Measurement of Pancreatic Polypeptide and Its Peptide Variant in Human Serum and Plasma by Immunocapture-Liquid-Chromatography-Tandem Mass Spectrometry. Reference Intervals and Practical Assay Considerations.

Human PP (APLEPVYPGDNATPEQMAQYAADLRRYINMLTRPRY-NH2, ≥95%) was acquired from ANASPEC (Fremont, CA). Anti-PP polyclonal antibodies were purchased from Bachem (Torrance, CA), Phoenix Pharmaceuticals, Inc. (Burlingame, CA) and Miravista/SSI products (Indianapolis, IN). Dynabeads® M-280 Tosyl activated magnetic beads were purchased from Life Technologies (Carlsbad, CA). Sitagliptin was acquired from Biovision (San Francisco, CA) and phosphoramidon disodium salt from Sigma-Aldrich (St. Louis, MO). Serum quality control samples were pooled human serum or plasma samples depleted from PP with Anti-PP antibody. Stable isotope labeled PP was used as internal standard (PPIS): APL*EPVYPGDNATPEQMAQYAADL*RRYINML*TRPRY-NH2 (L*=[13 C6, 15 N]) presenting a mass shift of 21 Da.

The N-terminus cleaved PP (PP3-36) LEPVYPGDNATPEQMAQYAADLRRYINMLTRPRY-NH2 was used as standard. PPIS and PP3-36 were synthesized at United Peptide (Herndon, VA) acquired with a certificate of amino acid analysis and net weight per vial. All other reagents were of highest purity commercially available. Solvents were of HPLC grade, purchased from JT Baker (Phillipsburg, NJ).

Detailed description of the method is provided in the Supplementary Materials. All studies using human samples were approved by the Institutional Review Board (IRB) of the University of Utah.

Sample preparation was performed on a liquid handler Janus (Perkin Elmer, Waltham, MA). Quality control samples and calibrators for PP and PP3-36 were prepared in serum depleted from PP using anti-PP antibody capture. Coupling of the antibody to beads was performed according to the manufacturerâ�?�?s recommendations (see section 1.5 of the Supplementary Materials). 450 μL of serum or plasma was aliquoted and mixed with 100 pg of PPIS, and 250 μL of Trizma buffer 3X (150 mM Trizma, 300 mM NaCl, NaN3 0.09%, pH 7.6) into a lobind™ 96 well plate (Eppendorf, Hauppauge, NY). 1 μg antibody (Phoenix Pharmaceuticals) linked to beads was added to samples, quality controls and calibrators, and incubated with shaking for 1 h at 4°C. Beads were washed three times with Trizma buffer 1X (50 mM Trizma, 100 mM NaCl, NaN3 0.09%, pH 7.6). Peptides were eluted with 40 �?µL of buffer (5% acetic acid in 0.0025% BSA); the elution was transferred to a PCR 96-well plate, and analyzed by LC-MS/MS.

Detailed descrition of the method is provideded in section 1.1 of the Supplementary Materials. In brief, the HPLC system consisted of binary 1260 series HPLC pump (Agilent Technologies, Santa Clara, CA), and a HTC PAL auto sampler (LEAP Technologies, NC). Chromatographic separation was performed at 70°C on a Poroshell 300 SB-C18, 75 mmx2.1 mmx2.7 μm column with attached guard cartridge (Agilent, Santa Clara, CA). Mobile phases (A) 10 mM acetic acid in water and (B) 10 mM acetic acid in acetonitrile where delivered at 0.7 mL/min. Analysis was performed on an AB SCIEX TripleQuad™ 5500 Mass Spectrometer in positive ion mode using electrospray ionization (ESI) and MRM mode of acquisition (Supplementary Table 1). Data were processed using software Analyst® 1.6.1 (AB SCIEX). A parent ion was selected (Supplementary Figure 1) and two mass transitions were monitored for each compound as quantifier and qualifier respectively: m/z 837.3→953.0 and m/z 837.3→411.3 for PP; m/z 803.7→853.7 and m/z 803.7→197.2 for PP3-36 and m/z 841.3→953.0 and m/z 841.3→411.3 for the PPIS (Supplementary Table 2 and Figure 2).

1. Presence of the PP variants in serum samples was assessed using theoretical MRM transitions corresponding to immunoreactive PP fragments previously reported in literature: PP2-36, PP3-36 and PP4-36 [17]; in addition, mass transitions corresponding to the PP with two oxidized methionines were monitored [18].

