Direct Image-Based Enumeration of Clostridium phytofermentans Cells on Insoluble Plant Biomass Growth Substrates

Abstract

A dual-fluorescent-dye protocol to visualize and quantify Clostridium phytofermentans ISDg (ATCC 700394) cells growing on insoluble cellulosic substrates was developed by combining calcofluor white staining of the growth substrate with cell staining using the nucleic acid dye Syto 9. Cell growth, cell substrate attachment, and fermentation product formation were investigated in cultures containing either Whatman no. 1 filter paper, wild-type Sorghum bicolor, or a reduced-lignin S. bicolor double mutant (bmr-6 bmr-12 double mutant) as the growth substrate. After 3 days of growth, cell numbers in cultures grown on filter paper as the substrate were 6.0- and 2.2-fold higher than cell numbers in cultures with wild-type sorghum and double mutant sorghum, respectively. However, cells produced more ethanol per cell when grown with either sorghum substrate than with filter paper as the substrate. Ethanol yields of cultures were significantly higher with double mutant sorghum than with wild-type sorghum or filter paper as the substrate. Moreover, ethanol production correlated with cell attachment in sorghum cultures: 90% of cells were directly attached to the double mutant sorghum substrate, while only 76% of cells were attached to wild-type sorghum substrate. With filter paper as the growth substrate, ethanol production was correlated with cell number; however, with either wild-type or mutant sorghum, ethanol production did not correlate with cell number, suggesting that only a portion of the microbial cell population was active during growth on sorghum. The dual-staining procedure described here may be used to visualize and enumerate cells directly on insoluble cellulosic substrates, enabling in-depth studies of interactions of microbes with plant biomass.

INTRODUCTION

Microbial decomposition of plant biomass is central to nutrient cycling in numerous varied environments, and this process plays a key role in the cycling of carbon on the planet. In terrestrial environments, carbon storage occurs primarily in forests where soils are estimated to store 1,200 gigatons of carbon, approximately two thirds more carbon than found in the atmosphere (1). A major thrust of terrestrial microbial ecology is centered on understanding the composition and function of microbial communities in order to assess their influence on carbon cycling (2). In anoxic environments rich in decaying plant material, diverse communities of interacting microbes are responsible for the degradation of cellulose, complex polysaccharides, and other abundantly produced plant cell wall components. Given that most of these substrates are insoluble, microbial decomposition of plant biomass occurs extracellularly, and breakdown products may become available to other community members, forming a basis for multifarious interactions that occur in these environments (3).

Anoxic decomposition of plant biomass is tied to health and nutrition in animals by way of gastrointestinal tract microbial communities. Most animals lack the enzymatic capacity required to digest cellulose and many other components of plant cell walls and instead rely on plant biomass-decomposing microbial communities to provide nutrition from plant fiber. Ruminants, a group of herbivorous mammals, degrade forage in a specialized foregut organ, the rumen. The microbial community housed in the rumen plays an essential role in the development and health of the ruminant by decomposing and fermenting plant materials and forming products, such as volatile fatty acids, that serve as essential nutrients (4). In humans, complex plant carbohydrates, also known as dietary fiber, serve as the substrates for intestinal microbes that ferment these components of plant biomass to short-chain fatty acids, which are tied to colonic and systemic health (5–9).

Microbial interactions with plant biomass are also central to technologies for the conversion of lignocellulosic feedstocks into useful products. For example, ethanol is an important renewable energy source that may replace petroleum-based transportation fuels and reduce CO₂ emissions (10, 11). Development of more-sustainable processes for second-generation biofuels derived from lignocellulosic biomass is limited by the recalcitrance of the substrate, primarily due to its lignin content, and often requires expensive pretreatment and use of hydrolytic enzymes (12–14). An alternative strategy to convert lignocellulosic feedstocks into biofuels is consolidated bioprocessing, where microbes capable of producing a complex set of glycoside hydrolases decompose plant biomass and ferment the products in a single step, for a more cost-effective process with savings in capital and operational costs (15–20). However, as detailed by Lynd and colleagues (21), the specifics of microbe-biomass interactions and related impacts on cell growth and fermentation are incompletely understood due to methodological problems associated with quantifying microbial cells in the presence of plant biomass.

