Research Descriptions

The following is a brief description of some of the research projects in my laboratories which involve two groups located in the Department of Chemistry and Biochemistry at the University of Notre Dame, and at Omicron Biochemicals, Inc. in South Bend. Both laboratories collaborate on research projects through the sharing of complementary expertise, equipment and research resources. In the present global research environment, university-industry linkages are now commonly viewed as critical factors in maintaining and enhancing national research competitiveness. The Notre Dame-Omicron synergy has been extremely effective in this role over the past fifteen years.

One of the many positive consequences of this cooperation is the broader training students receive in my laboratory. At Notre Dame, my students pursue basic research problems, but they also have the opportunity to observe the translation of basic research into practical applications in the commercial workplace. Exposure to a wider range of experimental techniques is also facilitated, since the Notre Dame and Omicron groups use different, yet complementary, scientific methods to reach their research goals. We also engage in off-campus collaborations on a routine basis in order to secure research expertise not present in the group and/or to pursue joint research on problems of mutual interest.

The current NIH project period is May 2001-April 2005. The following is an abstract and description of the specific aims of this project, as stated in the grant proposal.

Oligosaccharides are of central importance in the development and maintenance of biological systems. They are found free in solution and as constituents of proteins (glycoproteins) and lipids (glycolipids) on the cell surface, and have been implicated in the mechanisms of action of diverse cellular processes such as inflammation, cell-cell recognition and adhesion, protein turnover, and bacterial infection. Oligosaccharides, or molecular mimics thereof, are becoming more frequent targets in drug design for the treatment of human illness and disease. The cellular functions of oligosaccharides are intimately related to their three-dimensional shapes or conformations; knowledge of these 3D structures, both free in solution and in the bound state, is essential to deciphering the underlying factors that control recognition between oligosaccharides and specific cellular receptors. This proposal aims to develop new probes of oligosaccharide conformation based on nuclear magne tic resonance (NMR) spin-couplings involving ¹³C. Major objectives are to refine recently reported dependencies of J_CH and J_CC scalar couplings on molecular structure using experimental and computational methods, and to apply these couplings to investigate O-glycoside linkage conformation in a range of biologically-important ¹³C-labeled di-, tri- and tetrasaccharides. Residual ¹³C-¹H and ¹³C-¹³C dipolar couplings, and residual ²H quadrupolar couplings in orienting media will be investigated as additional conformational constraints in solution. These coupling data will be complemented by ¹H-¹H NOE and ¹³C and ²H relaxation data to assign preferred conformation and identify internal motions in these molecules. Motional properties will also be investigated computationally using molecular dynamics methods. Having established the adva ntages and limitations of these tools in smaller, more defined systems, investigations will be extended to larger, biologically-important systems which include the complex high-mannose N-glycan found on CD2, engineered N-glycans on IgG, and DNA aptamer-oligosaccharide complexes.

Proposal Specific Aims

Conformational analysis of oligosaccharides in solution by NMR spectroscopy has been plagued by at least two well-recognized problems: a lack of sufficient experimental parameters from which to deduce a 3D structure, and averaging effects on parameters due to conformational flexibility. In the first case, present studies of oligosaccharide conformation rely heavily on ¹H-¹H NOEs for information on the proximity of carbon-bound protons, from which 3D structure is deduced. This approach works well when there are sufficient measureable NOEs that are sensitive to molecular shape, as is the case for proteins. In oligosaccharides, linkage conformation largely determines global shape. Inter-residue NOEs report on the torsion angles, phi and psi, which define conformation about the O-glycosidic linkages (in (1,6)-linkages, a third torsion angle is involved). Unfortunately, the number of ¹H-¹H NOEs observed in most oligosaccharides is small, and those that are observed are often weak due to unfavorable rotational correlation times in water at room temperature. While the latter factor can sometimes be remedied by changing solution viscosity and/or temperature or by collecting data in the rotating frame, it is nevertheless true that conformational assignments based on NOE are not unequivocal. Furthermore, the effect of conformational averaging on NOE is complex; due to the r^-6 dependence of NOE on ¹H-¹H internuclear distance, NOE averaging will not be linear, and this nonlinearity can lead to faulty interpretations of NOE data in oligosaccharides. It should be noted that many conformational assignments of oligosaccharides are made currently by combining NMR data (mainly ¹H-¹H NOEs) and computational tools in order to assign structure. In most cases, very limited experimental data are fit to a computer-derived conformational model, itself subject to uncertainty due to the limitations and assumptions of the calculations. Additional experimental structural constraints are needed in order to make further progress in this area. The major objective of this project is to develop new NMR probes to address this problem. These probes will be developed in smaller systems that are conducive to experimental and theoretical treatments. Once the fundamental structural dependencies of these probes are established in these smaller systems, we will extend their use to larger, biologically-important oligosaccharides. The specific aims of the project are:

(2) To develop chemi-enzymic routes for the synthesis of isotopically-labeled disaccharides using glycosidases (transglycosylation), and to explore the use of this methodology, in conjunction with other enzymic and chemical glycosylation methods, for the synthesis of larger, isotopically-labeled oligosaccharides.