Presence of the truncated PP was assessed using theoretical MRM transitions of the expected peptides (Supplementary Table 2) by analysis of 24 plasma samples of healthy donors (fasting and postprandial); theoretical MRM transitions were determined using Protein Prospector software v 5.12.0 (http://prospector.ucsf.edu/prospector/mshome.htm). Potential enzymatic action of DPP-IV or naprilysin during blood processing was evaluated using plasma samples of five donors by assessing degradation of PP and PP3-36 untreated and treated with sitagliptin and phosphoramidon (DPP-IV and naprilysin inhibitors, respectively), a detailed description of these enzymatic assays has been included in the Supplementary material, section 1.2.

2. Method validation experiments were performed according to The Clinical Laboratory Improvement Amendments (CLIA) guidelines. Performance of the assay was established based on evaluation of imprecision, Lower Limit Of Quantification (LLOQ), Limit Of Detection (LOD), Upper Limit Of Linearity (ULOL), recovery, carryover, ion suppression, interference, accuracy, evaluation of the anti-PP antibodies from commercial companies. Within-run, between run per day and total imprecision of the method was evaluated by analyzing triplicates of 5 samples containing 30, 50, 100, 500 and 1000 pg/mL of PP and PP3-36 repeated over a period of 5 days. Plasma pools containing 30 and 500 pg/mL of PP and PP3-36 were used as quality control samples. Pools of human plasma patient samples depleted from PP and PP3-36 using PP antibody were prepared spiking known standard concentration of PP and PP3-36 to contain 5, 10, 20, 30 and 40 pg/μL of each analyte to evaluate the LLOQ and LOD of the method analyzed over a 4 days period in three replicates each per day. The lowest concentration for which precision was within 15% and accuracies were within 20% of the expected value, was set as the lower limit of quantitation. Limit of detection was determined as the lowest concentration at which the peaks of the analytes were present in both mass transitions at the expected retention time and signal to noise ratio for the quantitative mass transition was ≥ 5. Samples for linearity test were pools of serum samples (remaining aliquots of patient samples submitted to ARUP laboratories for testing) depleted from PP using anti-peptide antibody and spiked with a defined concentration of PP and PP3-36 for seven levels total: 50, 250, 1000, 2500, 5000, 50000 pg/mL. The samples were aliquoted in tubes, stored at â�?�?70°C and thawed prior to the analysis. They were analyzed in duplicates over a period of 5 days. The highest concentration at which precision was within 10% and accuracy within 20% of the expected values was considered to be the Upper Limit of Linearity (ULOL) of the method. Method recovery was determined with 11 serum samples analyzed as is and spiked with 650 pg/mL of PP and PP3-36. Samples were tested in duplicate. Difference between the observed and expected concentrations gave a measure of recovery of the sample preparation. Ion suppression was evaluated using post column infusion method [19]. A serum sample spiked with 50 pg of PP and other with PP3-36 were processed and injected into a flow of analyte with 1000 ng/mL at flow rate of 7 μL/min. Presence of potential interferences was evaluated by monitoring ratios of primary and secondary MRM transitions for PP and PP3-36; the acceptability range for the ratio was �?± 30% (0.7-1.3) based on comparison of the quantitative results using the two MS/MS transitions. Effects of icterus (bilirubin), hemolysis and lipemia were assessed in serum spiked with PP and PP3-36 mixed with the interfering substances: red blood cells (86 mg/L), lipids (320 mg/L) and bilirubin (173 mg/L). Method accuracy was evaluated by comparing the IC-LC-MS/MS method with a RIA [20]. Regression analysis was performed for 61 plasma samples using Deming regression. Commercial antibodies were purchased from three Vendors: Bachem (Torrance, CA); Phoenix Pharmaceuticals, Inc. (Burlingame, CA) and Miravista/SSI Products (Indianapolis, IN). Suitability of the antibodies for the method was assessed by analysis of serum samples containing 30 pg/mL of PP and PP3-36. The samples were analyzed in triplicate and concentrations were calculated based on calibration curve generated with calibration standards analyzed with the set of samples. The observed concentrations were evaluated using a one way ANOVA. The antibody incubation time was evaluated with one antibody selected by analysis of serum pooled from patients with a concentration of PP and PP3-36 at 100 pg/mL with incubation times of 1, 2, 3, 4 and 5 hours. Samples were analyzed in duplicate. The Incubation time corresponding to the maximum recovery was selected for the method. Carryover of the method was evaluated using experimental design outlined in EP Evaluator (Data Innovations, Burlington, VT). The method was evaluated by analyzing serum samples containing low (L) level (30 pg/mL) and high (H) level (100000 pg/mL) concentrations of PP and PP3-36 in the following sequence: L1, L2, L3, H1, H2, L4, H3, H4, L5, L6, L7, L8, H5, H6, L9, H7. Interferences were evaluated as described in the Supplementary material section 1.2.2

Serum samples spiked with PP and PP3-36 were evaluated for storage stability at room temperature, refrigerator (4°C) and freezer (-20°C) after 1, 3, 7, 14, 21 and 28 days. Serum controls were stored in -70°C freezer and analyzed over 28 days. A within day evaluation was also performed for 2 h, 4 h, 8 h and 24 h at room temperature, refrigerator (4°C), and on ice (0°C). Samples were placed in -70°C freezer and analyzed in a single batch. Freeze-thaw cycles were evaluated. Blood collection tubes-type suitability evaluation was performed using potassium EDTA, lithium-heparin, serum, Serum Separator (SST) and PPACK (thrombin inhibitor) following manufacture guidelines.

