Anal. Chem. 2006, 78, 1636-1643
Identification of Carboxyl-Terminal Peptide Fragments of Parathyroid Hormone in Human Plasma at Low-Picomolar Levels by Mass Spectrometry Chao-Xuan Zhang,* Brittney V. Weber, Jerdravee Thammavong, Thomas A. Grover, and David S. Wells
NPS Pharmaceuticals, 383 Colorow Drive, Salt lake City, Utah 84108
For decades, researchers have tried to identify the primary structures of circulating carboxyl-terminal parathyroid hormone (C-PTH) peptide fragments that may be present at only picomolar levels in human plasma. Although immunoassays and radiosequencing techniques have provided valuable fragment characterizations, no analysis has successfully determined their exact primary structures. In this work, for the first time, four human C-PTH peptide fragments, hPTH(34-84), hPTH(37-84), hPTH(38-84), and hPTH(45-84), have been identified from human plasma using MS-based methods. C-PTH peptide fragments were isolated from plasma samples by immunoaffinity extraction. The eluate was analyzed by capillary LC fractionation followed by MALDI-TOF-MS or by online coupling of nano-LC with ESI-TOF-MS. Both the MALDI- and the ESI-based approaches were capable of detecting C-PTH peptide fragments in human plasma at 99% by capillary electrophoresis-UV absorbance at 214 nm. All other synthesized C-PTH peptide fragments were purchased as lyophilized powders from Bachem (Torrance, CA). The purity and peptide content ranges of these fragments were 96-99 and 72-91%, respectively. Goat anti-hPTH(39-84) antibody gel suspension (10 mg Ab/mL) was purchased from Immutopics (San Clemente, CA) and kept at 4 °C. MALDI matrixes [R-cyano-4-hydroxycinnamic acid (HCCA), 2,5-dihydroxybenzoic acid (DHB), sinapinic acid (SA)] were purchased from Bruker Daltonics (Billerica, MA). Nitrocellulose membrane (0.45 µm) was from Bio-Rad (Hercules, CA). Trifluoroacetic acid (TFA) was from Pierce (Rockford, IL) and formic acid was from Sigma (St. Louis, MO). All other reagents were of HPLC or analytical grade. Human Plasma. Control plasma, pooled from healthy human donors, was purchased from Lampire Biological (Pipersville, PA). Plasma samples from patients with chronic renal insufficiency were obtained from Massachusetts General Hospital. Plasma samples were also collected from healthy postmenopausal women who were given 100 µg of recombinant hPTH by subcutaneous injection or by a 15-min intravenous infusion. All samples were shipped frozen on dry ice and stored at -80 °C until use. Immunoaffinity Extraction. The standard protocol, resulting from exploratory studies, was as follows. Prior to extraction, NaCl was added to the plasma (117 mg/mL, 2 M). The plasma samples were centrifuged at 14 000 rpm and then filtered using a 10-µm filter (USB Corp., Cleveland, OH). C-PTH peptides were extracted with the goat anti-hPTH(39-84) antibody gel packed in-house on a disposable C4 ZipTip (Millipore, Bedford, MA). A 50-µL aliquot of the antibody gel suspension was loaded onto the ZipTip using a PHD 2000 syringe pump (Harvard Apparatus, Holliston, MA) at a flow rate of 20 µL/min. Plasma (3 mL) was loaded onto the antibody gel-filled ZipTip at 20 µL/min. The gel-filled ZipTip was sequentially washed with 400 µL each of phosphate-buffered saline (PBS, pH 7.2), 2 M NaCl, and deionized water at 100 µL/min. C-PTH peptides (C-PTH peptide fragments and intact PTH) were eluted by pipetting 20 µL of 10 mM TFA aqueous solution through the gel-filled ZipTip. The eluate was kept at -20 °C until further analysis. Analytical Chemistry, Vol. 78, No. 5, March 1, 2006
1637