Here we present a simple and effective method for visualizing and quantifying microbes interacting with insoluble plant substrates. The method involves separately staining cells and substrate with two different fluorescent dyes. To evaluate our method, we used the cellulolytic microbe Clostridium phytofermentans, which produces ethanol as its primary fermentation product and grows using a wide array of simple and complex carbohydrate components of plant biomass as the substrates (20, 22). We demonstrate that the dual-staining procedure may be used to determine characteristics of C. phytofermentans growth and substrate attachment with insoluble cellulosic substrates.

MATERIALS AND METHODS

Bacterial strain and culture conditions.

Clostridium phytofermentans ISDg (ATCC 700394) was cultured using the anaerobic techniques of Hungate in a modified form of GS-2C medium (22) containing the following ingredients (in grams per liter): yeast extract, 6.0; urea, 2.1; KH₂PO₄, 2.9; K₂HPO₄, 1.5; Na₂HPO₄, 6.5; trisodium citrate dihydrate, 3.0; l-cysteine hydrochloride monohydrate, 2.0; resazurin, 1. The pH of the medium was adjusted to 7.0 using KOH. This basal medium was supplemented with 0.3% (wt/vol) of the specific substrate (Whatman no. 1 filter paper, wild-type Sorghum bicolor, or the brown midrib [bmr] reduced-lignin double mutant S. bicolor, bmr-6 bmr-12 double mutant) in the form of a slurry, obtained by wet-ball milling as described previously (23, 24). Cultures were incubated at 30°C under anaerobic conditions (100% N₂) by the method of Hungate et al. (25).

Dual-staining procedure.

Samples were prepared using a dual-staining procedure similar to that described previously (26) as follows. Prior to staining, each culture was diluted 1:1 with basal medium by adding 1 ml of culture to 1 ml of sterile basal medium and vortexed at maximum speed for 15 s. This diluted culture was further diluted 1:1 (50 μl in 50 μl of sterile basal medium) and vortexed. Eight microliters of a 10-μg/ml solution of calcofluor white (fluorescent brightener 28; Sigma-Aldrich) was added to the diluted culture sample and incubated for 8 min in the dark. Subsequently, 0.3 μl of Syto 9 (LIVE/DEAD BacLight bacterial viability kit; Thermo Fisher Scientific) was added to the sample and incubated for 7 min in the dark. A volume of 16 μl of stained sample was placed on a slide and covered with a coverslip (22 by 22 mm). Figure 1 provides a diagrammatic representation of the dual-staining protocol.

FIG 1.

Diagrammatic overview of dual-staining protocol.

Fluorescence microscopy and image processing.

A Nikon Eclipse E600 fluorescence microscope, equipped with a Diagnostic Instruments Spot-RT charge-coupled-device (CCD) camera and a 40× objective, was used to obtain images. For each sample, differential interference contrast and epifluorescence micrographs were obtained. Images were acquired and processed using SPOT Advanced imaging software. Cellulosic growth substrates stained with calcofluor white dye were visualized with light at a wavelength of 385 nm, and cells stained with Syto 9 dye were visualized with light at a wavelength of 480 nm. For three-dimensional (3D) imaging, multifocal plane images with a depth of 1 μm were obtained and combined to generate an image 21 μm deep. Subsequently, the Z-stack was deconvolved using Autoquant 2.1 (Media Cybernetics), and the three-dimensional image was rendered using Imaris 6.0 (Bitplane) software.

Cell quantification.

Micrographs were taken at 10 different locations on each slide. A custom-written script applied image dimensions, and cell counts were performed manually using Image J (http://rsb.info.nih.gov/ij/). Substrate-attached cells and free cells were separately counted. Vegetative and sporulating cells were included in cell counts. Cell numbers are the averages of triplicate determinations.

Determination of fermentation products.

Nongaseous fermentation products were determined by high-performance liquid chromatography (HPLC). Ethanol and acetate concentrations in culture supernatant fluids were measured using a Bio-Rad Aminex HPX-87H column (300 by 7.8 mm) at 30°C with 5 mM H₂SO₄ as the mobile phase at a flow rate of 0.60 ml/min in a Shimadzu LC-20AD HPLC with a RID-10A refractive index detector.

Data analysis.

