by Dr A. K. Korir and Prof. C. Larive
The isolation and purification of sufficient amounts of material for characterisation by NMR spectroscopy is a bottleneck in many analyses. Developments in NMR microcoil technology allow acquisition of spectra with much smaller sample amounts. This article describes the application of online cITP-NMR for the separation and analysis of trace pharmaceutical degradants and heparin-derived oligosaccharides.
Along with mass spectrometry, NMR spectroscopy is one of the most powerful techniques for structure elucidation. However, its relatively low sensitivity compared with other techniques hinders its use for the study of compounds available in limited quantities. Various strategies have been employed to enhance NMR limits of detection including use of higher field superconducting magnets, novel polarisation transfer techniques, [1] cryogenically cooled probes [2] and miniaturised radiofrequency (rf) coils with reduced diameters{3]. Compared with the other approaches aimed at improving NMR sensitivity, the use of reduced-diameter coil probes (microcoil probe technology) is a relatively simple and inexpensive method. In this article, we focus on the practical aspects of microcoil probe technology and its application for trace impurity analysis and in the structure elucidation of mass-limited natural products such as the glycosaminoglycans heparin and heparan sulphate.
Microcoil NMR technology
Detailed reviews of the theory, design and applications of rf microcoils for high-resolution NMR spectroscopy have been previously reported [4-6]. Microcoils with solenoidal geometry intrinsically have 2-3 times higher mass sensitivity than Helmholtz coils. In the design of solenoidal microcoils, a high signal-to-noise ratio is obtained by optimising the coil geometry with respect to wire diameter, number of turns and the interturn spacing. Also, since coil sensitivity is inversely proportional to the coil diameter, additional gains can be achieved with decreased coil diameter. In general, the performance of a microcoil probe depends on a number of factors including the quality of the rf coil, overall circuitry and properties of the sample.
Small diameter rf microcoils with nanolitre to microlitre observe volumes (Vobs) are especially important in the analysis of biological samples that are either volume- or mass-limited. Compared with conventional 5 mm NMR probes (Vobs ~ 220 µL), rf microcoil probes (Vobs ~ 5 nL to 1 µL) have demonstrated up to 40-fold enhancement in mass sensitivity [3]. Scaling NMR spectroscopy to smaller detection volumes for improved mass sensitivity provides an additional advantage for coupling to various microscale separation methods such as capillary liquid chromatography, capillary electrophoresis, and capillary isotachophoresis (cITP).
In our laboratory, we construct rf microcoils with Vobs of 25 nL. This volume was selected to closely match the volume of the focused analyte band formed in the on-line cITP separations described in the next section. Figure 1 shows a typical solenoidal rf microcoil having a length of roughly 1 mm and a diameter of 430 µm. This coil is wound around a piece of polyimide tubing (430 µm O.D., 370 µm I.D.) through which the separation capillary (360 µm O.D.) is inserted. One limitation of the use of microcoil probes is their relatively poorer concentration sensitivity compared with conventional probes. Furthermore, the efficient delivery of nanolitre volumes of samples to the microcoil detector can be difficult. Losses from sample handling and dilution during transfer of such small amounts of material are some of the challenges that have to be addressed. Many of these limitations can be overcome by coupling cITP, a sample-stacking technique, with online microcoil NMR detection.
Online cITP-NMR spectroscopy
A difference between cITP and the more familiar method of capillary electrophoresis is that cITP uses capillaries specially coated to reduce electroosmotic flow and employs a discontinuous buffer system that consists of a leading electrolyte (LE) of higher effective electrophoretic mobility (µeff) and a trailing electrolyte (TE) of lower µeff than that of the analyte[7]. The sample is injected between the LE and the TE and a high electric field (10-20 kV) applied across the capillary. The sample ions are separated according to their µeff, and at equilibrium stable boundaries are formed between zones of ions with different µeff. The focused bands remain in contact with each other and travel at constant velocity. Using cITP, a sample can be injected in a relatively large volume (microlitres), separated, and focused in a volume that closely matches the microcoil Vobs. In this way, online cITP-NMR compensates for the poor concentration sensitivity of the rf microcoils, and allows simultaneous separation and acquisition of NMR spectra of the separated components. This technique can be used to separate complex mixtures of related compounds provided they have different values of µeff [8]. Figure 2 shows a schematic of the instrumental cITP-NMR setup.
