Impact of pretreated Switchgrass and biomass carbohydrates on Clostridium thermocellum ATCC 27405 cellulosome composition: a quantitative proteomic analysis

Abstract

Background: Economic feasibility and sustainability of lignocellulosic ethanol production requires the development of robust microorganisms that can efficiently degrade and convert plant biomass to ethanol. The anaerobic thermophilic bacterium Clostridium thermocellum is a candidate microorganism as it is capable of hydrolyzing cellulose and fermenting the hydrolysis products to ethanol and other metabolites. C. thermocellum achieves efficient cellulose hydrolysis using multiprotein extracellular enzymatic complexes, termed cellulosomes.

Methodology/principal findings: In this study, we used quantitative proteomics (multidimensional LC-MS/MS and (15)N-metabolic labeling) to measure relative changes in levels of cellulosomal subunit proteins (per CipA scaffoldin basis) when C. thermocellum ATCC 27405 was grown on a variety of carbon sources [dilute-acid pretreated switchgrass, cellobiose, amorphous cellulose, crystalline cellulose (Avicel) and combinations of crystalline cellulose with pectin or xylan or both]. Cellulosome samples isolated from cultures grown on these carbon sources were compared to (15)N labeled cellulosome samples isolated from crystalline cellulose-grown cultures. In total from all samples, proteomic analysis identified 59 dockerin- and 8 cohesin-module containing components, including 16 previously undetected cellulosomal subunits. Many cellulosomal components showed differential protein abundance in the presence of non-cellulose substrates in the growth medium. Cellulosome samples from amorphous cellulose, cellobiose and pretreated switchgrass-grown cultures displayed the most distinct differences in composition as compared to cellulosome samples from crystalline cellulose-grown cultures. While Glycoside Hydrolase Family 9 enzymes showed increased levels in the presence of crystalline cellulose, and pretreated switchgrass, in particular, GH5 enzymes showed increased levels in response to the presence of cellulose in general, amorphous or crystalline.

Conclusions/significance: Overall, the quantitative results suggest a coordinated substrate-specific regulation of cellulosomal subunit composition in C. thermocellum to better suit the organism's needs for growth under different conditions. To date, this study provides the most comprehensive comparison of cellulosomal compositional changes in C. thermocellum in response to different carbon sources. Such studies are vital to engineering a strain that is best suited to grow on specific substrates of interest and provide the building blocks for constructing designer cellulosomes with tailored enzyme composition for industrial ethanol production.

Figure 1. Simplified schematic representation of Clostridium thermocellum cellulosomal architecture.

CipA is the backbone scaffoldin protein containing 9-Type I cohesins and can accommodate up to 9-Type I dockerin bearing catalytic units. CipA also contains a Type II dockerin for cell-surface attachment via anchor proteins and a Cellulose Binding Domain for attachment to the growth substrate. C. thermocellum genome encodes five proteins with Type II cohesins, four with S-layer homology domain (SdbA, OlpB, Orf2p and Cthe0735) for cell-surface anchoring of Type II dockerin bearing CipA and one without the SLH domain (Cthe0736). Also shown is the only subunit containing a Type II dockerin (Cthe1806) – (Figure adapted with permission from Carlos Fontes, CIISA, Portugal).

Figure 2. Clostridium thermocellum growth profile on various substrates, based upon protein levels.

Pellet protein (top panel) and supernatant protein (bottom panel) profiles of Clostridium thermocellum during growth on various biomass carbohydrates, as estimated by BCA and Bradford protein assays, respectively. Data represented is the average of two biological replicate fermentations of C. thermocellum in minimal medium containing 5 g/L total of the following carbon substrates: Cellulose with ¹⁵N labeled nitrogen source, Cellulose with ¹⁴N labeled nitrogen source, Cellulose-Xylan (3∶2 weight ratio), Cellulose-Pectin-Xylan (3∶1∶1), and Cellulose-Pectin (3∶2). Data is not shown for Z-Trim®, pretreated switchgrass and cellobiose fermentations.

Figure 3. Clostridium thermocellum cellulosomal protein profile during growth on various substrates, analyzed by protein gel electrophoresis.

SDS-PAGE (4–20% Tris-HEPES-Glycine gel, coommassie stain) separation of Clostridium thermocellum cellulosomal fractions isolated using the affinity digestion method from cell-free broth of duplicate fermentations during growth on pretreated switchgrass and other biomass carbohydrates. BR1, 2 = Biological Replicate 1, 2, respectively.

