Dual-Monoclonal, Sandwich Immunoassay ... - Semantic Scholar

Clinical Chemistry 57:6 849–855 (2011)

Endocrinology and Metabolism

Dual-Monoclonal, Sandwich Immunoassay Specific for Glucose-Dependent Insulinotropic Peptide1-42, the Active Form of the Incretin Hormone Jason S. Troutt,1† Robert W. Siegel,1† Jinbiao Chen,1 John H. Sloan,1 Mark A. Deeg,1 Guoqing Cao,1 and Robert J. Konrad1*

BACKGROUND: Glucose-dependent insulinotropic peptide (GIP) is an incretin peptide secreted by intestinal K cells that stimulates insulin secretion in a glucosedependent manner. It is secreted as an active, intact 42–amino acid peptide GIP1-42, which is rapidly degraded by dipeptidyl peptidase 4 to GIP3-42, which is inactive. There is currently no described monoclonal antibody– based sandwich immunoassay to quantify concentrations of GIP1-42, the active form of the peptide.

understanding of the role of GIP1-42 in regulating glucose-dependent insulin secretion. © 2011 American Association for Clinical Chemistry

CONCLUSIONS: The use of an N-terminal–specific monoclonal antibody in a sandwich ELISA format provides a robust and convenient method for measuring concentrations of GIP1-42, the active form of the incretin hormone. This ELISA should help to improve our

Glucose-dependent insulinotropic peptide (GIP)2 is an incretin hormone secreted by intestinal K cells in response to dietary intake of fat and carbohydrate (1–3 ). Together with glucagon-like peptide-1 (GLP-1), which is secreted from intestinal L cells, GIP plays a major role in stimulating glucose-induced insulin secretion (4, 5 ). Following ingestion of fat and carbohydrate, GIP is secreted into the plasma as an intact, active 42 amino acid peptide (GIP1-42), which augments glucose-induced insulin secretion from pancreatic ␤ cells (1–3 ). The circulating half-life of GIP1-42, however, is relatively short because it is cleaved in the circulation by dipeptidyl peptidase 4 (DPP4) at its N-terminal alanine residue to give rise to GIP3-42, which is inactive (6 – 8 ). GIP1-42 plays a critical role in the incretin effect (1–3 ). It is believed that GIP1-42 secretion by K cells of the small intestine in response to an oral glucose load may partly account for the increased insulin secretion observed compared to what occurs when a comparable glucose load is delivered intravenously (9 ). There have also been suggestions that GIP1-42 may improve ␤-cell survival and regulate ␤-cell viability and antiapoptotic effects (10 –14 ). Because of the recognized crucial role of GIP in augmenting glucose-induced insulin secretion, there have been efforts to develop robust immunoassays to measure GIP1-42. Unfortunately, because GIP1-42 differs from the inactive form of the hormone, GIP3-42, by the loss of only 2 amino acids at its N-terminus, it has proved difficult to develop monoclonal antibodies selective for GIP1-42. As a result, existing ELISAs measure both GIP1-42 and GIP3-42 and

1

2

METHODS:

To create a sandwich ELISA for GIP1-42, we generated a monoclonal antibody specific for the intact N-terminus of the peptide, which was further optimized to increase its affinity. We used this antibody as a conjugate antibody in a sandwich ELISA and paired it with an anti–total GIP capture monoclonal antibody to create a dual monoclonal sandwich ELISA for GIP1-42.

RESULTS:

The sandwich ELISA was highly specific for GIP1-42 and did not recognize GIP3-42. The ELISA demonstrated a broad dynamic range and a lower limit of quantification of 5 ng/L. Using the ELISA, we were able to show that GIP1-42 concentrations in healthy volunteers increased dramatically in the postprandial state compared to the fasting state. GIP1-42 values were correlated with total GIP values overall; however, there was substantial interindividual variation.

Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN. J.S. Troutt and R.W. Siegel contributed equally to the work, and both should be considered as first authors. * Address correspondence to this author at: Eli Lilly and Company, Indianapolis, IN 46285, USA. Fax 317-276-5281; e-mail [email protected]. Received November 23, 2010; accepted March 24, 2011. Previously published online at DOI: 10.1373/clinchem.2010.159954

†

Nonstandard abbreviations: GIP, glucose-dependent insulinotropic peptide; GLP-1, glucagon-like peptide-1; DPP4, dipeptidyl peptidase 4; VH, Ig heavy chain variable region; VL, Ig light chain variable region; scFv: single chain fragment variable; CDR, complementarity-determining region; MSD, MesoScale Discovery; TBST, Tris buffered saline ⫹ tween; GIPR: GIP receptor; OGTT, oral glucose tolerance test.

