidal chitin (SmF) under varying temperatures and incubation times. (d) Microscopy images showing mCherry expression in droplets with 0.5% and 1% glucose. (e) Fluorescent microscopy revealing alpha-amylase activity in parental strain at 21 °C after 72 h of incubation using EnzCheck Ultra Amylase as substrate and 1% glucose media. Error bars represent standard error (n = 3); scale bar = 100 µm. CTDF represents corrected total droplet fluorescence (see Methods for equation).

Fig. 2

Open in a new tab

The mechanisms by which the substrate promotes hyphal confinement were not yet fully understood. Hence, we first evaluated whether the same colloidal chitin droplet formulation used in our previous work for cultivation of C. rosea,16 could also support the growth and containment of A. oryzae. This initial experiment allowed us to determine whether the previously established conditions were transferable to a different filamentous fungus before testing alternative solid substrates. We used 1% glucose in the medium to culture A. oryzae in droplets containing colloidal chitin at two temperatures, 21 °C and 30 °C, and we quantified the hyphal exit through the droplets as a proxy for droplet stability. Based on preliminary microscopy observations, substantial hyphal escape was operationally defined as the presence of >5 hyphae extending beyond the boundary of an individual droplet. This threshold was selected empirically, as droplets with ≤5 exiting hyphae generally retained a clearly defined and discrete droplet interface, whereas extensive hyphal escape (>5 exiting hyphae) was frequently associated with visible droplet merging or loss of droplet integrity (Fig. S5a). Using this as the threshold, the A. oryzae droplets did not burst due to hyphae piercing for at least 168 hours at 21 °C, compared to only 48 h in droplets without colloidal support (Fig. 2b). At 30 °C, the droplets with solid support experienced faster hyphae bursting (72 h) (Fig. S5a). When we evaluated the fungal retention in the droplets at 30 °C, many of the droplets showed overgrown hyphae and excessive hyphal piercing through the droplets, causing multiple droplets to merge. This caused many of the droplets to be non-quantifiable, making it challenging to measure their protein production (Fig. S5b). Using the optimal temperature and glucose concentration, we measured the enzymatic activity of alpha-amylase and showed that we can measure the enzymatic activity of the parental and mutant strains inside the droplets (Fig. S6). For these measurements, we used microscopy imaging to evaluate growth and droplet stability via hyphal exit (Fig. 2c), α-amylase production and activity via droplet mCherry and EnzCheck assay fluorescence, respectively (Fig. 2d and e).

Next, to evaluate the impact of colloidal chitin on cellulase activity for T. reesei, we first performed a well-plate assessment of enzymatic activity with and without colloidal chitin. For T. reesei, cellulase activity rather than mCherry fluorescence was used as the screening because the mCherry reporter reflects expression of the heterologous β-lactoglobulin fusion and does not directly represent the production of the native cellulases. The activity assay therefore provides a more functionally relevant measure for identifying mutants with enhanced cellulolytic performance. After 5 days of culture, A. oryzae maintained high alpha-amylase activity both in the presence and absence of colloidal chitin (Fig. S7a). However, T. reesei showed a significant decrease in cellulase activity when cultured with colloidal chitin compared with the condition without any colloidal chitin (Fig. S7b). These results show that although colloidal chitin provides a suitable solid support for prolonged cultivation, its suitability is species- and enzyme assay-dependent.

This initiated further investigation into the conditions required for in-droplet culture of T. reesei. As an alternative to colloidal chitin, we evaluated two other solid supports for T. reesei: (i) hydrogel beads and (ii) colloidal cellulose. Hydrogel beads and cellulose have been used as individual substrates28 or together29 to provide a biocompatible matrix in growth media to support fungal growth, indicating their potential as a solid support in droplet fungal culture. In addition, cellulose is a known natural inducer of cellulases in T. reesei.30 When co-encapsulating with single spores, T. reesei did not show any attachment to the hydrogel beads under any of the tested media compositions (Fig. S8). Although hydrogels provide a porous three-dimensional structure, this result suggests that the presence of a physical support alone is not sufficient to promote localized fungal growth or prevent hyphal extension toward the droplet interface. The hydrogel beads used here are also not readily metabolized by T. reesei and therefore do not provide a local nutrient source that could promote growth within or around the matrix.31

In contrast, colloidal cellulose supported continuous fungal growth and protein expression for up to 96 h (Fig. 3a) while minimizing hyphae exit from the droplets (Fig. 3b). Similar to colloidal chitin, colloidal cellulose provides a dispersed solid substrate with a large surface area that can support localized hyphal growth. Moreover, because cellulose can be enzymatically degraded by T. reesei, it may provide a localized nutrient source and induce cellulase production, potentially reducing hyphal foraging toward and across the droplet interface.32,33 We hypothesize that hyphal retention results from a combination of physical and biological effects: the solid substrate provide surfaces for localized hyphal growth and form an internal substrate that supports the developing mycelial network, while metabolizable substrates may additionally provide localized nutrient availability. Changes in the local droplet microenvironment may also contribute to retention, although these effects were not independently resolved in the present study.

Fig. 3. Optimization of culture and cellulase production conditions for T. reesei. Optimization of carboxymethyl cellulose minimal media composition at 21 °C in 2 nL droplets with and without colloidal cellulose for fungal growth via (a) mCherry fluorescence and (b) droplet stability. (c) Microscopy images showing mCherry expression in droplets with 2% CMC with and without CCEL after 72 h incubation at 21 °C. (d) Bright field images of T. reesei growing in droplets with 0.5% and 2% CMC with colloidal cellulose after 96 h of incubation at 21 °C. Error bars represent standard error (n = 25); scale bar = 100 µm. CTDF represents corrected total droplet fluorescence.

Fig. 3

Open in a new tab

These observations are consistent with our previous work,16 in which colloidal chitin was used to support the in-droplet cultivation of C. rosea. Both C. rosea and A. oryzae produce chitinolytic enzymes,34–36 enabling them to degrade chitin, whereas T. reesei efficiently degrades cellulose through the production of cellulases. These observations suggest that the effectiveness of a solid support is species- and substrate-dependent and that successful materials should be selected based on both their physical properties and compatibility with fungal physiology. Although some hyphal escape was still observed during extended cultivation, the colloid concentration was increased from that used in our previous work to 4%, providing a greater density of solid material within each droplet and increasing the available substrate for fungal growth.

Based on these observations, we propose preliminary general design considerations for selecting solid supports for in-droplet cultivation of filamentous fungi. First, the material is biocompatible, remain sufficiently insoluble and stable throughout cultivation, and compatible with droplet generation and sorting. Particle size and concentration should permit relatively uniform incorporation into droplets without excessive sedimentation, aggregation, channel obstruction, or disruption of droplet stability. Second, the material should provide sufficient surface area and an appropriate morphology or porosity to support fungal association and the formation of a three-dimensional substrate capable of accommodating the developing hyphal network. Third, the material is compatible with the metabolic physiology of the organism and enzyme assay of interest. A metabolizable substrate may provide the additional benefit of localized nutrient availability and, for inducible enzyme systems, stimulation of the desired enzyme-production phenotype. These considerations may enable the strategy to be adapted to other filamentous fungi, including rapidly growing species such as Neurospora, for which substrates such as wheat bran could potentially provide both structural support and a nutrient source.37 However, faster growing or more aggressively branching species may additionally require growth-limiting media, increased substrate concentration, larger droplets, or shorter incubation periods to maintain droplet integrity. Systematic evaluation of each parameter will nevertheless need to be performed to ensure compatibility with target analytes and organisms.

