Presented at ASMS 2015

IntroductionNanoparticles (NP) are defined as particles between 1 and 100 nanometers in size. In the pharmaceutical field, NP technology is applied in colloidal drug delivery systems, which can improve drug bioavailability, increase circulation time, and deliver drugs to specific locations in the body. Polyethylenimine (PEI) is an amino-based hydrophilic polymer with repeating units composed of an amine group and an ethyl aliphatic (CH2CH2) spacer. There are three types of commercial PEI: linear PEI, branched PEI and dendrimer PEI (Figure 1). PEI can be formulated as nanoparticles through electrostatic interactions with anionic polysaccharides. Due to its large molecular weight and cationic character, it is challenging to quantify PEI in biological matrices. In this paper we developed a fast, reliable and high-throughput LC-MS/MS method for quantification of a large branched PEI (25 kDa) in dog plasma.

ConclusionA selective, fast and robust assay quantifying PEI in dog plasma was developed using a protein precipitation extraction and an LC-MS/MS analysis method. This is the first published LC-MS/MS assay for quantitative analysis of PEI in dog plasma.

AB SCIEX is a registered trademark of Life Technologies Corporation

A Fast, Selective and High-Throughput LC-MS/MS Assay for the Quantification of Polyethylenimine in Dog PlasmaAihua Liu, Troy Voelker, Bryce Ashby, Benjamin Johnson, Cassidy Hatch, Jeremy Thompson, Scott Reuschel and Min MengCovance, Salt Lake City, Utah

Covance is the drug development business of Laboratory Corporation of America® Holdings (LabCorp®). Findings in this poster weredeveloped by scientists who at the time were affiliated with the LabCorp Clinical Trials or Tandem Labs brands, now part of Covance.

Figure 2. Q1 MS full scan of 25k Da PEI.

Figure 3. Same product ions (m/z 113.1, 130.0 and 156.1) from different precursor ions of PEI (m/z 345.3, 388.4 and 431.5, respectively).

Figure 4. LC column comparison.

Figure 5: Non-specific binding issue.

Figure 9. Typical calibration curve of PEI in dog plasma K2EDTA.

Figure 6. Chromatogram of extracted blank matrix.

Figure 7. Chromatogram of IS Only Sample.

MethodologyLC-MS/MSMass Spec: AB SCIEX™ 5000 Source and ionization: ESI under positive modeColumn: Inertsil® ODSFlow Rate: 0.400 mL/minMobile Phase: A: mixture of FA, ammonium formate and water B: mixture of FA, IPA and MeCN LC Program: GradientSource Temperature: 500°C

MS Monitoring Parameters:

Analyte MS/MS transition tR (min)

PEI 388.5→133.0 0.92

Internal Standard (IS) 595.5→227.2 1.60

Protein Precipitation Extraction (PPE)Aliquot: 100 μL dog plasma K2EDTABuffer: Formic acidPrecipitation Solvent: mixture of IPA and MeOHDilution Solvent: 5% FA in water:MeOH (2:1 v/v)

ResultsMethod Development ▶ MS tuning. Unlike small molecules, PEI is a polymer and therefore

has no unique molecular weight (MW), but rather, a distribution of MW. Under Q1 full scan, a cluster of mass spectra was observed (Figure 2) due to the MW distribution and multiple charge states, making MS tuning challenging. During early method development, it was identified that iso-propanol was a necessary component of the mobile phase in order to produce a cluster of mass spectra, and PEI needed a higher DP value relative to small molecules. The most abundant Q1 ions observed were at m/z 302.3, 345.3 and 388.4; in this cluster of mass spectra, the 43 amu mass difference between adjacent ions is due to the C2H5N unit (Figure 2), and these different precursor ions all produce the same product ions (Figure 3). The transition of 388.4→133.0 was selected for quantification because it had a cleaner background relative to the other two transitions.

Figure 1. The chemical structures of Linear PEI; Branched PEI and Dendrimer PEI (Note: The PEI used in this paper is a branched PEI with an average MW of 25k Da).

