Quantitative Analysis of Polyethylene Glycol (PEG) and PEGylated

Jun 25, 2014 - Quantitative Analysis of Polyethylene Glycol (PEG) and PEGylated Proteins in Animal Tissues by LC-MS/MS Coupled with In-Source CID...
1 downloads 0 Views 847KB Size
Download PDF
Article pubs.acs.org/ac

Quantitative Analysis of Polyethylene Glycol (PEG) and PEGylated Proteins in Animal Tissues by LC-MS/MS Coupled with In-Source CID Jiachang Gong,*,† Xiaomei Gu,† William E. Achanzar,‡ Kristina D. Chadwick,‡ Jinping Gan,† Barry J. Brock,† Narendra S. Kishnani,† W. Griff Humphreys,† and Ramaswamy A. Iyer† †

Pharmaceutical Candidate Optimization, Bristol-Myers Squibb Company, Lawrenceville, New Jersey 08540, United States Drug Safety Evaluation, Bristol-Myers Squibb Company, New Brunswick, New Jersey 08901, United States

‡

S Supporting Information *

ABSTRACT: The covalent conjugation of polyethylene glycol (PEG, typical MW > 10k) to therapeutic peptides and proteins is a wellestablished approach to improve their pharmacokinetic properties and diminish the potential for immunogenicity. Even though PEG is generally considered biologically inert and safe in animals and humans, the slow clearance of large PEGs raises concerns about potential adverse effects resulting from PEG accumulation in tissues following chronic administration, particularly in the central nervous system. The key information relevant to the issue is the disposition and fate of the PEG moiety after repeated dosing with PEGylated proteins. Here, we report a novel quantitative method utilizing LC-MS/MS coupled with in-source CID that is highly selective and sensitive to PEG-related materials. Both 40KPEG and a tool PEGylated protein (ATI-1072) underwent dissociation in the ionization source of mass spectrometer to generate a series of PEG-specific ions, which were subjected to further dissociation through conventional CID. To demonstrate the potential application of the method to assess PEG biodistribution following PEGylated protein administration, a single dose study of ATI-1072 was conducted in rats. Plasma and various tissues were collected, and the concentrations of both 40KPEG and ATI-1072 were determined using the LC-MS/MS method. The presence of 40kPEG in plasma and tissue homogenates suggests the degradation of PEGylated proteins after dose administration to rats, given that free PEG was absent in the dosing solution. The method enables further studies for a thorough characterization of disposition and fate of PEGylated proteins.

T

are commonly used in PEGylation have long half-lives in humans and animals. While beneficial from a pharmacokinetic perspective, this could lead to significant PEG accumulation in tissues with chronic administration and potentially result in adverse effects.18,19 In a study where rats were administered daily IV doses of 20-kDa PEGylated proteins for 3 months, vacuolation was observed in the renal tubular epithelial cells.20 These vacuoles were formed over time in the macrophages and/or secretory cells that were responsible for the removal of PEG.21 Although no adverse effect or necrosis of epithelial cells was observed in the affected organs, the long-term impact of PEG accumulation remains unknown. Furthermore, the tissue distribution, the extent of accumulation in tissues after multiple doses, and the clearance of PEG following administration of PEGylated proteins have not been adequately characterized. With more and more PEGylated peptides and proteins either being approved in the market or entering clinical development, it is highly desirable to fully understand the disposition of PEGylated proteins after dose administration. The primary reason for the poor characterization of PEG disposition is the lack of analytical tools that are sensitive and

herapeutic peptides and small proteins are highly specific to the biological targets, and they generally have fewer side effects, compared to conventional small molecule drugs.1,2 However, their short half-lives after dose administration become a major challenge for clinical use. The primary reasons responsible for their rapid clearance are proteolytic degradation, deactivation by immune system, and renal excretion.3−5 The pharmacokinetic profiles of peptides and proteins can be improved dramatically after covalently coupling to polyethylene glycols (PEGs) with molecular weights (MWs) typically greater than 20 kDa: this is a process called PEGylation.6−9 PEGylation increases the MW and overall size of the peptides and proteins. This alters their physical chemical properties, resulting in reduced glomerular filtration in the kidney and increased resistance to proteolytic digestion.7,10,11 In addition, the potential for immunogenicity that could potentially lead to adverse drug reaction is often diminished, because of the shielding effect from the bulky PEG branches.12−14 PEG is generally considered biologically inert and safe in animals and humans.15,16 They are generally not extensively metabolized, and the main route of elimination is through glomerular filtration into the kidney. Therefore, the half-life and disposition of PEG in animals is mainly dependent on the MW of PEG, with higher MW PEGs exhibiting a decreased rate of clearance.17 Large PEGs, with MWs greater than 20 kDa, that © 2014 American Chemical Society

