AmericanChemicalSociety.com

For knowledge-based engineering of the enzymes to increase efficiency of biofuels from biomass, it is very important to understand the kinetics and mechanisms of the protein / cellulose interactions. Our focus in this project is on the kinetics of interactions between cellulose and the cellulose-binding domains (CBDs) of the cellulase enzymes. We are studying CBDs from Clostridium cellulovorrans cellulase 5A (Cel5A) and Cellulomonas fimi cellulase 9B (Cel9B). These CBDs bind noncrystalline cellulose and cello-oligosaccharides (smaller fragments of cellulose), and play important role to keep the enzymes in close proximity from the substrate. As described in the original proposal, we are pursuing two Specific Aims: 1) analysis of the kinetics of protein translocation between cello-oligosaccharide molecules, and 2) determination of the kinetic rate constants for protein / cello-oligosaccharide dissociation and association. NMR spectroscopy is our primary means to pursue these Specific Aims.

In this year, we have successfully carried out studies for Specific Aim 2. A postdoctoral fellow, Krishnarao Doddapuneni, and a graduate student, Debashish Sahu, made the DNA plasmids for E. coli to express the Cel5A and Cel9B CBDs (their genes were chemically synthesized), and prepared ¹⁵N-labeled and ¹³C/¹⁵N-labeled proteins by using cellulose-affinity chromatography, anion-exchange chromatography, and gel-filtration. Using an 800-MHz NMR spectrometer, a graduate student, Levani Zandarashvili, analyzed the kinetics of the association and dissociation processes involving the Cel9B CBD and cello-oligosaccharides (cellopentaose, 5 sugar rings; cellohexaose, 6 sugar rings). The NMR line-shape analysis based on the Ruben-Fiat approximation permits determination of the dissociation rate constant k_off and the association rate constant k_on for relatively fast processes that are difficult to analyze by other techniques. Using unlabeled cello-oligosaccharides (purchased from Associates of Cape Cod, Inc) and ¹⁵N-labeled Cel9B CDB, we analyzed ¹H- and ¹⁵N-lineshapes as a function of population of the bound state by line-shape fitting. Under our current experimental conditions (50 mM NaCl and pH 7 at 30ûC), the rate constants k_on and k_off were determined from the line-shape data as follows: k_on = (2.1 ± 0.5) x 10⁸ M^-1 s^-1 and k_off = (4.2 ± 0.7) x 10³ s^-1 for cellopentaose; k_on = (2.0 ± 0.6) x 10⁸ M^-1 s^-1 and k_off = (2.0 ± 0.2) x 10³ s^-1 for cellohexaose. The dissociation constants (K_d) were calculated to be (2.0 ± 0.3) x 10^-5 M and (1.0 ± 0.3) x 10^-5 M for cellopentaose and cellohexaose, respectively. Relatively fast dissociation is likely to be crucial for rapid turnover in the enzymatic process. Interestingly, the different values of K_d (= k_off / k_on) for cellopentaose and cellohexaose are mainly due to their different dissociation rate constants k_off. Our speculation is that the cellohexaose provides additional interaction that makes the lifetime of the complex longer (the half-lifetime is given by ln2 / k_off). The association rate constants k_on indicate that the association process is limited by diffusion. While rate constants k_on of an order of 10⁸ M^-1 s^-1 have been observed for some other protein-ligand interactions, the fast association is usually enhanced by long-range electrostatic interactions. However, considering that there are no electrically charged groups in the cellulose molecules, the mechanism for the fast association of the CBD to cellulose is an interesting subject for further investigations. 

For the Cel5A CBD, we have finished data collection of three-dimensional ¹H/¹³C/¹⁵N triple resonance spectra HNCO, HN(CA)CO, HNCA, HN(CO)CA, CBCA(CO)NH, HNCACB, C(CO)NH, HCCH-TOCSY, HCCH-COSY, HA(CACO)NH, and HBHA(CBCACO)NH. Judging from the data quality, we expect that we can finish ¹H/¹³C/¹⁵N resonance assignment for this 171-residue protein in a few weeks. Although the resonance assignment has not been finished, we have started experiments for the kinetic investigations of the interactions between the Cel5A CBD and cello-oligosaccharides. Our preliminary NMR data suggest that under current experimental conditions, the dissociation of the Cel5 CBD / cellulose complex is too fast to get quantitative information by NMR line-shape analysis. We plan to reduce temperature for the experiment so that the rate constant k_off becomes within the applicability range of our current approach.

To successfully perform studies for Specific Aim 1, we are optimizing the experimental conditions. In this experiments, the translocation processes of the CBD proteins bound to different cello-oligosaccharides at relatively high concentrations should be analyzed. This turned out to be experimentally challenging because of low solubility for cello-oligosaccharides. In particular, the maximum concentration of cellohexaose is only ~2 mM, which makes the study difficult. We are trying to solve this problem by using derivatives of cello-oligosaccharides with higher solubility and comparable affinities, because there are fortunately many ways to modify a reducing end of the cello-oligosaccharides. We expect that the use of such derivatives will allow us to carry out the experiments for Specific Aim 1 more easily.

In the context of this project, we have also developed the NMR pulse sequences to analyze amplitude and timescale of the motions of the side-chain NH₃⁺ groups, because dynamics of lysine side-chains seems relevant to the kinetic properties of the CBD proteins. A graduate student, Alexandre Esadze, and our NMR facility manager, Tianzhi Wang, were involved in this work. We have submitted a manuscript on this work.