When glucose was used as the carbon source, a significant decrease in cellulase activity was observed (Fig. S7c). Despite the overall cellulase activity being lower with carboxymethylcellulose (CMC), the presence of colloidal cellulose did not significantly change enzymatic activity. As the T. reesei strain is engineered to also express a second protein of interest fused to mCherry, this fluorescence can be used to monitor protein expression in-droplet similarly to A. oryzae. From the mCherry fluorescence, the use of CMC did not appear to negatively impact protein expression compared to glucose (Fig. S7d), suggesting that the fungi remained metabolically active and capable of sustaining recombinant protein production while growing on a cellulose-derived substrate. This is in agreement with the in-droplet results for droplets without colloid up to 72 h (Fig. 3a). Based on these results, to perform high-throughput screening of mutant populations, we incubated the mutant library in droplets in minimal medium containing 1% glucose with colloidal chitin at 21 °C for A. oryzae and 2% CMC with colloidal cellulose at 21 °C for T. reesei. To our knowledge, this is the first report of culturing these fungi in droplets with a solid substrate to measure enzyme activity.

Droplet sorting of mutant fungi using colloidal material

A critical requirement for sorting and selecting filamentous fungi in droplets is designing a device that preserves the integrity of both the droplets and the strains. For this purpose, we used our recently described co-planar electrode sorter system that generates an electric field gradient between the activated and grounded electrodes, guiding the droplets along the field.16,38–40 The design allows for low-voltage sorting which was optimized to handle fragile droplet libraries (e.g., droplets with fungal hyphae exiting) without frequent adjustments to sorter parameters, such as applied voltage and flow rate, despite variations in droplets. Droplets were generated at a size of 2 nL allowing our droplets to accommodate a larger quantity of colloidal substrate compared to our previous work. Larger droplets provide more space for the hyphae to grow before piercing the droplet and extended incubation time. However, increasing droplet size relative to the microfluidic channel dimensions can increase droplet confinement and deformation during reinjection and passage through the sorting junction, potentially affecting droplet stability, spacing, and sorting reliability.15,41

The key features of our device include: (i) a withdrawing system added to the waste outlet, and (ii) an electrode configuration containing an actuation electrode and a ground electrode (Fig. S9). We characterized the performance by changing the spacer oil flow rates and the withdrawal flow rates to ensure all droplets are directed to the waste channel without electrode actuation while maintaining adequate spacing between screened droplets. Experimental observations indicated that a withdrawal flow rate exceeding 50 nL s−1 in the sorting channel and a spacer oil flow rate above 200 nL s−1 are necessary for ideal droplet spacing and waste channel flow (Table S2).

Next, we evaluated the functionality of sorting by activating the electrodes to deflect fluorescein droplets toward the sorting channel. By pulsing the actuation electrode potential and varying the spacer oil and keeping the withdrawal flow rates, optimal sorting occurred with a spacer oil flow rate of 400 nL s−1, a withdrawing flow rate of 50 nL s−1, and potentials above 70 VRMS (Fig. S10a), consistent with results from our previous electrostatic sorter.16,40 Using these optimal parameters, we assessed the binary sorting efficiency with a mixture of 1 mM fluorescein and blue dye droplets. The integration time was set at 50 ms to obtain a reliable signal-to-noise ratio, while allowing for the acquisition 20 measurements per second. The sorting frequency was measured to be approximately 10 Hz (Fig. S11), reaching the upper limit for reliable droplet detection at this sampling rate. Higher sorting rates would require a shorter integration time and faster signal acquisition. However, reducing the integration time decreases the number of photons collected per measurement, resulting in a lower signal-to-noise ratio and potentially reducing the sensitivity and the accuracy of droplet sorting. Therefore, the 10 Hz sorting rate used here represents a compromise between throughput and reliable fluorescence-based detection. From these autonomous sorting experiments, we achieved accurate sorting (>99.2%) of the fluorescein droplets and a low false negative rate (∼3.37%) (Fig. S10b). These results are comparable with our previous electrostatic sorter, which demonstrated ∼96% efficiency and ∼3% false negatives (i.e. fluorescein droplets in the waste) for 1 nL droplets at a lower throughput of 7 Hz.

Prior to sorting, we assessed whether the fluorescence heterogeneity observed during screening could be distinguished from the intrinsic variability of the parental strains. We compared the fluorescence distributions of individual droplets containing the parental strains with those of their corresponding mutant libraries (Fig. S12a). The parental populations exhibited baseline droplet-to-droplet variability, with CVs of 16% (48 h) and 45% (72 h) for A. oryzae and 24% (48 h) and 36% (72 h) for T. reesei. The A. oryzae mutant library shows a variability similar to that of its parental strain based on their CVs, however, the T. reesei mutant library exhibited greater variability than its parental strain. Interestingly, for both organisms, the mutant fluorescence distributions extended to higher values than those typically observed for the corresponding parental populations (Fig. S12b). A subset of droplets within each mutant library exhibited high fluorescence values beyond the predominant range of the parental population. These results suggest that, despite the intrinsic heterogeneity observed in the parental strains, mutagenesis generated highly active enzymes that is different from the baseline population and are not masked by the intrinsic variability of the screening system.

Using our sorter system, we screened a UV-mutated library of A. oryzae and T. reesei based on alpha-amylase and cellulase production, respectively. We sorted based on the release of a green fluorophore from the enzymatic hydrolysis of a fluorescence substrate. For A. oryzae, single spores were encapsulated in droplets containing 1% glucose, colloidal chitin, and the EnzCheck Ultra Amylase fluorescent substrate, followed by incubation for 72 h at 21 °C. For T. reesei, spores were encapsulated in droplets with 2% carboxymethyl cellulose using colloidal cellulose and incubated for 48 h at 21 °C with the fluorescein di-beta-D-cellobioside. A total of approximately 400 000 droplets were screened per experiment at a throughput of 10 Hz, with ∼12 000 droplets sorted based on the 97th percentile. The sorted droplets were collected, and single fungal colonies were first isolated on PDA plates containing antibiotics before being re-inoculated on PDA plates. Approximately 80 mutants were selected for post-sorting analysis, and their enzymatic activity and specific enzyme activity were measured. Fig. 4a shows a scatter plot comparing the enzymatic activity of the mutant to the parental strain to identify the best-performing mutants. For A. oryzae, sorted mutants were analyzed in glucose-based media for alpha-amylase activity, showing that approximately 55% of post-sorted strains exhibited improved specific enzyme activity, while ∼47% showed enhanced total enzyme activity. In the case of T. reesei, the mutants were evaluated on both glucose-based media (Fig. 4b) and carboxymethyl cellulose-based media (Fig. 4c). The droplet sorting based on cellulase activity using colloidal cellulose yielded >90% of mutants with both increased specific enzyme activity and enhanced cellulolytic activity compared to the parental strain. We were effectively able to sort droplets containing the active mutants that have been cultured with colloidal chitin or cellulose.

Fig. 4. Post-screening analysis of A. oryzae and T. reesei mutants. Post-HTS fold improvements compared to the parental strain (dotted line at enzyme and specific enzyme activity of 1.0) in: (a) A. oryzae (colloidal chitin, BMDY with 2% glucose, 21 °C); (b) T. reesei (colloidal cellulose, TrMM with 4% glucose, 21 °C); (c) T. reesei (colloidal cellulose, TrMM with 2% carboxymethyl cellulose, 21 °C). Highlighted mutants indicate top-performing mutants.

Fig. 4

Open in a new tab

Droplet screening to scale-up: evaluation of alpha-amylase and cellulase production

A problem with microfluidic screening experiments is the potential for “hits” that do not always translate to experiments performed at conventional scales.42,43 To validate the practical relevance of droplet hits, we assessed whether the enzymatic performance observed in droplets translates to larger-scale cultivation. Scale-up experiments serve as a critical step to confirm the robustness and consistency of droplet-based selections. We evaluated the top five mutants (M) identified by droplet-based sorting for enhanced α-amylase and cellulase production, respectively, under scaled-up cultivation conditions and compared to their parental strain (PS). Specifically, for T. reesei, the strains were selected from a mix of the top performers from the glucose and CMC scale-up validation assays, namely TRM6, TRM20, TRM24, TRM25 and TRM37 (Fig. 4b and c). After 7 days of growth, protein content, enzyme activity, and biomass accumulation were quantified to assess strain performance (Fig. 5) along with statistical analyses of these results, which are provided in Tables S2–S4.