Received: April 24, 2014 Accepted: June 25, 2014 Published: June 25, 2014 7642

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Article

specific to PEG analysis.22 Radio-labeling is a common approach to address the disposition for small molecule drugs. However, labeling PEGylated proteins is challenging and costly, and the methodology of direct labeling on the PEG has not been established.22 Another approach is to use specific assays to detect PEG-related materials in the biological matrix. Nuclear magnetic resonance (NMR) spectroscopy was used previously to assess the metabolic fate of the PEG moieties.23 However, the technology had low sensitivity and was not suitable for complex matrices such as plasma and tissue homogenates, because of matrix interference. In addition, NMR measured total PEG concentration and did not differentiate free PEG and PEGylated proteins. As such, potential degradation of PEGylated proteins in vivo could not be addressed unless additional analytical tools such as Western blotting were employed.23 Enzyme-linked immunosorbent assay (ELISA), using monoclonal PEG antibody, provided sensitive detection for PEG-related materials but did not separate signals from different PEG components.22,24 An LC-MS/MS method that separates and quantifies each PEG-related component simultaneously would be an ideal tool to fully understand the disposition of PEGylated proteins. Because of the heterogeneous nature of PEG, direct measurement of PEG or PEGylated proteins with LC-MS/MS has proven challenging.25,26 Recently, it was demonstrated that PEG underwent collision-induced dissociation in the ionization source (in-source CID) of the mass spectrometer to generate PEG-specific daughter ions that can be used to identify, and potentially quantify, PEG.27 The work presented here reports a novel LC-MS/MS method, coupled with in-source CID, for the quantitative analysis of 40kPEG and PEGylated protein simultaneously. A PEGylated adnectin (ATI-1072, conjugated with 40-kDa 2-branched PEG) that is devoid of pharmacological activity was used as a tool protein to demonstrate the feasibility of this approach. After in-source CID, both 40kPEG and ATI-1072 generated similar fragment ions that were characteristic to PEG. Each PEG component was quantified through a conventional CID of the PEG-specific surrogate ions.

were provided by Bristol-Myers Squibb Research & Development (Princeton, NJ). Preparation of Calibration Standards and Quality Control (QC) Samples. Calibration standards and QC samples were prepared by spiking 40kPEG or ATI-1072 into blank rat plasma or tissue homogenates with final concentrations of 0.05, 0.1, 0.25, 0.50, 1.0, 2.5, 5.0, 10, 20, 30, and 50 μg/mL. All samples were stored at approximately −20 °C after aliquoting. Extraction of PEG and ATI-1072 from Plasma or Tissue Homogenates. To optimize the extraction efficiency, blank rat plasma, blank rat tissue homogenates (brain, liver, kidney, spleen, and skeletal muscle, including quadriceps and diaphragm), and a buffer solution (control sample) were spiked with either 40kPEG or ATI-1072 to give final concentrations of 2, 5, 10, and 20 μg/mL. The resulting mixtures were extracted with three volumes of different organic solvents, including methanol, ethanol, acetonitrile, iso-propanol, and n-butanol. The mixtures were vortex mixed for 2 min, and centrifuged at 4000 g for 15 min. The supernatant was collected into a new tube, and the pellet was subject to an additional extraction with 2 mL of organic solvent/water mixture (3:1 v/v). The supernatants were combined and a portion of supernatant (50−80 μL) was injected into LC-MS/MS for mass-spectral analysis as described below. Extraction recovery was estimated by comparing the mass response of plasma extracts to that of the control. Chromatography and Mass Spectrometry. Qualitative Analysis. Qualitative analysis of 40kPEG and ATI-1072 was performed on an LC-ESI-MS system that consisted of an Agilent 1100 HPLC system (Agilent Technologies, Santa Clara, CA) and an LTQ/Orbitrap Classic mass spectrometer (Thermo Fisher Scientific, Waltham, MA). HPLC was performed at a constant flow rate of 0.8 mL/min with a binary solvent system. Solvent A was consisted of 0.1% formic acid in water, and solvent B was acetonitrile. The HPLC gradient started with 10% B and was linearly increased to 55% B in 30 min. After that, the gradient was increased to 75% B in 2 min and maintained at 75% B for 4 min prior to column reequilibration. Chromatographic separation was achieved with a reverse-phase PLRP-S polymer column (4.6 mm × 250 mm, 5 μm, 1000 Å pore size) (Agilent Technologies, Santa Clara, CA). The LTQ/Orbitrap settings were as follows: sheath gas at 50, auxiliary gas at 25, sweep gas at 5, spray voltage at 5 kV, and tube lens at 100 V. MS full scans were carried out with a mass range of 250−2000 m/z. LC-MS/MS with In-Source CID. In-source CID spectra of 40k PEG and ATI-1072 were generated with a Shimadzu Class VP HPLC system interfaced to an API 4000 Q-trap mass spectrometer (Sciex, Toronto, Canada) that was equipped with Turboionspray source. The HPLC system was equipped with two pumps (Model LC-10AT), an HTC PAL autosampler (Leap Technologies, Cary, NC), and a diode array detector (Model SPC-M10A). Chromatographic separation was achieved with a reverse-phase PLRP-S polymer column (2.1 mm × 150 mm, 8 μm, 1000 Å pore size) (Agilent Technologies, Santa Clara, CA) at room temperature. HPLC was performed at a constant flow rate of 0.4 mL/min using a binary solvent system. Solvent A consisted of 0.1% formic acid in water and solvent B was acetonitrile. The gradient started at 30% B for 1 min and was increased linearly to 42% over 10 min. After that, the gradient was increased to 75% in 1 min and held at 90% B for 2 min prior to column re-equilibration. The Q-