Capillary LC Fractionation Followed by MALDI-TOF-MS. LC fractionation was performed on an Agilent 1100 binary capillary LC system equipped with a six-port manual injector (Agilent, Palo Alto, CA) and a C18 column (300-µm i.d., 25-cm length, Vydac, Hesperia, CA). Mobile phase A was deionized water containing 150 µM TFA, and mobile phase B was 100 µM TFA in acetonitrile. The sample injection volume was 20 µL and the gradient profile was 5% B for 10 min followed by a linear ramp to 100% B over 20 min (5 µL/min). One-minute fractions (5 µL) were collected manually from 22 to 25 min. A sample (0.5-1 µL) from each LC fraction was directly applied to a MALDI target and analyzed with an OmniFLEX MALDI-TOF-MS instrument equipped with a 337nm nitrogen laser (Bruker Daltonics, Billerica, MA). The MALDI target was coated in advance with 120 µL of matrix solution (15 mg/mL HCCA in 2-propanol-acetone (7:3, v/v) containing 5 mg/ mL nitrocellulose). The remainder of the LC fraction was further cleaned up and concentrated using a C18 ZipTip (Millipore) and the Millipore recommended washing and elution solutions, except that a PHD 2000 syringe pump was used for peptide elution at 0.3 µL/min. The liquid drop eluting from the ZipTip was applied directly onto a matrix-coated MALDI target. Mass spectra were collected in positive ion reflectron mode over a mass-to-charge (m/z) range of 1000-6000. Each spectrum was the sum of 50 laser shots at multiple positions within each spot. The resulting spectra were mass-calibrated using a two-point external calibration scheme with hPTH(70-84), m/z 1588, and hPTH(53-84), m/z 3512. Nano-LC On-Line Coupling with ESI-TOF-MS. Sample analyses were performed on an Ultimate nano-LC system equipped with a Famos autosampler and a Switchos column switching device (LC Packings-Dionex, Amsterdam, The Netherlands). The nano-LC was connected on-line to a LCT Premier ESI-TOF-MS system fitted with a nanospray interface and lock spray (WatersMicromass, Manchester, U.K.). The aqueous mobile phase A contained 0.5% formic acid in water-acetonitrile (98:2, v/v), and the organic mobile phase B contained 0.5% formic acid in acetonitrile. Sample (20 µL) was loaded, using a user-programmed mode without any sample loss, on a Symmetry C18 trap column (0.18-mm i.d., 22.5-mm length, Waters, Milford, MA) at a flow rate of 10 µL/min for 10 min with mobile phase A. Peptides were eluted from the trap column to a C18 analytical column (75-µm i.d., 25-cm length, Vydac) with a linear gradient of 0-23% B over 7 min followed by isocratic elution with 23% B (flow rate 300 nL/ min). LC effluent was delivered to the LCT Premier through a nanospray emitter (fused silica, 90-µm o.d., 20-µm i.d.). MS data were acquired in positive ion W (double reflectron) mode over an m/z range of 450-1500 at an acquiring rate of 1 spectrum/s. The LCT Premier was calibrated over the same m/z range using a sodium formate solution (water-propanol (1:9, v/v) containing 5 mM NaOH and 0.5% (v/v) formic acid) infused by syringe pump through the lock spray device at a flow rate of 100 nL/min. During sample data acquisition, the LCT Premier acquired data from the lock spray device for 1 s out of every 10 s to effect real-time mass calibration. RESULTS AND DISCUSSION Immunoaffinity Extraction of C-PTH Peptides (C-PTH Peptide Fragments and Intact PTH) from Plasma. Our goal was to be able to detect C-PTH peptide fragments at the 10 pM 1638
Analytical Chemistry, Vol. 78, No. 5, March 1, 2006
