The Role of Complimentary Methods in Analytical Quality Control

• GlaxoSmithKline Pharmaceuticals

It has long been recognized that complete resolution of all components of a complex mixture is extremely challenging using a single chromatographic method¹. In addition, complete resolution of all of the additional components, i.e. related products or impurities, from the major peak (the drug substance) is an ancillary challenge^2,3,4. Therefore, the development of impurity profiling methodologies is often challenging because of these selectivity considerations (structural similarity of analytes) and the low levels of impurities present (often >0.1%). The identity and number of impurities is often unknown and hence reference standards cannot be used to facilitate optimal separations⁵. As such, it is often desirable to evaluate one or more orthogonal (or complimentary) methods to improve the likelihood that all impurities are resolved and ultimately quantified^6,7. Orthogonality can be defined as “systems that differ significantly in chromatographic selectivity”⁸. The orthogonality of two systems are usually described in terms of a correlation matrix of individual retention times or peak capacities (N), i.e. the maximum number of peaks that can be resolved within the available retention space. This is then expressed as Pearson correlation coefficients (r), which vary from r=1 (perfect correlation) to r=0 (dissimilar or orthogonal)^5,9,10.

There appears to be two main strategies for developing orthogonal methods². The first approach, exemplified in two-dimensional HPLC, is to optimize the selectivity differences between the two HPLC methods. This can be done by using a wide variety of variables including stationary phase, buffer pH, organic modifier composition and temperature^2,7,8. However, the huge proliferation of stationary phases for use in reverse phase HPLC can make the choice of orthogonal methodologies a formidable undertaking. Columns with seemingly indistinguishable selectivity properties can produce significantly different separations^11,12.

The second approach is to adopt a different type of methodology involving completely different separation mechanisms². These might include normal phase chromatography or its modern equivalent, hydrophilic interaction chromatography (HILIC), supercritical fluid chromatography (SFC), capillary electrophoresis (CE), capillary zone electrophoresis (CZE), and size exclusion chromatography (SEC).

CE or the related capillary electro-chromatographic (CEC) techniques, exhibit many advantages compared to HPLC. These include better intrinsic specificity, reduced solvent consumption, inexpensive stationary phases and higher kinetic performance¹³. Vassort et al.⁶ demonstrated the orthogonality of three generic CZE-MS (mass spectrometry) methods with the established HPLCMS methods for the impurity profiling of six different drugs. They subsequently developed open tubular CEC-ESI (electro-spray ionization)-MS and non-aqueous CE (NACE)-ESI-MS methods and compared their orthogonality with their previous assessment¹⁴. The NACE-ESI-MS method was seen as orthogonal to both the CZE-MS and HPLC-MS methods. Bushey and Jorgensen¹⁵ utilized both HPLC and CE to characterize fluorescently labelled peptide degradation products arising from tryptic digestion. The authors demonstrated that these two techniques were complimentary, separating different sub-sets of this complex analytical matrix.

HILIC is also a good choice for a second method, as it utilizes polar stationary phases and aqueous-polar organic mobile phases, which contain high levels (>60%) of an organic modifier, typically acetonitrile¹⁶. This mobile phase has an inherently low back pressure and the enhanced volatility due to the high organic modifier content improves droplet formation and desolvation efficiencies, leading to significantly enhanced sensitivity gains in ESI-MS¹⁷. However, although HILIC and HPLC do show a high degree of orthogonality, the combination of the two techniques isn’t always effective due to the poor peak capacities (N) exhibited by the HILIC approach. D’Attoma et al.¹⁸ indicated that HPLC/HPLC may offer better resolutions due to enhanced peak capacities of both complimentary phases.

Another, ‘normal’ phase chromatographic technique, which is complimentary to HPLC, is SCF. The introduction of UPLC hardware with increased pressure ranges (≤1300 bar), the development of very precise back-pressure regulators (BPR) and increased accuracy pumping modules has led to the resurgence of SFC applications in impurity profiling, particularly for orthogonal usages^19,20.

