Mass Spectrometry in The Quantitative Analysis of Therapeutic Intracellular Nucleotide Analogs

MASS SPECTROMETRY IN THE QUANTITATIVE ANALYSIS OF THERAPEUTIC INTRACELLULAR NUCLEOTIDE ANALOGS
Robert S. Jansen,1* Hilde Rosing,1 Jan H.M. Schellens,2,3 and Jos H. Beijnen1,3 1 Department of Pharmacy & Pharmacology, Slotervaart Hospital/ The Netherlands Cancer Institute, Louwesweg 6, 1066 EC Amsterdam, the Netherlands 2 Department of Clinical Pharmacology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, the Netherlands 3 Science Faculty, Department of Pharmaceutical Sciences, Utrecht University, Utrecht, the Netherlands
Received 27 April 2009; accepted 29 September 2009
Published online 8 July 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/mas.20280
Nucleoside analogs are widely used in anti-cancer, anti(retro)viral, and immunosuppressive therapy. Nucleosides are prodrugs that require intracellular activation to mono-, di-, and nally triphosphates. Monitoring of these intracellular nucleotides is important to understand their pharmacology. The relatively involatile salts and ion-pairing agents traditionally used for the separation of these ionic analytes limit the applicability of mass spectrometry (MS) for detection. Both indirect and direct methods have been developed to circumvent this apparent incompatibility. Indirect methods consist of dephosphorylation of the nucleotides into nucleosides before the actual analysis. Various direct approaches have been developed, ranging from the use of relatively volatile or very low levels of regular ion-pairing agents, hydrophilic interaction chromatography (HILIC), weak anion-exchange, or porous graphitic carbon columns to capillary electrophoresis and matrix-assisted light desorptiontime of ight (MALDI-TOF) MS. In this review we present an overview of the publications describing the quantitative analysis of therapeutic intracellular nucleotide analogs using MS. The focus is on the different approaches for their direct analysis. We conclude that despite the technical hurdles, several useful MS-compatible chromatographic approaches have been developed, enabling the use of the excellent selectivity and sensitivity of MS for the quantitative analysis of intracellular nucleotides. # 2010 Wiley Periodicals, Inc., Mass Spec Rev 30:321343, 2011 Keywords: nucleotides; mass spectrometry; quantitative; bioanalysis; intracellular
I. INTRODUCTION
Nucleoside analogs form a class of drugs that is widely used in anti-cancer, anti-(retro)viral, and immunosuppressive therapy. The structural formulas of nucleosides are composed of a sugar and a base moiety (Fig. 1). The base moiety can consist of a pyrimidine base (cytosine (C), thymine (T), uracil (U)) or a
*Correspondence to: Robert S. Jansen, Department of Pharmacy & Pharmacology, Slotervaart Hospital/The Netherlands Cancer Institute, Louwesweg 6, 1066 EC Amsterdam, the Netherlands. E-mail: robert.jansen@slz.nl
purine base (adenine (A), guanine (G)), whereas the sugar moiety consists of ribose or deoxyribose. Nucleosides are the natural building blocks of DNA and RNA. Nucleoside analogs, however, are modied at the base moiety or sugar moiety (Figs. 2 and 3). Nucleoside analogs are metabolized through the same pathways as their natural counterparts, but also block these pathways. The analogs are, therefore, called anti-metabolites. Through these metabolic pathways they are consecutively converted into their monophosphate (MP), diphosphate (DP), and nally triphosphate (TP) in cells. The phosphorylation of nucleoside analogs is depicted in Figure 4, with zidovudine (AZT), the rst active anti-human immunodeciency virus (HIV) drug, as example. Other analogs are administered as nucleoside analog prodrug (a prodrug rst requires conversion to the actual drug) (abacavir (ABC)), base (6-mercaptopurine (6-MP)), or nonhydrolyzable MP (tenofovir (PMPA), adefovir (PMEA)) and undergo slightly different activation pathways. The TP form inhibits human and viral DNA and RNA polymerases and is incorporated into nucleic acids, slowing down cell proliferation or viral replication. Therefore, the TP form is considered to be the main active metabolite. The MP and DP, however, can be pharmacologically active as well. The intracellular pharmacokinetics of the nucleotides are different from the plasma pharmacokinetics of their parent nucleoside analogs. The anti-retroviral agent emtricitabine (FTC), for example, has a plasma half-life of 810 hr, whereas its intracellular TP has a half-life of 39 hr (Wang et al., 2004). Information on the intracellular half-life has, therefore, been used to support once-daily dosing (Wang et al., 2004; Back et al., 2005). Moreover, intracellular drugdrug interactions occurring at the nucleotide level have been explained by quantitation of intracellular nucleotide analogs (Barreiro et al., 2004). Finally, intracellular nucleotide analog levels can vary from patient to patient, resulting in undertreatment or toxicities (Anderson et al., 2003). For example, in the case of the nucleotides of 6-thioguanine (6-TGN), an analog widely used against inammatory bowel disease, this variation can result in active disease on the one hand and hepatic toxicity on the other. Therapeutic drug monitoring to optimize the dose of the individual patients can result in better clinical outcomes (Osterman et al., 2006). Thus, monitoring of intracellular nucleotide analogs is imperative to understand their pharmacology (Rodriguez Orengo et al., 2000). Nucleotide analogs are, however, often present in
Mass Spectrometry Reviews, 2011, 30, 321343 # 2010 by Wiley Periodicals, Inc.
&
JANSEN ET AL.
FIGURE 1. Structural elements of nucleosides.
cells at very low levels, whereas their natural variants are present at very high levels. Sensitive and selective analytical techniques are thus required. A myriad of analytical methods has been developed for the quantitative determination of nucleotide analogs in cells. A review on the analysis of nucleoside reverse transcriptase inhibitors (NRTIs) and their phosphorylated metabolites in HIV-infected matrices has recently been published by Lai, Wang, and Cai (2008). The analytical methods can be divided into direct and indirect methods.
Indirect methods consist of a de-phosphorylation step, in which the nucleotides are re-converted into nucleosides, followed by quantication (Fig. 5). The de-phosphorylation can be performed with or without prior fractionation of the nucleotides on an anion-exchange column, resulting in separate results for the MP, DP, and TP or in a total nucleotide amount. The subsequent quantication has been performed using radioimmunoassays (Kuster et al., 1991; Slusher et al., 1992) and highperformance liquid chromatography (HPLC) with ultraviolet (Solas et al., 1998) or uorescence detection (Pike et al., 2001). More recently, mass spectrometric (MS) detection was introduced for subsequent quantication. Direct methods do not use a de-phosphorylation step and analyze the nucleotides as such (Fig. 5). Immunoassays have been used without any sample separation (Goujon et al., 1998; Le Saint et al., 2004), but most methods require chromatographic separation before detection. Traditionally, strong anion-exchange (SAX) and ion-pairing liquid chromatography (IP-LC) have been used to separate these ionic analytes, which were then quantied using an enzymatic assay or radioactivity (Robbins et al., 1994), ultraviolet (Reichelova, Albertioni, & Liliemark, 1996), or uorescence (Dai et al., 2001; Tidd & Dedhar, 1978) detection. The use of relatively involatile salts and IP agents, however, hampers the applicability of MS for detection. Nevertheless, various MS methods have now been developed for the direct analysis of nucleotides ranging from the use of relatively volatile or very low levels of regular IP agents, weak anion-exchange (WAX) or porous graphitic carbon (PGC) columns to capillary
FIGURE 2. Chemical structures of pyrimidine analogs and their natural counterparts.
322
Mass Spectrometry Reviews DOI 10.1002/mas
QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS
&
FIGURE 3. Chemical structures of purine analogs and their natural counterparts.
electrophoresis (CE) and matrix-assisted light desorptiontime of ight (MALDI-TOF) MS. In this review we present an overview of the publications describing the quantitative analysis of intracellular nucleotide analogs of therapeutic nucleosides using MS, with the focus being on approaches for their direct analysis.
II. SAMPLE COLLECTION A. Matrix

Monitoring intracellular nucleotide analogs is more informative than monitoring nucleoside analogs because the active metabolites are determined, rather than their prodrug. Moreover, these active metabolites are determined at the target site. For this
reason, nucleotide analogs active against hepatitis, acquired immunodeciency syndrome (AIDS), and leukemia have been measured (both in vitro and in vivo) in hepatocytes, peripheral blood mononuclear cells (PBMCs), and leukemic blasts, respectively (Claire, 2000; Ray et al., 2004; Kalhorn et al., 2005). Matrices associated with toxicity have also been analyzed, such as erythrocytes in ribavirin therapy (Yeh et al., 2007). Target tissue can, however, be difcult to obtain in vivo. Therefore, surrogate matrices such as erythrocytes and PBMCs have been analyzed often to monitor therapy against inammatory bowel disease and solid tumors (Dervieux et al., 2005; Veltkamp et al., 2006). Nucleotide levels may, however, differ between cell types (Bergan et al., 1997; Durand-Gasselin et al., 2007), and even cell subtypes, as was shown for AZT and lamivudine (3TC) nucleotides in PBMCs (Anderson et al., 2007). Therefore, careful
FIGURE 4. Phosphorylation route of zidovudine (AZT).
323
&
JANSEN ET AL.
C. Extraction of Nucleotides from Cells

The nucleotide extraction step should be performed with a solution that completely extracts nucleotides, precipitates and inactivates proteins, and in which the analytes are stable (Becher et al., 2003a). Moreover, the solution should not interfere with subsequent chromatography or MS detection. The extraction recovery can be determined by spiking a suspension of intact cells with the analyte before and after lysis and centrifugation (Jansen et al., 2009b). Whether the thus obtained recovery reects the actual recovery is, however, questionable since the analytes are present inside the cells in biological samples. Determination of the recovery after incubation of cells with radioactive material is of limited value since the compounds are extensively metabolized. The determined radioactive recovery is thus the overall recovery of all metabolites. Traditionally, extraction of nucleotides from cells has been executed with strong acids like perchloric acid. Although extraction yields with perchloric acid are good (for example, 90% for cladribine (2CdA)-TP (Reichelova, Albertioni, & Liliemark, 1996)), analytes can be unstable at low pH and the perchloric acid and low pH can deteriorate peakshapes and reduce retention times in IP-LC (Au, Su, & Wientjes, 1989). Moreover, involatile acid salts pollute the MS. Although some MS methods still use perchloric acid, the vast majority of the described methods use MS-friendly 6080% methanol (MeOH) or acetonitrile (ACN) with or without buffers and metal chelators. Mixtures of organic solvents and water are hypotonic and disrupt cell membranes, thus effectively extracting nucleotides. For adenosine triphosphate (ATP), for example, recoveries of 80104% were observed for several waterorganic solvent combinations (Palmer & Cox, 1994; Grob et al., 2003). Furthermore, the analytes are stable in MeOH lysate during lysis (Kimball & Rabinowitz, 2006), and when stored at 708C (Thevanayagam, Jayewardene, & Gambertoglio, 2000). Finally, organic solvent lysates can easily be concentrated by evaporation. Solutions containing 6080% MeOH or ACN thus fulll all requirements of an MS-compatible extraction solvent.
FIGURE 5. Schematic representation of direct and indirect methods for the analysis of nucleotides (N, nucleoside; NMP, NDP, and NDP, nucleoside mono-, di-, and triphosphate).
selection and isolation of the matrix is pivotal for obtaining relevant results.
B. Cell Isolation
Most methods use PBMCs as matrix. PBMCs can be isolated from relatively large volumes (525 mL) of whole blood by density gradient centrifugation. Whole blood is layered over a density gradient such as Ficoll-paque or directly drawn into commercially available cell preparation tubes containing a density gradient. Although rich in PBMCs, other cell types can be present as well. Special attention should be given to erythrocyte contamination. Samples with a pink or red color, an indication of erythrocyte contamination, showed a larger and more variable matrix effect (Shi et al., 2002). Endogenous nucleotides originating from the red blood cells most likely caused ionization suppression observed as the changed matrix effect. More importantly, some analogs are phosphorylated in erythrocytes. Contamination with erythrocytes containing phosphorylated nucleoside analogs thus severely alters the amount of analyte present in a sample. Indeed, deliberately increasing the number of erythrocytes in a sample increased the amount of PMPA-DP. Clinical samples contained up to sevenfold more PMPA-DP with erythrocyte contamination compared to the same sample without erythrocyte contamination (Durand-Gasselin et al., 2007). To prevent contamination an extra erythrocyte lysis step can be executed (Ray et al., 2005; Durand-Gasselin et al., 2007). Because cells are still intact during isolation, ongoing analyte metabolism ex vivo must be prevented. Isolation should therefore be performed as quickly as possible, and preferably on ice (Pruvost et al., 2001). Finally, to express the determined concentration per cell, the number of cells isolated should be quantitated. The most straightforward approach is to count the number of cells using a hemocytometer, microscope, or ow cytometry, but protein (Bradford, 1976) and DNA (Benech et al., 2004) determinations have also proven to be useful. Concluding, the cell isolation should remove interfering cells, after which the cells of interest can be processed further.
324
III. BIOANALYTICAL METHODS

