Development and validation of a stochastic molecular model of cellulose hydrolysis by action of multiple cellulase enzymes

Deepak Kumar1,2, Ganti S. Murthy1*

1 Biological and Ecological Engineering, Oregon State University, Corvallis, OR, USA.

2Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, USA.

Deepak Kumar:

Ganti S. Murthy:

*Corresponding Author; , phone: 541-737-6291

Table S1: Values of parameters used for EG I, CBH I and CBH II action

The values of various parameters used in model simulations of cellulose hydrolysis by action of EG I, CBH I and CBH II are given in table below.

Parameters / EG I / CBH I / CBH II
Prob_EF_bind1 (%) / 75 / 75 / 75
Nnb2 / 0.2 / 0.02 / 0.015
NnP3 / 0.3 / 0.2 / 0.15
Ninhib _G24 / 0.3 / 0.15 / 0.05
Ninhib _G 5 / 0.015 / 0.004 / 0.0015
Prob_act_Crystal (%)6 / 25 / NA / NA
Prob_bind_longchain7 (%) / NA / 95 / 95
Prob_desorb_crystal8 (%) / NA / -Number of bonds broken in run <25: 5
-Number of bonds broken in run 25-50: 15
-Number of bonds broken in run >75: 35 / -Number of bonds broken in run <25: 5
-Number of bonds broken in run 25-50: 15
-Number of bonds broken in run >75: 35
Prob_desorb_amorphous9 (%) / NA / -Number of bonds broken in run <25: 10
-Number of bonds broken in run 25-50: 30
-Number of bonds broken in run >75: 70 / -Number of bonds broken in run <25: 10
-Number of bonds broken in run 25-50: 30
-Number of bonds broken in run >75: 70

1 Probability of binding on bonds that are on elementary fibril surface but not on microfibril surface

2 Increment in the counter during hydrolysis simulation if bindings conditions were not met

3 Increment in the counter during hydrolysis simulation when binding occurred but hydrolysis conditions were not met

4 Increment in the counter during hydrolysis simulation when randomly chosen bond is a cellobiose molecule

5 Increment in the counter during hydrolysis simulation when randomly chosen bond is a glucose molecule

6 Probability of endogluancase to act in crystalline region (endoglucanases is active mainly in amorphous regions only)

7 Probability of CBH I and II to bind on long chain (cellobiohydrolase prefer short chains)

8 Probability of CBH I and II to break from chain in crystalline region

9 Probability of CBH I and II to break from chain in amorphous region

Fig. S1: Comparison of model simulations (previous and new version) with experimental data from literature

Fig. S1aComparison of model simulations (current and previous version) and experimental data of cellobiose production during hydrolysis of Avicel (50 g/L) at various enzyme loadings. The data points are from literature studies (Bezerra and Dias 2004; Bezerra et al. 2011), solid lines are from the new model predictions, and dotted lines are for predictions from previous version of model.

Fig. S1bComparison of model simulations (current and previous version) and experimental data of cellobiose production during hydrolysis of Avicel (25 g/L) at various enzyme loadings. The data points are from literature studies (Bezerra and Dias 2004; Bezerra et al. 2011), solid lines are from the new model predictions, and dotted lines are for predictions from previous version of model.

Fig. S2: Comparison of model simulations (old version and current model) with experimental data during hydrolysis of filter paper by CBH I

Fig. S2Comparison of sugar profile among model simulations (both from old version and current model) and experimental data during hydrolysis of filter paper (25 g/L) at CBH I loading of 10 mg/g cellulose. Solid lines are for simulation results from new model and dotted lines are from simulation results of old model.

The previous version of model did not account for the glucose formation during cellulose hydrolysis by cellobiohydrolases, hence, the glucose concentrations were zero throughout hydrolysis, and data did not fit with the experimental data. Predictions of cellobiose concentrations were similar from both models, however, final cellobiose concentration was predicted less by old model. This could be because of assumption of that oligomers with DP>6 are insoluble, which would limit the hydrolysis of these oligomers. In new model, these oligomers were assumed partially soluble, and could be hydrolyzed by the catalytic domain of enzyme, even if enzyme is not adsorbed on the surface.

Fig. S3: Endoglucanases action on substrates with different crystallinity

The figure A6.1 illustrates the hydrolysis profile of filter paper (semi-crystalline cellulose, CrI 0.4-0.5) and Cotton cellulose (highly crystalline cellulose) by action of endoglucanases. After 48 h of hydrolysis, conversion of cotton cellulose was found 57.2 % lower than that of filter paper.

Fig.S3Model predictions: Action of endoglucanases (EG I) enzyme on filter paper and cotton cellulose

Fig. S4: Glucose production profile during action of endoglucanases on filter paper

Fig. S4Model predictions of sugar concentrations during hydrolysis of Filter paper by endoglucanases(filter paper, 25 g/L; EGI, 10mg/g glucans) [DP6+ on primary axis to the left and rest on secondary axis to the right].