level. Therefore, prior to MS analysis, C-PTH peptides present at picogram per milliliter levels in plasma need to be concentrated. At the same time, high-abundance plasma matrix proteins and peptides (∼65 mg/mL in total, including 45 mg/mL albumin)17,19,33 need to be reduced. Immunoaffinity extraction was chosen for plasma pretreatment because it has higher selectivity than solidphase or liquid-liquid extraction approaches. Immunoaffinity procedures have been used to pretreat tryptic digests of model proteins34,35 and blood samples22-25 to improve the detection sensitivity of MALDI-MS and LC-ESI-MS for specific proteins and peptides. In this work, immunoaffinity extraction was initially performed in a bulk-phase format. Plasma (0.5 mL) was incubated with 100 µL of antibody gel suspension for 2 h. Filter centrifugation, in a tube fitted with a 10-µm filter, separated the antibody gel with the extracted C-PTH peptides from bulk plasma. The antibody gel was then washed with PBS buffer and water. C-PTH peptides were eluted with 100 µL of 10 mM TFA. A 20-µL aliquot of the eluate was injected into the capillary LC-UV system. Capillary LC-UV analysis showed that C-PTH peptides were recovered from spiked plasma. While, in theory, only C-PTH peptides containing residues in the range of 39-84 of the PTH molecule should be captured with the antibody, a large quantity of matrix proteins and peptides were coextracted via nonspecific binding and eluted 2 min after C-PTH peptides (data not shown). These matrix components could occupy antibody binding sites and ultimately reduce the MS signals obtained for the peptides of interest. Selective removal of major plasma proteins, like albumin, could improve the sensitivity for low-abundance peptides.36 However, when a protein precipitation step with ammonium sulfate was incorporated prior to the immunoaffinity extraction, PTH peptide peaks also decreased significantly in the capillary LC-UV chromatograms. These results are consistent with those reported in the literature.21 Various approaches to remove plasma matrix proteins, including ethanol precipitation, urea disruption of protein-peptide interactions, and solid-phase capture of plasma proteins have been reported but have actually led to decreases in signal intensity of low-abundance peptides in the MALDI-MS analysis of blood samples.21 In fact, many proteins and peptides are removed along with the albumin using depletion techniques.37,38 In an alternative approach, solid NaCl was added to the plasma prior to extraction, and the antibody gel was washed additionally with NaCl solution. With this approach, the large protein peak was significantly reduced in capillary LC-UV chromatograms. Thus, NaCl was efficient in suppressing nonspecific binding. The incubation time of plasma with antibody gel, ranging from 1 h to overnight, did not markedly affect the recovery of C-PTH peptides. The concentration of NaCl added to plasma and in the washing (33) Anderson, N. L.; Anderson, N. G. Mol. Cell. Proteomics 2002, 1, 845-867. (34) Fenaille, F.; Tabet, J.-C.; Guy, P. A. Anal. Chem. 2002, 74, 6298-6304. (35) Raska, C. S.; Parker, C. E.; Sunnarborg, S. W.; Pope, R. M.; Lee, D. C.; Glish, G. L.; Borchers, C. H. J. Am. Soc. Mass Spectrom. 2003, 14, 10761085. (36) Ramstrom, M.; Hagman, C.; Mitchell, J. K.; Derrick, P. J.; Hakansson, P.; Bergquist, J. J. Proteome Res. 2005, 4, 410-416. (37) Adkins, J. N.; Varnum, S. M.; Auberry, K. J.; Moore, R. J.; Angell, N. H.; Smith, R. D.; Springer, D. L.; Pounds, J. G. Mol. Cell. Proteomics 2002, 1, 947-955. (38) Gundry, R. L.; Fu, Q.; Van Eyk, J. E.; Cotter, R. J. In 53rd ASMS Conference, TP544.
Figure 1. MALDI-TOF-MS spectra of spiked plasma after immunoaffinity extraction and capillary LC fractionation. (A) Plasma (0.5 mL) spiked with 0.79 nM hPTH(70-84) using bulk-phase immunoaffinity extraction. (B) Same as (A) except immunoaffinity extraction by column (200-µL pipet tip) format. (C) Same as (B) except the plasma volume was 2.5 mL and the concentration of hPTH(70-84) was 79 pM (10% the concentration of (B)). (D) Same as (C) except the plasma volume was 1 mL and the column was a C4 ZipTip. Other conditions are given in the Experimental Section. Peak: (2) hPTH(70-84), m/z 1588.