In this section the bioanalytical methods divided into indirect and direct methods are described. Detailed information of each method is given in the accompanying tables. The lower limits of quantitation (LLQs) of the methods are alternately expressed as amount on column, concentration in lysate, amount per million PBMCs, or amount per sample. Although desirable for a comparison, re-expressing the LLQs as a single equivalent unity is not possible because the necessary information is not always available. With the extra information in the footnotes a comparison will, however, be possible in most cases.
A. Indirect Methods 1. Sample Pretreatment

Indirect methods have been applied to determine the total as well as the speciated nucleotide concentrations. Fractionation required for the speciation has been performed on WAX (NH2)and SAX (QMA)-solid-phase extraction (SPE) columns, using high salt concentrations for elution. Proper separation of the
&
nucleotides is pivotal for their correct quantication but cannot be conrmed for each sample. As a consequence, the fractionation must be reproducible and well characterized to prevent coelution of the nucleotides (Robbins, Waibel, & Fridland, 1996; Robbins et al., 2007). De-phosphorylation has mainly been performed using acid or alkaline phosphatase. Methods for the quantication of phosphorylated 6-mercaptopurine riboside (6-MPR) and 6-methylmercaptopurine riboside (6-MMPR) have, however, used hydrolysis (1008C, 60 min), degrading the nucleotide analogs to their base and a derivative thereof (Dervieux & Boulieu, 1998). Finally, an SPE purication step is performed to remove salts and enzymes from the crude de-phosphorylated samples. Although the fractionation and de-phosphorylation are critical in the analytical process, most methods do not use an internal standard during these steps. Moreover, systematic differences have been observed between hydrolysis techniques (Shipkova et al., 2003), and between indirect and direct methods (Vikingsson et al., 2009). It is evident that indirect methods are labor intensive because samples have to be fractionated, de-phosphorylated, and puried before the actual analysis. Also, each fraction has to be analyzed separately, requiring extra instrument time. Moreover, the extensive and relatively uncontrolled sample pretreatment is a potential source of errors. The nal analysis is, however, relatively simple, fast, and selective. Concluding, de-phosphorylation simplies the nal analysis but increases analysis time and introduces potential pitfalls in sample pretreatment. Indirect methods have also been used to analyze modied nucleotides that are naturally present in RNA or are formed after damage to DNA. Since these nucleotides are incorporated in nucleic acids, liberation from these nucleic acids is rst required. After enzymatic or thermal pretreatment the resultant nucleosides, bases, or in some cases nucleotides can be quantitated (Andrews, Vouros, & Harsch, 1999). As opposed to many applications described in this review, speciation between nucleotides is not required because only MPs are present. Moreover, many of the analytical methods for modied nucleotides are more qualitative in nature (Banoub et al., 2005; Singh & Farmer, 2006).
natural abundance. For 14C determinations the sample should rst be converted to elemental carbon to produce a stable signal and reduce memory effects. After negative ionization of the sample, the ions are accelerated, recharged to positive, and again accelerated. Finally, unwanted ions are removed using electric and magnetic elds, followed by detection with nuclear detection techniques (Hellborg & Skog, 2008). The method can be used to specically detect 14C-containing drugs and their metabolites at very low concentrations. All structural information is, however, lost because samples need to be graphitized before analysis. The method has been used to measure 14 C-labeled AZT in PBMCs (Vuong et al., 2008), but because no separation was performed, the found concentrations were the sum of all AZT metabolites in the PBMCs, including those incorporated in nucleic acids. After isolation and de-phosphorylation of AZT-TP the methodology could, however, be used for the specic quantication of 14C-labeled AZT-TP in PBMCs (Chen et al., 2009). The results obtained with this AMS method were in very good agreement with conventional indirect LCMS/MS (King et al., 2006a), but the AMS method approximately 30,000-fold more sensitive, enabling 14C detection in the atto- to zeptogram range (1018 1021 g). Therefore, AMS can be used to reduce sample size. This can be benecial in, for example, pediatric and animal studies, or studies involving tissue biopsy. Moreover, AMS can be used to answer specic research questions requiring extreme sensitivity. The limited availability and high costs, however, are serious drawbacks of AMS.
B. Direct Methods
In this section methods for the direct determination of nucleotide analogs in cells are described. The validation parameters that should be assessed for these methods include the stability during cell preparation and the effect of the number of cells and have been described in detail by Becher et al. (2003a).
1. Sample Pretreatment
After isolation and lysis of cells, further sample pretreatment is often required to allow direct analysis. When cells are lysed with an organic solvent the lysate can be concentrated by evaporation of the organic phase. When cells are lysed with an acid the lysate should be neutralized. More elaborate sample pretreatment has also been performed. For example, IP-SPE (Claire, 2000), WAX-SPE (Crauste et al., 2009; Pucci et al., 2009), and regular SPE (Bartley, Begley, & Clark, 2007) have been applied for the analysis of FTCTP, ara-CTP, and PMPA-DP, respectively. Using immobilized metal afnity chromatography (IMAC), Tuytten et al. (2004) were able to trap nucleotides on a column. The electron-donating phosphate groups of the nucleotides complexated with immobilized Fe3 ions on the column and could subsequently be eluted using a basic phosphate buffer. In this pilot study the methodology was used for on-line sample pretreatment in a column switching set-up. A 1 mL loop lled with the eluted nucleotides was injected onto an analytical reversed phase column and separated using the IP agent N,N-dimethylhexylamine (N,N-DMHA). Others successfully used an on-line WAX trap-column followed by IP chromatography for the quantication of several nucleotide analogs in PBMCs (Kuklenyik et al., 2008). The nucleotides were loaded and trapped on the WAX column using a neutral ammonium acetate solution. The analytes were then eluted onto an analytical column with a phosphate buffer of
325
2. Separation and Detection

a. LCMS. After de-phosphorylation the nucleosides have been separated on normal (e.g., Cahours et al., 2001; Yuen et al., 2004) and polar-modied (e.g., King et al., 2006b; Moyle et al., 2009) C18 columns using isocratic elution or straightforward gradients of ACN or MeOH in formate or acetate buffers. The nucleosides are most effectively ionized in the positive ionization mode, as illustrated by the fact that all methods applied the positive ion mode for detection, except one method for AZT (Williams et al., 2003). The developed methods have been used to support relatively large clinical studies (Anderson et al., 2003; Moore et al., 2007). These studies showed that PBMCs from women contained higher AZT-TP and 3TC-TP levels than men and that AZT-TP and 3TC-TP levels were associated with treatment effect. The indirect methods for the quantication of intracellular nucleotide analogs are summarized in Table 1. b. Accelerator mass spectrometry. Accelerator mass spectrometry (AMS) is an ultra sensitive technique to specically measure isotopes such as 14C, a carbon isotope with a very low
&
JANSEN ET AL.
TABLE 1. Indirect methods for the quantication of intracellular nucleotide analogs
Mean of 28 106 cells/sample. Usually 16 106 cells/sample. c 10 106 cells/400 mL. d Approximately 1.1 109 cells/sample (100 mL erythrocytes). e 10 106 cells/sample. f 5 106 cells/sample. g Mean of 30 106 cells/sample. h 100 mL erythrocytes/sample. i 10 106 cells/sample. j Minimum yield of 36 106 cells/sample. k Using 500 mL PBMC extract from approximately 8 mL whole blood.
b
pH 10, followed by their separation using DMHA. The approach resulted in peaks that had an area approximately 20 times as high as the same technique without the WAX precolumn. Special sample pretreatments have also been performed for the challenging analysis of AZT nucleotides, which is discussed in detail in Section IVC.
2. Separation and Detection