SEC often provides complementary data to other separation techniques, particularly if high molecular weight degradant(s) are anticipated, e.g. dimers or formation of adducts with a polymeric excipient. An intermediate was observed to have an extra peak using SEC-CNLD and the authors postulated that reactive impurities, e.g. aldehydes and ketones, in the binder, hydroxypropyl cellulose (HPC) were reacting with the API. This information wasn’t achievable using HPLC-MS²¹.

Argentine et al.² highlighted that hyphenation of analytical methodologies with different detection methods often enhanced differences in selectivity. Mass spectrometry can facilitate automated peak tracking procedures that are absolutely pivotal for comprehensive impurity profiling²².

Up to now we have discussed the use of complementary or orthogonal methodologies as an aid to complete resolution of all impurity peaks present in the sample. However, for certain complex samples this basic tenant of complete resolution of related compounds is unattainable, i.e. in proteins, complex peptides and oligonucleotides. This is predicated on a different understanding of these impurities, in that they do have some levels of ‘efficacy’ compared to the parent molecule and don’t need to be controlled to the same strict levels required for small molecule impurities (ICH Q3A²³, ICH Q3B²⁴). There is some impurity guidance provided in ICH Q5C²⁵ and to some extent ICH Q6B²⁶; however, these guidance indicate that absolute purity in these products is extremely difficult to determine. Hence, there is a requirement to justify the scientific approach taken, rather than prescribing certain limits, unlike the approach taken in ICH Q3A²³/ICH Q3B²⁴.

But the key question still remains, how much specificity is required for these orthogonal methods?

Therefore, for biological molecules, a slightly different concept of orthogonality needs to be applied, where each method provides only part of the picture. Interestingly, this requirement for comparability of method performance isn’t covered in the existing method guidance²⁷ or the more recent supplementary guidance²⁸. In the recent FDA28 guidance for method validation of drugs and biologics even though there is a section on analytical method comparability which deals with alternative analytical procedures there is no mention made of the requirements for methods working in this type of complimentary manner.

Hence, a full understanding of the physicochemical properties of the analytes and their related impurities is therefore required to facilitate a successful separation strategy. For instance, oligonucleotides are linear assemblies of RNA or DNA base pairs and the individual base pairs are connected via phosphothiate (PS) moieties that link the 3’ and 5’ oxygen groups of ribose sugars. The ribose/phosphate backbone is highly charged and hydrophilic in nature, whereas, the base pairs which are attached to the C1 position of the ribose sugar are relatively hydrophobic in nature. Additionally, the single stranded (singlet or un-paired) or duplex (paired) oligonucleotides show pronounced differences in their abilities to interact with the surfaces of standard chromatographic stationary phases²⁹.

The typical impurities arising from the solid-stage synthesis of oligonucleotides include shortened sequences, i.e. N-x, where N is the full length product and x = 1, 2, 3, etc., these are typically termed ‘shortmers’ and it is also possible to form ‘longmers’, i.e. N+x and oxidized variants, whereby the backbone consists of PO rather than PS linkages. Additionally, isomerization is commonly encountered in oligonucleotide synthesis, with 2’,5’ isomers (rather than the desired 3’,5’ arrangement) commonly encountered with basic pHs and/or high temperatures³⁰.

Ion exchange (IEX), ion-pair reversed phase HPLC (IP-RP) and SEC are all used in a complimentary fashion with oligonucleotides. In addition, the former two methods are used in both non-denaturing and denaturing fashions. The denaturing approach is intended to prevent non-covalent interactions, i.e. base pairing to produce the duplex form.

Ion exchange (IEX) chromatography, in particular anion exchange chromatography (AEX), is capable of separating oligonucleotides based on very minor differences in charge, where the total charge increases as a function of oligonucleotide chain length (for both single stranded and duplexes). This mainly involves interactions between the charged, substituted phosphate back bone and the functional groups on the IEX column, but there are also secondary interactions³¹.