solution to reduce nonspecific binding was in the broad range of 1-4 M. The recoveries of C-PTH peptides following immunoaffinity extraction of spiked plasma (77 nM each peptide) were 57 and 56% for hPTH(53-84) and hPTH(70-84), respectively, as determined by capillary LC-UV chromatography. There was no noticeable difference in the recoveries from plasma or PBS, suggesting that plasma proteins did not affect recovery. The antibody capacity of 100 µL of gel suspension was sufficient to capture at least 43 pmol of C-PTH peptides (i.e., 0.5 mL plasma × 77 nM × 56% × 2 ) 43 pmol), even though some binding sites were likely occupied by plasma proteins and peptides. The recovery of C-PTH peptides, as determined by capillary LC-UV alone, is not sufficient for the evaluation of the immunoaffinity extraction step since coextracted components may suppress or complicate MS detection. The goal for the extraction step was to achieve the greatest MS signal intensity for C-PTH peptides, not necessarily the highest yield for the extraction of these peptides. Therefore, MALDI-TOF-MS was used as a readout for further optimization of the immunoaffinity extraction procedures. The eluate from the immunoaffinity extraction step was injected into the capillary LC-UV system, and 5-µL fractions were collected. Aliquots (0.5 µL) from each fraction were applied to a MALDI target. The MS spectrum in Figure 1A clearly indicates that hPTH(70-84) could be detected from a plasma sample spiked at 0.8 nM using the bulk-phase extraction format. However, a detection limit of 0.8 nM was significantly higher than our target
goal of 10 pM. In an effort to enhance sensitivity, the extraction step was changed from the bulk format to a column format. A 200-µL pipet tip with a 10-µm filter disk at the bottom served as a column for the affinity gel. Peptides were extracted as described in the Experimental Section. PTH peptide signal intensity increased dramatically (compare Figure 1A to B, bulk format vs column format). Note that, in the protocol for Figure 1A, only one-fifth of the immunoaffinity eluate (20 µL out of 100 µL) was injected into the capillary LC, while in the protocol for Figure 1B, the entire eluate (20 µL) was injected. Immunoaffinity extraction in column format allows peptides to be eluted in a small volume, and therefore, the entire sample volume can be used in subsequent steps. However, the 10-fold gain in signal intensity cannot be ascribed only to sample concentration (i.e., smaller volume). The increased signal intensity obtained with the column format was also from improved antibody capturing efficiency for C-PTH peptides and improved washing efficiency to reduce the retention of nonspecific peptides/proteins. The MS spectrum in Figure 1C indicates that hPTH(70-84) could be unambiguously detected at the 79 pM level in plasma using the column format for extraction. However, in this sample, the sample volume had been increased from 0.5 to 2.5 mL and plasma matrix components became more prominent in the MS spectrum. The subsequent use of a C4 ZipTip instead of a pipet tip for preparation of the antibody column increased the PTH peptide MS signal intensity and reduced nonspecific binding (Figure 1D). The tiny amount of C4 packing Analytical Chemistry, Vol. 78, No. 5, March 1, 2006
1639
material (0.6 µL) in the ZipTip served as a filter. Its original function (i.e., capturing large proteins, capacity 3 µg) is negligible for the plasma volume used. MALDI-TOF-MS Analysis of Capillary LC Fractions. Capillary LC of the immunoaffinity eluant was used as an additional purification and concentration step for C-PTH peptides. In MALDIMS analysis, the sample is often mixed with an equal volume of matrix solution in a vial. Then a portion of the mixture (0.5-1 µL) is spotted on the MALDI target.39 This common practice resulted in signal loss due to sample dilution and handling, which is not desirable in the analysis of trace peptides. To avoid sample loss, a two-step spotting method (i.e., application of 0.2-0.5 µL of matrix solution to each target spot followed by the application of 0.5 µL of sample) was tested. In another method, 120 µL of matrix solution was applied to the entire target, forming a homogeneous coating over the surface, followed by the application of 0.5 µL of sample. Coating the entire target with matrix resulted in a 2-3fold increase in MS signal intensity when compared to the twostep spotting method (data not shown). MALDI matrixes (HCCA, DHB, SA) were tested at various concentrations (5-50 mg/mL), with or without nitrocellulose at various concentrations (2-15 mg/ mL). A matrix solution prepared with 2-propanol-acetone (7:3) containing 15 mg/mL HCCA and 5 mg/mL nitrocellulose was found to provide the best sensitivity for C-PTH peptides. The lowest detectable concentration (LDC, S/N ) 3) for samples prepared in deionized water (no extraction) was in the range of 2-5 nM (1-3 fmol) for hPTH(73-84), hPTH(70-84), hPTH(6484), hPTH(53-84), and hPTH(35-84). With the optimized immunoaffinity extraction, capillary LC fractionation, and MALDI-TOF-MS procedures detailed in the Experimental Section, hPTH(73-84), hPTH(70-84), hPTH(6484), and hPTH(53-84) could be detected in plasma at lowpicomolar levels (LDC 5-8 pM) (Figure 2). The analytical procedure started with a 3-mL plasma sample and ended with a 5-µL LC fraction. Thus, the sample was concentrated 600-fold based on volume. When compared with the MS signal intensity obtained for C-PTH peptides prepared in deionized water, the overall recovery for hPTH(73-84) and hPTH(70-84) at lowpicomolar levels in plasma was ∼80%. The overall recovery for hPTH(64-84) and hPTH(53-84) was ∼50%. Overall recovery depends not only on the number of analyte molecules physically extracted in the immunoaffinity extraction (i.e., physical recovery) but also on the coextracted plasma matrix components that may suppress MS signal intensity. It is the overall recovery, rather than the physical recovery, that determines the signal intensity of C-PTH peptides in MS analysis. Note that, in the above analyses, an aliquot of only 0.5-1 µL of the 5-µL LC fraction was applied to the MALDI target. The repeated application of an LC fraction on a MALDI target spot (0.5 µL each time) did not increase signal intensity. Evaporation of a 5-µL LC fraction down to 0.5-1 µL and the application of the entire fraction to a MALDI target actually decreased the signal intensity. Treatment of an LC fraction with a C18 ZipTip improved the signal intensity of C-PTH peptides (Figure 3). After ZipTip treatment, the MS signal of plasma matrix peaks around m/z 4100 and 5900 disappeared (Figure 3A and B), while the C-PTH peptide (39) Napper, S.; Kindrachuk, J.; Olson, D. J.; Ambrose, S. J.; Dereniwsky, C.; Ross, A. R. Anal. Chem. 2003, 75, 1741-1747.