a. Reversed phase chromatography. Nucleoside TPs, and to a lesser extent nucleoside DPs, are only minimally retained, and separated from other nucleotides, on reversed phase stationary
326
phases. The retention of ATP using an ACN gradient in 10 mM ammonium acetate (pH 10) on several reversed phase materials is shown in Figure 6. This minimal retention is only obtained when using a highly aqueous mobile phase, which results in poor electrospray ionization (ESI)-MS sensitivity (Wang et al., 2009). Moreover, many interferences in biomatrices also elute early and compromise the specicity and selectivity of a method. Although nucleoside DPs and TPs require IP agents for retention on C18 columns, nucleoside MPs often show sufcient retention to allow accurate and precise bioanalysis. Like for their analysis in plasma (see, for example, PMPA in Nirogi et al., 2008), simple gradients starting with a low
&
percentage of MeOH or ACN in ammonium acetate or formate buffers have been applied (Lynch, Eisenberg, & Kernan, 2001; Kim et al., 2004, 2005; Bousquet et al., 2009). Although these methods are relatively uncomplicated, the lack of retention for DPs and TPs restricts their applicability. The methods using reversed phase chromatography are summarized in Table 2. b. Ion-pairing chromatography. Both low amounts of involatile tetraalkylammonium salts and relatively volatile alkylamines have been used as IP agents to separate nucleotide analogs. Here we describe the methods using these IP agents. Tetrabutylammonium salts. Witters et al. were the rst to explore the use of tetrabutylammonium (TBA) in combination with MS. Despite the relative involatility of TBA they were able to separate and detect endogenous cyclic nucleotides without a loss in sensitivity over the tested period of several days (Witters et al., 1997).
The rst method using TBA for the analysis of therapeutic nucleotide analogs was used for the quantication of FTC-TP (Claire, 2000). TBA was selected because it was effective in relatively low concentrations. By using 1.0 mm internal diameter columns with a ow of 50 mL/min and by using a divert valve, the amount of involatile mobile phase constituents (2 mM TBA and 10 mM ammonium phosphate) delivered to the MS was kept to a minimum. Furthermore, a ow of 20% MeOH was continuously pumped through the source of the MS at a ow of 10 mL/min to prevent clogging of the ESI capillary. No residue was detected on the ESI capillary, but despite the precautions a brownish-red viscous liquid was observed on the sample cone. The same method was used later for the simultaneous quantication of 3TC-TP, carbovir (CBV)-TP, and PMPA-DP (Hawkins et al., 2005). Because the MS signal was strongly related to the amount of salts in the mobile phase, lower amounts have also been tested (Wang et al., 2004). The amount of TBA could be decreased by increasing the amount of TBA in the injection solvent up to 200 mM. Later, the method was made
FIGURE 6. Retention of nucleotides on reversed phase stationary phases. LCESI-MS chromatograms of ATP retained on various reversed-phase HPLC columns: (A) Zorbax SB-C8 (150 mm 2.1 mm, 4 mm); (B) Symmetry Shield TM RP8 (150 mm 2.1 mm, 3.5 mm); (C) Synergi Max-RP 80A (150 mm 2.0 mm, 4 mm); (D) Synergi Hydro-RP (150 mm 2.0 mm, 4 mm); and (E) Alltima HP HILIC (150 mm 2.0 mm, 3 mm). Mobile phases consisted of 10 mM ammonium acetate (pH 10) (A) and ACN. (B) Gradient elution was performed at 300 mL/min with 10% B at 02 min, increasing to 50% B from 2 to 8 min. (Reprinted from Wang et al., 2009, Copyright 2009, with permission from Elsevier.)
327
&
JANSEN ET AL.
TABLE 2. Direct methods for the quantication of intracellular nucleotide analogs using reversed phase chromatography
suitable for the separation of MPs, DPs, and TPs by introducing an ACN gradient, as shown for PMEA and its phosphorylated anabolites (Ray et al., 2004). Similar methods were applied for several other nucleotide analogs (Ray et al., 2005, 2008; BorrotoEsoda et al., 2006; Delaney et al., 2006; Reiser et al., 2008). Despite all improvements, the use of TBA in the injection solvent still caused signicant ion suppression during the rst 9 10 min of the analysis (Fig. 7). After this time most of the bulk of the TBA is most likely eluted from the column. Moreover, phosphate clusters and their ammoniated adducts were observed that had masses similar to those of the analytes and could thus cause interferences (Vela et al., 2007). Concluding, the use of TBA in combination with phosphates has some important drawbacks such as pollution of the MS source and ion suppression. Notwithstanding these drawbacks, many applications have been developed. Only one of these methods was extensively validated. All methods using TBA are summarized in Table 3. Alkylamines. Trialkylamines are generally more volatile than tetraalkylammonium salts, which make them more suitable for a combination with MS detection. Therefore, IP agents such
FIGURE 7. Ion suppression caused by the ion-pairing agent TBA. Ion suppression during chromatography was observed by infusing 10 mL/min 1 mM solution of PMEA-DP postcolumn during a normal ion-pairing LC gradient. The trace of PMEA-DP signal (C) was then overlaid on the LCMS/MS chromatogram of an injection of 200 nM (2 pmol on column) PMEA (A), PMEA-MP (B), PMEA-DP (C), and PMPA-DP (D) spiked into cellular matrices from 1 105 Hep G2 cells and separated on an Xterra column to illustrate the position of ion suppression relative to the elution of analytes. Similar ion suppression proles were observed during infusion of all analytes and under all column and gradient conditions studied. (Adapted from Vela et al., 2007, Copyright 2007, with permission from Elsevier.)
as tripropylamine (Gaus et al., 1997; Fung et al., 2001) and triethylamine (Apffel et al., 1997) have been used to analyze nucleotides in combination with MS detection. However, all but a few methods described in this section use N,N- or 1,5-DMHA as an IP agent. The type of DMHA used is not always specied in the referred articles. If the type is specied in an article, or if it can be assumed based on other publications it is specied in this review. If the type of DMHA is not specied in a referred publication it is referred to as DMHA in this review. The different types, however, possess different chromatographic and ion-suppression properties. The rst reported use of DMHA for the LCMS analysis of nucleotide analogs dates back to 1996, when Smyrnis et al. (1996) applied it to penciclovir (PCV) and its nucleotides. Little was, however, reported on the tested matrix or sensitivity. Later, DMHA was employed for the identication (Auriola et al., 1997) and quantication (Monkkonen et al., 2000) of bisphosphonate nucleotides. Bisphosphonates are synthetic pyrophosphate analogs that are used in the treatment of metabolic bone disorders. The nucleotide metabolites of these drugs are formed by the addition of a nucleoside to the bisphosphonate (Fig. 8), rather than by the addition of phosphates to a nucleoside, which is the normal route for nucleotide formation (Fig. 4). Tuytten et al. (2002) systematically assessed the effect of N,N-DMHA on the separation and MS detection of 12 endogenous nucleosides and nucleotides and concluded that a gradient of MeOH in 5 mM aqueous N,N-DMHA resulted in good HPLC selectivity and MS detection. Fung et al. (2001) were the rst to fully describe a method for the determination of a nucleoside analog and its nucleotides. They separated ABC and four of its nucleotides using 20 mM N,N-DMHA in an acetate buffer. The same method was later used for the quantication of a prodrug of PMPA and its metabolites (Lynch, Eisenberg, & Kernan, 2001). A similar method, using 1,5-DMHA with a formate buffer (pH 10.5), was developed for the anabolites of d4T (Pruvost et al., 2001). The method, however, suffered from a long cycle time (26 min) and poor robustness; the column had to be replaced every 2 weeks because of the high pH. The robustness was later improved by using a more pH stable column. The improved method was adapted for the quantication of several other TPs (Becher et al., 2002a; Pruvost et al., 2005) and was successfully used to support clinical studies (Becher et al., 2003b, 2004). These studies revealed signicant interpatient variability in the d4T-TP and ddA-TP levels and showed that treatment with AZT could result in d4T-TP formation. This might explain the cross resistance observed between AZT and d4T. By increasing the amount of ACN in the mobile phase Lee et al. (2006) were able to quantitate clevudine (L-FMAU)-TP. Interfering peaks, however, prevented its application to AZT-TP (see Section IVC). AZT-TP could, on the other hand, be analyzed with the same method after an immunoafnity extraction (Becher
328
&
TABLE 3. Direct methods for the quantication of intracellular nucleotide analogs using tetrabutylammonium salts as ion-pairing agent
Approximately 10 106 cells/sample. Approximately 29 106 cells/sample. c 0.1 106 cells/sample. d 15 106 cells/sample. e 10 mL injection containing 0.1 106 cells. f 23 106 cells/sample.
b
et al., 2002b), or by selecting a less abundant, but specic transition for AZT-TP (Compain et al., 2007). The resultant loss in sensitivity was unconventionally corrected for by spiking all samples with a small amount of analyte. The addition of the analyte lifted small chromatographic peaks above the background and allowed their quantication after correction for the amount of analyte added. Moreover, the run time could be reduced to 9.5 min by using a shorter column (150 mm vs. 50 mm) with a faster gradient. A slightly adjusted version of this method was used to determine AZT and 3TC metabolite levels in newborns (Durand-Gasselin et al., 2008). The study showed that the TP levels were higher in
FIGURE 8. Clodronate and its nucleotide metabolite.
the rst 15 days of life and that 3TC-TP could be measured in children from 3TC-treated mothers up to 1 week after birth. Using the faster gradient, the method was again validated for AZT-MP, AZT-TP, 3TC-TP, PMPA-DP, and CBV-TP (DurandGasselin et al., 2007; Pruvost et al., 2008). The concentration of the IP agents was constant in most methods, except for the method described by HernandezSantiago et al. (2004) who used a gradient with a decreasing N,N-DMHA concentration for the quantication of N-hydroxycytidine (NHC)-TP. Others used a decreasing hexylamine concentration and pH in combination with increasing MeOH (Coulier et al., 2006, 2008). Although 1,5- and N,N-DMHA are more volatile than TBA and phosphate, many presented methods still require ushing of the ESI source to prevent clogging and a loss of signal (Hernandez-Santiago et al., 2004; Pruvost et al., 2008). Pollution could be repressed by reducing the column diameter and ow. Compared to the methods using TBA, relatively large column diameters are used (2.02.1 mm). The use of microbore columns was investigated by Becher et al. (2002a), but no improvement was achieved. Cai et al. on the other hand could reduce the N,NDMHA concentration to 10 and 5 mM by changing the column diameter to 1.0 and 0.5 mm, respectively. Concomitantly, the MS signal increased 5- to 10-fold (Cai, Song, & Yang, 2002). In contrast to previous reports (Lai, Wang, & Cai, 2008), a clear relation between column diameter and applied N,N-DMHA
329
&
JANSEN ET AL.
concentration cannot be inferred from the methods described in this review. Even though both N,N- and 1,5-DMHA have successfully been used by some groups, others using N,N-DMHA were less successful due to ion suppression, source pollution, strong binding of the IP agent to the analytical system, carryover, and reproducibility problems (Melendez et al., 2006; King et al., 2006a; Pucci et al., 2009). Problems with 1,5-DMHA have, on the other hand, not been reported. Despite the fact that DMHA is more volatile than TBA, and phosphate is not used, at least N,N-DMHA has still been associated with source pollution and ion suppression. Nevertheless, the method has proven its value in many applications, making it one of the methods of choice. In view of ion suppression and source pollution 1,5-DMHA seems to be preferred over N,N-DMHA. The methods using alkylamines as IP agents for the separation of nucleotide analogs are summarized in Table 4. Concluding, by reducing the amount of IP agents and salts delivered to the MS, source pollution and ion suppression have been reduced to an acceptable level. Several IP methods have been developed for the quantication of nucleotide analogs, and their use has been shown in numerous applications. Still, pollution of the source remains a major point of care.
sides can be regarded as a drawback. The described WAX methods are summarized in Table 5. d. Porous graphitic carbon chromatography. Although PGC only consists of carbon sheets, polar and ionic compounds are retained on its surface (Gu & Lim, 1990). Polar or polarizable analytes cause a charge-induced dipole at the graphite surface resulting in retention of negatively charged analytes like nucleotides. Charged analytes can be retained on PGC without the need of IP agents and can subsequently be eluted from the column without the need of high salt concentrations. PGC is thus very suitable for the analysis of nucleotides with MS detection. Moreover, it has shown a remarkable selectivity for structurally similar compounds, which is most likely caused by differences in analyte contact area with the planar carbon sheets (Xia et al., 2006). PGC has therefore been used to separate cytarabine (ara-C) and its MP from their endogenous stereoisomers cytidine and its MP (Gouy et al., 2006). Moreover, Xing et al. (2004) have explored the use of PGC for the simultaneous analysis of several endogenous nucleosides and their mono-, di-, and TPs. Various buffers were tested for their ability to separate the analytes and for their effect on MS signal. Addition of diethylamine (DEA) was required to reduce peak tailing. Finally, separation of 16 nucleosides and nucleotides was achieved in just 15 min using a gradient of ACN in water with 50 mM ammonium acetate and 0.1% DEA. More recently, PGC was used in combination with the IP agent hexylamine to separate ara-CMP, -DP, and -TP from their endogenous stereoisomers cytidine-MP, -DP, and -TP (Crauste et al., 2009). A chromatogram showing the separation of CTP and ara-CTP using this method is shown in Figure 10. Again, however, as much as 0.4% DEAwas required to reduce peak tailing. Using an ammonium bicarbonate gradient (025 mM) in 15% ACN, Jansen et al. were able to separate dFdC, dFdU, and their MP, DP, and TP, without the need of DEA addition, in 15 min. Strict control of pH and redox state of the column was required to maintain this separation during multiple runs (Jansen et al., 2009a). The developed method was later validated for quantication in PBMCs (Jansen et al., 2009b). A typical chromatogram is shown in Figure 11. Concluding, exceptional separations have been achieved using PGC. Because of its unconventional selectivity, PGC is especially suitable for the separation of isomers or isobaric nucleotides. The unpredictable and variable retention properties are, however, an important drawback. The methods using PGC for the quantitative determination of nucleotide analogs are presented in Table 6. e. Capillary electrophoresis. Capillary electrophoresis can be used to separate ionic analytes with a high resolution, which makes it very suitable for the analysis of nucleotides (Uhrova, Deyl, & Suchanek, 1996; McKeown, Shaw, & Barrett, 2001; Zinellu et al., 2008). A combination with MS makes the technique even more powerful, as shown by many applications for nucleotide analysis (Willems et al., 2005). Hyphenation of CE with MS has, however, initially posed instrumental problems mainly caused by the separated electrical circuits required for both techniques (Banks, 1997). Moreover, the electrophoresis run buffer should be compatible with MS detection. Volatile buffers such as ammonium bicarbonate and acetate, however,
c. Anion-exchange chromatography. Although anion exchange has traditionally been the method of choice to analyze nucleotides, the use of involatile competing ions for elution hampered hyphenation with MS. Shi et al. (2002), however, used an alternative elution mechanism. By applying a pH gradient (pH 610.5) to a WAX column they changed the charge of the basic functional groups (pKa $ 8) of the column. As a consequence the capacity of the column decreased at a higher pH, thereby eluting the anionic nucleotides. Therefore, as opposed to classical anion-exchange methods, this method did not require high concentrations of involatile competing ions for elution, making it directly applicable to MS detection. Separation between MPs, DPs, and TPs was achieved within only 2 min by simultaneously applying an inverse ammonium acetate gradient (101 mM). A typical chromatogram is shown in Figure 9. The method was fully validated for the quantication of dexelvucitabine (D-D4FC)-TP in PBMCs. A similar method was later applied to its enantiomer elvucitabine (D4FC)-TP (Nguyen et al., 2006). Methods with such short runtimes and limited separation are, however, sensitive to matrix effects caused by other sample constituents. A less steep, 13-min gradient was used for the quantication of dFdC-TP in canine melanoma cells (Freise & Martin-Jimenez, 2006). A similar gradient, separating dFdC-MP, -DP, and -TP in 10 min, was used for the validated quantication of dFdC-TP in PBMCs (Veltkamp et al., 2006) and was later used for the quantication of its de-aminated metabolite diuoro deoxyuridine (dFdU)-TP as well (Veltkamp et al., 2008). Simultaneous quantication of MPs, DPs, and TPs was also feasible with this gradient, as shown for 2CdA-MP, -DP, and -TP in cells and culture medium (Jansen et al., 2007). If short runtimes are not of pivotal importance, the longer gradients are the methods of choice. The different gradients result in slight changes in selectivity. These WAX methods are robust, very sensitive, and MS-friendly, but the lack of retention of nucleo330
&
TABLE 4. Direct methods for the quantication of intracellular nucleotide analogs using alkylamines as ion-pairing agent
10 106 cells/mL. 50 106 cells/mL. c Determined in water. d Around 14 106 cells/sample. e Approximately 710 106 cells/sample. f Approximately 10 106 cells/sample. g 10 mL injection of approximately 100 mL containing 2 106 cells. h 25 106 cells/sample. i LOD; 25 mL injection of 100 mL lysate containing 1 106.
b
resulted in good CE separation and MS signal in the analysis of endogenous nucleotides (Feng & Zhu, 2006; Friedecky et al., 2006; Soga et al., 2007). Cai et al. used CEMS for the simultaneous quantication of intracellular ABC, CBV-MP, CBV-DP, and CBV-TP (Cai, Fung, & Sinhababu, 2003). The nucleotides could be separated using an ammonium acetate
buffer on bare fused silica. Similar methods were used to study the metabolism of 3TC in HepG2 cells (Liu et al., 2005) and d4T- and ddA-MP and -TP in PBMCs (Bezy et al., 2006). Bezy et al., however, applied a coaxial sheath liquid composed of decauoroheptanoic acid in 2,2,2,-triuoroethanol to improve sensitivity and spray stability.
331
&
JANSEN ET AL.
FIGURE 9. Separation of nucleotides using weak anion-exchange chromatography. Overlaid chromatogram of D-D4FC-MP, -DP, and -TP (100 nM each) in human PBMC extract. (Reprinted from Shi et al., 2002, Copyright 2002, with permission from Wiley-Blackwell.)
The high resolution of CE makes CEMS a good alternative to conventional LCMS. Table 7 summarizes the methods using CEMS for nucleotide analog quantication. f. Hydrophilic interaction chromatography. In hydrophilic interaction chromatography (HILIC) polar columns are used in combination with mobile phases rich in organic solvents. Polar analytes are eluted from the column by increasing the water content of the mobile phase. Because the applied mobile phases are highly volatile, hyphenation with MS is favorable (Hsieh, 2008).
Hydrophilic interaction chromatography (HILIC)MS detection was applied to endogenous nucleotides by Bajad et al. (2006). Similarly, silica hydride was used in the normal phase mode (Pesek et al., 2009). Currently, only one HILICMS method for the determination of a nucleotide analog has been described (Pucci et al., 2009). The method, using an aminopropyl column, was preferred over several other HILIC, PGC, and IP methods because of its sensitivity and selectivity. Still, a 30 min run time was required. The method is summarized in Table 8. The value of HILIC for the analysis of nucleotide analogs should be conrmed in future applications.
TABLE 5. Direct methods for the quantication of intracellular nucleotide analogs using weak anion-exchange chromatography
3 106 cells/sample. Approximately 0.66 106 cells/mL. c 3.8 106 cells/sample. d Approximately 0.6 106 cells/mL.
b
332
&
FIGURE 10. Separation of nucleotides using PGC with interference
from endogenous nucleotides. Ion chromatogram (transition of m/z 482 159) of cell extract spiked with araCTP (0.3 mg/mL) showing the large CTP interference. (Reprinted from Crauste et al., 2009, Copyright 2009, with permission from Elsevier.)
g. MALDI-TOF. Only one article describes the use of MALDITOF MS for the quantication of nucleotide analogs (van Kampen et al., 2004). Using anthranilic acid mixed with nicotinic acid as matrix, the investigators were able to detect AZT-TP, as well as several endogenous TPs, in amounts as low as 0.5 fmol per sample. Endogenous nucleotides (ATP and deoxyguanosine TP (dGTP)) having the same mass as AZT-TP could, however, only be distinguished using postsource decay, which resulted in less sensitivity (see Section IVC). Since no validation of the method was performed, the article should be regarded as explorative. The MALDI-TOF method used for the quantication of AZT-TP is summarized in Table 9.
FIGURE 11. Separation of nucleotides on PGC with interferences
IV. MASS SPECTROMETRY A. Ionization