Fig. 5. Droplet scale-up validation in 100 mL liquid cultivation shake flasks. Production kinetics of (a) alpha-amylase by A. oryzae in glucose-based medium and of (c) cellulase T. reesei in carboxymethyl cellulose medium over 7 days for the top five hits (and the parental strain) from the droplet screen. (b) Alpha-amylase production of A. oryzae culture per dry weight after 5 days of cultivation in 2% glucose- or 1% starch-based medium. (d) Cellulolytic enzyme production (50–75 kDa) of T. reesei culture per dry weight after 5 days of cultivation in 2% CMC- or 4% glucose-based medium. Error bars represent standard deviation (n = 3 for a, c and d, n = 2 for b). Statistical significance was determined using one-way ANOVA with a 95% confidence interval, with p-values indicated as follows: 0.1234 (ns), 0.0332 (*), 0.0021 (**), 0.0002 (***).

Fig. 5

Open in a new tab

For A. oryzae, three mutants – AOM34, AOM67, and AOM69 – showed higher enzyme activity (Fig. 5a), with up to 25% increase after five days of cultivation in glucose media. Examining the specific activity, AOM67 and AOM69 exhibited 6% and 13% higher catalytic efficiency, respectively (Fig. S13a), per unit of protein produced (or per biomass; Fig. S13b), without any significant difference in biomass yield compared to the parental strain (Fig. S13c). In addition, we evaluated the performance of the mutants in starch-based media by measuring enzyme production via SDS-PAGE (Fig. S14). The results mirrored those observed in glucose-based media, with AOM67 and 69 outperforming both the parental strain and other mutants. Specifically, AOM67 and AOM69 mutants showed a 47% and 61% increase in enzyme production, respectively. The enhancement in enzyme production when using starch as the carbon source is likely due to the strong induction of the amylase promoter by starch and maltose.43 These findings show the trends observed in the flasks follow those identified in the droplets, suggesting that the effect scales from micro- to flask-scale.

In contrast, the total cellulase activity of T. reesei shows an increase in enzymatic activity for all mutants (except TRM20) (Fig. 5c). While baseline strains remained in the ∼10–13 mU mL−1 of cellulose activity, most mutants show enhanced activity, with TRM6 and TRM37 reaching ∼2.9- and ∼3.4-fold higher signal than the parental when normalized by dry biomass (Fig. S15). As the cellulolytic enzyme system comprises multiple enzymes, we estimated the cellulase production by quantifying the bands within the 50–75 kDa, which contain cellobiohydrolase 2 (CBH2), endoglucanases (EGs) 1, 2 and 444 as benchmarks for cellulase production. Analyzing these strains in glucose-based media, TRM24 and TRM25 showed approximately 41% and 22% improvements in cellulase activity normalized by dry biomass, respectively (Fig. S16). The SDS-PAGE gel (Fig. S17) showed elevated enzyme expression in the higher-performing mutants. TRM6 and TRM37 had 60% and 210% higher expression, respectively in CMC-based medium, while TRM24 and TRM25 were at 126% and 199% increase over the parental strain, respectively (Fig. 5d). TRM20 in CMC-based medium and TRM37 in glucose-based medium also showed elevated enzyme expression, but their biomass normalized cellulolytic activities were at similar levels to that of the parental strain. This indicates that the observed activity improvements may reflect both increased enzyme production and changes in the activity of the cellulolytic enzymes.

The strong and consistent performance of AOM67 and AOM69, in both glucose- and starch-based media, for A. oryzae and TRM6, TRM24 and TRM25 in CMC- and glucose-based media, for T. reesei highlights their potential as robust production strains for industrial applications. These mutants exhibited increased enzyme titers and catalytic efficiency without compromising biomass, making them ideal for cost-effective, large-scale fermentation. Several studies suggest that enhanced alpha-amylase production is due to genetic mutations in secretion-related pathways such as vesicle trafficking or protein folding.45–48 The starch-inducible amylase promoter showed strong activation, offering a native, tightly regulated system that can be harnessed to drive heterologous protein expression. The strain's ability to express increased productivity across different media also indicates valuable metabolic flexibility. The enhanced cellulase production may reflect alterations in key transcriptional regulators such as xyr1 or relief from carbon catabolite repression, which enables sustained enzyme production in mixed or non-natural substrates.49–51 Overall, these mutants exhibited increased enzyme titers and catalytic efficiency without compromising biomass, making them ideal for cost-effective, large-scale fermentation using our newly designed solid-phased and droplet-based microfluidic system.

Conclusion

In this work, we demonstrated a robust and scalable strategy to enable solid-state fermentation of filamentous fungi within droplet-based microfluidic systems for high-throughput screening applications. By systematically matching the solid substrate to the fungal species and/or the screened enzyme, we optimized colloidal chitin for A. oryzae and colloidal cellulose for T. reesei. With those solid substrates, we achieved stable confinement of single fungal spores within water-in-oil droplets, extending incubation from 48 h without solid support to 168 h, while allowing hyphal growth and protein production over multiple days. Beyond identifying which substrates worked, our comparison across colloidal chitin, colloidal cellulose, and hydrogel beads also provided further insight into why substrate choice matters, pointing to three general preliminary design criteria for solid support selection: the matrix must remain insoluble and uniformly distributable during droplet generation, contain pores large enough to accommodate hyphal growth, and represent a metabolizable resource for the organism to limit outward hyphal growth. These criteria offer a rationale that should generalize to substrate selection for other fungal species beyond those tested here.

Building on this extended incubation window, our low-voltage electrostatic sorter screened approximately 400 000 droplets per experiment at 10 Hz with greater than 99.2% sorting accuracy, and the recovered mutants showed genuine enrichment for improved producers: up to 90% of sorted clones exhibited higher enzyme activity than the parental strain. This enrichment was not limited to well plate assays, the top-performing mutants were cultivated at conventional shake-flask scale, where A. oryzae strains AOM67 and AOM69 retained up to a 1.3-fold increase in biomass-normalized α-amylase activity, and T. reesei strains showed up to a 3.4-fold cellulolytic enzyme activity increase in CMC-based medium, without any loss in biomass yield. It is this scale-up consistency, rather than the droplet-scale followed by well plate assay results alone, that supports our conclusion that the platform identifies robust and industrially relevant producer strains.

These results show that filamentous fungi can be reliably cultured, screened, and selected using solid-support droplets, overcoming a major barrier in applying microfluidics to fungal biotechnology. We propose that our approach offers a screening tool for strain improvement campaigns ahead of larger-scale fermentation validation in industrial bioprocessing.

Materials and methods

Reagents, materials, and equipment

For microfabrication, the equipment used included the Harrick Plasma PDC-001 (Ithaca, NY, USA), Quintel Q-4000 mask aligner (Neutronix Quintel, Morgan Hill, CA, USA), Laurell spin coater (model WS-650Mz-8NPPB, Laurell Technologies Corporation, North Wales, PA, USA), and the UV-KUB 3 mask aligner (Kloe, Montpellier, France).

Reagents and materials for device fabrication included chromium-coated glass slides with AZ-1500 positive photoresist (Telic, Valencia, CA, USA), MF-321 developer (Kayaku Advanced Materials, Westborough, MA, USA), CR-4 chromium etchant (OM Group, Cleveland, OH, USA), AZ-300T photoresist stripper (Integrated Micro Materials, Argyle, TX, USA), silicon wafers (Silicon Valley Microelectronics Inc., Santa Clara, CA, USA), SU-8 2075 and SU-8 developer (Kayaku Advanced Materials, Westborough, MA, USA), DI water (resistivity of 15 MΩ cm−1), cleanroom lab grade acetone and isopropanol (Sigma-Aldrich, Oakville, ON, Canada), chlorotrimethylsilane (Sigma-Aldrich), Polydimethylsiloxane (PDMS) 184 Sylgard (Dow, Toronto, ON, Canada), and 3M Novec 1720 (M.G. Chemicals, Burlington, ON, Canada). The photomasks used for electrodes and channels layers were designed in AutoCAD and printed by Artnet Pro Inc., Bandon, OR, USA.