Figure 4. Weighted-Normalized Spectral Abundance Factor (Weighted-NSAF) of cellulosomal components in mass spectrometry analysis.

Weighted-NSAF data for ¹⁴N isotopologs across seven different samples is arranged in descending order of values for cellulose sample. For each cellulosomal sample, the average NSAF value of the various components was divided by the value for the scaffoldin protein CipA to obtain weighted-NSAF values. Heat plot representation shows weighted-NSAF distribution (Red, highest; Green, lowest) of 54 cellulosomal subunits. The proteins with top 10, and the next 10, highest weighted-NSAF values in each sample are highlighted in Yellow and Orange, respectively. Locus entries of newly detected cellulosome components are highlighted in Blue. Substrate legend: C = Cellulose, CX = Cellulose-Xylan, CP = Cellulose-Pectin, CPX = Cellulose-Pectin-Xylan, SWG = Pretreated Switchgrass, Cb = Cellobiose, ZT = Z-Trim®.

Figure 5. Quantitative changes in Clostridium thermocellum cellulosome composition in response to carbon substrate – Part I

(see Figure 6 caption).

Figure 6. Quantitative changes in Clostridium thermocellum cellulosome composition in response to carbon substrate – Part II.

Quantitative data shown is combined from two biological and two technical replicates and expressed as Log₂Ratio (LowerCI, UpperCI) of cellulosomal component X during growth on substrate Y over that in ¹⁵N-labeled Cellulose cellulosome sample. Substrate key: C = Crystalline Cellulose (with ¹⁴N or ¹⁵N labeled nitrogen source), C-X = Cellulose-Xylan (5 g/L total in 3∶2 weight ratio), C-P = Cellulose-Pectin (3∶2), C-P-X = Cellulose-Pectin-Xylan (3∶1∶1), SWG = Pretreated Switchgrass (50% glucan, 8% xylan, 24% lignin), Cb = Cellobiose, Z-T = Z-Trim® (60% amorphous cellulose, 16% hemicellulose). Data was normalized to the scaffoldin CipA protein. Locus entries highlighted in Blue have not been observed experimentally prior to this study. Quantitation data highlighted in Yellow (increased expression in Substrate Y) and Orange (decreased expression in Substrate Y) satisfy the cut-off criteria. Criteria for differential expression was based on control comparison of ¹⁴N- and ¹⁵N-labeled cellulose cellulosome samples and were, (1) Log₂Ratio should be >+0.4 or <−0.4 and (2) Upper/Lower Confidence Intervals should exclude Log₂Ratio = 0. Structural, catalytic and/or binding module information was obtained from the following sources: http://pfam.sanger.ac.uk/; http://www.cazy.org/; http://genome.ornl.gov/microbial/cthe/. Domain key: GH = Glycoside Hydrolase, CE = Carbohydrate Esterase, PL = Polysaccharide Lyase, CBM = Carbohydrate Binding Module. Legend key: MW = Molecular Weight, Log₂Ratio (LowerCI, UpperCI) - protein was quantified in both biological replicates (BR), with overlapping confidence intervals; Blank - protein not identified or quantified; d - protein quantified in only one BR, but result indicates down-regulation (UpperCI<0, Log₂ratio<−0.4); D - protein quantified in both BR, but confidence intervals don't overlap. Both confidence intervals indicate down-regulation (UpperCI<0, Log₂ratio<−0.4); i - identified in one BR, but no quantification result; I - identified in both BR, but no quantification result; n - protein quantified in only one BR, with result indicating no change in expression (−0.4<Log₂ratio<+0.4); N - protein quantified in both BR, but confidence intervals don't overlap. Both confidence intervals indicate no change in expression (−0.4<Log₂ratio<+0.4); u - protein quantified in only one BR, but result indicates up-regulation (LowerCI>0, Log₂ratio>+0.4); U - protein quantified in both BR, but confidence intervals don't overlap. Both confidence intervals indicate up-regulation (LowerCI>0, Log₂ratio>+0.4); A indicates that the protein was identified in both ¹⁴N and ¹⁵N forms in both BR; A single underscore indicates that the protein was identified in both ¹⁴N and ¹⁵N forms in only one BR; No underscore indicates that we did not identify both ¹⁴N and ¹⁵N forms in either BR.