(3) To establish, using experimental and theoretical tools, a complete understanding of the structural dependencies of trans-O-glycoside scalar couplings (²J_COC, (³J_COCH, (³J_COCC) in oligosaccharides for use as structural constraints in conformational analysis. All major types of biologically-relevant linkages will be investigated, including those involving aminosugars and sialic acid.

(4) To establish, using experimental and theoretical tools, a complete understanding of scalar couplings involving hydroxyl protons in carbohydrates for use as probes of hydrogen bonding in solution. Similarly, J_CH, J_CC, J_NH and/or J_NC will be investigated as probes of N-acetyl group conformation in N-acetylated sugars, and exocyclic hydroxymethyl group conformation in aldohexopyranosyl rings, especially those involved in (1,6)glycoside linkages.

(5) To establish, using experimental and theoretical tools, a complete understanding of ²H nuclear quadrupolar coupling constants in saccharides, for use as independent structural probes and for use in the interpretation of ²H relaxation in oligosaccharides.

(6) To establish, using experimental and theoretical tools, a complete understanding of ¹³C chemical shift anisotropy in saccharides, for use as independent probes of saccharide structure and in the analysis of ¹³C relaxation data.

(7) To develop new experimental approaches to assessing molecular motion in oligosaccharides, using ¹³C and ²H nuclear spin relaxation and new motional models that may distinguish motions on differernt timescales. To achieve this objective, a generalized peptide-based tethering system will be developed to control overall molecular reorientation rates and thereby render the deconvolution of relaxation times into overall motion and different internal (anisotropic) motional regimes.

(8) To develop methodology for the integration of the above-noted new experimental tools with molecular dynamics computations in a effort to completely characterize the structures and dynamics of a wide range of oligosaccharides.

(9) To expand the database of crystal structures of underivatized oligosaccharides for use in statistical analyses of bond lengths, angles and torsions, and for use as reference structures in the development of quantum mechanical methods of structure prediction.

(10) To identify and investigate several important biologically relevant problems involving oligosaccharide conformation to which can be applied the above new structural methodologies.

SHORT DESCRIPTIONS OF SOME PROBLEMS AND RESULTS

Under Revision: February 2001

We have been applying QM computational methods since the mid-1980s to study the structures and conformations of carbohydrates, with most of this effort focused on biologically-relevant compounds containing furanosyl rings. We have applied mainly HF and MP methods with various basis sets, but more recent calculations are being conducted using density functional theory (DFT); the latter method has proven very useful to study more complex carbohydrate systems, especially when the aim is to calculate NMR J-couplings (see below). We have recently conducted a study to compare DFT results with those derived from the SCF HF/MP methods, thereby calibrating the former and setting the stage for more extensive applications of the DFT method.

We have an ongoing collaboration with Dr. Ian Carmichael of the Notre Dame Radiation Laboratory which promotes the use of molecular orbital data (from Gaussian) to compute NMR spin-spin coupling constants (J-couplings) with a high degree of accuracy in carbohydrates. We have exploited this unique opportunity to study ¹³C-¹H and ¹³C-¹³C spin-coupling constants (J_CH and J_CC) in saccharides. The furanose systems examined both computationally and experimentally include the beta-D-ribofuranosyl ring found in RNA and the 2-deoxy-beta-D-erythro-pentofuranosyl ring found in DNA. We have examined the effect of ring structure and conformation (and most recently hydroxymethyl group conformation)on the magnitudes of ¹J_CH, ²J_CH and ³J_CH values in these model structures, with the intent to learn which of the available J-couplings are sensitive to ring and hydroxymethyl group conformation. Armed with this information, we aim to apply these parameters in structural studies of larger RNAs and DNAs (seebelow).

Similar computational tools have been applied to better understand ¹J_CC values in saccharides; for example, we have learned that ¹J_CC displays a Karplus-like dependency in HO-C-C-OH systems common to saccharides. We have completed a detailed computational and experimental NMR study of ¹³C-¹³C spin-couplings of importance to oligosaccharide conformational analysis, namely, ²J_COC and ³J_COCC values across O-glycoside linkages. This recent work is presently being extended to other biologically and pharmaceutically relevant glycosidic linkages in the NIH-sponsored project discussed briefly above.