■

EXPERIMENTAL DETAILS Chemical Reagents and Materials. Eight kDa linear methoxyl poly(ethylene glycol) (8kPEG, >99% purity) and 40kDa 2-branched methoxyl poly(ethylene glycol) (40KPEG, >99% purity) were purchased from NOF Corporation (Tokyo, Japan). PEGylated adnectin (ATI-1072, 15 mg/mL, 99% purity) was supplied by Bristol-Myers Squibb Research & Development (Waltham, MA). The purity of ATI-1072 was determined with size-exclusion HPLC, and there was no free PEG detected in the stock solution. The protein sequence of ATI-1072 was illustrated in a previous publication, with a cysteine residue near the C-terminus.28 Type I reagent- grade water was prepared with a Milli-Qplus ultrapure water system (Millipore Corp, Bedford, MA). HPLC-grade acetonitrile, ethanol, methanol, isopropanol, and n-butanol were purchased from Burdick & Jackson, Inc. (Muskegon, MI). Human serum albumin (HSA) and all standard buffer solutions were purchased from Sigma−Aldrich Co. (St. Louis, MO). MALDI matrixes, including sinapinic acid, 4-hydroxy-alpha-cyanocinnamic acid, 2,5-dihydroxy benzoic acid, and 2-cyano-hydroxylsinipic acid, were purchased from Bruker Daltonics (Billerica, MA), with purities of >99%. Blank rat plasma and brain, liver, kidney, spleen, and skeletal muscle (quadriceps and diaphragm) 7643

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Article

Figure 1. Representative structure of ATI-1072, showing the protein, linker, and PEG components.

suspension was centrifuged at 14 000g for 10 min and a portion of supernatant (50−80 μL) was injected into LC-MS/MS for analysis. MALDI-TOF Mass Spectrometry Analysis. Multiple injections of rat plasma samples were made into an Agilent 1100 HPLC system, and the peaks corresponding to 40kPEG and ATI-1072 were collected. The HPLC column and solvent gradient are described in the previous section for the qualitative LC-MS analysis. The combined fractions were concentrated under nitrogen flow to ∼5 μL, and aliquots (1 μL) were taken for the analysis with MALDI-TOF mass spectrometry (Bruker Daltonics, Germany). Several matrix materials were tested, including sinapinic acid, 4-hydroxy-alpha-cyanocinnamic acid, 2,5-dihydroxy benzoic acid, and 2-cyano-hydroxyl-sinipic acid. 2-Cyano-hydroxylsinipic acid appeared to give the most-satisfactory sensitivity for both 40kPEG and ATI-1072, and therefore was used as the matrix for the entire MALDI analysis. For each measurement, the sample of interest (1 μL) was well mixed with the matrix solution (1 μL of 10 mM solution in acentonitrile/water) and the mixture was loaded onto a conductive plate. After drying the sample spots in air, the plate was inserted in the MALDITOF mass spectrometry. Desorption was triggered by a 50 Hz, 337-nm N2 laser. The laser power, ion extraction delay, and other relevant parameters were optimized to achieve the maximum sensitivity.