As described previously,38,40,52,53 the hardware for sorting included low-pressure neMESYS pump system (Cetoni, Korbussen, DE), a TREK high-voltage amplifier PZD700A (Advanced Energy Inc., Denver, CO, USA), 33210A Keysight function/arbitrary waveform generator 10 MHz (Keysight Technologies, Santa Rosa, CA, USA), GW Instek GPE-4323 Linear DC Power Supply with 4 Channels (GW Instek America Corp., Montclair, CA, USA), inverted microscope (Olympus IX73; Tokyo, Japan) mounted on a vibration-dampening bench (Thorlabs; Newton, NJ, USA), filter block (FOFMS-UV, Thorlabs; Newton, NJ, USA), 1 mW, 470 nm fiber-coupled LED light source (Ocean Optics; Orlando, FL, USA), bandpass filters (FB450-40 and FL457.9-10), fiber optics (FG105UCA – Ø 105 µm, FG200UEP – Ø 200 µm, and FG200LCC – Ø 200 µm) (Thorlabs), and a Hamamatsu ORCA-Flash 4.0 LT+ (C11440-42U30) (Township, NJ, USA), and Flame-S-XR1-ES spectrometer (Ocean Optics). The software to automate the sorting functions (i.e., controlling flow rates, measuring fluorescence, activating electrodes) were uploaded to our Bitbucket repository (https://bitbucket.org/shihmicrolab/uflowcontrol/).52

The materials for operating microfluidic devices included 1, 3, and 5 mL plastic syringes (VWR, Mississauga, ON, Canada) which were attached to probe needles (M922 Probe Needles, McMaster-Carr, Elmhurst, Illinois, USA) and connected to Tygon tubing ID 0.02 IN and OD 0.06 IN, 3M Novec 1720 (M.G. Chemicals, Burlington, ON, Canada), 008-fluoro-surfactant dissolved in HFE 7500 (2% wt) (Ran Biotechnologies, Beverly, MA, USA). All liquids were filtered using nylon filter cartridges (0.22 µm, Millex® GP, Millipore).

Enzyme substrates used for screening and sorting included the EnzChek™ Ultra Amylase Assay Kit (Thermo Fisher Scientific, Burlington, ON, Canada) and fluorescein-di-β-d-cellobioside (AAT Bioquest, Pleasanton, CA, USA). Hydrogel Dummy Beads (1 M) was purchased from Ran Biotechnologies, Beverly, MA, USA, the 4-methylumbelliferyl-based fluorescent substrates, chitin, Avicel, starch from corn, and other chemicals used for media formulation and assay reagents were purchased from Sigma-Aldrich (St. Louis, MO) unless specified otherwise.

Colloidal solid substrate preparation

Colloidal chitin and cellulose were prepared using HCl to break the crystal structures.16,54 Both materials were primarily mixed with HCl 1 N in a 1 : 10 ratio. For colloidal chitin, chitin from shrimp shells (practical grade, CAS No. 1398-61-4, Sigma-Aldrich) was incubated overnight at 37 °C with stirring, while for colloidal cellulose, Avicel was heated at 70 °C for 10 h with stirring. Following, cold ethanol (1 : 5 ratio) was added and mixed for 30 minutes at 4 °C. For the washing steps, the solutions were centrifuged at 3000 rpm for 30 minutes at 4 °C, then washed with sterile water and centrifuged, for a total of three washes. The pH was adjusted to neutral, and the solution was diluted with sterile water to a 4% (w/v) colloidal chitin and colloidal cellulose. For both colloids, the solutions were autoclaved, and aliquots of 1.5 mL were stored at 4 °C. Before use, the aliquots were centrifuged, discarding the supernatant, then diluted into the specific media for the fungi. More details are provided in the SI as Note S1.

Filamentous fungal strains

Two mCherry-expressing strains, e.g., T. reesei (M6895) and A. oryzae (M7091) were used in this study. Briefly, T. reesei (M6895) was created by introducing xyr1 mutation at V821F to T. reesei RUT-C30,49,55 followed by insertion of a codon-optimized β-lactoglobulin-mCherry fusion construct under cbh1 promoter, synthesized and cloned to a plasmid containing cbh1-upstream and downstream fragments by Genscript (Rijswijk, Netherlands) (Fig. S4). For A. oryzae, M7091 was created by deletion of ligD, pyrG and creA genes, respectively, followed by insertion of a codon-optimized amyC-mCherry fusion construct under a synthetic expression system.3 The deletion cassettes were constructed using 600–1300 bp upstream and downstream DNA fragments of the target genes containing an overlap of the selection marker pyrG from A. oryzae. The expression cassette of optimized amyC linked with mCherry was synthesized and cloned to a plasmid containing amyC-upstream and downstream fragments by Genscript (Fig. S4). The deletion or expression cassette was transformed through protoplast-mediated transformation as previously described.49 The transformants were confirmed by PCR and qPCR towards the target genes to confirm the correct integration. Spore suspensions were prepared by growing the fungus on potato-dextrose agar plates for five days. Spores were harvested and suspended in a spore suspension buffer (0.8% NaCl, 0.025% Tween 20, and 20% glycerol), through cotton and stored at −80 °C.

UV mutagenesis

For UV mutagenesis, the spore solution was adjusted to 1 × 107 spore mL−1 using spore suspension buffer. Two 600 µL of the spore suspensions were irradiated with different UV doses using a UV Stratalinker 2400 (Stratagene, La Jolla, CA, USA). After the UV mutagenesis, the spores were collected, covered with aluminium foil to prevent photoactivation, and kept frozen until sorting. The UV dose for generating the mutant population of T. reesei strain was 75 mJ cm−3 (survival rate of 7%) and 25 mJ cm−3 for A. oryzae (9.3% survival rate).

Macroscale fungal growth conditions

Growth and enzymatic profiles of parental strains in media with and without solid support were assessed by culturing 1 mL of media in 96 deep well plates (VWR, Mississauga, ON) at 30 °C and 200 rpm for five days. Following cultivation, the supernatant was collected for enzyme activity measurement. For T. reesei, the culture media was T. reesei Minimal Media (TrMM, pH 4.8) containing 1.5% (w/v) KH2PO4, 0.5% (w/v) (NH4)2SO4, 0.05% (w/v) proteose peptone, and varying concentrations of glucose and CMC (0.5–2%). For A. oryzae, the culture media contained minimal medium with 0.6% (w/v) NaNO3, 0.052% (w/v) KCl, 0.025% (v/v), 5% (v/v) 1M potassium–phosphate buffer (pH 6.0), 0.5% (w/v) MgSO4, and varying concentrations of glucose (0–2%).

After sorting and collecting mutant strains, they were cultured in potato dextrose agar (PDA) at 30 °C for ∼10 days or until fungal colonies were visible. Colonies were transferred to PDA treated with ampicillin and/or chloramphenicol (50 mg L−1 final concentration) for an additional 7 days. The colonies were re-suspended in 0.9% NaCl and 50 µL of spore solution was added to 4 mL of liquid media in a 24-well plate format. For T. reesei, two TrMM liquid media were used: one containing 4% glucose and the other containing 2% CMC, and for A. oryzae 2% glucose in BMDY liquid media was used, which contained 1% (w/v) yeast extract, 2% (w/v) Bacto peptone, 0.004% (v/v) biotin, 10% (v/v) 1M potassium–phosphate buffer (pH 6.0), and 1.34% YNB (Yeast Nitrogen Base with ammonium sulfate, without amino acids). The strains were cultivated for an additional five days at 200 rpm at 30 °C. After incubation, the supernatant was retrieved and stored at 4 °C for downstream evaluation of enzymatic activity and protein content.

To evaluate the top-sorted mutants on the bench scale, mutants were transferred to 250 mL Erlenmeyer flasks that contained 100 mL of 2% glucose or 1% starch BMDY media for A. oryzae and 2% CMC or 4% glucose TrMM media for T. reesei. Each flask was inoculated with 100 µL of spore solution (for a final spore concentration of 1 × 104 spores mL−1) from each mutant and cultivated for up to six days at 200 rpm and 30 °C, and aliquots of the samples were collected at different time points throughout the incubation time. At the end of the cultivation, the dry biomass was collected by filtering with a Whatman® qualitative filter paper and dried at 65 °C over 40 h.

