Search

Replacement of a non-bridging oxygen atom of the phosphate diester linkage of an oligonucleotide by sulfur conveys pharmacokinetic benefits, such as increased nuclease resistance and enhanced protein binding. Substitution renders the internucleotide linkages chiral, and so phosphorothioate diester (PS) oligonucleotides comprise complex mixtures of diastereoisomers. Currently, chromatographic separation of individual diastereoisomers is limited to oligonucleotides that contain no more than about four or five PS linkages. The development of therapeutic PS oligonucleotides, which often contain >15 PS linkages, would be greatly aided by methods useful for assessing batch-to-batch stereo-reproducibility. To this effect, the relative sensitivities of metal ion complexation chromatography (MICC), in-series reversed phase-strong anion exchange chromatography (RP-SAX), and 31P NMR toward changes in the diastereoisomeric distributions of therapeutic PS oligonucleotides were compared. Model oligonucleotides synthesized under conditions known to impact PS stereochemistry were used to evaluate the method performance, and all three methods showed excellent sensitivity toward changes in the diastereoisomeric composition. Interactions via the solvent-accessible areas and a combination of hydrophobic and electrostatic forces may be responsible for the selectivity demonstrated by MICC and in-series RP-SAX, respectively.

The separation of impurities in phosphorothioate diester (PS) oligonucleotides is complicated by (1) the presence of a very large number of diastereoisomers, e.g., 2

The relatively large molecular size, diastereoisomeric nature, and complex impurity profiles of therapeutic phosphorothioate oligonucleotides create significant analytical challenges for the quality control laboratory. To overcome the lack of selectivity inherent to traditional chromatographic approaches, an ion pair liquid chromatography-mass spectrometry (LCMS) method combining ultraviolet and mass spectrometry quantification was developed and validated for35 different oligonucleotide drug substances and products, including several commercialized drugs. The selection of chromatographic and spectrometric conditions, data acquisition and processing, critical aspects of sample and buffer preparation and instrument maintenance, and results from method validation experiments are discussed.

Phosphorothioate oligonucleotide drugs typically contain product-related impurities that are difficult to resolve chromatographically from the parent oligonucleotide due to the size of these compounds and the large number of stereoisomers that comprise the parent. The presence of co-eluting impurities hinders the process of determining assay based on chromatographic separation alone. A mass spectrometry-based purity assessment of the main chromatography peak can be used to quantify co-eluting impurities and enable the accurate determination of assay, but a more direct measure of assay was desired due to the complexity of measuring all co-eluting impurities by mass spectrometry. Therefore, we developed an assay method that utilizes the specificity of mass spectrometry to measure the amount of active pharmaceutical ingredient in a sample, which eliminates the need for chromatographic separation of impurities from the product. This procedure uses a single quadrupole mass spectrometer and incorporates an internal standard that is co-sprayed with the analyte to compensate for the drift commonly associated with mass spectrometry-based quantitation. Using the mass spectrometry response ratio for sample to internal standard enables the method to achieve excellent linearity (R2 = 0.998), repeatability (relative standard deviation = 0.5%), intermediate precision (0.6%), and accuracy, with measured assay values consistently within 2.0% of expected. The results indicate the method possesses the accuracy and precision required for measuring assay in clinical and commercial stage pharmaceutical products. Since the method is based on the specificity of the mass spectrometer, and does not rely on chromatographic separation of impurities, the procedure should be applicable to a wide variety of oligonucleotide therapeutics regardless of sequence or chemical modifications.

A novel analytical approach capable of measuring deaminated degradation products of oligonucleotide therapeutics is described. The method employs high-resolution mass spectrometry to assess the shift in isotopic distribution that accompanies deamination. Isotopic Distribution Factors (IDF), derived directly from the peak heights of the isotopic pattern, are employed to measure deamination levels of as little as 0.5%. Results obtained from application of the method to a phosphorothioate diester oligonucleotide exposed to various temperatures were used to determine deamination rates.