trap mass spectrometer was operated in positive electrospray ionization (ESI) mode. Nitrogen was used as the nebulizer and auxiliary gas. The desolvation temperature was 300 °C and the source temperature was 150 °C. The declustering potential (DP) of the ionization source was tested in the range of 50− 200 to optimize for the dissociation of 40kPEG and ATI-1072. For the quantitative analysis of 40kPEG and ATI-1072, the following MRM (multiple reaction monitoring) transitions of PEG-specific ions were selected: m/z 133.1 → m/z 89.1, m/z 177.1 → m/z 89.1, m/z 353.2 → m/z 89.1, m/z 397.2 → m/z 133.1, and m/z 441.2 → m/z 133.1. To improve assay sensitivity, the summation of the signals from five transitions was used for standard curve and sample analysis.29 The standard curves with the concentrations ranged from 0.05 to 20 μg/mL were fitted to a quadratic regression model using weighting factors 1/x.30,31 Concentrations of 40kPEG and ATI1072 were calculated using the Analyst software (Version 1.6, AB Sciex) based on their MRM responses and standard curves. Briefly, after the data acquisition of standards and unknown samples, a quantitative method was created in the quantitate mode for both PEG and ATI-1072. A calibration curve was then developed based on the MRM responses of the standards and concentrations of both analystes in the unknown sample were interpolated. Tissue Distribution of ATI-1072 in Rats. All animal studies were conducted under the standards recommended by the Guide for the Care and Use of Laboratory Animals, and were approved by the Institutional Animal Care and Use Committee. Male Sprague−Dawley (SD) outbred albino rats (n = 3) received a single subcutaneous injection of ATI-1072 at a target dose level of 15 mg/kg. Animals were fed Certified Rodent Diet 5002 (PMI Nutrition International, Inc.), provided water, and maintained on a 12-h light/dark cycle. Serial blood samples were collected from animals via tail vein into tubes containing K2EDTA before dosing and at 4, 24, and 48 h after dosing. All rats were euthanized after 48 h post dose, and brain, liver, kidney, spleen and skeletal muscle (quadriceps and diaphragm) were collected, split, and snap frozen. Preparation of Rat Tissue Samples for LC-MS/MS Analysis. Rat tissue samples (∼0.5 g of brain, liver, kidney, spleen, and skeletal muscle) were mixed with 2× weight of water and were homogenized with an OMNI-TH01 homogenizer (OMNI International, Kennesaw, GA). Plasma samples and tissue homogenates (0.5−1 mL) were spiked with 8kPEG as internal standard with a final concentration of 1 μg/mL and were extracted with n-butanol as described above. The combined supernatant was dried in a lyophilizer and the residue was suspended in 0.20 mL n-butanol/water (1:1 v/v) containing 200 ng/mL human serum albumin (HSA). The

■

RESULTS AND DISCUSSIONS LC-ESI-MS Analysis of PEG and PEGylated Protein. Initial qualitative characterization of 40kPEG and PEGylated protein was conducted with an LTQ/Orbi-trap mass spectrometer. An adnectin (MW 11 kDa) that has no pharmacologic target and is devoid of pharmacological activity was used as a tool protein.28 It was coupled to a 40-kDa 2branched PEG to give a PEGylated adnectin (ATI-1072) through the reaction between cysteine residue at the Cterminus of adnectin and the maleimide ring of 40kPEG (Figure 1).28 Figure 2 illustrates HPLC chromatogram and the corresponding mass spectra of 40kPEG and ATI-1072 under standard ESI-MS conditions. Both 40kPEG and ATI-1072 gave complex spectra in the mass range between m/z 650 and m/z 1800 with no single ion peak resolved, presumably because of the heterogeneous nature of PEG and multiple charges of each molecule. The spectra were consistent with what have been described previously25,26 and were not suitable for deconvolution. Thus, neither qualitative nor quantitative information could be extracted from the conventional LC-ESI-MS analysis. LC-ESI-MS Analysis of PEG and PEGylated Protein with In-Source CID. PEG and PEGylated proteins can undergo collision-induced dissociation in the ionization source 7644