The DNA/RNA double helix is unstable above pH 10.5 and therefore elevated pHs are often used in denatured IEX, to analyze singlet structure^32,33. Elevated temperatures (up to 95°C) have been used to facilitate singlet formation due to enhanced mass transfer rates within the polymeric stationary phase²⁹. Finally, the use of different counterions can have a profound effect on specificity of oligonucleotide separations. Separations can be significantly enhanced by changing the eluting power of the counterion (Cl<Br<I<perchlorate<thiocyante)^32,33.

Good separations of 2’, 5’ isomers have been generated using IEX chromatography^34,36. Good separations of the phosphodiester (PO) impurities from the phosphorothioate backbone can also be affected using IEX. The PO impurities tend to be less well retained on the stationary phase due to their reduced hydrophobic character.

Ion-Pair Reversed Phase Chromatography (IP-RP)

The use of IP-RP can facilitate very high resolution separations of oligonucleotides. Ion pairs are formed between the charged ionpairing reagent in the mobile phase and the oppositely charged analyte resulting in different retention rates on the column. IP-RP can be interfaced with ESI-MS³⁵. IP-RP have been successfully employed in both the non-denaturing³⁶, as well as the denaturing³⁷ modes. Many different ion-pairs have been utilized, but hexafluoro-2-propanol (HFIP) is particularly attractive due to its high volatility, compatibility with ESI-MS and its ability to reduce the influence of the hydrophobicity of the single stranded oligonucleotides³⁸.

Both silica and polymeric columns have been used, but the latter has greater tolerability of elevated pHs and temperatures. IP-RP using denaturing settings is routinely employed in evaluations of the more common oligonucleotide impurities, e.g. shortmers, longmers, 2’,5’ isomers, as well as lipophilic adducts²⁹.

Size Exclusion Chromatography (SEC)

SEC is widely used in oligonucleotide and protein analysis, separating peaks according to relative size and diffusion rates into the resin pores. Thus, larger analytes which are excluded from the resin pores elute first, i.e. the elution order is the inverse of other chromatographic approaches³⁹, making it a very complimentary technique to other chromatographic approaches. Although, SEC doesn’t have the same resolving power as IEX and IP-RP, it can readily separate duplexes from singlets and higher order structures (including aggregates) and consequently, SEC is better suited to non-denaturing analysis²⁹. It is typically used with mild operating conditions, i.e. neutral pH, physiological buffers, ambient temperature, etc. Reducing flow rates can enhance separations, but buffer concentration and ionic strength have limited effect. SEC is particularly suited for non-denaturing analysis of proteins, where the mild conditions facilitate analysis of weakly held aggregates.

Conclusion

It is often desirable to assess one or more orthogonal (or complimentary) methods to improve the likelihood that all impurities are resolved and ultimately quantified. However, for complex biological (proteins, peptides) or synthetic oligonucleotides the peak capacity (N), i.e. the maximum number of peaks that can be resolved within the available retention space, is significantly less than the likely number of related substances present. To address this deficiency, complimentary orthogonal methodologies are necessary, each method providing part of the full picture. In the case of oligonucleotides, several different methods are utilised in parallel; ion-pair reversed phase HPLC (IPRP) and IEX are used to separate shortmers, longmers, isomers and oxidation products (PO). In addition, SEC is used to separate singlet and duplexes, as well as higher order multiplexes and aggregates.

In summary, as there is no regulatory guidance in this area, analysts need to clearly articulate within the analytical target profile (ATP) the role that each of the analytical methods performs in providing an overview of the total quality of the product.