Trace impurity characterisation
NMR spectroscopy is widely used for the structure elucidation of pharmaceutical impurities. The FDA and ICH guidelines require structural identification of impurities and degradation products present in a formulated drug above the identification threshold of 0.1%. This analysis is complicated because pharmaceutical impurities typically occur at low amounts in the presence of high levels of the parent drug and matrix components. One approach for structure elucidation of impurity samples is to isolate and collect quantities sufficient for direct analysis, however, this increases the total analysis time and introduces the risk of further degradation of unstable impurities. Alternatively, online LC-NMR analysis may be used, although the poor sensitivity of NMR detection often means that obtaining sufficient amounts of the impurity for structure elucidation requires overloading of the column thus sacrificing the quality of the separation.
We have recently demonstrated the capability of cITP-NMR to separate and detect acetaminophen thermal degradation products[9].Pure acetaminophen is stable under dry conditions at temperatures up to 45 °C but upon exposure to excess heat and humidity, it thermally degrades to 4-aminophenol. Due to the nephrotoxic and teratogenic effects of 4-aminophenol, the United States, British, and European pharmacopeia limit for the impurity is 50 ppm (0.05% w/w) in the drug substance.
The 1H NMR spectrum of a degradation product is often similar to that of the parent compound, which is typically present in a large excess. For example, the main difference in the 1H NMR spectra of acetaminophen and 4-aminophenol is the presence of an acetyl resonance peak at 2.13 ppm in the acetaminophen spectrum. Because the chemical shifts of the aromatic resonances of these compounds are very similar, it is impossible to detect the resonances of the 4-aminophenol impurity in a static 1H- NMR of acetaminophen. The presence of matrix components can further complicate the interpretation of the NMR spectrum of the degradation sample. These problems were solved using cITP-NMR to selectively concentrate and detect 4-aminophenol in acetaminophen samples.
Differences in the pH chemistry of the parent compound and degradation product can be exploited to gain selectivity in the cITP-NMR analysis. Because the amino group of 4-aminophenol has a pKa of 5.28, selecting a pH less than 4 for the separation buffers creates conditions in which 4-aminophenol is positively charged whereas acetaminophen, present in a large excess, remains neutral. In a cITP-NMR experiment in which 4-aminophenol was spiked into an acetaminophen tablet solution at the 0.1% level, separation and detection of 0.213 µg of the impurity was accomplished with no interference from the parent drug or formulation excipients. Because the neutral acetaminophen was not focused by the electric field only 4-aminophenol was focused and detected, demonstrating the potential of cITP-NMR for pharmaceutical trace impurity analysis. Figure 3 shows a portion of the on-line cITP-NMR spectra of an acetaminophen forced degradation sample showing resonances of the 4-aminophenol. Unexpectedly, resonances of other unknown components were also detected providing spectral evidence of additional degradation products. The chemical shifts and splitting patterns of the resonances of the unknown components suggested they were dimers formed during the forced degradation. This assertion was subsequently confirmed by LC-MS/MS [9].
Separation and characterisation of heparin oligosaccharides by cITP-NMR
Heparin and heparan sulphate (HS) are structurally related compounds of the glycosaminoglycan (GAG) family that exhibit a wide variety of biological activities [10]. Heparin, discovered as an inhibitor of blood coagulation in 1916, is the most commonly used anticoagulant drug today. Heparin and HS are acidic polysaccharides that are ubiquitous in the extracellular matrix and at cell surfaces. Despite their biological significance, the microheterogeneity and structural complexity of GAGs has prevented the elucidation of their complete structure. Thus far, investigators have determined the primary sequence of very few heparin and HS oligosaccharide subunits important for protein binding. The lack of sufficiently sensitive and specific analytical techniques has been a major impediment to the elucidation of GAG structure and to attempts to relate structure and biological function. The ability of modern NMR experiments to reveal elements of structure makes it an essential tool for the chemical analysis of heparin oligosaccharides. However a major limitation of NMR spectroscopy is its relatively poor limits of detection compared with other analytical techniques. Therefore, near milligram quantities of purified material are typically required for NMR analysis, amounts that can be difficult to obtain, especially for rare GAG oligosaccharide subunits.