Rationale Quantitative Ion-Pair (IP)-HPLC MS methods are employed to determine the complex impurity profiles of oligonucleotide therapeutics. While impurities that co-elute with the main product are routinely monitored, the large number of early and late eluting impurities makes their individual measurements tedious and time-consuming. An improved method is needed for routine analyses. Methods A Combined Ranking (CR) index is developed to provide a composite value for both qualitative and quantitative impurity changes. Positive and Negative Impurity Change (PIC/NIC) indices are developed to determine the degree and direction of change. Optimized experimental conditions are determined for the characterization of trace levels of impurities. Replicate analysis, blank subtraction, and signal processing approaches are used to enhance the S/N. Dot-product and Euclidean distance equations monitor spectral changes. Results Spiking experiments with individual and complex impurity mixtures show the method can distinguish among samples differing in impurity content by 0.2% wt. The method has been applied to monitor changes in impurity profile among different batches of the same oligonucleotide and changes in the same batch over time (stability analysis). The method permits rapid determination of changes in the types and amounts (increases and decreases) of impurities present. Conclusions A novel approach for the rapid determination of changes in the impurity profile of oligonucleotide therapeutics has been developed. The straightforward data treatment and the speed and simplicity of the approach make the method easy to implement and use. Possible quality control applications include drug substance and drug product stability studies, and the assessment of batch-to-batch variability.

Ion pair-reversed phase (IP-RP) HPLC is one of the most widely used methods for the analysis of oligonucleotide impurities. The method is compatible with mass spectrometry and has been used to guide the development of improved synthesis and purification approaches. The ability to detect and characterize impurities depends on the reagents and the IP buffer system employed, as each can directly affect the degree of chromatographic separation and the sensitivity of detection by MS. Previous work in our laboratory has shown that small alkyl amines are suitable IP reagents for the analysis of impurities in phosphate diester oligonucleotides and can be used to differentiate among individual members of composite impurity families. The addition of an alkyl acid often further enhances peak separation, but at the detriment of ion signal. An improved method with increased chromatographic performance and sensitivity of detection is presented here. Improvements were mainly realized through the use of lower concentrations of small alkyl amine (i.e., 5 mM) and acid (0.5 mM) IP reagents, and ammonium bicarbonate (20 mM) as a buffer. The improved capabilities of the new method are demonstrated by separation of the individual components of the composite n − 1 impurity in a set of four production-scale batches of a single oligonucleotide. Addition of the alkyl acid resulted in resolution of most individual n − 1 impurities. The observed enhanced sensitivity of detection allowed multiple reaction monitoring (MRM) experiments, which were used to differentiate among unresolved impurities.

A new method is presented for the automated determination of early eluting (E.E.) oligonucleotide impurities analyzed by IP-RP HPLC HRMS. E.E. are impurities shorter than the main drug component and are primarily formed by the mechanisms of coupling failure, and depurination. The method is based on the detection of the theoretically derived most abundant mass of an impurity in the experimental data. An exhaustive list of candidate impurities and their formulas is automatically generated using the parent sequence and the known mechanisms of impurity formation. The approach accounts for possible modifications in the individual oligonucleotide sequence moieties (e.g., linkage, sugar, and base, 3', and 5' ends). The detected ion signal is summed to provide four nested levels of impurity breakdown information. The approach allows for the rapid determination of relationships and trends of impurities in samples generated by different manufacturing processes and conditions. Representative examples are given to illustrate the capabilities and utility of the approach in synthesis and purification process optimization applications.

In an effort to improve separation of impurities in oligonucleotide drugs, alkyl amines of different length and carbon content were evaluated as reagents in ion pair-reversed phase (IP-RP) HPLC with mass spectrometric detection. A range of columns was tested in combination with different buffers, ion-pair modifiers and varying pH adjustments. For phosphorothioate oligonucleotides, larger amines, like tributyl and hexyl amine provided the best chromatography, as small amines tended to broaden peaks due to the separation of diastereoisomers. For phosphate diester oligonucleotides, the best separations were obtained using small alkyl amines, like propyl-, isopropyl- and diethylamine. Conditions optimized for oligonucleotide sequence and type of impurity enabled full separation of the individual components of composite impurities, such as n-1, N3-(2-cyanoethyl)thymine (CNET), deaminated and 3-(2-oxopropyl)imidazopyrimidinone (OPC) impurities. The addition of long-chain alkyl acids like hexanoic acid to the IP buffer resulted in further improvements in peak separation.

Continued improvements in understanding RNA biology have led to significant success in the development of antisense oligonucleotide therapeutics, and several oligonucleotide drugs have now been approved. Manufacturing of oligonucleotides may be associated with the production of impurities. Current methods for quantification of impurities that co-elute with the main drug component rely on manual ion extraction and integration of the characteristic mass signal of each impurity. For certain applications however, especially those involving large sets of samples such as those generated in the optimization of oligonucleotide manufacturing processes, a rapid method that provides semi-quantitative determination of impurity levels would be sufficient. In this work, an automated approach has been developed to rapidly determine the relative amounts of co-eluting impurities in oligonucleotide samples. The most abundant mass in the isotopic distribution is automatically calculated from the impurity formula and used to detect the presence of the impurities. The principles of the approach are described, and representative examples are given. Impurities determined in different manufacturing lots are compared directly, and by principal component analysis. The ability of the method to determine impurity levels across large sample sets is illustrated for an oligonucleotide drug purification study.