References

1. Gray, M., Dennis, G.R., Wormell, P., Shalliker, R.A., Salonecker, P. 2002. Two dimensional reversed-phase reversed-phase separations. Isomeric separations incorporating C18 and carbon clad zirconium stationary phases. J. Chromatogr. A, 975, 285-297.
2. Argentine, M.D., Owens, P.K., Olsen, B.A. 2007. Strategies for the investigation and control of process-related impurities in drug substance. Adev. Drug Deliv. Revs., 59, 12-28.
3. Davis, J.M., Giddings, J.C. 1983. Statistical theory of component overlap in multicomponent chromatograms. Anal. Chem. 55, 418-424.
4. Martin, M., Herman, D.P., Guiochon, G. 1983. Probability distributions of the number of chromatographically resolved peaks and resolvable components in mixtures. Anal. Chem., 58, 2200-2207.
5. Dumarey, M., Put, R., Van Gyseghem, E., Vander Heyden, Y. 2008. Dissimilar or orthogonal reversed-phase chromatographic systems: A comparison of selection techniques. Anal. Chimica Acta, 609, 223-234.
6. Venkatramani, C.J., Xu, J., Philips, J.B. 1996. Separation orthogonality in temperatureprogrammed comprehensive two-dimensional gas chromatography. Anal. Chem., 68, 1486-1492.
7. Vassort, A., Barrett, D.A., Shaw, P.N., Ferguson, P.D., Szucs, R. 2005. A generic approach to the impurity profiling of drugs using standardized and independent capillary zone electrophoresis coupled to electro-spray ionisation mass spectrometry. Electrophoresis, 26, 1712-1723.
8. Van Gyseghem, E., Dejaegher, B., Put, R., Forlay-Frick, P., Elkihel, A., Daszykowski, M., Héberger, K., Massart, D.L., Vander Heyden, Y. 2006. Evaluation of chemometric techniques to select orthogonal chromatographic systems. J. Pharm. Biomed. Anal., 41, 141-151.
9. Liu, Z., Patterson, D.G., Lee, M.L. 1995. Geometric approach to factor analysis for the estimation of orthogonality and practical peak capacity in comprehensive twodimensional separations. Anal. Chem., 67, 3840-3845.
10. Slonecker, P.J., Li, X., Ridgway, T.H., Dorsey, J.G. 1996. Anal. Chem., 68, 682-689.
11. Johnson, A.R., Johnson, C.M., Stoll, D.R., Vitha, M.F. 2012. Identifying orthogonal and similar reverse phase liquid chromatography stationary phases using the system selectivity cube and the hydrophilic substraction model. J. Chromatogr. A. 1249, 62-82.
12. Zhang, Y., Carr, P.W. 2009. A visual approach to stationary phase selectivity classification based on the Snyder–Dolan Hydrophobic-Subtraction Model. J.Chromatogr. A. 1216, 6685-6694.
13. Guillarme, D., Nguyen, D.T.T., Rudaz, S., Veuthey, J-L. 2008. New trends in fast and highresolution chromatography: A Critical comparison of existing approaches. Eur. J. Pharm. Biopharm. 68, 430-1082.
14. Vassort, A., Shaw, P.N., Ferguson, P.D., Szûcs, R., Barret, D.A. 2008. Comparison of CZE, open tubular CEC and non-aqueous CE coupled to electrospray MS for the impurity profiling of drugs. Electrophoresis 29, 3563-3574.
15. Bushey, M.M., Jorgensen, J.W. 1990. Automated instrumentation for comprehensive twodimensional high-performance liquid chromatography/capillary zone electrophoresis. Anal. Chem. 62, 978-984.
16. Periat, A., Kohler, I., Bugey, A., Bieri, S., Versace, F., Staaub, C., Guillarme, D. 2014. Hydrophilic interaction chromatography versus reversed phase liquid chromatography coupled to mass spectrometry: Effect of electrospray ionization source geometry. J.Chromatogr. A. 1356, 211-220.
17. Nguyen, H.P., Schug, K.A. 2008. The advantages of ES-MS detection in conjunction with HILIC mode separations: Fundamentals and applications. J. Sep. Sci. 31, 1465-1480.
18. D’Attoma, A., Grivel, C., Heinisch, H. 2013. On-line comprehensive two-dimensional separations of charged compounds using reversed-phase high performance liquid chromatography and hydrophilic interaction chromatography. Part I: Orthogonality and practical peak capacity considerations. J.Chromatogr. A. 1262, 148-159.