1640 Analytical Chemistry, Vol. 78, No. 5, March 1, 2006
Figure 2. MALDI-TOF-MS spectra of plasma spiked with 8.3 pM hPTH(73-84), 7.9 pM hPTH(70-84), 7.4 pM hPTH(64-84), and 7.7 pM hPTH(53-84). Conditions as in Figure 1D except the plasma volume was 3 mL. Peaks: (1) hPTH(73-84), m/z 1274; (2) hPTH(70-84), m/z 1588; (3) hPTH(64-84), m/z 2231; (4) hPTH(53-84), m/z 3512. hPTH(73-84) and hPTH(70-84) coeluted in the same LC fraction (A) while hPTH(64-84) and hPTH(53-84) coeluted (B).
peaks (marked as P1-3) were enhanced. Mass peaks commonly occurring in the control plasma of healthy subjects were identified to differentiate them from potential C-PTH peptide mass peaks in patient plasma and postdosing samples. The C-PTH peptide peaks were not detected in control plasma and predose plasma samples with the MALDI-based approach. Using the FindPep Tool program at http://au.expasy.org/ tools/findpept.html, mass peaks P1, P2, and P3 were identified solely as hPTH(34-84), hPTH(37-84), and hPTH(38-84), respectively. The differences between the observed mass (average) and the calculated average value of the assigned peptides were in the range of 0.2-2.5 Da. To achieve the sensitivity needed, high laser power was used, and no clear isotopic resolution was achieved for peptides of MW >4000. The mass accuracy (2 Da for peptides of MW over 5000, 400 ppm) and mass resolution (5000) were suitable for the preliminary matching of PTH peptide masses. Two of the C-PTH peptide fragments found in patients with chronic renal insufficiency, hPTH(37-84) and hPTH(3884), were also detected in the plasma of a healthy subject receiving recombinant hPTH (Figure 3D). There was an additional mass peak (P4 in Figure 3D), identified as hPTH(45-84). The four identified C-PTH peptide fragments were confirmed by reanalyses of the plasma samples using nano-LC-ESI-TOF-MS with much higher mass accuracy (10 ppm; see next section).
Figure 3. MALDI-TOF-MS spectra (A, B) of a plasma sample from a patient with chronic renal insufficiency and (C, D) plasma from a healthy male volunteer (pooled collections of 0.5, 1.0, and 1.5 h after subcutaneous administration of 100 µg of hPTH). (A, C) Direct application of 0.5 µL from an LC fraction (5 µL) to a MALDI target. (B, D) Same LC fractions (amount remaining from (A) and (C), respectively) treated with C18 ZipTips and then reanalyzed by MALDI-TOF-MS. P1, P2, P3, and P4 were identified as hPTH(34-84) m/z 5473, hPTH(37-84) m/z 5156, hPTH(38-84) m/z 5043, and hPTH(45-84) 4380, respectively.