A 2D-LC system coupled with a TOF mass spectrometer has been evaluated for its capabilities to provide enhanced characterization of oligonucleotide impurities. To address loop limitations in the total volume (40 μL) sampled across 1D peaks, a column trap was incorporated between the 1D and 2D columns. The main advantages of the column trap include reduction of the total number of sequential 2D runs required to fully sample broad 1D peaks, and most importantly, reduction of the error in quantitative determination of the components in broad 1D peaks by avoiding the numerical stitching of data from several 2D runs. Comprehensive RP x IP provided orthogonal separation despite its lower 1D resolution. In contrast, IP x IP did not provide orthogonal separation. RP x IP using the direct on-line extended heart-cutting system with the column trap showed additional benefits, in the elimination of off-line fractionation and sample handling errors and was successfully applied in a pH stability study of a crude oligonucleotide. SAX x IP successfully separated the isobaric “n+16” doublet of the “n + O” and “[n + S-O]” impurity species, a feat not currently possible by mass spectrometry alone or 1D-LC, demonstrating the importance of the added capabilities of the 2D-LC approach.

Quantitative Ion-Pair (IP)-HPLC MS methods are employed to determine the complex impurity profiles of oligonucleotide therapeutics. While impurities that co-elute with the main product are routinely monitored, the large number of early and late eluting impurities makes their individual measurements tedious and time-consuming. An improved method is needed for routine analyses.A Combined Ranking (CR) index is developed to provide a composite value for both qualitative and quantitative impurity changes. Positive and Negative Impurity Change (PIC/NIC) indices are developed to determine the degree and direction of change. Optimized experimental conditions are determined for the characterization of trace levels of impurities. Replicate analysis, blank subtraction, and signal processing approaches are used to enhance the S/N. Dot-product and Euclidean distance equations monitor spectral changes.Spiking experiments with individual and complex impurity mixtures show the method can distinguish among samples differing in impurity content by 0.2% wt. The method has been applied to monitor changes in impurity profile among different batches of the same oligonucleotide and changes in the same batch over time (stability analysis). The method permits rapid determination of changes in the types and amounts (increases and decreases) of impurities present.A novel approach for the rapid determination of changes in the impurity profile of oligonucleotide therapeutics has been developed. The straightforward data treatment and the speed and simplicity of the approach make the method easy to implement and use. Possible quality control applications include drug substance and drug product stability studies, and the assessment of batch-to-batch variability.

It is demonstrated that solid support containing a novel universal linker could be efficiently used to synthesize both phosphorothioate oligodeoxyribonucleotides and second-generation 2′- O -methoxyethyloligoribonucleotides with high yield and quality as judged by ion-pair-liquid chromatography–electrospray mass spectroscopy, 31 P NMR and reversed phase HPLC. Analysis of oligonucleotides shows quality being superior to that produced with standard succinyl-linker solid supports, without contamination of materials resulting from linker or support backbone decomposition.

The impuritiy profiles of acetonitrile solutions of the four standard O-cyanoethyl-N,N-diisopropyl-phosphoramidites of 5'-O-dimethoxytrityl (DMT) protected deoxyribonucleosides (dG(ib), dA(bz), dC(bz), T) were analyzed by HPLC-MS. The solution stability of the phosphoramidites decreases in the order T, dCdAdG. After five weeks storage under inert gas atmosphere the amidite purity was reduced by 2% (T, dC), 6% (dA), and 39% (dG), respectively. The main degradation pathways involve hydrolysis, elimination of acrylonitrile and autocatalytic acrylonitrile-induced formation of cyanoethyl phosphonoamidates. Consequently, the rate of degradation is reduced by reducing the water concentration in solution with molecular sieves and by lowering the amidite concentration. Acid-catalyzed hydrolysis could also be reduced by addition of small amounts of base.

Efficiency of phosphorothioate oligonucleotide syntheses could be improved by increasing the concentration of dichloroacetic acid in toluene to 10% from 3% (v/v) during the detritylation step. It was also found that dichloroacetic acid is better than dichloropropionic acid when compared at similar concentrations.