Data analysis was performed using Prism 6.0 (GraphPad Software) one-way analysis of variance (ANOVA) with a Tukey's posttest to measure variance. Error bars represent standard deviations of the means.

RESULTS

Visualization of C. phytofermentans cells growing on insoluble substrates using the dual-staining procedure.

Cellulolytic bacteria, such as C. phytofermentans, are known to adhere to their cellulosic substrates (23, 27–29), and many species form biofilms on this nutritive surface (26, 30–33). However, differentiating cellulolytic microbial cells from their insoluble growth substrates, enumerating cells, and quantifying cell population growth are major challenges as detailed by Lynd et al. (21). Therefore, the use of light microscopy and fluorescent dyes to distinguish plant-decomposing microbes from their growth substrates is appealing, since it would enable researchers to measure microbial growth on plant biomass in terms of an increase in cell number while also observing microbe-substrate interactions. Representative images showing cell-substrate interactions are presented in Fig. 2. A nucleic acid dye, Syto 9 (LIVE/DEAD BacLight bacterial viability kit) (34), was used to stain bacterial cells, and calcofluor white, a cellulose/chitin dye, stained plant material and filter paper. In order to demonstrate the difficulty of visualizing cells growing on insoluble plant substrates by light microscopy, differential interference contrast (DIC) micrographs were also obtained (Fig. 2A). However, when viewed by fluorescence microscopy at 385 to 400 nm, the filter paper and plant fibers were readily visualized due to the intense blue fluorescence of calcofluor white staining, but bacterial cells were not apparent (Fig. 2B). When samples were illuminated at 480 nm, bacterial cells fluoresced green due to Syto 9 staining of their nucleic acids (Fig. 2C). The two fluorescent images (Fig. 2B and C) were merged in Fig. 2D to demonstrate cell-substrate differentiation with bacterial cells stained green by Syto 9 and the paper and plant fiber stained blue with calcofluor white.

FIG 2.

Differential interference contrast (DIC) and epifluorescence micrographs of C. phytofermentans cultured for 3 days with filter paper, wild-type S. bicolor (WT sorgum), or reduced-lignin S. bicolor double mutant (bmr-6 bmr-12 sorghum). Cultures were prepared for microscopy using the dual-staining procedure. For each sample, one field is shown imaged by DIC microscopy (A), and fluorescence microscopy at 385 to 400 nm showing cellulose fibers stained with calcofluor white (B) and at 480 nm showing cells stained with Syto 9 (C). The epifluorescence images in rows B and C are merged in the row D images. Stained culture preparations were diluted with sterile basal culture medium prior to slide preparation: the filter paper culture was diluted 1:3, and sorghum cultures were diluted 1:1. Bar, 10 µm.

Cell morphology and substrate interactions visualized by 3D imaging.

In order to visualize interactions of C. phytofermentans cells with an insoluble growth substrate, cultures grown on filter paper as the growth substrate were stained using the dual-staining procedure, and three-dimensional images were obtained as illustrated in Fig. 3. Green fluorescing rod-shaped cells of various lengths were observed, some with swellings at one end, presumably due to endospore formation as described previously by Warnick et al. (22). Some C. phytofermentans cells were aligned parallel to cellulose fibers as observed by Zuroff et al. (26), while many cells appeared to be attached perpendicularly to finer threads of the cellulose substrate with the putative endospores located opposite the site of attachment as observed in Clostridium thermocellum (33). In the XY cross-sectional planes, cells were visualized within the substrate. Such 3D images indicate the potential of the dual-staining protocol to image bacterial cells interacting with insoluble plant biomass substrates.

FIG 3.

Two-dimensional rendering of a 3D image of C. phytofermentans cultured for 5 days on medium with filter paper as the growth substrate. The culture was prepared for microscopy using the dual-staining procedure. Multiple focal plane images at a depth of 1 μm were obtained and used to generate 3D images.

Cell growth and ethanol production during fermentation of insoluble substrates by C. phytofermentans.

Measurements of growth of bacteria on their insoluble substrates are problematic for several reasons. Viable counts on plates of agar media (35–37) are complicated by the propensity of cellulose-decomposing cells to adhere to their substrate, frustrating efforts to accurately dilute cultures. Also, viable counts are time-consuming and challenging for fastidious anaerobes. Problems associated with protein determinations as a means to quantify microbial cell mass in studies involving cellulosic substrates have been addressed (38). Our method, which allows rapid, direct enumeration of cellulolytic bacterial cells, is facilitated by the use of two fluorescent dyes that enhance the contrast between microbes and cellulosic substrates.