Nano-LC-ESI-TOF-MS Analysis. The direct infusion of PTH peptide standards at different water-acetonitrile percentages indicated a significant effect on MS signal intensity. MS signal intensity was optimal when the acetonitrile content was in the range of 20-25% (v/v). Therefore, the LC gradient given in the Experimental Section was used to elute C-PTH peptides at optimal acetonitrile percentages. It was also found that better sensitivity was achieved with 0.5% (v/v) formic acid in the mobile phase instead of 0.1%. In direct infusion analysis of C-PTH peptide standards prepared with the optimal solvent (25% acetonitrile, 0.5% formic acid), the LDCs were 2-4 nM for hPTH(70-84), hPTH(53-84), and hPTH(35-84) using a 0.6-µL sample volume (mass sensitivity, 1-2 fmol). The detection sensitivity for C-PTH peptide standards using ESI-TOF-MS was similar to that obtained with MALDI-TOF-MS. However, the on-line coupling of nano-LC with the ESI-TOF-MS minimized sample transfers, fully used the sample, concentrated C-PTH peptides in narrow LC peaks, and therefore improved the detection sensitivity (Figure 4). The LDCs in plasma were 3, 2, 5, and 6 pM for hPTH(70-84), hPTH(5384), hPTH(35-84), and hPTH(1-84), respectively. Mass resolution and accuracy were 10 000 and 10 ppm, respectively. Isotopic clusters were clearly separated, even for hPTH(1-84) (MW 9425). Nano-LC-ESI-TOF-MS spectra of plasma from a patient with chronic renal insufficiency revealed the C-PTH peptide fragments hPTH(34-84), hPTH(37-84), hPTH(38-84), and hPTH(45-84) (Figure 5). Each peptide was detected as several ions of different
Figure 4. Nano-LC-ESI-TOF-MS spectra of human plasma spiked with 8 pM each of hPTH(70-84) and hPTH(53-84), 7 pM hPTH(35-84), and 10 pM hPTH(1-84). Plasma (3 mL) was pretreated with immunoaffinity extraction. Each spectrum was the sum of scans over the chromatographic peak of the C-PTH peptide. Each of these C-PTH peptides eluted at different retention times. Each PTH peptide had several ion peaks of different charge states. Only the most intense charge-state ion is presented.
charge states. The observed monoisotopic masses were highly reproducible as calculated from the peptide’s different chargedstate ions (RSD hPTH(37-84) > hPTH(38-84) > hPTH(45-84). Plasma concentrations for these peptides were estimated to be in the range of 10-100 pM based on the MS signal intensity of a known concentration of hPTH(53-84) spiked in plasma. These four C-PTH peptide fragments (40) Spengler, B. J. Am. Soc. Mass Spectrom. 2004, 15, 703-714.
1642 Analytical Chemistry, Vol. 78, No. 5, March 1, 2006
Figure 6. Nano-LC-ESI-TOF-MS spectra of pooled plasma from six healthy postmenopausal women who received 100 µg of hPTH by subcutaneous injection. The plasma samples (0.5 mL from each subject) were collected 2 h after administration. hPTH(45-84), hPTH(38-84), hPTH(37-84)m and hPTH(34-84) were identified. Several ions of different charge states were observed for each peptide. Only one charge state is presented.
were also found in a healthy postmenopausal woman who received 100 µg of recombinant hPTH by intravenous infusion. Plasma samples collected 30 and 35 min after the start of the infusion were pooled for analysis. The relative amounts of hPTH(34-84), hPTH(37-84), hPTH(38-84), and hPTH(45-84) were in the same order of relative abundance as found in patients with renal insufficiency (estimated concentration range 5-40 pM). Accurate quantification of C-PTH peptides in plasma is dependent on the ionization efficiency (MS) and percent recovery of each peptide during the extraction and chromatography steps. Synthetic standards of each peptide for calibration curves will be necessary for the accurate quantification of C-PTH peptides in plasma. These four C-PTH peptide fragments were also found in plasma samples from healthy postmenopausal women following the subcutaneous injection of 100 µg of recombinant hPTH (Figure 6). Two hours after the injection, the most abundant peptide fragment was hPTH(45-84) and the least abundant was hPTH(34-84) (estimated concentration range 5-20 pM). This is opposite to that found in patients with chronic renal insufficiency and a subject receiving an intravenous dose of hPTH. This may be due to differences in renal clearance of the C-PTH fragments in the patients with impaired kidney function or greater importance of peripheral metabolism to the circulating ratio of C-PTH peptides following subcutaneous injection of hPTH. Pooled human plasma samples were purchased from a commercial source in several lots and analyzed as controls. Plasma samples collected from healthy women prior to dosing with hPTH were also analyzed. hPTH(45-84) was not found in any control or predosing samples. hPTH(34-84), hPTH(37-84), and hPTH(38-84) were found in some control or predosing samples as small peaks (estimated concentrations