Microfluidic devices fabrication and assembly

The devices were fabricated by photolithography methods using previously described methods.16,40,52 Briefly, channels were patterned onto a 100 mm Si-wafer that was cleaned with acetone, isopropyl alcohol, and DI water prior to photoresist deposition. The wafers were treated under oxygen plasma for 2 min 30 s prior to spin coating. SU-8 2075 was used for fabricating 120 µm channel height for droplet generators and 160 µm channel height for sorters. The spin coating protocol was executed in three steps: for 120 µm height (10 s at 500 rpm, 30 s at 1850 rpm, and 10 s at 500 rpm), for 160 µm height (10 s at 500 rpm, 30 s at 1500 rpm, and 10 s at 500 rpm); baking and exposure time followed the manufacturer's instructions.

The master mold was exposed to chlorotrimethylsilane vapor deposition in a vacuum desiccator for 45 min. Next, PDMS (1 : 10 w/w ratio curing agent to prepolymer), degassed, and poured over the mold and left to cure in an oven (65 °C, 2 h). PDMS layers were excised with an X-Acto knife, inlets and outlets were made using a disposable biopsy punch of 1.50 mm diameter (Robbins Instruments, TX, USA). The PDMS was washed with isopropyl alcohol, DI water, air dried, and cleaned with tape to remove dust before device assembly – PDMS to glass (for droplet generators) and PDMS to PDMS (for sorters).

Droplet generator devices consisting of a PDMS channel layer and a 50 × 75 mm glass slide were treated with oxygen plasma for 45 s then bonded together and baked at 90 °C for 30 min. Device channels were then treated with 3M Novec 1720 fluorosilane polymer surfactant, followed by a baking step of 160 °C for 30 min.

The sorter device contained an electrode and a dielectric layer prior to incorporating a channel layer, as previously reported.16,38–40,52 To fabricate the electrodes, chromium-coated 2″ × 3″ mm glass slides (Telic, Valencia, CA, USA) with AZ-1500 positive were UV exposed (7 s at 38–50 mW cm−2), then developed in MF-321 developer, etched with CR-4 chromium etchant, and stripped with AZ-300T photoresist stripper. For the dielectric layer, ∼7 µm PDMS was spin coated, the dielectric layer was then baked at 65 °C for 2 h. The PDMS channel layer (i.e. sorting channels) was manually aligned with the dielectric coated electrodes using an inverted microscope using a 4× objective (Olympus IX73; Evident Scientific, Mississauga, ON, Canada). After bonding. channels were treated with 3M Novec 1720 fluorosilane polymer surfactant for 30 minutes at room temperature and baked at 165 °C for 30 minutes. Two flat cleaved multi-mode optical fibers were prepared for droplet excitation (100 µm core, 0.22 N.A.) and detection (200 µm core, 0.49 N.A.) (Thorlabs, NJ, USA). Fibers were fixed in place with tape directly on the device.

Microscale culture for optimization of solid substrates in droplets

To culture parental strains expressing mCherry fusion proteins of A. oryzae and T. reesei in droplets, spore solutions were diluted following the Poisson distribution (λ = 0.1, ∼1 × 105 spore mL−1). The spores were diluted into the different media compositions and 2 nL droplets were formed by using a T-junction droplet generator with one inlet containing 2% fluorosurfactant in HFE 7500 oil and another inlet containing the spores and media at flow rates 800 nL per sand 400 nL s−1 respectively. For co-encapsulation of T. reesei spores with hydrogel beads, we used λ = 1 for Poisson distribution (1 × 106 beads per mL−1) to maximize droplets containing at least one bead, the spore concentration was the same as mentioned above.

Droplets were collected in a microcentrifuge tube with 30 µL 2% fluorosurfactant in HFE 7500 oil and incubated at 21 and 30 °C in the dark for both A. oryzae and T. reesei. Droplets were transferred by pipetting from the tube to cell counting slides (Bio-Rad, Cat. Number 1450011) to obtain fluorescence and bright field images. Measurements were taken every 24 h to measure hyphae growth and mCherry expression. To determine the optimal incubation time for enzyme activity detection, fluorescent substrates were used: EnzCheck Ultra Amylase kit for alpha-amylase and Fluorescein di-β-d-cellobioside for cellulase. Protein expression was assessed via mCherry fluorescence, enzyme activity via fluorescent substrate assays, and hyphal growth and droplet integrity by counting hyphae piercing the droplets. The fluorescence intensities were measured using Fiji (ImageJ) based on previously reported technique (see Note S2, SI),40,56 and calculated through corrected total droplet fluorescence (CTDF) according to eqn (1).

graphic file with name d6cb00197a-t1.webp 1

This approach normalizes fluorescence by droplet size, background fluorescence, and baseline levels. Droplet fluorescence (DF; a.u.) is the measured density of fluorescent droplets, using the function IntDen in Fiji-ImageJ. DA (in2) is the droplet area, measured as area in Fiji-ImageJ. Mean BR (AU/in) is the mean background fluorescence, which is the measured mean value in Fiji-ImageJ (AU/in). Initial conditions (t0) for these parameters: DFt0, DAt0 and mean BRt0, were obtained immediately after droplet generation.

Solid substrate encapsulation validation

To estimate the colloidal substrate encapsulation distribution, two sets of 2 nL droplets were generated; control (empty) droplets and encapsulation of 4% colloidal chitin droplets. Using a custom script (https://bitbucket.org/shihmicrolab/droplet-colloid-content-distribution-analyzer), the areas of the droplets were detected and analyzed for pixel intensity. The average intensity of a pixel within the empty control droplets was set as the threshold value. In a bright field image, the solid support appears as a darker matrix and as such, we estimated the droplet area coverage of the substrate by calculating the percentile composition of pixels with an intensity lower than threshold (eqn (2)).

graphic file with name d6cb00197a-t2.webp 2

Npixel in droplet < threshold is the number of pixels within a given droplet with an intensity value lower than the set threshold (average pixel intensity value for an empty droplet).

Sorting operation and validation

As previously described,16,40,52 plastic syringes of volumes 1-, 3- and 5 mL were connected to Tygon tubing and installed on a low-pressure neMESYS pump system (Cetoni, Korbussen, DE). The sorting device with installed optical fibers was fixed in a 3D printed holder and clamped in place with a pogo pin PCB that directly contacts the electrode pads for electrical actuation. The holder base plate was designed to be placed in the scanning stage (XYZ Tango, Marzhauser, Wetzlar, DE) of an inverted epi-fluorescence microscope (Olympus IX73, Olympus, Tokyo, Japan). Next, the SMA end of the excitation fiber was coupled to a 500 nm short-pass filter (Thorlabs, NJ, US) connected to a high power (1 mW) 470 nm fiber coupled LED light source. The SMA end of the emission fiber was coupled to a portable mini-spectrometer (FLAME-S-XR1-ES, Ocean Insight, NY, USA) with the integration time set to 50 ms (20 measurements per second). The spectrometer, flow rates, and electrode actuation were controlled using an in-house automation system and graphical user interface.52

The sorter design was evaluated by generating 2 nL droplets containing 4% colloidal chitin, which were collected in a 3 mL plastic syringe followed by re-injection into the sorter. The inlets were connected to the 3 mL syringe containing the droplets and to a 5 mL spacer oil (3M Novec – HFE 7500 without surfactant). The waste outlet contained a withdrawing 5 mL syringe, and the sorting channel contained Tygon tubing that was connected to a microcentrifuge tube with 100 µL of spacer oil. To test flow operational conditions (Table S2), we evaluated different spacer oil flow rates with no electrode actuation. The operational success was measured by determining the number of droplets being injected and collected into the waste channel.