Growth of C. phytofermentans on filter paper, wild-type S. bicolor, or reduced-lignin S. bicolor bmr-6 bmr-12 double mutant was measured (Fig. 4A). With filter paper as the substrate, cell numbers increased over 5 days, whereas cell numbers on both wild-type and double mutant sorghum reached a maximum at day 3 with 1.1 × 10⁸ cells/ml and 2.7 × 10⁸ cells/ml, respectively (Fig. 4A). However, the ethanol concentration in cultures with all of the three substrates steadily increased for 5 days (Fig. 4B). Consistent with previously reported results (39), cultures with double mutant sorghum as the substrate produced significantly more ethanol than cultures grown on wild-type sorghum. Interestingly, cultures with filter paper as the substrate produced significantly less ethanol than cultures with an equivalent amount of either wild-type or double mutant sorghum (Fig. 4B), indicating the important chemical differences in these substrates. When filter paper served as the substrate, after day 2, ethanol production per cell decreased (Fig. 4C) as cell numbers increased (Fig. 4A). Since both vegetative and sporulating cells were included in cell counts, reduced ethanol production per cell may reflect cell differentiation and the onset of sporulation, which would suggest that sporulating cells are less metabolically active. Unexpectedly, ethanol production per cell in cultures grown with either wild-type or double mutant sorghum increased after 3 days of growth (Fig. 4C) when cell numbers were decreasing (Fig. 4A). Possibly, with an early onset of sporulation, a portion of the microbial cell population was less metabolically active during the first 3 days of growth on sorghum, and upon completion of sporulation, mature spores, which were not counted, were lost from the recorded cell counts. These results also suggest that limiting sporulation may be an effective strategy to increase metabolic activity and ethanol production in C. phytofermentans.

FIG 4.

Cell growth (A) and ethanol production (B and C) by C. phytofermentans, and cell numbers (D), cell attachment to substrate (E) and ethanol concentration (F) after 3 days of growth with filter paper, wild-type S. bicolor, or reduced-lignin S. bicolor bmr-6 bmr-12 double mutant as the growth substrate. The concentration of ethanol in cultures is expressed as millimolar concentration (B and F), and the amount of ethanol produced per cell is expressed as nanomoles per cell (C). Error bars represent standard deviations of the means (n = 3). Statistically significant differences are represented with bars and asterisks; one asterisk for P ≤ 0.05, two asterisks for P ≤ 0.01, and three asterisks for P ≤ 0.001.

Figure 4 also includes a direct comparison of cell number (Fig. 4D), substrate attachment (Fig. 4E), and ethanol production (Fig. 4F) after 3 days of C. phytofermentans growth on the three different biomass substrates. Although cell growth on filter paper exceeded growth on either sorghum substrate (Fig. 4D), there was no statistically significant difference in the percentage of cells attached to filter paper (88%) and to reduced-lignin double mutant sorghum (89%) (Fig. 4E). However, cell attachment to wild-type sorghum was significantly lower than to either filter paper or reduced-lignin double mutant sorghum (Fig. 4E). The increased microbial cell attachment to the double mutant sorghum as compared to wild-type sorghum may reflect greater access to cellulose and other polysaccharide components of biomass in the double mutant sorghum. While cell numbers in cultures with either sorghum substrate were low compared to cultures growing on filter paper (Fig. 4D), the concentration of ethanol produced with wild-type sorghum (17.5 mM) or double mutant sorghum (20.7 mM) was significantly greater than that produced with filter paper as the substrate (13.7 mM) after 3 days of growth (Fig. 4F), possibly due to a higher content of readily fermentable polysaccharides in the sorghum substrates.

These experiments indicated that the C. phytofermentans growth cycle varies greatly depending on the cellulosic substrate. With filter paper as the growth substrate, ethanol production was strongly correlated with cell number (Fig. 5A). Surprisingly, little or no correlation was found between cell number and ethanol production with either wild-type sorghum (Fig. 5B) or reduced-lignin double mutant sorghum (Fig. 5C) as the growth substrate. As shown in Fig. 4, ethanol production occurred later in the growth cycle on both sorghum substrates when C. phytofermentans cell numbers were decreasing, indicating that some cells remained active while the overall cell population decreased, possibly due to sporulation.