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Article

In-source CID of ATI-1072 gave a mass spectrum similar to that of 40kPEG in which a series of PEG-specific fragment ions were observed (Figure 3b). No fragment ions from the dissociation of the protein moiety were observed, even at high DP values, which is surprising, because, in the previous studies, only the ions from the dissociation of peptides rather than PEG were observed.26,32 This could be due to the different charge distribution on the PEGylated proteins, which was known to affect the dissociation.32 The similar in-source CID spectra of 40k PEG and ATI-1072 allowed us to identify and quantify both species simultaneously with mass spectrometer after HPLC separation. Given that metabolic cleavage of PEGylated protein following dose administration to rats has been confirmed,23 simultaneous detection and quantification of PEG and PEGylated protein is desired in order to understand their pharmacokinetics and tissue distribution. Development of LC-MS/MS Method for 40kPEG and ATI-1072. Chromatography Development. Since the quantitative analysis of both 40kPEG and ATI-1071 relies on the PEGspecific ions, an accurate measurement requires adequate chromatographic resolution of the analytes. Any overlap of the peaks on chromatography would influence concentration determination of each species. A variety of chromatography techniques have been applied in the area of polymer and protein separation, including ion-exchange, size-exclusion, and reverse-phase chromatography.33 Ion-exchange and size-exclusion chromatography typically require ion-containing mobile phases that are not suitable for direct injection in mass spectrometry. Thus, our screening effort mainly focused on reverse-phase columns where the analytes can be directly analyzed by mass spectrometry after separation. It was identified that large particle pore size (1000 Å) of an HPLC reverse-phase column was essential for good separation, satisfying peak shape and minimal carryover.25 As detailed in the Experimental Section, the current HPLC method can resolve not only 40kPEG and ATI-1072, but also PEGs with different MW (including 20K, 10K, 8k, and 5k PEGs), allowing for the detection and quantitation of PEG degradants (Figure 2b). Sample Preparation. A bioanalytical method with low LLOQ is highly desired in order to fully characterize the distribution and fate of PEG components following dose administration to animals. Some tissues might have a low distribution of PEG, and it will be challenging to accurately measure the PEG concentration with a low-sensitivity assay. These tissues include brain and choroid plexus that generally have low exposures to drugs but draw most concerns on the potential toxicity of PEG accumulation.21 An extraction method with high recovery of PEGylated protein is important to improve assay sensitivity. Solid-phase extraction,34 protein precipitation,35 and immunoaffinity precipitation36 have been successfully utilized previously to extract PEGylated proteins from a biological matrix. During method development, we found that protein precipitation gives satisfactory recovery for both 40kPEG and ATI-1072, from either rat plasma or rat tissue homogenates of various organs. This benefits from the fact that PEG has good solubility in both organic and aqueous solutions. Other methods for enrichment either gave low recovery (SPE) or required reagents that were not readily available (immunoaffinity). A series of organic solvents including methanol, acetonitrile, ethanol, iso-propanol, and n-butanol, were tested for the extraction of 40kPEG and ATI-1072 from

Figure 2. LTQ/Orbi total ion chromatograms of (A) 40kPEG and ATI1072 and (B) PEGs with different MW; (C) a representative LTQ/ Orbi mass spectrum of 40kPEG.

(in-source CID) of a mass spectrometer.26,27,32 The technique has been utilized previously for the elucidation of PEGylation site32 and identification of PEG-related peaks.27 In particular, it was employed in the quantitative analysis of PEGylated peptide, where peptide fragments generated from in-source CID were used as the surrogate ions.26 Since all PEG molecules essentially have the same structure except for the different numbers of the oxyethlene repeating units, in theory, all PEGs and PEGylated proteins could lead to the same PEG-specific fragments. Indeed, when applying in-source CID with increasing declustering potentials (DP) on an API 4000 Qtrap mass spectrometer, the 40kPEG generated a spectrum of singly or doubly charged ions in the range of 80−1000 m/z at DP ≥ 100 (Figure 3a). On the other hand, no PEG-specific ions were formed from in-source CID when DP value was below 50. The m/z difference between singly charged ions was 44 Da, which is characteristic of PEG, and the number of oxyethylene units in each ion can be easily determined based on the m/z (e.g., m/z 353.1 has 8 oxyethylene units). All these ions were generated from the dissociation of PEG, and their intensity was dependent on the DP values at the ionization source. A few source-formed ions were tested as the precursor ions for the conventional CID and for potential use as surrogate ions for the quantitation of PEG. All selected ions were degraded further to multiple daughter ions by losing one or more oxyethylene (Figure 4). Thus, the concentration of 40kPEG can be determined through the MRM of these surrogate ions. 7645

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Figure 3. In-source CID/Q1 mass spectra of (A)

Article

40k

PEG and (B) ATI-1072 at DP 200 with an API 4000 Q-trap mass spectrometer.

biological matrixes, with n-butanol giving the best extraction recovery (>80%). Assay Evaluation. Standards and QCs were prepared by spiking 40kPEG or ATI-1072 in blank rat plasma or tissue homogenates. No signals were observed either in the blank rat plasma or tissue homogenates around the retention times of 40k PEG and ATI-1072, suggesting no interference from endogenous substances. Assay linearity was evaluated within the concentration range of 0.05−50 μg/mL. However, using standard procedures of protein precipitation, linearity could not be established either at low (20 μg/mL) concentrations for both 40kPEG and ATI-1072. The nonlinearity at low concentration was probably due to the nonspecific binding of PEG with sample vials or injection needles, which could be solved with the addition of HSA (200 ng/mL) during sample reconstitution, which is a common approach to prevent nonspecific binding.37 No ion suppression to 40kPEG or ATI-1072 was observed from the co-injection of HSA during mass detection. The nonlinearity at high concentrations of 40kPEG was likely due to the saturation during in-source CID; therefore, the upper limit was set at 20 μg/mL. The coefficient of determination (R2) was 0.99 with a quadratic regression model (weighing factor 1/x), using 9 concentrations of calibration standards in the range between 0.05 and 20 μg/mL (see Figures S1 and S2 in the Supporting Information).