Intuitively the anionic nucleotides should be detected in the negative ion mode. Early LCMS methods for the quantication of nucleotides therefore started method development with the negative ion mode, but surprisingly nally used the positive ion mode (Claire, 2000; Fung et al., 2001). Fung et al. (2001) argued that N,N-DMHA could mask the negative charges on the phosphate moiety, but Shi et al. (2002) also used the positive mode, without any IP agents. Table 10 shows that most nucleotide analogs can be detected in the positive as well as in the negative ion mode, even at a high pH (10.5). No increase in signal intensity was observed upon lowering the pH of the mobile phase by Vela et al. (2007), whereas Shi et al. even noticed an increase in analyte signal in mobile phase B (pH 10.5) over mobile phase A (pH 6.0). Some used postcolumn addition of formic or acetic acid (Nguyen et al., 2006), but for others it did not change (Jansen et al., 2007), or even decreased sensitivity (Claire, 2000; Shi et al., 2002). Finally, dFdU was not detectable in the positive ion mode whereas its nucleotides, bearing anionic phosphate groups, were (Jansen et al., 2009a). The source of these contradictory results are not clear but might be explained by several factors such as mobile phase composition other than pH (e.g., IP agents) or differences in the analytes. Although some reports are contraMass Spectrometry Reviews DOI 10.1002/mas
caused by other nucleotide analogs. Chromatograms of PBMC lysate spiked with dFdC (4.29 nM), dFdU (42.1 nM), dFdCMP (29.0 nM), dFdUMP (25.4 nM), dFdCDP (31.4 nM), dFdUDP (43.2 nM), dFdCTP (36.9 nM), and dFdUTP (52.7 nM). The arrows indicate the signals of dFdC nucleotides (containing a naturally occurring 13C- or 15N-isotope) in the mass transitions of the dFdU nucleotides. (Reprinted from Jansen et al., 2009, Copyright 2009, with permission from Wiley-Blackwell.).
dictory most publications show that a high pH has a positive effect on the sensitivity. It is unexpected that analytes with such acidic groups show more sensitivity in basic environments using the positive ion mode. Good MS responses using the positive ionization mode in combination with basic environments have previously been reported, and gas-phase protonation of the analytes by NH4 has been proposed as a possible underlying mechanism (Peng & Farkas, 2008). This does, however, not explain the effective protonation of the acidic nucleotides. Interestingly, calculations of the gas-phase proton afnities showed that the most likely site of protonation of some uridine and thymidine MPs is the phosphate group (Green-Church & Limbach, 2000). Others calculated that the addition of a negatively charged phosphate group increases the proton afnity of the nearby base (Pan, Verhoeven, & Lee, 2005). Still, nucleotides containing a base with an external amino function (A, G, and C) are protonated more effectively than those
333
&
JANSEN ET AL.
TABLE 6. Direct methods for the quantication of intracellular nucleotide analogs using porous graphitic carbon chromatography
Assuming 4 106 cells/sample.
that do not possess this function (T and U) (Bezy et al., 2006; Pruvost et al., 2008). These results are in good agreement with the determined and calculated gas-phase proton afnities of (deoxy)nucleoside MPs (A > G > C T/U) (Green-Church & Limbach, 2000). Thymine and uracil containing nucleosides and nucleotides can, however, be detected in the negative ion mode, with the deprotonated base as a common fragment (Table 10). Thus, most nucleotides are efciently ionized in a wide pH range using both the negative and the positive ionization mode. The LCMS detection of nucleotides can also be improved by derivatization. Derivatization of dGMP and adducted variants thereof with hexamethyleneimine resulted in a 3- to 10-fold increase in detection limits using the positive ion mode. This approach has, however, never been used for the quantication of nucleotide analogs.
B. Ionization Suppression
Ion suppression is the reduced ionization of analytes. It can be caused by compounds that are present during ionization which interfere with droplet formation or analyte ionization. Ion-pairing (IP) agents not only pollute the MS but are also a notorious cause of ion suppression. Ion suppression has clearly been observed with both TBA and N,N-DMHA (Witters et al., 1997; Claire, 2000; Tuytten et al., 2002). An example of ion suppression caused by TBA is shown in Figure 7. Both agents, however, showed an increase in signal intensity at low concentrations. This initial increase is most likely caused by lowering of the surface tension, thereby facilitating the spray process. The concentration at which ion suppression starts is
differently reported in publications and experiments, ranging from 1 mM DMHAformate (Monkkonen et al., 2000) to 5 mM (using LC system) and >20 mM (ow injections) N,N-DMHA (Tuytten et al., 2002) and 0.05 mM TBAbromide (Witters et al., 1997). Fung et al. (2001) still detected a higher [ATP H] signal with 10 mM N,N-DMHA compared to no N,N-DMHA. It is therefore unclear whether the IP agents actually cause ion suppression at the concentrations used in the described methods. Most likely the ion-suppressing effect of IP agents is inuenced by other factors, like pH and analyte concentration. Ion suppression involving 1,5-DMHA has not been reported, but a comparison between N,N-DMHA and 1,5-DMHA regarding ion suppression has never been published. N,N-dimethylhexylamine (N,N-DMHA) has also caused the formation of adducts (Tuytten et al., 2002). These adducts have been used as quantifying ions by some (Cai, Song, & Yang, 2002; Qian, Cai, & Yang, 2004), whereas others, using 1,5-DMHA, found that the product ions of these adducts were not sufciently specic because only the adduct was lost upon fragmentation (Pruvost et al., 2008). Unless the intensity of the non-adduct is unacceptable, the use of adducts should be avoided in quantitative bioanalysis. In addition to mobile phase components, components originating from the samples that co-elute with the analytes, such as endogenous nucleotides, can also have an important contribution to the ion suppression. A principal factor in this matrix effect is the number of cells present in a sample. Because the number of cells can vary between samples, the number of cells should either be kept constant (Lee et al., 2006; Robbins et al., 2007) or a range should be validated (Becher et al., 2003a; Veltkamp et al., 2006).
TABLE 7. Direct methods for the quantication of intracellular nucleotide analogs using capillary electrophoresis
50 106 cells/mL. 2.05 106 cells/mL. c 20 106 cells/mL.

b
334
&
TABLE 8. Direct methods for the quantication of intracellular nucleotide analogs using hydrophilic interaction chromatography
A, ammonium; A, mobile phase A; Ac, acetate; ACN, acetonitrile; ACP, acid phosphatase; ALP, alkaline phosphatase; AMS, accelerator mass spectrometry; B, mobile phase B; BCa, bicarbonate; DMHA, dimethylhexylamine; DP, de-phosphorylation; DS, de-salting; EDTA, ethylenediamine tetraacetic acid; EGTA, ethyleneglycol tetraacetic acid; F, fractionation; Fo, formate; H-Ac, acetic acid; H-Fo, formic acid; ID, internal diameter; IT, ion trap; LIT, linear ion trap; LLQ, lower limit of quantication; LOD, limit of detection; MALDI-TOF, matrix-assisted light desorptiontime of ight; MeOH, methanol; NS, not specied; OD, outer diameter; OH, hydroxide; PBMC, peripheral blood mononuclear cell; PCA, perchloric acid; Ph, phosphate; Q, single quadrupole; QqQ, triple quadrupole; SPE, solid phase extraction; TBA, tetrabutylammonium; TRIS, tris(hydroxymethyl)aminomethane; W, water; WAX, weak anion exchange.
In short, IP agents can cause ion suppression but also ion enhancement, mainly depending on their concentration. The number of cells in a sample has a profound impact on the ionization of the analytes and should be validated thoroughly.
2. Other Nucleoside Analogs