To optimize voltage conditions, the colloidal chitin droplets were re-injected into the sorter with the successful operational conditions (Table S2). The applied AC signal amplitude varied (35–210 VRMS, 15 kHz sine, 5 levels), and the droplet re-injection rate was set to 20 nL s−1. Droplets were sorted by actuating the sorting electrodes (SE), and the sorting response was measured by the number of droplets that flowed into the activated sorting channel, and the droplet travel time and actuation time were also experimentally recorded. For data analysis, images were recorded using a Hamamatsu ORCA-Flash LT+ 4.0 camera.

To validate autonomous sorting, the sorting efficiency was measured by generating droplets of fluorescein 1 mM and blue dye at a 1 : 10 ratio and collected into a 3 mL syringe. These droplets were injected into the sorter (20 nL s−1 droplet flow rate, 400 nL s−1 spacer oil flow rate, −350 nL s−1 withdrawing flow, 30 ms actuation delay and 20 ms actuation time). The fluorescence intensity was detected by the optical fibers, and autonomous sorting was executed by electrode actuation (140 VRMS; 15 kHz) controlled via in-house software. The efficiency was measured by counting the percentage of fluorescent droplets that were sorted (see eqn (3)), and the false positive rate was calculated by counting the number of blue dye droplets that have been sorted (eqn (4)). The false negative rate was determined by the number of fluorescent droplets in the waste channel (i.e. missed during the sorting event) based on eqn (5). Images of the droplets in the sorting and waste channel were recorded using a Hamamatsu ORCA-Flash4.0 LT+ camera at 30 frames per second. Sorting frequency was determined through frame analysis (Fig. S10), based on the frame interval at which droplets reach the sorting junction (eqn (6)).

graphic file with name d6cb00197a-t3.webp 3
graphic file with name d6cb00197a-t4.webp 4
graphic file with name d6cb00197a-t5.webp 5
graphic file with name d6cb00197a-t6.webp 6

TSD (total sorted sorted), SFD (sorted fluorescent droplets), TWD (total waste droplets) and UFD (unsorted fluorescent droplets).

Fungal enzymatic activity and protein evaluation

To measure fungal enzyme activity in a 96 well plate format, three different substrates were used: EnzChek™ Ultra Amylase Assay Kit (for A. oryzae amylase) and 4-methylumbelliferyl β-d-cellobioside (for T. reesei cellulase).

For cellulase, the enzyme activity was measured by mixing 50 µL of the enzyme substrate (750 µM final concentration) and 50 µL of the supernatant culture, followed by incubation at 30 °C for 2 h. After incubation, sodium carbonate (0.4 M) stop buffer (pH 11) was added and the fluorescence (λex = 360 nm, λem = 450 nm) was measured in a CLARIOstar Plus (BMG LABTECH, Ortenberg, Germany), and activity was calculated using a methylumbelliferone standard curve.

For amylase activity, 50 µL of the enzyme substrate and 50 µL of the supernatant culture were mixed followed by incubation at room temperature for 30 min. The fluorescence (λex = 477/500 nm, λem = 525/520 nm) was measured in a CLARIOstar Plus (BMG LABTECH, Ortenberg, Germany), and activity was calculated using a BODIPY™ FL standard curve.

Total protein was quantified using the Bradford assay by combining 150 µL of Bradford reagent with 150 µL of culture supernatant. When absorbance exceeded the detection limit of the standard curve, appropriate dilutions were prepared to ensure accurate measurement. The absorbance (λ = 595 nm) was measured in a CLARIOstar Plus (BMG LABTECH, Ortenberg, Germany), and protein content was calculated using a BSA standard curve.

The molecular mass of proteins in the supernatants was analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) using 4–20% polyacrylamide gel (CriterionTM TGXTM Precast Protein Gels, Bio-Rad) and the standard marker, Precision Plus protein All blue standard (Bio-Rad) with PageBlue™ Protein Staining Solution (Thermo Scientific).

Mutant droplet generation, sorting, and recovery

The UV-generated mutant population of T. reesei strains (75 mJ cm−3 – 7% survival rate) and A. oryzae (25 mJ cm−3 – 9.3% survival rate) were prepared and encapsulated in droplets as the protocol above. Droplets were collected into a 3 mL plastic syringe covered with parafilm and incubated for 72 h (for colloidal chitin) and 48 h (for colloidal cellulose). After incubation, the droplets were re-injected into the sorter through a flow rate of 20 nL s−1, spacer oil flow rate of 400 nL per sand withdrawing flow rate of −350 nL s−1. The fluorescence of the population was measured and graphed into a histogram to set up the sorting gates (97th percentile) and the droplets were sorted by applying potentials to the sorting electrodes (140 VRMS; 15 kHz). The droplets were collected using a Tygon tubing connected between the sorting outlet and an open Eppendorf tube containing 300 µL of HFE7500. The sorted droplets were plasma treated and vortexed to break up the emulsion followed by centrifugation at 15 000 rpm for 5 minutes. The oil phase was separated from the aqueous and the aqueous phase was resuspended into 150 µL NaCl 0.9% solution, and 10 µL aliquots were plated on PDA plates. The plates were incubated at 30 °C for 10 days, after which ∼80 single colonies were taken for post-sorting mutant comparison with the parental strain, analyzing enzyme activity (α-amylase for A. oryzae, cellulase for T. reesei) and total protein content as described in the macroscale culture conditions, and in the enzymatic activity and protein content sections.

Data analysis

Data and statistical analysis were performed using GraphPad Prism v8.4.3 and the significance p-values for each plot are shown in Tables S3–S5. Image analysis and droplet sizes calculations were done using Fiji (ImageJ). The applied electrical signal was measured using an oscilloscope. The graphical figures were generated with Adobe Illustrator and Biorender. All in-house codes were previously described by our group.16,52

Author contributions

Experiments were designed by CLA, AD, MV, and SCCS. CLA designed and built the microfluidic devices and conducted all droplet microfluidic experiments. JZW and ZD also provided guidance with the design and fabrication of the microfluidic devices. CLA, AD, BP, and MV designed and performed the A. oryzae culturing experiments along with the enzyme analysis. JZW, EK, CLA, AD, BP, and MV designed and performed the T. reesei culturing experiments along with the heterologous protein and enzyme analysis. AD, MV, BP, and APA engineered the mCherry cassettes in the fungal strains and performed the UV mutagenesis. CLA, JZW, EK, AD, MV, and SCCS analyzed the data and organized the data/figures. NR aided with culturing spores and spore preparation for the experiments. All authors contributed to the writing of the manuscript, and all authors reviewed the final version of the manuscript before submission.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

CB-OLF-D6CB00197A-s001

Acknowledgments

We thank the Natural Sciences and Engineering Research Council (NSERC), the Canadian Foundation of Innovation (CFI), VTT Technical Research Center of Finland for funding. CLA, NR, and EK thank Fonds de Recherche Québec - Nature et Technologies (FRQ-NT) for a PhD scholarship. CLA, EK, JZW, ZD, NR, thank Concordia University Faculty of Engineering for FRS funding. SCCS thanks Concordia University for a Research Chair. AD, APA, MV thank funding from CERAFIM (2022–2024), a co-innovation project supported by Business Finland (grant number 9392/31/2021).

Data availability

All data are located in a controlled access data storage at Concordia University. The data is readily available on request.

Supplementary information (SI) provides additional experimental methods, platform characterization, and validation data supporting the droplet-based screening workflow. Supplementary notes describe the preparation of colloidal cellulose and image-based quantification of droplet and fungal fluorescence. Supplementary tables compare the current platform with our previously reported fungal screening system, detail optimization of droplet spacing and withdrawal flow conditions, and provide statistical analyses of scaled-up α-amylase and cellulase production by selected A. oryzae and T. reesei mutants. Supplementary figures provide additional characterization of droplet generation and solid-substrate loading; fungal growth, hyphal containment, and enzyme expression under different substrate, media, temperature, and incubation conditions; electrostatic sorter design, performance, and throughput; droplet-level fluorescence heterogeneity of parental and mutant populations; and shake-flask validation and SDS-PAGE analysis of selected high-producing strains. See DOI: https://doi.org/10.1039/d6cb00197a.