A collaboration with Professor John Brady at Cornell University has been initiated recently using molecular dynamics simulations of various saccharide-containing systems with the inclusion of solvent water. These studies aim to address the effect of saccharide structure (e.g., configuration) on ring conformational dynamics, and to learn how water explicitly interacts with these sugars (hydrogen-bonding networks).

Since 1988, we have been systematically preparing ribonucleosides and 2'-deoxyribonucleosides selectively labeled with ¹³C and ²H at various sites within the furanose ring. Selective ¹³C-labeling has been confined mainly to C1' and C2' in these compounds. We are presently extending this work to C3'-, C4'- and C5'-labeling. These compounds will be used initially to obtain precise measurements of the magnitudes and signs of ¹³C-¹H spin-coupling constants involving the labeled carbons. Various J_CC values will also be measured. These coupling data have not yet been reported for these compounds, and thus their dependencies on molecular structure remain largely unclear. Since we have built, from past model studies, a relatively extensive knowledge base on which to base structural interpretations of J_CH and J_CC values in furanose rings, these new spin-coupling constants will provide a te st of structure/J-magnitude correlations and will potentially lead to firmer conclusions about furanose ring geometry/dynamics in solution.

In 1996 we proposed a new projection resultant (PR) method to interpret ²J_COC values across O-glycoside linkages of oligosaccharides. This rule has been tested experimentally with the use of triply-¹³C-labeled compounds and ¹³C-¹³C COSY-45 methods to verify sign predictions made by this method. In 1998, we reported a comprehensive investigation of ³J_COCC values across O-glycoside linkages. Current theoretical work will extend these calculations to a wide range of biologically-important linkages. At the same time, experimental work is being extended to measurements of ²J_COC and ³J_COCC values across the O-glycoside linkages in a wide range of ¹³C-labeled biologically-important oligosaccharides. These labeled oligosaccharides are being prepared by both chemical and enzymic methods.

This research area has received less attention recently, yielding to higher priority studies of NMR scalar couplings and other fundamental NMR phenomema. Several selectively ¹³C-labeled DNA oligomers using solid-phase synthesis (Omicron maintains a solid-phase DNA synthesizer) were prepared in 1993 to demonstrate synthetic feasibility, and we have developed the ability to prepare a wide range of specific-sequence ¹³C- and/or ²H-labeled oligomers. Recent efforts have focused on obtaining a quantitative understanding of J-coupling behavior, and other NMR parameters, in simple furanoses and nucleosides, since we reasoned that this fundamental knowledge is a prerequisite to J-coupling measurement and interpretation in oligomers.

We plan to return to labeled oligomer studies, having defined the behavior of these couplings in simpler, more defined model systems. Further studies of CCAAT-binding sites in DNA may be pursued, although other biologically-important DNA or RNA recognition sequences are under consideration. A combination of site-specific and uniform labeling strategies will be applied to assist in the analysis of spectral data. This work will be conducted in collaboration with the Omicron group and with Dr. Jaroslav Zajicek (ND NMR facility) who will assist in 2D/3D data collection.

Plant systems are being investigated as vehicles to manufacture uniformly-labeled biomolecules. Our present focus is on potato. We have constructed a plexiglass growth chamber (3'x 3'x 4') for the growth of potato plants in a ¹³CO₂ atmosphere. The chamber is engineered to monitor and control gas pressure, temperature, humidity and CO₂ concentration, and is illuminated with high-intensity light for optimal growth. A number of trial growths have been conducted to optimize environmental conditions, and studies with labeled ¹³CO₂ are anticipated. This project is conducted in collaboration with the Omicron group.

We are interested in several enzymes for use as reagents for saccharide synthesis (triokinase) and in studies of catalytic mechanisms (mutarotase). We plan to isolate triokinase from porcine kidney and to examine its use in the synthesis of labeled carbohydates.

We have studied the structure and catalytic mechanism of mutarotase from E. coli. The role of tryptophan in the binding site was examined, and the effect of fluorine substitution in the carbohydrate substrate on binding and catalysis has been explored. Since we have been interested historically in factors affecting rates of spontaneous carbohydrate anomerization in solution, it is useful to extend this previous work to the enzyme-catalyzed case. Since mutarotase has been cloned and overexpressed, mutant forms are accessible to support mechanistic studies.