The DP value was optimized to produce maximum in-source CID signals for 40kPEG and ATI-1072. In-source ions (m/z 133.1, 177.1, 353.1, 397.2, 441.2) with relatively high abundance were subjected to further dissociation in the positive ESI-MRM mode and the summation of five transitions (see the Experimental Section) were employed to quantify both 40kPEG and ATI-1072 (LLOQ 0.05 μg/mL for both analytes). The approach of using the summation of multiple MRM transitions to increase the sensitivity is now increasingly considered in the LC-MS/MS analysis.29 In the current study, using the total signals from 5 MRM transitions instead of 1 transition increased the assay sensitivity by at least 3-fold. The intraday and interday assay validations were tested for 40kPEG or ATI1072 at six concentrations of QC samples in rat plasma, ranging from 0.05 μg/mL to 15 μg/mL. As shown in Tables S1−S4 in the Supporting Information, the assay accuracy (% Dev) was within 16% of the nominal values for the analytes at various concentrations, and the precision was within 18% CV for both 40k PEG and ATI-1072 either in the intraday or interday assay. Quantitative Analysis of 40kPEG and ATI-1072 in Rat Plasma and Tissues. The LC-MS/MS method, coupled with in-source CID, has significant advantages in the characterization of the disposition and fate, as well as pharmacokinetics of PEGylated proteins. The method was highly sensitive to PEGrelated species with LLOQ of 0.05 μg/mL, ∼200-fold lower than the previously reported NMR method.23 This allows for the quantitative analysis of PEG components in various tissues, 7646

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Article

(Figure 5) matched with the reference standards. Both the retention time and MALDI spectra indicated that the peaks

Figure 5. MALDI mass spectra of the HPLC fractions eluded at (A) 7.6 min and (B) 8.8 min, collected from rat plasma following a single subcutaneous dose of ATI-1072. MALDI spectrum was obtained using 2-cyano-hydroxyl-sinipic acid as the matrix material.

Figure 4. Representative in-source CID/MRM chromatograms of 40k PEG, ATI-1072, and internal standard 8kPEG (1 μg/mL) in plasma with an API 4000 Q-trap mass spectrometer: (A) plasma LLOQ, 0.05 μg/mL concentration for both 40kPEG and ATI-1072, mass transition m/z 177.1 to m/z 89.1; (B) 5 μg/mL concentration for both 40kPEG and ATI-1072, mass transition m/z 441.2 to m/z 133.1; and (C) 5 μg/ mL concentration for both 40kPEG and ATI-1072, mass transition m/z 177.1 to m/z 89.1.

eluted at 7.6 and 8.8 min were indeed 40KPEG and intact ATI1072, respectively. However, MALDI data could not clearly determine whether or not the linker was attached to the PEG after protein degradation, because of the heterogeneous nature of the PEG polymer. Large PEG molecules with slight difference on the MW cannot be differentiated on MALDI, given the broad PEG peak. The concentrations of 40kPEG and ATI-1072 are summarized in Figure 6 and in Table S5 in the Supporting Information. In plasma, intact ATI-1072 comprised the majority of PEG-related materials, with the level increasing continuously up to, and possibly beyond, 48 h post-dose, suggesting slow adsorption after subcutaneous dose administration. The 40kPEG was present in plasma at 4 h post-dose, with the relative concentration increasing over time. In tissues collected at 48 h post-dose, relatively high concentrations of PEG-related materials were observed in the liver, kidney, and spleen. In contrast to plasma, 40kPEG was the predominant species in these tissues, with tissue/plasma concentration ratios of 6.2, 1.2, and 1.8 for the kidney, liver, and spleen, respectively, suggesting a moderate to high degree of accumulation of 40kPEG. Low amounts of 40kPEG and ATI-1072 were observed in brain and diaphragm muscle. In quadriceps, levels of PEG-related materials were below LLOQ. The rat study was not designed to fully characterize the disposition of ATI-1072, but rather to demonstrate the feasibility of utilizing PEG-specific method in addressing the disposition of PEG-related species. The presence of 40kPEG in plasma and tissues suggests the degradation of ATI-1072 following dose administration to rats, given that free PEG was absent in the dosing solution. This is consistent with a previous study where the dissociation of PEGylated insulin was observed following an intravenous dose to rats.23 It was hypothesized