References

  1. Oliveira L. R. et al., Genetic Engineering of Filamentous Fungi: Prospects for Obtaining Fourth-Generation Biological Products. Appl. Microbiol. 2024;4:794–810. doi: 10.3390/applmicrobiol4020055. [DOI] [Google Scholar]
  2. Punt P. J. et al., Filamentous fungi as cell factories for heterologous protein production. Trends Biotechnol. 2002;20:200–206. doi: 10.1016/S0167-7799(02)01933-9. [DOI] [PubMed] [Google Scholar]
  3. Rantasalo A. et al., A universal gene expression system for fungi. Nucleic Acids Res. 2018;46:e111. doi: 10.1093/nar/gky558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Denardi de Souza T. Leal C. A. Massarolo K. C. Badiale-Fulong E. Profile of Phenolic Compounds Released from Rice Bran by Rhizopus oryzae and Trichoderma reesei: Their Relation with Hydrolases Activity. J. Food Sci. 2019;84:1382–1389. doi: 10.1111/1750-3841.14646. [DOI] [PubMed] [Google Scholar]
  5. Gupta R. Kumari A. Syal P. Singh Y. Molecular and functional diversity of yeast and fungal lipases: Their role in biotechnology and cellular physiology. Prog. Lipid Res. 2015;57:40–54. doi: 10.1016/j.plipres.2014.12.001. [DOI] [PubMed] [Google Scholar]
  6. Balakrishnan M. Jeevarathinam G. Kumar S. K. S. Muniraj I. Uthandi S. Optimization and scale-up of α-amylase production by Aspergillus oryzae using solid-state fermentation of edible oil cakes. BMC Biotechnol. 2021;21:33. doi: 10.1186/s12896-021-00686-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Häkkinen M. et al., Screening of candidate regulators for cellulase and hemicellulase production in Trichoderma reesei and identification of a factor essential for cellulase production. Biotechnol. Biofuels. 2014;7:14. doi: 10.1186/1754-6834-7-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Correia J. Borges A. Simões M. Simões L. C. Beyond Penicillin: The Potential of Filamentous Fungi for Drug Discovery in the Age of Antibiotic Resistance. Antibiotics. 2023;12:1250. doi: 10.3390/antibiotics12081250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Lu Z. et al., Morphological Engineering of Filamentous Fungi: Research Progress and Perspectives. J. Microbiol. Biotechnol. 2024;34:1197–1205. doi: 10.4014/jmb.2402.02007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gibbs P. A. Seviour R. J. Schmid F. Growth of Filamentous Fungi in Submerged Culture: Problems and Possible Solutions. Crit. Rev. Biotechnol. 2000;20:17–48. doi: 10.1080/07388550091144177. [DOI] [PubMed] [Google Scholar]
  11. Cebrián M. and Ibarruri J., Filamentous fungi processing by solid-state fermentation, Current Developments in Biotechnology and Bioengineering, Elsevier, 2023, pp. 251–292 [Google Scholar]
  12. Brand A. Gow N. A. Mechanisms of hypha orientation of fungi. Curr. Opin. Microbiol. 2009;12:350–357. doi: 10.1016/j.mib.2009.05.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Montenecourt B. S. Eveleigh D. E. Preparation of mutants of Trichoderma reesei with enhanced cellulase production. Appl. Environ. Microbiol. 1977;34:777–782. doi: 10.1128/aem.34.6.777-782.1977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Montenecourt M. and Eveleigh D. E., Selective Screening Methods for the Isolation of High Yielding Cellulase Mutants of Trichoderma reesei, in Hydrolysis of Cellulose: Mechanisms of Enzymatic and Acid Catalysis, 1979, vol. 181, pp. 289–301 [Google Scholar]
  15. Beneyton T. et al., High-throughput screening of filamentous fungi using nanoliter-range droplet-based microfluidics. Sci. Rep. 2016;6:27223. doi: 10.1038/srep27223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Samlali K. Alves C. L. Jezernik M. Shih S. C. C. Droplet digital microfluidic system for screening filamentous fungi based on enzymatic activity. Microsyst. Nanoeng. 2022;8:123. doi: 10.1038/s41378-022-00456-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. He R. Ding R. Heyman J. A. Zhang D. Tu R. Ultra-high-throughput picoliter-droplet microfluidics screening of the industrial cellulase-producing filamentous fungus Trichoderma reesei. J. Ind. Microbiol. Biotechnol. 2019;46:1603–1610. doi: 10.1007/s10295-019-02221-2. [DOI] [PubMed] [Google Scholar]
  18. Zhang C. et al., Core–Shell Droplet-Based Microfluidic Screening System for Filamentous Fungi. ACS Sens. 2023;8:3468–3477. doi: 10.1021/acssensors.3c01018. [DOI] [PubMed] [Google Scholar]
  19. Luu X. C. et al., A novel high-throughput approach for transforming filamentous fungi employing a droplet-based microfluidic platform. New Biotechnol. 2022;72:149–158. doi: 10.1016/j.nbt.2022.11.003. [DOI] [PubMed] [Google Scholar]
  20. Li S. et al., Establishment of picodroplet-based co-culture system to improve erythritol production in Yarrowia lipolytica. Biochem. Eng. J. 2023;198:109036. doi: 10.1016/j.bej.2023.109036. [DOI] [Google Scholar]
  21. Li Q. et al., High-throughput droplet microfluidics screening and genome sequencing analysis for improved amylase-producing Aspergillus oryzae. Biotechnol. Biofuels. 2023;16:185. doi: 10.1186/s13068-023-02437-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Miyazawa K. Umeyama T. Hoshino Y. Abe K. Miyazaki Y. Quantitative Monitoring of Mycelial Growth of Aspergillus fumigatus in Liquid Culture by Optical Density. Microbiol. Spectrum. 2022;10:e00063–21. doi: 10.1128/spectrum.00063-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Novy V. Schmid M. Eibinger M. Petrasek Z. Nidetzky B. The micromorphology of Trichoderma reesei analyzed in cultivations on lactose and solid lignocellulosic substrate, and its relationship with cellulase production. Biotechnol. Biofuels. 2016;9:169. doi: 10.1186/s13068-016-0584-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Liu D. Garrigues S. de Vries R. P. Heterologous protein production in filamentous fungi. Appl. Microbiol. Biotechnol. 2023;107:5019–5033. doi: 10.1007/s00253-023-12660-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Cherry J. R. Fidantsef A. L. Directed evolution of industrial enzymes: an update. Curr. Opin. Biotechnol. 2003;14:438–443. doi: 10.1016/S0958-1669(03)00099-5. [DOI] [PubMed] [Google Scholar]
  26. Li Y. et al., Rapid filamentous fungi gene knockout identification through high-throughput droplet microfluidics. Anal. Chim. Acta. 2025;1358:344094. doi: 10.1016/j.aca.2025.344094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Derntl C. et al., Mutation of the Xylanase regulator 1 causes a glucose blind hydrolase expressing phenotype in industrially used Trichoderma strains. Biotechnol. Biofuels. 2013;6:62. doi: 10.1186/1754-6834-6-62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Schwarze F. W. M. R. Gholam Z. Nyström G. Heeb M. Microfibrillated cellulose as growth media for cultivation and maintenance of wood decay fungi. J. Microbiol. Methods. 2022;203:106615. doi: 10.1016/j.mimet.2022.106615. [DOI] [PubMed] [Google Scholar]
  29. Reyes C. et al., Cocultivation of White-Rot Fungi and Microalgae in the Presence of Nanocellulose. Microbiol. Spectrum. 2022;10:e03041–22. doi: 10.1128/spectrum.03041-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Bischof R. H. Ramoni J. Seiboth B. Cellulases and beyond: the first 70 years of the enzyme producer Trichoderma reesei. Microb. Cell Fact. 2016;15:106. doi: 10.1186/s12934-016-0507-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Barrios-González J. Solid-state fermentation: Physiology of solid medium, its molecular basis and applications. Process Biochem. 2012;47:175–185. doi: 10.1016/j.procbio.2011.11.016. [DOI] [Google Scholar]