▶ Optimization of LC conditions. Development of LC conditions that produced acceptable chromatography was a major challenge that was encountered and overcome. In order to achieve acceptable chromatography, various HPLC columns, including ODS, C18, PFP, C8, Oligo-RP, Polar-RP, Polymer, Phenyl-Hexyl, Hypercarb, Ultra-IBD, Phenyl, C18-PFP, Hypersil GOLD Dash and HILIC columns, were screened. PEI showed either no elution or no retention on most columns, and only the ODS column showed both acceptable peak shape and retention of PEI (Figure 4). Various mobile phases were also screened, and it was determined that iso-propanol is a necessary mobile phase component to obtain symmetrical peaks. Because a stable labeled internal standard was not available for this polymeric compound, various analogue internal standards were also screened. The analogue TL-025-PMBS demonstrated sufficient ability to track PEI through both extraction and LC-MS/MS analysis, and no significant contributions were observed between the analyte and IS.

▶ Optimization of extraction methods. Severe non-specific binding issues were also encountered during the early stages of method development. Although neat solutions exhibited good sensitivity by LC-MS/MS (Figure 5), very poor sensitivity was observed after spiking the neat solution into blank matrix extact (Figure 5). Different extraction conditions (water-bath sonication, heating, etc) and different pH conditions (acidic, basic and neutral) were investigated and it was found that only a strong acidic environment can prevent instantaneous binding and a severe loss of sensitivity (Figure 5). A variety of extraction techniques including PPE, LLE and SPE were screened to maximize extraction recovery; however, only PPE provided sufficient recovery for analysis.

Method PerformanceThe method performance was evaluated through the determination of linearity, reproducibility, accuracy and precision, and selectivity of this method.

▶ Selectivity. Interferences were detected in blank dog plasma at the retention time of 0.92 min (Figure 6), which increased the assay LLOQ from 100 ng/mL (LLOD) to 500 ng/mL so that the interference peak was less than 20% of LLOQ. There was no contribution between the analyte and the IS (Figure 7). Representative chromatograms for dog plasma samples containing 100 ng/mL (LLOD), 500 ng/mL (LLOQ) and 100,000 ng/mL (ULOQ) of PEI are presented in (Figure 8).

▶ Linearity, precision and accuracy. The calibration curve was successful validated over the concentration range of 500-100,000 ng/mL in dog plasma (Figure 9). Correlation coefficients for all calibration curves were greater than 0.99. The deviations of the back-calculated concentrations from their nominal values were within ± 15% and the relative standard deviations were below 10% (Table 1). The accuracy (% bias) observed at the LLOD, LLOQ, Low, Low-Medium, Medium and High QC levels was -12.7 %, -4.0 %, 2.8 %, -1.9 %, 2.0% and 0.8 %, respectively, and the precision (% CV) at these levels was 6.14 %, 3.0 %,, 1.8 %,, 2.6%, 2.5 % and 5.2 %, respectively (Table 2).

Figure 8. Chromatograms of LLOD (100 ng/mL), LLOQ (500 ng/mL) and ULOQ (100,000 ng/mL)

Table 1. Back-Calculated Concentrations of Calibration Standards

Standards Std 0 Std 1 Std 2 Std 3 Std 4 Std 5 Std 6 Std 7 Std 8

Nominal Concentration

100 200 500 1,000 2,000 10,000 60,000 90,000 100,000

Concentration Found

114 195 464 895 1820 8440 61200 87500 91600

90.3 195 496 1140 2010 11500 67000 92500 99100Mean 102 195 480 1020 1915 10500 64100 90000 95400

%Theoretical 102 97.5 96.0 102 95.8 105 107 100 95.4%Bias -2.0 -2.5 -4.0 2.0 4.2 5.0 7.0 0.0 -4.6

Note: All concentrations are expressed as ng/mL.

Table 2. Accuracy and Precision of Quality Control Samples

Quality Control

LLOD QC (100)

LLOQ QC (500)

Low QC (5000)

Low-Medium QC (10000)

Medium QC (40000)

High QC (80000)

85.8 495 5150 9670 39400 76700

84.0 489 5160 9740 40000 8230083.6 474 5010 9790 41300 8180082.3 456 5060 9560 42100 8530094.6 477 5210 9800 40700 8300093.5 490 5260 10300 41500 74200

Mean 87.3 480 5140 9810 40800 80600S.D. 5.36 14.4 92.8 256 1000 4200

%CV 6.14 3.00 1.8 2.6 2.5 5.2%Theoretical 87.3 96.0 102.8 98.1 102 100.8

%Bias -12.7 -4.0 2.8 -1.9 2.0 0.8n 6 6 6 6 6 6

%Theoretical 87.3 96.0 102.8 98.1 102 100.8%Bias -12.7 -4.0 2.8 -1.9 2.0 0.8

n 6 6 6 6 6 6 Note: All concentrations are expressed as ng/mL.