In addition to endogenous nucleosides and nucleotides, other nucleoside and nucleotide analogs can also interfere with MS detection. For example, apricitabine (ATC) and 3TC are isomers and produce the same daughter ions when subjected to tandem mass spectrometry (Holdich, Shiveley, & Sawyer, 2007). Moreover, dFdC (263 Da) and its de-aminated metabolite dFdU (264 Da) differ only 1 Da in molecular weight. A dFdC molecule containing a naturally occurring 13C- or 15N-isotope (approximately 11%) will therefore have the same nominal mass as dFdU, resulting in a signal in the dFdU mass transition window unless a high-resolution mass analyzer (such as an orbitrap or Fourier transform ion cyclotron resonance (FTICR) mass analyzer) is used. Correspondingly dFdC-MP, -DP, and -TP give signals in the transition windows of dFdU-MP, -DP, and -TP (Fig. 11) (Jansen et al., 2009a). Likewise, ara-C (243 Da) and ara-uridine (ara-U; 244 Da) and their nucleotides differ only 1 Da in molecular weight. Furthermore, ara-C and ara-U are also stereoisomers of cytosine and uridine (Gouy et al., 2006; Crauste et al., 2009). The same is true for ribavirin (244 Da) and its prodrug viramidine (243 Da), which also share their mass transitions with cytosine and uridine (Yeh et al., 2007). Another problem occurring during the simultaneous analysis of nucleotides is the up-front fragmentation of higher phosphates into lower phosphates, which, for example, results in a MS signal in the DP mass transition at the retention time of the TP. The described forms of interference can be summarized in the following example: when directly analyzing ara-UMP the MS can detect UMP (endogenous isomer), ara-UTP and ara-UDP (up-front fragmentation to ara-UMP), 13C/15N-araCMP and 13C/15N-CMP (isotope signals), despite the selection of a parent and product ion. If a high-resolution mass analyzer
C. Selectivity
Endogenous nucleosides and their analogs are all based on the same molecular structures (Fig. 1). As a result their molecular weights and fragmentation patterns are alike. Because of these similarities interfering peaks are often observed. The various interferences will be discussed below, followed by approaches to improve the selectivity.
1. Endogenous Compounds
Many, often high abundant, endogenous nucleotides that can potentially interfere with MS detection are present in cells. Unidentied peaks were observed in the analysis of dFdC-TP (Veltkamp et al., 2006), OPC-TP (Pucci et al., 2009), and PMPADP and CBV-TP (Pruvost et al., 2008; Delinsky et al., 2006). Some interfering compounds were identied, such as deoxycytidine TP (dCTP) in the analysis of dD4FC-TP (Shi et al., 2002), ATP and dGTP (MW 507.2 for both) in the analysis of AZT-TP (Becher et al., 2002a; van Kampen et al., 2004; King et al., 2006a), and the Na-adduct of 3TC-TP in the analysis of deoxyadenosine TP (dATP) (Pruvost et al., 2008). Cytosine and its nucleotides were observed in the analysis of their isomers araC and its nucleotides (Gouy et al., 2006; Crauste et al., 2009), and cytidine and uridine in the mass transitions of viramidine and ribavirine, respectively (Yeh et al., 2005, 2007). Figure 10 illustrates the interference of CTP in the LCMS/MS analysis of ara-CTP.
TABLE 9. Direct methods for the quantication of intracellular nucleotide analogs using MALDI-TOF
0.85 106 cells/sample.
335
&
JANSEN ET AL.
TABLE 10. Mass transitions used for quantication of the nucleoside and nucleotide analogs
336
&
TABLE 10. (Continued )
(B), base ion; N, nucleoside; Nt, nucleotide; Deriv., derivative.
is not available, these interferences make selective sample pretreatment or chromatographic separation of the analytes indispensable for accurate quantication.
3. Tools to Improve the Selectivity

As discussed in Section IVA, nucleotides can be ionized in the positive as well as in the negative ion mode. Common fragments in the negative ion mode are the MP (m/z 79) and DP (m/z 159) ion, whereas the protonated base is frequently observed in the positive ion mode (Table 10). Although abundant, the DP fragment is formed from every nucleoside DP or TP, thus relying solely on the parent mass for MS selectivity. Though still the same for many nucleosides, the base fragment obtained in the positive ion mode offers much more selectivity (Pruvost et al., 2008). Switching from the negative to the positive ion mode has therefore solved many of the problems related to endogenous peaks (Claire, 2000; Shi et al., 2002; Pruvost et al., 2008). Switching of polarity is, however, not feasible for all analytes. For example, AZT nucleotides show a lack of signal in the positive ion mode (Compain et al., 2007). Several sample pretreatments have been investigated in an attempt to solve this problem. The separation of nucleosides is easier than the separation of their corresponding nucleotides. Therefore, AZT nucleotides were successfully quantitated after de-phosphorylation (Font et al., 1999; Moore et al., 2000; Rodriguez et al., 2000; Williams
et al., 2003; King et al., 2006a). Others experimented with direct approaches. A selective periodate-methylamine reduction of ribonucleotides (such as ATP and GTP) was used to deplete the interfering ATP (Becher et al., 2002b; Hennere et al., 2003; Compain et al., 2007), but the remaining dGTP, which has the same mass, still prevented quantication. Both ATP and dGTP were effectively removed using an immunoafnity extraction (Becher et al., 2002b), but the need to raise specic antibodies, lack of robustness, and low throughput hampered routine use. The best solution for the direct determination of AZT-TP to date is the selection of a low abundant, but selective, product ion (m/z 380) (van Kampen et al., 2004; Compain et al., 2007). Others, however, were less successful in using a selective fragment (King et al., 2006a). For the simultaneous quantication of cytidine and uridine analogs, the best approach is to use a chromatographic system that can separate the analyte such as recently described (Jansen et al., 2009a) or, if available, to use high-resolution mass analyzers. The risk of interferences in mass transitions of internal standards has been limited by using Br- and Cl-substituted bases, which have two naturally occurring isotopes and, therefore, two similar mass transitions, or by using thiophosphates that do not result in a normal DP (m/z 159) fragment (Becher et al., 2002a). Both these effects are seen for the clodronate metabolite.
337
&
JANSEN ET AL.
Despite the use of tandem MS, interfering peaks have often been observed. MS selectivity can be improved by using the positive ionization mode. Otherwise, specic sample preparation or selective chromatographic systems are required.
D. Adsorption to Metal
Throughout the literature on the analysis of nucleotides with MS peak tailing is reported. Some groups found that ammonium phosphate improved peak shape (Claire, 2000) and MS signal (Tuytten et al., 2002), whereas others observed that peak tailing was reduced by using a shorter stainless steel spray capillary in the ESI source (Shi et al., 2002). Moreover, variation in the MS signal and peak tailing was decreased by replacing the stainless steel spray capillary by a fused silica spray capillary (Shi et al., 2002; Jansen et al., 2007). Two groups extensively reported on the effect and concluded that it could, at least in part, be due to adsorption of phosphate groups to metal surfaces (Wakamatsu et al., 2005; Tuytten et al., 2006). Wakamatsu et al. (2005) noticed extensive peak tailing when analyzing adenosine MP (AMP). The same effect was seen for guanosine MP (GMP) and uridine MP (UMP), but not for adenosine, guanosine, or uridine. When phosphate ions were added to the mobile phase, the peak shapes improved considerably. After pretreatment of the stainless steel capillary with phosphoric acid, peak shapes improved and remained acceptable for several hours. Tuytten et al. (2006) systematically examined the effect and concluded that phosphate-containing analytes, including nucleotides, can be trapped on any steel part of the analytical system (in the following order: MP < DP < TP). In addition to identifying the injector needle, loop, and valve as culprits, they also noticed an extensive loss of analytes on the inner wall of the capillary. Corrosion by prolonged use of the needle made the effect even more pronounced. The nucleotides could only be eluted from the metal surfaces under very basic conditions. An alternative approach to prevent peak tailing was provided by Asakawa et al. (2008), who found that (bi)carbonate was as effective as phosphate in reducing peak tailing. The MS compatibility of (bi)carbonate offers a great advantage over phosphate. As opposed to undesired peak tailing, Tuytten et al. (2004) have used the effect to their advantage by performing sample clean up and concentration using an IMAC column. Since the nucleotide analysis has traditionally been performed with mobile phases containing phosphate buffers, the adsorption to metal is a problem that specically emerged in the LCMS analysis of nucleotides. Several solutions have, however, been provided. Concluding, phosphate ions (introduced before or during analysis), or replacement of the stainless steel capillary by fused silica are effective in reducing peak tailing, but the most promising approach seems to be the addition (bi)carbonate ions to the mobile phase.
Indirect and direct methods for the analysis of drug-related nucleotides show a comparable sensitivity and have successfully been validated and applied to biological samples. The methods, however, differ in their strengths and weaknesses. Indirect methods are labor intensive because samples have to be fractionated, de-phosphorylated, and puried before the actual analysis. Indirect methods are also instrument time intensive because each fraction has to be analyzed separately. Moreover, the extensive sample pretreatment is a potential source of errors. The nal analysis is, however, relatively simple, fast, and selective and can readily be used with MS detection. Direct methods require little to no sample pretreatment but sometimes lack selectivity because chromatographic separation is mainly based on the number of phosphate groups. Long run times were initially required, but those of recent methods compete with run times of indirect analyses. The advantages and drawbacks of both methods are illustrated by a recent discussion between two groups with long experience in the analysis of intracellular nucleotide analogs, using an indirect and a direct method (Becher et al., 2003b; Benech et al., 2006; Melendez et al., 2006). Taking everything into account, direct methods seem preferable over indirect methods. The use of IP agents is well established. Through the reduction of IP agents and salts delivered to the source, the method has become relatively MS-friendly. Many, however, still report problems with the approach. More recently it was shown that WAX, using a pH gradient instead of salts for elution, is a good and relatively simple alternative. PGC has shown unsurpassed selectivity, which makes it especially attractive for the separation of closely related nucleotides that cannot be fully distinguished with conventional mass spectrometers. CEMS has also proven its applicability to the analysis of intracellular nucleotides. Though sensitive, the lack of experience and low reproducibility make MALDI-TOF not the method of choice. More applications using HILIC are required to determine its value for the analysis of intracellular nucleotides. Due to its limited accessibility, the use of AMS is most likely only favored for specic research question requiring its ultra-sensitivity. Despite the initial difculties, MS has become the method of choice for the quantitative determination of intracellular nucleotide analogs because of its sensitivity and selectivity. With these advantages over other methods, nucleotide analogs can selectively be detected in even smaller concentrations and sample volumes, enabling further elucidation of the pharmacology of this important class of drugs.
REFERENCES
Anderson PL, Kakuda TN, Kawle S, Fletcher CV. 2003. Antiviral dynamics and sex differences of zidovudine and lamivudine triphosphate concentrations in HIV-infected individuals. AIDS 17:2159 2168. Anderson PL, Zheng JH, King T, Bushman LR, Predhomme J, Meditz A, Gerber J, Fletcher CV. 2007. Concentrations of zidovudine- and lamivudine-triphosphate according to cell type in HIV-seronegative adults. AIDS 21:18491854. Andrews CL, Vouros P, Harsch A. 1999. Analysis of DNA adducts using highperformance separation techniques coupled to electrospray ionization mass spectrometry. J Chromatogr A 856:515526. Apffel A, Chakel JA, Fischer S, Lichtenwalter K, Hancock WS. 1997. New procedure for the use of high-performance liquid chromatography-
V. CONCLUSION AND PERSPECTIVES