FIG 5.

Correlation of ethanol production with cell number during growth. Ethanol production is given as a function of the cell number in cultures with filter paper (A), wild-type S. bicolor (B), and reduced-lignin S. bicolor bmr-6 bmr-12 double mutant (C) as the growth substrate. Error bars represent standard deviations of the means.

DISCUSSION

The dual-staining procedure described here is a simple and efficient method to visualize and enumerate cells in the presence of insoluble plant biomass substrates. This method provides direct total cell counts and may replace or complement complex and indirect methods for measuring microbial growth on cellulosic substrates (38, 40). Also, growth characteristics of cocultures and other complex communities may be assessed using this method if community members are morphologically distinguishable. Importantly, the dual-staining procedure enables studies of interactions of microbial cells with insoluble substrates such as plant biomass feedstocks. It is possible to employ this method to study cell-substrate interactions using 3D visualization with little impact on the sample. Also, coupling the dual-staining procedure with image processing techniques, cell counts may be rapidly performed enabling high-throughput analyses. A possible limitation of this method might be interference due to extracellular polymeric substances (EPS) produced by biofilm formation on substrates, making cell visualization more difficult; however, EPS formation did not pose a problem in the current study.

Our observation of higher ethanol production by C. phytofermentans during fermentation of reduced-lignin double mutant sorghum compared to wild-type sorghum is consistent with the findings of Lee et al. (39). Also, we observed higher microbial cell attachment to the reduced-lignin double mutant sorghum compared to wild-type sorghum, correlating with higher ethanol production. This may be attributed to the fact that the lignin content of plant biomass adversely affects substrate attachment, decomposition, fermentation, and subsequently, ethanol production (21, 39, 41, 42).

Using the dual-staining procedure, we examined C. phytofermentans growth on different insoluble substrates. According to company literature, Whatman no. 1 filter paper is prepared from cotton linters that are treated to achieve an alpha cellulose content of at least 98%, and as such, it is a highly processed, relatively pure form of cellulose. We observed very different growth characteristics on purified cellulose (filter paper) compared to growth on wild-type and mutant sorghum. Both cell numbers and ethanol concentration increased for 5 days when filter paper served as the substrate. However, with wild-type and mutant sorghum as the substrate, cell numbers reached a maximum after 3 days of growth and then decreased while ethanol concentration continued to increase for 5 days. With filter paper as the growth substrate, ethanol production was correlated with cell number, but unexpectedly, little or no correlation was found between cell number and ethanol production with either wild-type or mutant sorghum as the growth substrate. This result indicated that only a portion of the microbial cell population remained active during growth on sorghum, possibly due to differentiation and the onset of sporulation in some cells, resulting in reduced metabolic activity and lower ethanol production.

The dual-staining procedure described here is an effective means for quantifying microbial cells growing on plant biomass substrates. Due to its simplicity, quantitative data may be obtained rapidly and efficiently. In the dual-staining procedure, individual fluorescent dyes are added at separate times. Syto 9, a nucleic acid dye, stains cells and emits a green fluorescence that is observed by fluorescence microscopy. Cellulosic biomass substrates are stained by calcofluor white, which emits a blue fluorescence when exposed to UV light. We demonstrated that visualization and enumeration of microbial cells growing on different cellulosic substrates may be performed by obtaining micrographs and counting cells using Image J. Statistical analyses may be used to evaluate differences in cell growth and substrate attachment, which in turn, may be correlated with fermentation product formation. Finally, use of the dual-staining procedure for three-dimensional visualization of microbes, such as C. phytofermentans, growing on insoluble plant substrates is a useful tool for gaining insight into plant-microbe interactions in a broad range of applications from biofuels to human health.

Funding Statement

This research was supported by Cooperative State Research, Extension, Education Service, U.S. Department of Agriculture, Massachusetts Agricultural Experiment Station project MAS00923, a Sponsored Research Agreement between Qteros and S.B.L. (UMA53405), and National Science Foundation grant NSF BBS8714235 to the University of Massachusetts Amherst Central Microscopy Facility. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.