  32. Tlalka M. Bebber D. P. Darrah P. R. Watkinson S. C. Fricker M. D. Quantifying dynamic resource allocation illuminates foraging strategy in Phanerochaete velutina. Fungal Genet. Biol. 2008;45:1111–1121. doi: 10.1016/j.fgb.2008.03.015. [DOI] [PubMed] [Google Scholar]
  33. Fukasawa Y. Ishii K. Foraging strategies of fungal mycelial networks: responses to quantity and distance of new resources. Front. Cell Dev. Biol. 2023;11:1244673. doi: 10.3389/fcell.2023.1244673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Tzelepis G. Dubey M. Jensen D. F. Karlsson M. Identifying glycoside hydrolase family 18 genes in the mycoparasitic fungal species Clonostachys rosea. Microbiology. 2015;161:1407–1419. doi: 10.1099/mic.0.000096. [DOI] [PubMed] [Google Scholar]
  35. Qu T. et al., A Novel GH Family 20 β-N-acetylhexosaminidase With Both Chitosanase and Chitinase Activity From Aspergillus oryzae. Front. Mol. Biosci. 2021;8:684086. doi: 10.3389/fmolb.2021.684086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Machida M. et al., Genome sequencing and analysis of Aspergillus oryzae. Nature. 2005;438:1157–1161. doi: 10.1038/nature04300. [DOI] [PubMed] [Google Scholar]
  37. Verma N. Kumar V. Impact of process parameters and plant polysaccharide hydrolysates in cellulase production by Trichoderma reesei and Neurospora crassa under wheat bran based solid state fermentation. Biotechnol. Rep. 2020;25:e00416. doi: 10.1016/j.btre.2019.e00416. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ahmadi F. Samlali K. Vo P. Q. N. Shih S. C. C. An integrated droplet-digital microfluidic system for on-demand droplet creation, mixing, incubation, and sorting. Lab Chip. 2019;19:524–535. doi: 10.1039/C8LC01170B. [DOI] [PubMed] [Google Scholar]
  39. Samlali K. Ahmadi F. Quach A. B. V. Soffer G. Shih S. C. C. One Cell, One Drop, One Click: Hybrid Microfluidics for Mammalian Single Cell Isolation. Small. 2020;16:2002400. doi: 10.1002/smll.202002400. [DOI] [PubMed] [Google Scholar]
  40. Leal-Alves C. et al., A Microfluidic Multiplex Sorter for Strain Development. Adv. Mater. Technol. 2025;10:2401209. doi: 10.1002/admt.202401209. [DOI] [Google Scholar]
  41. Saffar Y. Nobes D. S. Sabbagh R. Experimental Investigation of the Motion and Deformation of Droplets in Curved Microchannel. Ind. Eng. Chem. Res. 2023;62:17275–17286. doi: 10.1021/acs.iecr.3c01623. [DOI] [Google Scholar]
  42. De Stefano P. Bianchi E. Dubini G. The impact of microfluidics in high-throughput drug-screening applications. Biomicrofluidics. 2022;16:31501. doi: 10.1063/5.0087294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Cong H. Zhang N. Perspectives in translating microfluidic devices from laboratory prototyping into scale-up production. Biomicrofluidics. 2022;16:021301. doi: 10.1063/5.0079045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Arai T. et al., Inducer-free cellulase production system based on the constitutive expression of mutated XYR1 and ACE3 in the industrial fungus Trichoderma reesei. Sci. Rep. 2022;12:19445. doi: 10.1038/s41598-022-23815-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Tsuchiya K. et al., Deletion Analysis of the Taka-amylase A Gene Promoter Using a Homologous Transformation System in Aspergillus oryzae. Biosci., Biotechnol., Biochem. 1992;56:1849–1853. doi: 10.1271/bbb.56.1849. [DOI] [PubMed] [Google Scholar]
  46. Wang Q. Zhong C. Xiao H. Genetic Engineering of Filamentous Fungi for Efficient Protein Expression and Secretion. Front. Bioeng. Biotechnol. 2020;8:293. doi: 10.3389/fbioe.2020.00293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Sun Z. et al., Aspergillus oryzae as a Cell Factory: Research and Applications in Industrial Production. J. Fungi. 2024;10:248. doi: 10.3390/jof10040248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Yoon J. Aishan T. Maruyama J. Kitamoto K. Enhanced Production and Secretion of Heterologous Proteins by the Filamentous Fungus Aspergillus oryzae via Disruption of Vacuolar Protein Sorting Receptor Gene Aovps10. Appl. Environ. Microbiol. 2010;76:5718–5727. doi: 10.1128/AEM.03087-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sveholm E. et al., Transcriptomic and metabolic changes in trichoderma reesei caused by mutation in xylanase regulator 1 (xyr1) Biotechnol. Biofuels Bioprod. 2024;17:106. doi: 10.1186/s13068-024-02556-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Portnoy T. et al., Differential Regulation of the Cellulase Transcription Factors XYR1, ACE2, and ACE1 in Trichoderma reesei Strains Producing High and Low Levels of Cellulase. Eukaryotic Cell. 2011;10:262–271. doi: 10.1128/EC.00208-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Stricker A. R. Grosstessner-Hain K. Würleitner E. Mach R. L. Xyr1 (Xylanase Regulator 1) Regulates both the Hydrolytic Enzyme System and d-Xylose Metabolism in Hypocrea jecorina. Eukaryotic Cell. 2006;5:2128–2137. doi: 10.1128/EC.00211-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Ahmadi F. et al., An automated single-cell droplet-digital microfluidic platform for monoclonal antibody discovery. Small. 2024;20:2308950. doi: 10.1002/smll.202308950. [DOI] [PubMed] [Google Scholar]
  53. Little S. R. et al., A Tri-Droplet Liquid Structure for Highly Efficient Intracellular Delivery in Primary Mammalian Cells Using Digital Microfluidics. Adv. Mater. Technol. 2023;8:2300719. doi: 10.1002/admt.202300719. [DOI] [Google Scholar]
  54. Wang H. et al., Intracellular sugar transporters facilitate cellulase synthesis in trichoderma reesei using lactose. Biomolecules. 2023;13:295. doi: 10.3390/biom13020295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Ellilä S. et al., Development of a low-cost cellulase production process using Trichoderma reesei for Brazilian biorefineries. Biotechnol. Biofuels. 2017;10:30. doi: 10.1186/s13068-017-0717-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Shihan M. H. Novo S. G. Le Marchand S. J. Wang Y. Duncan M. K. A simple method for quantitating confocal fluorescent images. Biochem. Biophys. Rep. 2021;25:100916. doi: 10.1016/j.bbrep.2021.100916. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

CB-OLF-D6CB00197A-s001
CB-OLF-D6CB00197A-s001.pdf (3.3MB, pdf)

Data Availability Statement

All data are located in a controlled access data storage at Concordia University. The data is readily available on request.

Supplementary information (SI) provides additional experimental methods, platform characterization, and validation data supporting the droplet-based screening workflow. Supplementary notes describe the preparation of colloidal cellulose and image-based quantification of droplet and fungal fluorescence. Supplementary tables compare the current platform with our previously reported fungal screening system, detail optimization of droplet spacing and withdrawal flow conditions, and provide statistical analyses of scaled-up α-amylase and cellulase production by selected A. oryzae and T. reesei mutants. Supplementary figures provide additional characterization of droplet generation and solid-substrate loading; fungal growth, hyphal containment, and enzyme expression under different substrate, media, temperature, and incubation conditions; electrostatic sorter design, performance, and throughput; droplet-level fluorescence heterogeneity of parental and mutant populations; and shake-flask validation and SDS-PAGE analysis of selected high-producing strains. See DOI: https://doi.org/10.1039/d6cb00197a.

Articles from RSC Chemical Biology are provided here courtesy of Royal Society of Chemistry

RESOURCES