Detritylation of a 5‘-O-DMT-2‘-deoxyadenosine moiety attached to solid support under acidic conditions (3% dichloroacetic acid/solvent) leads to depurination during oligonucleotide synthesis. Depro...

By on-line addition of a central atom (for example, AgI, BIII, PdII, LiI) positively or negatively charged complexes of analytes can be formed for CIS-MS. This technique is applicable to both polar and nonpolar compounds—for example, for alcohols, ethers, and a large number of olefins, polyolefins, and arenes as well as steroids, vitamins of the D and E families, carotinoids, polystyrols, terpenes, and unsaturated fatty acids—and can be readily coupled with separation techniques.

A crude palm-oil extract rich in vitamin E homologues was investigated by HPLC-MS and HPLC-NMR coupling. For mass spectrometry a newly introduced ionization technique called Coordination Ion Spray (CIS) was used. Through the addition of silver ions to the HPLC eluent, the ionization process of nonpolar substances is facilitated. Chromatography and all coupling experiments were conducted on a C(30) column which exhibited an extraordinary shape selectivity and overwhelming sample-loading capability. Experiments were performed with pure methanol as an eluent which proved to be ideal for NMR spectroscopy as well as mass spectrometry. All necessary information for unambiguous structural assignment was collected within 45 min of the LC-NMR experiment and 15 min of the LC-MS experiment. Six compounds were identified, i.e., α-, β-, γ-, and δ-tocotrienol, α-tocoenol, and α-tocopherol.

[Chemical reaction: See text] Depurination is an important degradation pathway for antisense phosphorothioate oligonucleotides under conditions of thermal stress. We present evidence showing that depurinated oligonucleotides react with cytosine-containing sequences giving products containing a 6-(2-deoxy-beta-D-erythro-pentofuranosyl)-3-(2-oxopropyl)imidazo[1,2-c]pyrimidin-5(6H)-one residue. Further, we demonstrate that the same product is formed upon treatment of 2'-deoxycytidine with 4-oxo-2-pentenal, the latter being an expected byproduct of serial elimination reactions at apurinic sites. In addition to being important for synthetic oligonucleotides, apurinic site formation in cellular DNA is a common occurrence. Because repair of these sites can result in the production of 4-oxo-2-pentenal, it is interesting to speculate whether 6-(2-deoxy-beta-D-erythro-pentofuranosyl)-3-(2-oxopropyl)imidazo[1,2-c]pyrimidin-5(6H)-one residues can form in vivo.

Some commercial batches of dichloroacetic acid (DCA) contain traces of chloral (trichloroacetaldehyde). Using such DCA to effect detritylation during solid-phase oligonucleotide synthesis results in the formation of a family of process impurities in which the atoms of chloral (Cl3CCHO) are incorporated between the 5'-oxygen and phosphorus atoms of an internucleotide linkage. The structure was elucidated by HPLC with UV and MS detection, digestion of the oligonucleotide, synthesis of model compounds, and 1H and 31P NMR spectroscopy. By understanding the chemistry behind its formation, we are now able to limit levels of this impurity in synthetic oligonucleotides by limiting chloral in DCA.

In LC/MS, nonpolar substances in the majority of cases cannot be ionized by standard electrospray ionization (ESI) because they obviously lack a site for protonation or deprotonation. The ionization of carotenoids and tocopherols can be greatly enhanced by the addition of silver ions. The Ag(+)-carotenoid and Ag(+)-tocopherol adducts thus formed render these substances amenable to MS. alpha-, beta-, gamma-, and delta-tocopherol, alpha-tocopherol acetate, and the various isomers of lycopene and beta-carotene were separated by C30 RP-HPLC and could be identified by online ESI-MS. A mixture of six different carotenoids was analyzed by scanning the mass range from m/z 500 to 800. The mass spectra of the peaks revealed that all carotenoids and most tocopherols were partially oxidized to radical cations. The detection limit for canthaxanthin was approximately 500 fmol while that of beta-carotene was below 300 fmol. An increase in sensitivity in the MRM mode can be attained by monitoring ions formed by loss of elemental silver from the adducts in the CID cell. Dichloromethane extracts of tomato, carrot, and vegetable juices, a vitamin drink, and a commercial infant food product were analyzed by LC/MS. After postcolumn argentation, from the mass-selective extracts of the TIC, the carotenoids and tocopherols present could be identified by their masses and their retention times. For all studies, a silver perchlorate solution with an overall concentration of 50 micrograms/mL was used.