particularly in the central nervous system where low exposures to systemically administered PEG-related materials are usually expected. It also allows for tissue distribution studies of PEGylated proteins at therapeutically relevant low doses. In addition, since the current method is based on the dissociation of PEG moiety, it would be able to detect any PEG-related materials, including the degraded products following dose administration of PEGylated proteins, given that adequate chromatographic separation was achieved for PEGs with different size (Figure 2b). Finally, the method enables simultaneous quantification of different PEG-related materials in complex matrices, including plasma and tissue homogenates, and thus, a full spectrum of quantitative tissue distribution of each PEG component can be obtained. To demonstrate the potential application of the method, a single dose study of ATI-1072 was conducted in rats. Plasma and various tissues were collected and analyzed with LC-MS/ MS coupled with in-source CID. Two PEG-related peaks (eluded at 7.6 and 8.8 min; see Figure 4) were observed in plasma and tissue homogenates of brain, liver, spleen, kidney, and diaphragm, with the retention times consistent with free 40k PEG and ATI-1072, respectively. Since the MW of PEGrelated materials cannot be determined with ESI-MS, the HPLC peaks corresponding to PEG and ATI-1072 in the plasma samples were collected and analyzed with MALDITOF-MS. The resulting MALDI spectrum of each fraction 7647

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Article

Figure 6. Mean concentration of 40kPEG and ATI-1072 in rat plasma (4, 24, and 48 h post-dose) and tissues (48 h post-dose), following a single subcutaneous dose of ATI-1072 to rats. (2) Woodnutt, G.; Violand, B.; North, M. Curr. Opin. Drug Discovery Dev. 2008, 11, 754−761. (3) Lin, J. H. Curr. Drug Metab. 2009, 10, 661−691. (4) Vaddady, P. K.; Meibohm, B. Biopharmaceutics, Pharmacokinetics, And Pharmacodynamics of Protein Therapeutics; John Wiley & Sons: New York, 2010; pp 689−700. (5) Bailon, P.; Won, C.-Y. Expert. Opin. Drug Delivery 2009, 6, 1−16. (6) Pasut, G.; Guiotto, A.; Veronese, F. M. Expert. Opin. Ther. Pat. 2004, 14, 859−894. (7) Jevsevar, S.; Kunstelj, M. Half-Life Extension through PEGylation; Wiley−Blackwell, 2012; pp 41−61. (8) Jevsevar, S.; Kunstelj, M.; Porekar, V. G. Biotechnol. J. 2010, 5, 113−128. (9) Jevsevar, S.; Kusterle, M.; Kenig, M. Methods Mol. Biol. 2012, 901, 233−246. (10) Kang, J. S.; Deluca, P. P.; Lee, K. C. Expert Opin. Emerg. Drugs 2009, 14, 363−380. (11) Hamidi, M.; Rafiei, P.; Azadi, A. Expert Opin. Drug Discovery 2008, 3, 1293−1307. (12) Descotes, J.; Gouraud, A. Expert Opin. Drug Metab. Toxicol. 2008, 4, 1537−1549. (13) Nechansky, A.; Kircheis, R. Expert Opin. Drug Discovery 2010, 5, 1067−1079. (14) Veronese, F. M.; Mero, A. BioDrugs 2008, 22, 315−329. (15) Pelham, R. W.; Nix, L. C.; Chavira, R. E.; Cleveland, M. V.; Stetson, P. Aliment. Pharmacol. Ther. 2008, 28, 256−265. (16) Hamidi, M.; Azadi, A.; Rafiei, P. Drug Delivery 2006, 13, 399− 409. (17) Yamaoka, T.; Tabata, Y.; Ikada, Y. J. Pharm. Sci. 1994, 83, 601− 606. (18) Gaberc-Porekar, V.; Zore, I.; Podobnik, B.; Menart, V. Curr. Opin. Drug Discovery Dev. 2008, 11, 242−250. (19) Webster, R.; Didier, E.; Harris, P.; Siegel, N.; Stadler, J.; Tilbury, L.; Smith, D. Drug Metab. Dispos. 2007, 35, 9−16. (20) Bendele, A.; Seely, J.; Richey, C.; Sennello, G.; Shopp, G. Toxicol. Sci. 1998, 42, 152−157. (21) Johanson, C.; Stopa, E.; McMillan, P.; Roth, D.; Funk, J.; Krinke, G. Toxicol. Pathol. 2011, 39, 186−212. (22) Cheng, T.-L.; Chuang, K.-H.; Chen, B.-M.; Roffler, S. R. Bioconjugate Chem. 2012, 23, 881−899. (23) Elliott, V. L.; Edge, G. T.; Phelan, M. M.; Lian, L.-Y.; Webster, R.; Finn, R. F.; Park, B. K.; Kitteringham, N. R. Mol. Pharmaceutics 2012, 9, 1291−1301. (24) Su, Y.-C.; Chen, B.-M.; Chuang, K.-H.; Cheng, T.-L.; Roffler, S. R. Bioconjugate Chem. 2010, 21, 1264−1270. (25) Huang, L.; Gough, P. C.; DeFelippis, M. R. Anal. Chem. 2009, 81, 567−577.