19. Grand-Guillaume Perrenoud, A., Veuthey, J-L., Guillarme, D. 2012. Comparison of ultrahigh performance super critical fluid chromatography and ultra-high performance liquid chromatography for the analysis of pharmaceutical compounds. J.Chromatogr. A. 1266, 158-167.
20. Grand-Guillaume Perrenoud, A., Veuthey, J-L., Guillarme, D., Fekete, S. 2013. Maximising kinetic performance in super critical fluid chromatography using state-of-the-art instruments. J.Chromatogr. A. 1314, 288-297.
21. Pan, C, Harmon, F., Toscano, K., Liu, F., Vivilecchia, R. 2008. Strategy for identification of leachables in packaged pharmaceutical liquid formulations. J. Pharm. Biomed. Anal. 46, 520-527.
22. Xue, G., Bendick, A.D., Chen, R., Sekulic, S.S. 2004. Automated peak tracking for comprehensive impurity profiling in orthogonal liquid chromatographic separation using mass spectrometric detection. J. Chromatogr. A. 1050, 159-171.
23. ICH Q3A(R2): Impurities in new drug substances, current step 4 version, 25th October 2006.
24. ICH Q3B(R2): Impurities in new drug products, current step 4 version, 02nd June 2006.
25. ICH Q5C: Quality of biotechnological products: stability testing of biotechnological/ biological products, current step 4 version, 30th November 1995.
26. ICH Q6B: Specifications, test procedures and acceptance criteria for biotechnological/ biological products, current step 4 version, 10th March 1999.
27. ICH Q2(R1): Validation of Analytical Procedures: Text and Methodology, current step 4 version, incorporated in November 2005.
28. Analytical procedures and methods validation for drugs and biologics, FDA guidance for industry, July 2015.
29. Noll, B., Roehl, I. 2013. Separation of oligonucleotides and related substances. In Analysis of Oligonucleotides and their Related Substances, Eds. Okafo, G., Elder, DP., Webb, M. ILM publications, 2013.
30. Mikkola, S., Kaukinen, U., Lonnberg, H. 2001. The effect of secondary structure on the cleavage of the phosphodiester bonds of DNA. Cell Biochem. Biophysics. 34, 95-119.
31. Warren, W.J., Vella, G. 1995. Principles and methods for the ansalysis and purification of synthetic deoxyribonucleotides by high-performance liquid chromatography. Mol. Biotechnol. 4, 179-199.
32. Thayer, J.R., Barreto, V., Rao, S., Pohl, C. 2005. Control of oligonucleotide retention on a pHstabilized strong anion exchange column. Anal. Biochem. 361, 132-139.
33. Thayer, J.R., Pohl, C., Puri, N., Burnett, C.A., Young, M. 2007. Identification of aberrant 2’-5’ RNA linkage isomers by pellicular anion exchange chromatography. Anal. Biochem. 361, 132-139.
34. Seiffert, S., Debelak, H., Hadwiger, P., Jahn-Hofmann, K., Roehl, I., Vornlocher, H.P. Noll, B. 2011. Characterization of side reactions during the anneraling of small interfering RNAs. Anal. Biochem. 414, 47-57.
35. Huber, C.G., Oberacher, H. 2001. Analysis of nucleic acids by on-line liquid chromatographymass spectrometry. Mass Spectromr. Rev. 20, 310-343.
36. Beverley, M., Hartsough, K., Machemer, L. 2005. Liquid chromatography/electrospray mass spectrometric analysis of metabolites from an inhibitory DNA duplex. Rap. Comm. Mass Spectromr. 19, 1675-1682.
37. Gilar, M. 2001. Analysis and purification of synthetic oligonucleotides by reversed-phase liquid chromatography with photodiode array and mass spectrometry detection. Anal. Biochem. 298, 196-206.
38. Beverley, M., Hartsough, K., Machemer, L., pavco, P., Lockridge, J. 2006. Liquid chromatography electrospray ionization mass spectrometry analysis of the ocular metabolites from a short interfering duplex. J. Chromatogr. B, 835, 62-70.
39. Chan, D., Roymoulik,I. 2011. Purity analysis and molecular weight determinations by size exclusion HPLC. In: Bonilla and Srivasta (Eds.), Handbook of Analysis of Oligonucleotides and Related Products, CRC Press, Boca Raton, FL, US.