We have developed a unique strategy that employs microcoil NMR for the analysis of microgram quantities of heparin oligosaccharides [11,12]. Figure 4 is the 1H NMR spectrum obtained with online cITP focusing of approximately 2.5 µg of an oligosaccharide isolated from porcine intestinal heparin. From this 1H NMR spectrum we can identify four anomeric protons, labelled with asterisks, confirming that the unknown oligosaccharide is a tetrasaccharide. Tertiary-butyl alcohol (1.24 ppm) was used as the chemical shift reference allowing tentative resonance assignment by comparison of chemical shifts with literature values. For example, other than the characteristic H-4 proton resonance (5.98 ppm) of the iduronic acid double bond at the non-reducing end of the oligosaccharide, only three anomeric proton resonances are well-resolved at 5.56, 5.50 and 5.44 ppm. The fourth anomeric proton is shifted upfield to 4.60 ppm as shown in the expansion of the cITP-NMR spectrum inset. This unique chemical shift allows tentative assignment to a non-sulphated iduronic acid (IdoA) residue, which is expected to have a frequency near 4.6 ppm.
Our results demonstrated the potential of on-line cITP-NMR for evaluation of sample purity and tentative identification of novel oligosaccharides from inspection of the 1H cITP-NMR spectrum using only 1-3 µg of the analyte. If these preliminary results suggest that the isolated oligosaccharide is of sufficient interest, subsequent analysis using 2D NMR spectroscopy experiments for complete structure elucidation can be carried out with 20-30 µg of the purified oligosaccharide and a commercially available CapNMR probe (Protasis/MRM Corp., Savoy, IL). The dual analytical approach of cITP-NMR for compound dereplication followed by 2D analysis of unknown oligosaccharides is expected to significantly advance the determination of oligosaccharide primary structure by reducing the mass of isolated unknown oligosaccharides required for NMR analysis by as much as 50-100 fold.
Conclusion
The improved mass sensitivity possible with microcoil NMR technology makes this technique well suited for characterisation of mass-limited analytes. In conjunction with cITP, microcoil NMR allows for measurement of 1H spectra of separated analytes. In the characterisation of pharmaceutical impurities, we demonstrated the separation and detection of the acetaminophen degradant 4-aminophenol in the presence of a thousand fold excess of the parent drug in the sample matrix. We have also used microcoil NMR to streamline the structure elucidation of heparin and heparan sulphate-derived oligosaccharides using cITP-NMR to generate 1H survey spectra for dereplication and complete 2D NMR characterisation using a commercially
available CapNMR microcoil probe.
Acknowledgements
C. K. L. gratefully acknowledges financial support from the National Science Foundation grant CHE-0616811
References
1. Long HW et al. J Am Chem Soc 1993; 115: 8491.
2. Spraul M et al. Anal Chem 2003; 75: 1536.
3. Olson DL et al. Science 1995; 270: 1967.
4. Webb AG. Prog Nucl Magn Reson Spectrosc 1997; 31: 1.
5. Schroeder FC & Gronquist M. Angew Chem Int Ed 2006; 45: 7122.
6. Webb AG. J Pharm Biomed Anal 2005; 38: 892.
7. Bocek P et al. Analytical Isotachophoresis; VCH: New York, 1988.
8. Kautz RA et al. J Am Chem Soc. 2001; 123: 3159..
9. Eldridge SL et al. Anal Chem 2007; 79: 8446.
10. Rabenstein DL. Nat Prod Rep 2002; 19: 312.
11. Korir AK et al. Anal Chem 2005; 77: 5998.
12. Korir AK & Larive CK. Anal Bioanal Chem 2007: 388: 1707.
The authors
Prof. Cynthia K. Larive
Department of Chemistry
University of California
Riverside, CA 92521, USA
Tel + 1 951 8272990
clarive@ucr.edu
Dr Albert Korir
Department of Chemistry
University of California
Riverside, CA 92521, USA
Tel + 1 951 827 3080
akorir@gmail.com