Both indirect and direct methods have been developed to overcome the apparent incompatibility of nucleotide separation and MS detection.
338
&
electrospray ionization mass spectrometry for the analysis of nucleotides and oligonucleotides. J Chromatogr A 777:321. Asakawa Y, Tokida N, Ozawa C, Ishiba M, Tagaya O, Asakawa N. 2008. Suppression effects of carbonate on the interaction between stainless steel and phosphate groups of phosphate compounds in high-performance liquid chromatography and electrospray ionization mass spectrometry. J Chromatogr A 11981199:8086. Au JL, Su MH, Wientjes MG. 1989. Extraction of intracellular nucleosides and nucleotides with acetonitrile. Clin Chem 35:4851. Auriola S, Frith J, Rogers MJ, Koivuniemi A, Monkkonen J. 1997. Identication of adenine nucleotide-containing metabolites of bisphosphonate drugs using ion-pair liquid chromatography-electrospray mass spectrometry. J Chromatogr B Biomed Sci Appl 704:187 195. Back DJ, Burger DM, Flexner CW, Gerber JG. 2005. The pharmacology of antiretroviral nucleoside and nucleotide reverse transcriptase inhibitors: Implications for once-daily dosing. J Acquir Immune Dec Syndr 39(Suppl 1):S1. Bajad SU, Lu W, Kimball EH, Yuan J, Peterson C, Rabinowitz JD. 2006. Separation and quantitation of water soluble cellular metabolites by hydrophilic interaction chromatography-tandem mass spectrometry. J Chromatogr A 1125:7688. Banks JF. 1997. Recent advances in capillary electrophoresis/electrospray/ mass spectrometry. Electrophoresis 18:22552266. Banoub JH, Newton RP, Esmans E, Ewing DF, Mackenzie G. 2005. Recent developments in mass spectrometry for the characterization of nucleosides, nucleotides, oligonucleotides, and nucleic acids. Chem Rev 105: 18691915. Barreiro P, Garcia-Benayas T, Rendon A, Rodriguez-Novoa S, Soriano V. 2004. Combinations of nucleoside/nucleotide analogues for HIV therapy. AIDS Rev 6:234243. Bartley S, Begley JA, Clark TN. 2007. Development and validation of a novel automated LC/MS/MS assay for the direct quantication of tenofovir diphosphate in peripheral blood mononuclear cells. Proceedings of the 55th ASMS Conference on Mass Spectrometry and Allied Topics. Becher F, Pruvost A, Goujard C, Guerreiro C, Delfraissy JF, Grassi J, Benech H. 2002a. Improved method for the simultaneous determination of d4T, 3TC and ddl intracellular phosphorylated anabolites in human peripheral-blood mononuclear cells using high-performance liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom 16:555565. Becher F, Schlemmer D, Pruvost A, Nevers MC, Goujard C, Jorajuria S, Guerreiro C, Brossette T, Lebeau L, Creminon C, Grassi J, Benech H. 2002b. Development of a direct assay for measuring intracellular AZT triphosphate in human peripheral blood mononuclear cells. Anal Chem 74:42204227. Becher F, Pruvost A, Gale J, Couerbe P, Goujard C, Boutet V, Ezan E, Grassi J, Benech H. 2003a. A strategy for liquid chromatography/tandem mass spectrometric assays of intracellular drugs: Application to the validation of the triphosphorylated anabolite of antiretrovirals in peripheral blood mononuclear cells. J Mass Spectrom 38:879890. Becher F, Pruvost AG, Schlemmer DD, Creminon CA, Goujard CM, Delfraissy JF, Benech HC, Grassi JJ. 2003b. Signicant levels of intracellular stavudine triphosphate are found in HIV-infected zidovudine-treated patients. AIDS 17:555561. Becher F, Landman R, Mboup S, Kane CN, Canestri A, Liegeois F, Vray M, Prevot MH, Leleu G, Benech H. 2004. Monitoring of didanosine and stavudine intracellular trisphosphorylated anabolite concentrations in HIV-infected patients. AIDS 18:181187. Benech H, Theodoro F, Herbet A, Page N, Schlemmer D, Pruvost A, Grassi J, Deverre JR. 2004. Peripheral blood mononuclear cell counting using a DNA-detection-based method. Anal Biochem 330:172174. Benech H, Becher F, Pruvost A, Grassi JJ. 2006. Is stavudine triphosphate a natural metabolite of zidovudine? Antimicrob Agents Chemother 50:28992901.
Bergan S, Bentdal O, Sodal G, Brun A, Rugstad HE, Stokke O. 1997. Patterns of azathioprine metabolites in neutrophils, lymphocytes, reticulocytes, and erythrocytes: Relevance to toxicity and monitoring in recipients of renal allografts. Ther Drug Monit 19:502 509. Bezy V, Chaimbault P, Morin P, Unger SE, Bernard MC, Agrofoglio LA. 2006. Analysis and validation of the phosphorylated metabolites of two anti-human immunodeciency virus nucleotides (stavudine and didanosine) by pressure-assisted CE-ESI-MS/MS in cell extracts: Sensitivity enhancement by the use of peruorinated acids and alcohols as coaxial sheath-liquid make-up constituents. Electrophoresis 27: 24642476. Borroto-Esoda K, Vela JE, Myrick F, Ray AS, Miller MD. 2006. In vitro evaluation of the anti-HIV activity and metabolic interactions of tenofovir and emtricitabine. Antivir Ther 11:377384. Bousquet L, Pruvost A, Guyot AC, Farinotti R, Mabondzo A. 2009. Combination of tenofovir and emtricitabine plus efavirenz: In vitro modulation of ABC transporter and intracellular drug accumulation. Antimicrob Agents Chemother 53:896902. Bradford MM. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248254. Cahours X, Tran TT, Mesplet N, Kieda C, Morin P, Agrofoglio LA. 2001. Analysis of intracellular didanosine triphosphate at sub-ppb level using LC-MS/MS. J Pharm Biomed Anal 26:819827. Cai Z, Song F, Yang MS. 2002. Capillary liquid chromatographichigh-resolution mass spectrometric analysis of ribonucleotides. J Chromatogr A 976:135143. Cai Z, Fung EN, Sinhababu AK. 2003. Capillary electrophoresis-ion trap mass spectrometry analysis of Ziagen and its phosphorylated metabolites. Electrophoresis 24:31603164. Chen J, Seymour M, Lee L, Fuchs E, Hubbard W, Parsons T, Pakes G, Fletcher C, Garner C, Flexner C. 2009. Accelerator mass spectrometry: A new way to measure intracellular NRTI triphosphates. 16th Conference on Retroviruses and Opportunistic Infections. Claire RL III. 2000. Positive ion electrospray ionization tandem mass spectrometry coupled to ion-pairing high-performance liquid chromatography with a phosphate buffer for the quantitative analysis of intracellular nucleotides. Rapid Commun Mass Spectrom 14:1625 1634. Compain S, Durand-Gasselin L, Grassi J, Benech H. 2007. Improved method to quantify intracellular zidovudine mono- and triphosphate in peripheral blood mononuclear cells by liquid chromatography-tandem mass spectrometry. J Mass Spectrom 42:389404. Coulier L, Bas R, Jespersen S, Verheij E, van der Werf MJ, Hankemeier T. 2006. Simultaneous quantitative analysis of metabolites using ion-pair liquid chromatography-electrospray ionization mass spectrometry. Anal Chem 78:65736582. Coulier L, van Kampen JJ, de Groot R, Gerritsen HW, Bas RC, van Dongen WD, Brull LP, Luider TM. 2008. Simultaneous determination of endogenous deoxynucleotides and phosphorylated nucleoside reverse transcriptase inhibitors in peripheral blood mononuclear cells using ionpair liquid chromatography coupled to mass spectrometry. Proteomics Clin Appl 2:15571562. Crauste C, Lefebvre I, Hovaneissian M, Puy JY, Roy B, Peyrottes S, Cohen S, Guitton J, Dumontet C, Perigaud C. 2009. Development of a sensitive and selective LC/MS/MS method for the simultaneous determination of intracellular 1-beta-d-arabinofuranosylcytosine triphosphate (araCTP), cytidine triphosphate (CTP) and deoxycytidine triphosphate (dCTP) in a human follicular lymphoma cell line. J Chromatogr B Analyt Technol Biomed Life Sci 877:14171425. Dai F, Kelley JA, Zhang H, Malinowski N, Kavlick MF, Lietzau J, Welles L, Yarchoan R, Ford H, Jr. 2001. Fluorometric determination of 20 -beta-uoro-20 ,30 -dideoxyadenosine 50 -triphosphate, the active metabolite of a new anti-human immunodeciency virus drug, in human lymphocytes. Anal Biochem 288:5261.
339
&
JANSEN ET AL.
Delaney WE, Ray AS, Yang H, Qi X, Xiong S, Zhu Y, Miller MD. 2006. Intracellular metabolism and in vitro activity of tenofovir against hepatitis B virus. Antimicrob Agents Chemother 50:2471 2477. Delinsky DC, Hernandez-Santiago B, Schinazi RF. 2006. Simultaneous quantication of tenofovir diphosphate and emtricitabine triphosphate in human and macaque primary lymphocytes by ion exchange LC-MS/ MS. Proceedings of the 54th ASMS Conference on Mass Spectrometry and Allied Topics. Dervieux T, Boulieu R. 1998. Identication of 6-methylmercaptopurine derivative formed during acid hydrolysis of thiopurine nucleotides in erythrocytes, using liquid chromatography-mass spectrometry, infrared spectroscopy, and nuclear magnetic resonance assay. Clin Chem 44: 25112515. Dervieux T, Meyer G, Barham R, Matsutani M, Barry M, Boulieu R, Neri B, Seidman E. 2005. Liquid chromatography-tandem mass spectrometry analysis of erythrocyte thiopurine nucleotides and effect of thiopurine methyltransferase gene variants on these metabolites in patients receiving azathioprine/6-mercaptopurine therapy. Clin Chem 51: 20742084. Durand-Gasselin L, Da Silva D, Benech H, Pruvost A, Grassi J. 2007. Evidence and possible consequences of the phosphorylation of nucleoside reverse transcriptase inhibitors in human red blood cells. Antimicrob Agents Chemother 51:21052111. Durand-Gasselin L, Pruvost A, Dehee A, Vaudre G, Tabone MD, Grassi J, Leverger G, Garbarg-Chenon A, Benech H, Dollfus C. 2008. High levels of zidovudine (AZT) and its intracellular phosphate metabolites in AZT- and AZT-lamivudine-treated newborns of human immunodeciency virus-infected mothers. Antimicrob Agents Chemother 52: 25552563. Feng YL, Zhu J. 2006. On-line enhancement technique for the analysis of nucleotides using capillary zone electrophoresis/mass spectrometry. Anal Chem 78:66086613. Font E, Rosario O, Santana J, Garcia H, Sommadossi JP, Rodriguez JF. 1999. Determination of zidovudine triphosphate intracellular concentrations in peripheral blood mononuclear cells from human immunodeciency virus-infected individuals by tandem mass spectrometry. Antimicrob Agents Chemother 43:29642968. Freise KJ, Martin-Jimenez T. 2006. Pharmacokinetics of gemcitabine and its primary metabolite in dogs after intravenous bolus dosing and its in vitro pharmacodynamics. J Vet Pharmacol Ther 29:137145. Friedecky D, Bednar P, Prochazka M, Adam T. 2006. Analysis of intracellular nucleotides by capillary electrophoresis-mass spectrometry. Nucleosides Nucleotides Nucleic Acids 25:12331236. Fung EN, Cai Z, Burnette TC, Sinhababu AK. 2001. Simultaneous determination of Ziagen and its phosphorylated metabolites by ionpairing high-performance liquid chromatography-tandem mass spectrometry. J Chromatogr B 754:285295. Gaus HJ, Owens SR, Winniman M, Cooper S, Cummins LL. 1997. On-line HPLC electrospray mass spectrometry of phosphorothioate oligonucleotide metabolites. Anal Chem 69:313319. Goujon L, Brossette T, Dereudre-Bosquet N, Creminon C, Clayette P, Dormont D, Mioskowski C, Lebeau L, Grassi J. 1998. Monitoring of intracellular levels of 50 -monophosphate-AZT using an enzyme immunoassay. J Immunol Methods 218:1930. Gouy M, Fabre H, Blanchin M, Peyrottes S, Perigaud C, Lefebvre I. 2006. Quantication of 50 -monophosphate cytosine arabinoside (Ara-CMP) in cell extracts using liquid chromatography-electrospray mass spectrometry. Anal Chim Acta 566:178184. Green-Church KB, Limbach PA. 2000. Mononucleotide gas-phase proton afnities as determined by the kinetic method. J Am Soc Mass Spectrom 11:2432. Grob MK, OBrien K, Chu JJ, Chen DD. 2003. Optimization of cellular nucleotide extraction and sample preparation for nucleotide pool analyses using capillary electrophoresis. J Chromatogr B 788:103 111.
Gu GH, Lim CK. 1990. Separation of anionic and cationic compounds of biomedical interest by high-performance liquid chromatography on porous graphitic carbon. J Chromatogr 515:183192. Hawkins T, Veikley W, Claire RL III, Guyer B, Clark N, Kearney BP. 2005. Intracellular pharmacokinetics of tenofovir diphosphate, carbovir triphosphate, and lamivudine triphosphate in patients receiving triplenucleoside regimens. J Acquir Immune Dec Syndr 39:406411. Hellborg R, Skog G. 2008. Accelerator mass spectrometry. Mass Spectrom Rev 27:398427. Hennere G, Becher F, Pruvost A, Goujard C, Grassi J, Benech H. 2003. Liquid chromatography-tandem mass spectrometry assays for intracellular deoxyribonucleotide triphosphate competitors of nucleoside antiretrovirals. J Chromatogr B 789:273281. Hernandez-Santiago BI, Beltran T, Stuyver L, Chu CK, Schinazi RF. 2004. Metabolism of the anti-hepatitis C virus nucleoside beta-D-N4hydroxycytidine in different liver cells. Antimicrob Agents Chemother 48:46364642. Holdich T, Shiveley LA, Sawyer J. 2007. Effect of lamivudine on the plasma and intracellular pharmacokinetics of apricitabine, a novel nucleoside reverse transcriptase inhibitor, in healthy volunteers. Antimicrob Agents Chemother 51:29432947. Hsieh Y. 2008. Potential of HILIC-MS in quantitative bioanalysis of drugs and drug metabolites. J Sep Sci 31:14811491. Jansen RS, Rosing H, de Wolf CJ, Beijnen JH. 2007. Development and validation of an assay for the quantitative determination of cladribine nucleotides in MDCKII cells and culture medium using weak anionexchange liquid chromatography coupled with tandem mass spectrometry. Rapid Commun Mass Spectrom 21:40494059. Jansen RS, Rosing H, Schellens JH, Beijnen JH. 2009a. Retention studies of 20 -20 -diuorodeoxycytidine and 20 -20 -diuorodeoxyuridine nucleosides and nucleotides on porous graphitic carbon: Development of a liquid chromatography-tandem mass spectrometry method. J Chromatogr A 1216:31683174. Jansen RS, Rosing H, Schellens JH, Beijnen JH. 2009b. Simultaneous quantication of 20 -20 -diuoro-deoxycytidine and 20 -20 -diuoro-deoxyuridine nucleosides and nucleotides in white blood cells using porous graphitic carbon chromatography coupled with tandem mass spectrometry. Rapid Commun Mass Spectrom 23:30403050. Kalhorn TF, Ren AG, Slattery JT, McCune JS, Wang J. 2005. A highly sensitive high-performance liquid chromatography-mass spectrometry method for quantication of udarabine triphosphate in leukemic cells. J Chromatogr B 820:243250. Kim J, Chou TF, Griesgraber GW, Wagner CR. 2004. Direct measurement of nucleoside monophosphate delivery from a phosphoramidate pronucleotide by stable isotope labeling and LC-ESI(-)-MS/MS. Mol Pharm 1:102111. Kim J, Park S, Tretyakova NY, Wagner CR. 2005. A method for quantitating the intracellular metabolism of AZT amino acid phosphoramidate pronucleotides by capillary high-performance liquid chromatographyelectrospray ionization mass spectrometry. Mol Pharm 2:233241. Kimball E, Rabinowitz JD. 2006. Identifying decomposition products in extracts of cellular metabolites. Anal Biochem 358:273280. King T, Bushman L, Anderson PL, Delahunty T, Ray M, Fletcher CV. 2006a. Quantitation of zidovudine triphosphate concentrations from human peripheral blood mononuclear cells by anion exchange solid phase extraction and liquid chromatography-tandem mass spectroscopy; an indirect quantitation methodology. J Chromatogr B 831:248257. King T, Bushman L, Kiser J, Anderson PL, Ray M, Delahunty T, Fletcher CV. 2006b. Liquid chromatography-tandem mass spectrometric determination of tenofovir-diphosphate in human peripheral blood mononuclear cells. J Chromatogr B 843:147156. Kuklenyik Z, Martin A, Pau C, Garcia-Lerma G, Heneine W, Pirkle J, Barr J. 2008. Advanced HPLC-MS/MS methods for quantitation of nucleotide and nucleoside HIV reverse transcriptase inhibitor metabolites. Proceedings of the 56th ASMS Conference on Mass Spectrometry and Allied Topics.
340
&