Reference intervals for PP total PPT (PP plus PP3-36) were established using serum or plasma from 118 self-reported healthy adults, ages 20-60, with no known current or past pancreatic medical conditions. Ratio of the concentrations of PP and PP3-36 in the circulation fasting and postprandial were calculated. A detailed description of the methodology is included in section 1.3.1 of the Supplementary Materials.

Association of PP concentrations with Body Fat Percentage (BF %) and Body Mass Index (BMI) was performed on 104 participants. A description of the BMI and BF% measurements methodology is described in Supplementary material, section 1.3.2

Blood samples were collected from 13 adult volunteers (ages 20-59, fasting and postprandial on three separate days). Participants did not have previous pancreatic medical condition and were advised avoid any strenuous physical activity 24 h before blood collection. Effect of the following meals was evaluated: first day rich in proteins and fat, second day rich in carbohydrates, and third day five subjects from the group consumed a meal rich in carbohydrates and fat. A detailed description of the methodology of analysis is described in the section 1.4 of the Supplementary Materials.

Figure 1 shows a chromatogram with primary and secondary mass transitions of the analytes and the PPIS.

The search for the truncated fragments showed that PP3-36 was present in all analyzed samples in addition to the intact PP; while peaks of other variants were not detected (Supplementary Figure 3). The LLOQ and LOD for PP and PP3-36 was 10 and 5 pg/mL, respectively; the ULOQ was 1,000 pg/mL (Figure 2 and Supplementary Table 3). Data on the imprecision of the assay are listed in Tables 1 and 2.

We found PP and PP3-36 enrichment efficiency was comparable among three antibodies assessed for the method (Supplementary Table 5). One hour was selected as time of incubation (Supplementary Table 6). Recovery of PP was 83-98% and 85-119% for PP3-36 (Supplementary Table 7). No ion suppression was observed at retention time of both peptides.

Storage stability showed no significant degradation in samples stored frozen for a month at -20°C. Approximately 50% degradation of both peptides was observed after 24 h storage at 4°C. Complete degradation was observed after 24 h storage at room temperature. In brief, acceptable storage conditions for samples are: room temperature up to 2 h, refrigerated at 4°C for up to 8 h, on ice up to 24 h, and frozen (at -20°C or below) more than one month (Supplementary Figure 4). Controls kept at -70°C during 28 days and one freeze-thaw cycle showed no significant effect in peptide degradation.

PP and PP3-36 were stable in frozen serum and plasma samples collected in Na-Heparin, SST, PPACK or EDTA tubes (Supplementary Table 8). The results for PPACK also demonstrated no improvement in stability of PP and PP3-36.

No interference was observed in samples icteric, hemolyzed and lipidemic (Supplementary Table 9). Method comparison between this IC-LC-MS/MS method and a commercial RIA method (n= 61) showed poor agreement between the methods using Deming regression (Figure 3): fasting, LC-MS/MS=0.063 RIA+24.3, R2=0.027; postprandial, LC-MS/MS=0.661 RIA ��? 24.5, R2=0.630.

Ratio of PPT concentrations for the same samples analyzed by RIA and LC-MS/MS ranged between 1.0 -79.0 and 0.8 - 6.0 in samples from individuals fasting and postprandial, respectively.

The nonparametric reference intervals established as central 95% are: fasting: <265 pg/mL of PPT and postprandial PPT (30 minutes after meal): 20-639 pg/mL. Supplementary Figure 5 shows chromatograms for PP and PP3-36 in a fasting healthy donor and postprandial donor, and in a negative control. The ratio PP/PP3-36 fasting was 0.08-6.45 and postprandial 0.19-2.01 (Supplementary Table 10). No differences in concentration of PP and PP 3-36 were found between collected blood samples treated and untreated with inhibitors for DPP-IV and neprilysin, enzymes involved in the fragmentation of other peptides of the NPY family (Supplementary Table 11). BMI and percent fat were calculated and compared with the PPT using a simple regression analysis (Supplementary Figure 6). Associations of PPT with BMI and body fat index in women and man were evaluated, no correlation between PPT and both indexes was observed.