that the degradation of PEGylated insulin was mediated by proteases after being uptaken into various organs to give free PEG. The PEG was then released slowly into circulation and eventually excreted in urine.23 Even though ATI-1072 and PEGylated insulin have different protein sequences and linker structures, the degradation to PEG, which is presumably mediated by proteases, appears to be a major clearance pathway for both conjugates.

■

CONCLUSIONS In summary, we have developed a robust and sensitive LC-MS/ MS method for the quantitative analysis of PEG-related materials. Both 40kPEG and ATI-1072 underwent in-source CID and MRM, and their concentrations were determined simultaneously in biological samples. A single dose rat study was conducted to test the feasibility of utilizing the method in addressing PEG-related issues. Both 40kPEG and intact ATI1072 were observed in plasma and tissue homogenates, and the identity of each species in plasma was further confirmed with MALDI analysis. The presence of 40kPEG in plasma and tissues suggests the degradation of ATI-1072, following dose administration to rats. The method allows further studies for a thorough characterization of tissue distribution and metabolic fate, as well as pharmacokinetic profiles of PEGylated proteins.

■

ASSOCIATED CONTENT

S Supporting Information *

Representative standard curves of 40kPEG and ATI-1072, interday and intraday validation results, and plasma and tissue concentrations of 40kPEG and ATI-1072. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: (609)-252-3767. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Leader, B.; Baca, Q. J.; Golan, D. E. Nat. Rev. Drug Discovery 2008, 7, 21−39. 7648

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649

Analytical Chemistry

Article

(26) Li, H.; Rose, M. J.; Holder, J. R.; Wright, M.; Miranda, L. P.; James, C. A. J. Am. Soc. Mass Spectrom. 2011, 22, 1660−1667. (27) Warrack, B. M.; Redding, B. P.; Chen, G.; Bolgar, M. S. Anal. Bioanal. Chem. 2013, 405, 4283−4287. (28) Maxwell, B. D.; Cao, K.; Bonacorsi, S.; Sivaraman, L.; Wang, B.; McNerney, M.; Wang, L.; Christopher, L. J.; Fung, N. K. E.; Wang, J.; Arnold, M. E. J. Labelled Compd. Radiopharm. 2013, 56, 492−494. (29) Manjunath Swamy, J.; Kamath, N.; Radha Shekar, A. K.; Srinivas, N. R.; Kristjansson, F. Biomed. Chrom. 2010, 24, 528−534. (30) D’Avolio, A.; Simiele, M.; Siccardi, M.; Baietto, L.; Sciandra, M.; Oddone, V.; Stefani, F. R.; Agati, S.; Cusato, J.; Bonora, S.; Di Perri, G. J. Pharm. Biomed. Anal. 2011, 54, 779−788. (31) de Matos Barbosa, M. Comput. Methods Prog. Biomed. 1994, 43, 239−245. (32) Lu, X.; Gough, P. C.; DeFelippis, M. R.; Huang, L. J. Am. Soc. Mass Spectrom. 2010, 21, 810−818. (33) Howard, J. W.; Kay, R. G.; Pleasance, S.; Creaser, C. S. Bioanalysis 2012, 4, 2971−2988. (34) Yang, Z.; Ke, J.; Hayes, M.; Bryant, M.; Tse, F. L. S. J. Chromatogr., B 2009, 877, 1737−1742. (35) Wu, S. T.; Ouyang, Z.; Olah, T. V.; Jemal, M. Rapid Commun. Mass Spectrom. 2011, 25, 281−290. (36) Xu, Y.; Mehl, J. T.; Bakhtiar, R.; Woolf, E. J. Anal. Chem. 2010, 82, 6877−6886. (37) Sun, B. F.; Kobayashi, H.; Le, N.; Yoo, T. M.; Drumm, D.; Paik, C. H.; McAfee, J. G.; Carrasquillo, J. A. J. Nucl. Med. 2000, 41, 318− 326.

7649

dx.doi.org/10.1021/ac501507g | Anal. Chem. 2014, 86, 7642−7649