Kuster H, Vogt M, Joos B, Nadai V, Luthy R. 1991. A method for the quantication of intracellular zidovudine nucleotides. J Infect Dis 164:773776. Lai J, Wang J, Cai Z. 2008. Nucleoside reverse transcriptase inhibitors and their phosphorylated metabolites in human immunodeciency virus-infected human matrices. J Chromatogr B 868:112. Le Saint C, Terreux R, Duval D, Durant J, Ettesse H, Dellamonica P, Guedj R, Vincent JP, Cupo A. 2004. Determination of ddATP levels in human immunodeciency virus-infected patients treated with dideoxyinosine. Antimicrob Agents Chemother 48:589595. Lee J, Yoo BC, Lee HS, Yoo HW, Yoo HH, Kang MJ, Kim DH. 2006. Rapid quantitative determination of L-FMAU-TP from human peripheralblood mononuclear cells of hepatitis B virus-infected patients treated with L-FMAU by ion-pairing, reverse-phase, liquid chromatography/ electrospray tandem mass spectrometry. Ther Drug Monit 28:131 137. Liu CC, Huang JS, Tyrrell DL, Dovichi NJ. 2005. Capillary electrophoresiselectrospray-mass spectrometry of nucleosides and nucleotides: Application to phosphorylation studies of anti-human immunodeciency virus nucleosides in a human hepatoma cell line. Electrophoresis 26:14241431. Lynch T, Eisenberg G, Kernan M. 2001. LC/MS determination of the intracellular concentration of two novel aryl phosphoramidate prodrugs of PMPA and their metabolites in dog PBMC. Nucleosides Nucleotides Nucleic Acids 20:14151419. McKeown AP, Shaw PN, Barrett DA. 2001. Electrophoretic behaviour of oligonucleotides and mono-, di- and triphosphate nucleotides by capillary zone electrophoresis. Electrophoresis 22:11191126. Melendez M, Blanco R, Delgado W, Garcia R, Santana J, Garcia H, Rosario O, Rodriguez JF. 2006. Lack of evidence for in vivo transformation of zidovudine triphosphate to stavudine triphosphate in human immunodeciency virus-infected patients. Antimicrob Agents Chemother 50:835840. Monkkonen H, Moilanen P, Monkkonen J, Frith JC, Rogers MJ, Auriola S. 2000. Analysis of an adenine nucleotide-containing metabolite of clodronate using ion pair high-performance liquid chromatographyelectrospray ionisation mass spectrometry. J Chromatogr B Biomed Sci Appl 738:395403. Moore JD, Valette G, Darque A, Zhou XJ, Sommadossi JP. 2000. Simultaneous quantitation of the 50 -triphosphate metabolites of zidovudine, lamivudine, and stavudine in peripheral mononuclear blood cells of HIV infected patients by high-performance liquid chromatography tandem mass spectrometry. J Am Soc Mass Spectrom 11:11341143. Moore JD, Acosta EP, Johnson VA, Bassett R, Eron JJ, Fischl MA, Long MC, Kuritzkes DR, Sommadossi JP. 2007. Intracellular nucleoside triphosphate concentrations in HIV-infected patients on dual nucleoside reverse transcriptase inhibitor therapy. Antivir Ther 12: 981986. Moyle G, Bofto M, Fletcher C, Higgs C, Hay PE, Song IH, Lou Y, Yuen GJ, Min SS, Guerini EM. 2009. Steady-state pharmacokinetics of plasma abacavir and intracellular carbovir triphosphate following abacavir 600 mg once daily and 300 mg twice daily in HIV-infected subjects. Antimicrob Agents Chemother 53:15321538. Nguyen T, Zhang X, Amin J, Marlor C, Mushtaq G, Buechter D. 2006. A novel method for quantitative analysis of nucleoside analogs by LC/ MS/MS. Proceedings of the 54th Conference on Mass Spectrometry and Allied Topics. Nirogi R, Bhyrapuneni G, Kandikere V, Mudigonda K, Komarneni P, Aleti R, Mukkanti K. 2008. Simultaneous quantication of a non-nucleoside reverse transcriptase inhibitor efavirenz, a nucleoside reverse transcriptase inhibitor emtricitabine and a nucleotide reverse transcriptase inhibitor tenofovir in plasma by liquid chromatography positive ion electrospray tandem mass spectrometry. Biomed Chromatogr 23:371381. Osterman MT, Kundu R, Lichtenstein GR, Lewis JD. 2006. Association of 6thioguanine nucleotide levels and inammatory bowel disease activity: A meta-analysis. Gastroenterology 130:10471053.
Palmer S, Cox S. 1994. Comparison of extraction procedures for highperformance liquid chromatographic determination of cellular deoxynucleotides. J Chromatogr A 667:316321. Pan S, Verhoeven K, Lee JK. 2005. Investigation of the initial fragmentation of oligodeoxynucleotides in a quadrupole ion trap: Charge level-related base loss. J Am Soc Mass Spectrom 16:18531865. Peng L, Farkas T. 2008. Analysis of basic compounds by reversed-phase liquid chromatography-electrospray mass spectrometry in high-pH mobile phases. J Chromatogr A 1179:131144. Pesek JJ, Matyska MT, Hearn MT, Boysen RI. 2009. Aqueous normal-phase retention of nucleotides on silica hydride columns. J Chromatogr A 1216:11401146. Pike MG, Franklin CL, Mays DC, Lipsky JJ, Lowry PW, Sandborn WJ. 2001. Improved methods for determining the concentration of 6-thioguanine nucleotides and 6-methylmercaptopurine nucleotides in blood. J Chromatogr B Biomed Sci Appl 757:19. Pruvost A, Becher F, Bardouille P, Guerrero C, Creminon C, Delfraissy JF, Goujard C, Grassi J, Benech H. 2001. Direct determination of phosphorylated intracellular anabolites of stavudine (d4T) by liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom 15:14011408. Pruvost A, Negredo E, Benech H, Theodoro F, Puig J, Grau E, Garcia E, Molto J, Grassi J, Clotet B. 2005. Measurement of intracellular didanosine and tenofovir phosphorylated metabolites and possible interaction of the two drugs in human immunodeciency virus-infected patients. Antimicrob Agents Chemother 49:19071914. Pruvost A, Theodoro F, Agrofoglio L, Negredo E, Benech H. 2008. Specicity enhancement with LC-positive ESI-MS/MS for the measurement of nucleotides: Application to the quantitative determination of carbovir triphosphate, lamivudine triphosphate and tenofovir diphosphate in human peripheral blood mononuclear cells. J Mass Spectrom 43:224233. Pucci V, Giuliano C, Zhang R, Koeplinger KA, Leone JF, Monteagudo E, Bonelli F. 2009. HILIC LC-MS for the determination of 20 -C-methylcytidine-triphosphate in rat liver. J Sep Sci 32:12751283. Qian T, Cai Z, Yang MS. 2004. Determination of adenosine nucleotides in cultured cells by ion-pairing liquid chromatography-electrospray ionization mass spectrometry. Anal Biochem 325:7784. Ray AS, Vela JE, Olson L, Fridland A. 2004. Effective metabolism and long intracellular half life of the anti-hepatitis B agent adefovir in hepatic cells. Biochem Pharmacol 68:18251831. Ray AS, Myrick F, Vela JE, Olson LY, Eisenberg EJ, Borroto-Esodo K, Miller MD, Fridland A. 2005. Lack of a metabolic and antiviral drug interaction between tenofovir, abacavir and lamivudine. Antivir Ther 10:451457. Ray AS, Vela JE, Boojamra CG, Zhang L, Hui H, Callebaut C, Stray K, Lin KY, Gao Y, Mackman RL, Cihlar T. 2008. Intracellular metabolism of the nucleotide prodrug GS-9131, a potent anti-human immunodeciency virus agent. Antimicrob Agents Chemother 52:648654. Reichelova V, Albertioni F, Liliemark J. 1996. Determination of 2-chloro-20 deoxyadenosine nucleotides in leukemic cells by ion-pair highperformance liquid chromatography. J Chromatogr B Biomed Appl 682:115123. Reiser H, Wang J, Chong L, Watkins WJ, Ray AS, Shibata R, Birkus G, Cihlar T, Wu S, Li B, Liu X, Henne IN, Wolfgang GH, Desai M, Rhodes GR, Fridland A, Lee WA, Plunkett W, Vail D, Thamm DH, Jeraj R, Tumas DB. 2008. GS-9219A novel acyclic nucleotide analogue with potent antineoplastic activity in dogs with spontaneous non-Hodgkins lymphoma. Clin Cancer Res 14:28242832. Robbins BL, Rodman J, McDonald C, Srinivas RV, Flynn PM, Fridland A. 1994. Enzymatic assay for measurement of zidovudine triphosphate in peripheral blood mononuclear cells. Antimicrob Agents Chemother 38:115121. Robbins BL, Waibel BH, Fridland A. 1996. Quantitation of intracellular zidovudine phosphates by use of combined cartridge-radioimmunoassay methodology. Antimicrob Agents Chemother 40:26512654.
341
&
JANSEN ET AL.
Robbins BL, Poston PA, Neal EF, Slaughter C, Rodman JH. 2007. Simultaneous measurement of intracellular triphosphate metabolites of zidovudine, lamivudine and abacavir (carbovir) in human peripheral blood mononuclear cells by combined anion exchange solid phase extraction and LC-MS/MS. J Chromatogr B 850:310 317. Rodriguez JF, Rodriguez JL, Santana J, Garcia H, Rosario O. 2000. Simultaneous quantitation of intracellular zidovudine and lamivudine triphosphates in human immunodeciency virus-infected individuals. Antimicrob Agents Chemother 44:30973100. Rodriguez Orengo JF, Santana J, Febo I, Diaz C, Rodriguez JL, Garcia R, Font E, Rosario O. 2000. Intracellular studies of the nucleoside reverse transcriptase inhibitor active metabolites: A review. P R Health Sci J 19:1927. Shi G, Wu JT, Li Y, Geleziunas R, Gallagher K, Emm T, Olah T, Unger S. 2002. Novel direct detection method for quantitative determination of intracellular nucleoside triphosphates using weak anion exchange liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom 16:10921099. Shipkova M, Armstrong VW, Wieland E, Oellerich M. 2003. Differences in nucleotide hydrolysis contribute to the differences between erythrocyte 6-thioguanine nucleotide concentrations determined by two widely used methods. Clin Chem 49:260268. Singh R, Farmer PB. 2006. Liquid chromatography-electrospray ionizationmass spectrometry: The future of DNA adduct detection. Carcinogenesis 27:178196. Slusher JT, Kuwahara SK, Hamzeh FM, Lewis LD, Kornhauser DM, Lietman PS. 1992. Intracellular zidovudine (ZDV) and ZDV phosphates as measured by a validated combined high-pressure liquid chromatography-radioimmunoassay procedure. Antimicrob Agents Chemother 36:24732477. Smyrnis E, Sacks S, Burton R, Wall R, Abbott F. 1996. Reverse phase-ion paired HPLC/electrospray MS-MS detection of penciclovir and its phosphorylated derivatives. Proceedings of the 44th ASMS Conference on Mass Spectrometry and Allied Topics, p. 189. Soga T, Ishikawa T, Igarashi S, Sugawara K, Kakazu Y, Tomita M. 2007. Analysis of nucleotides by pressure-assisted capillary electrophoresismass spectrometry using silanol mask technique. J Chromatogr A 1159:125133. Solas C, Li YF, Xie MY, Sommadossi JP, Zhou XJ. 1998. Intracellular nucleotides of ()-20 ,30 -deoxy-30 -thiacytidine in peripheral blood mononuclear cells of a patient infected with human immunodeciency virus. Antimicrob Agents Chemother 42:29892995. Thevanayagam LN, Jayewardene AL, Gambertoglio JG. 2000. The stability of intracellular zidovudine and its anabolites in extracts of peripheral blood mononuclear cells (PBMCs). J Pharm Biomed Anal 22:597603. Tidd DM, Dedhar S. 1978. Specic and sensitive combined high-performance liquid chromatographic-ow uorometric assay for intracellular 6-thioguanine nucleotides metabolites of 6-mercaptopurine and 6-thioguanine. J Chromatogr 145:237246. Tuytten R, Lemiere F, Dongen WV, Esmans EL, Slegers H. 2002. Short capillary ion-pair high-performance liquid chromatography coupled to electrospray (tandem) mass spectrometry for the simultaneous analysis of nucleoside mono-, di- and triphosphates. Rapid Commun Mass Spectrom 16:12051215. Tuytten R, Lemiere F, Van Dongen W, Slegers H, Newton RP, Esmans EL. 2004. Investigation of the use of immobilised metal afnity chromatography for the on-line sample clean up and pre-concentration of nucleotides prior to their determination by ion pair liquid chromatography-electrospray mass spectrometry: A pilot study. J Chromatogr B 809:189198. Tuytten R, Lemiere F, Witters E, Van Dongen W, Slegers H, Newton RP, Van Onckelen H, Esmans EL. 2006. Stainless steel electrospray probe: A dead end for phosphorylated organic compounds? J Chromatogr A 1104:209221.
Uhrova M, Deyl Z, Suchanek M. 1996. Separation of common nucleotides, mono-, di- and triphosphates, by capillary electrophoresis. J Chromatogr B Biomed Appl 681:99105. van Kampen JJ, Fraaij PL, Hira V, van Rossum AM, Hartwig NG, de Groot R, Luider TM. 2004. A new method for analysis of AZT-triphosphate and nucleotide-triphosphates. Biochem Biophys Res Commun 315:151159. Vela JE, Olson LY, Huang A, Fridland A, Ray AS. 2007. Simultaneous quantitation of the nucleotide analog adefovir, its phosphorylated anabolites and 20 -deoxyadenosine triphosphate by ion-pairing LC/MS/ MS. J Chromatogr B 848:335343. Veltkamp SA, Hillebrand MJ, Rosing H, Jansen RS, Wickremsinhe ER, Perkins EJ, Schellens JH, Beijnen JH. 2006. Quantitative analysis of gemcitabine triphosphate in human peripheral blood mononuclear cells using weak anion-exchange liquid chromatography coupled with tandem mass spectrometry. J Mass Spectrom 41:16331642. Veltkamp SA, Jansen RS, Callies S, Pluim D, Visseren-Grul CM, Rosing H, Kloeker-Rhoades S, Andre VA, Beijnen JH, Slapak CA, Schellens JH. 2008. Oral administration of gemcitabine in patients with refractory tumors: A clinical and pharmacologic study. Clin Cancer Res 14:3477 3486. Vikingsson S, Carlsson B, Almer SH, Peterson C. 2009. Monitoring of thiopurine metabolites in patients with inammatory bowel disease What is actually measured? Ther Drug Monit 31:345350. Vuong lT, Ruckle JL, Blood AB, Reid MJ, Wasnich RD, Synal HA, Dueker SR. 2008. Use of accelerator mass spectrometry to measure the pharmacokinetics and peripheral blood mononuclear cell concentrations of zidovudine. J Pharm Sci 97:28332843. Wakamatsu A, Morimoto K, Shimizu M, Kudoh S. 2005. A severe peak tailing of phosphate compounds caused by interaction with stainless steel used for liquid chromatography and electrospray mass spectrometry. J Sep Sci 28:18231830. Wang LH, Begley J, St CR III, Harris J, Wakeford C, Rousseau FS. 2004. Pharmacokinetic and pharmacodynamic characteristics of emtricitabine support its once daily dosing for the treatment of HIV infection. AIDS Res Hum Retroviruses 20:11731182. Wang J, Lin T, Lai J, Cai Z, Yang MS. 2009. Analysis of adenosine phosphates in HepG-2 cell by a HPLC-ESI-MS system with porous graphitic carbon as stationary phase. J Chromatogr B Analyt Technol Biomed Life Sci 877:20192024. Willems AV, Deforce DL, Van Peteghem CH, Van Bocxlaer JF. 2005. Analysis of nucleic acid constituents by on-line capillary electrophoresis-mass spectrometry. Electrophoresis 26:12211253. Williams LD, Von Tungeln LS, Beland FA, Doerge DR. 2003. Liquid chromatographic-mass spectrometric determination of the metabolism and disposition of the anti-retroviral nucleoside analogs zidovudine and lamivudine in C57BL/6N and B6C3F1 mice. J Chromatogr B Analyt Technol Biomed Life Sci 798:5562. Witters E, Van Dongen W, Esmans EL, Van Onckelen HA. 1997. Ion-pair liquid chromatography-electrospray mass spectrometry for the analysis of cyclic nucleotides. J Chromatogr B Biomed Sci Appl 694:5563. Xia YQ, Jemal M, Zheng N, Shen X. 2006. Utility of porous graphitic carbon stationary phase in quantitative liquid chromatography/tandem mass spectrometry bioanalysis: Quantitation of diastereomers in plasma. Rapid Commun Mass Spectrom 20:18311837. Xing J, Apedo A, Tymiak A, Zhao N. 2004. Liquid chromatographic analysis of nucleosides and their mono-, di- and triphosphates using porous graphitic carbon stationary phase coupled with electrospray mass spectrometry. Rapid Commun Mass Spectrom 18:15991606. Yeh LT, Nguyen M, Lourenco D, Lin CC. 2005. A sensitive and specic method for the determination of total ribavirin in monkey liver by highperformance liquid chromatography with tandem mass spectrometry. J Pharm Biomed Anal 38:3440. Yeh LT, Nguyen M, Dadgostari S, Bu W, Lin CC. 2007. LC-MS/MS method for simultaneous determination of viramidine and ribavirin levels in monkey red blood cells. J Pharm Biomed Anal 43:10571064.
342
&
Yuen GJ, Lou Y, Bumgarner NF, Bishop JP, Smith GA, Otto VR, Hoelscher DD. 2004. Equivalent steady-state pharmacokinetics of lamivudine in plasma and lamivudine triphosphate within cells following administration of lamivudine at 300 milligrams once daily and 150 milligrams twice daily. Antimicrob Agents Chemother 48: 176182.
Zinellu A, Sotgia S, Pasciu V, Madeddu M, Leoni GG, Naitana S, Deiana L, Carru C. 2008. Intracellular adenosine 50 -triphosphate, adenosine 50 -diphosphate, and adenosine 50 -monophosphate detection by shortend injection capillary electrophoresis using methylcellulose as the effective electroosmostic ow suppressor. Electrophoresis 29:3069 3073.
343

Mass Spectrometry in The Quantitative Analysis of Therapeutic Intracellular Nucleotide Analogs

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mass Spectrometry in The Quantitative Analysis of Therapeutic Intracellular Nucleotide Analogs

Uploaded by

Copyright:

Available Formats

MASS SPECTROMETRY IN THE QUANTITATIVE ANALYSIS OF THERAPEUTIC INTRACELLULAR NUCLEOTIDE ANALOGS

FIGURE 1. Structural elements of nucleosides.

FIGURE 2. Chemical structures of pyrimidine analogs and their natural counterparts.

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

FIGURE 3. Chemical structures of purine analogs and their natural counterparts.

II. SAMPLE COLLECTION A. Matrix

FIGURE 4. Phosphorylation route of zidovudine (AZT).

Mass Spectrometry Reviews DOI 10.1002/mas

C. Extraction of Nucleotides from Cells

III. BIOANALYTICAL METHODS

A. Indirect Methods 1. Sample Pretreatment

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

2. Separation and Detection

TABLE 1. Indirect methods for the quantication of intracellular nucleotide analogs

2. Separation and Detection

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

FIGURE 8. Clodronate and its nucleotide metabolite.

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

FIGURE 10. Separation of nucleotides using PGC with interference

FIGURE 11. Separation of nucleotides on PGC with interferences

IV. MASS SPECTROMETRY A. Ionization

Assuming 4 106 cells/sample.

50 106 cells/mL. 2.05 106 cells/mL. c 20 106 cells/mL.

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

2. Other Nucleoside Analogs

0.85 106 cells/sample.

Mass Spectrometry Reviews DOI 10.1002/mas

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

TABLE 10. (Continued )

(B), base ion; N, nucleoside; Nt, nucleotide; Deriv., derivative.

3. Tools to Improve the Selectivity

V. CONCLUSION AND PERSPECTIVES

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

Mass Spectrometry Reviews DOI 10.1002/mas

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

Mass Spectrometry Reviews DOI 10.1002/mas

Mass Spectrometry Reviews DOI 10.1002/mas

QUANTITATIVE NUCLEOTIDE ANALYSIS WITH MS

Mass Spectrometry Reviews DOI 10.1002/mas

You might also like