849

thus give an indication of total GIP concentrations (8, 15, 16 ), whereas a previously described RIA for GIP1-42 used a polyclonal antibody (8 ). LC-MS assays specific for GIP1-42 have been described, but these assays have relied on either direct LC-MS analyses or the use of immunoprecipitation of total GIP followed by LC-MS quantification of GIP1-42 (8, 15, 16 ). Although these assays have been shown to be both accurate and precise, their complexity and expense, and the high level of operator expertise required each time the assay is performed, have prevented their implementation into most laboratories. In addition, low concentrations of GIP1-42 observed during the fasting state (ⱕ20 ng/L) have made quantification a challenge for these assays (8, 15, 16 ). To develop a highly specific and robust immunoassay for active GIP, we generated a monoclonal antibody that specifically targeted the N-terminus of GIP1-42. Afterward, in vitro– directed evolution was applied to improve the affinity of the antibody while maintaining specificity for GIP1-42. We then paired this monoclonal antibody with an anti–total GIP monoclonal antibody to develop a dual monoclonal sandwich ELISA that specifically measures GIP1-42. Materials and Methods SAMPLES AND REAGENTS

Blood samples from 16 healthy volunteers were collected in P800 tubes (Becton Dickinson) containing EDTA, aprotinin, and DPP4 inhibitor. All volunteers were enrolled in the Eli Lilly volunteer blood donor program and gave informed consent. The degradation of GIP by DPP4 is not as rapid as for GLP-1, but it is still appreciable; hence the inclusion of the DPP4 inhibitor. After plasma was separated from cells, plasma samples were stored at ⫺70 °C before analysis of GIP1-42 concentrations. Blood samples were collected from volunteers under fasting conditions and at specific time points after they consumed a mixed meal challenge consisting of approximately 400 fat calories, 400 carbohydrate calories, and 100 protein calories. Synthesized GIP1-42 and GIP3-42 peptides were purchased from Ana Spec. Anti–total GIP monoclonal antibody was purchased from Millipore. The exact binding region of this antibody is unknown, and it recognizes both GIP1-42 and GIP3-42. We measured total GIP concentrations using a sandwich ELISA kit (Millipore). This total GIP ELISA uses the same anti–total GIP monoclonal antibody described above as the capture antibody and rabbit polyclonal anti–total GIP antibody as the detection antibody. The manufacturer states that this kit is specific for GIP and does not recognize GLP-1, GLP-2, or glucagon. 850 Clinical Chemistry 57:6 (2011)

Anti-GIP1-42 antibody generation and affinity maturation. The primary amino acid sequence of GIP1-42 is YAEG TFISDYSIAMDKIHQQDFVNWLLAQKGKKNDWKH NITQ and differs from that of GIP3-42 only by tyrosine and alanine residues at the N-terminus. To generate a monoclonal antibody specific for this N-terminus, we obtained anti-GIP hybridoma clones after fusing splenocytes from mice immunized with a peptide corresponding to the first 6 residues of human GIP1-42 conjugated to keyhole limpet hemocyanin. Next, IgGs from individual clones were screened for reactivity to both GIP1-42 and GIP3-42. An anti-GIP1-42–specific antibody was selected and optimized by using yeast cell surface display (17 ). We extracted mRNA and amplified the Ig heavy chain variable region (VH) and Ig light chain variable region (VL) genes from cDNA using a mouse Ig primer set, and determined sequences after insertion into a TA TOPO (vaccinia DNA topoisomerase I) cloning vector (Invitrogen). The wild-type single-chain fragment variable (scFv) gene, in which the antigen-specific VH and VL domains are linked into a single polypeptide, was created by single overlap extension–PCR and ligated into a display vector after digestion with SfiI restriction enzyme. Spiked mutagenesis with degenerate oligonucleotides was used for saturation mutagenesis of the VH and VL complementarity-determining regions (CDR). Individual CDR libraries were constructed by recombining the mutated scFv DNA pool into display vector digested with SfiI after cotransformation into EBY100 yeast (a kind gift from Dr. Dane Wittrup, Massachusetts Institute of Technology) by using lithium acetate as described (18, 19 ) and propagated in synthetic dextrose casamino acid media at 30 °C with shaking. The scFv expression was induced by transferring midlog growth cells into synthetic galactose casamino acid media (2% galactose) and grown at 20 °C for 12–24 h with shaking. An overrepresentation of each CDR library or sort output from a previous round of selection was incubated with GIP1-42-biotin peptide (CPC Scientific) and anti-V5 monoclonal antibody (1 mg/L) (Invitrogen) in selection buffer [PBS (157 mmol/L NaCl, 2.7 mmol/L KCl, 8.1 mmol/L Na2HPO4, 1.8 mmol/L KH2PO4), pH 7.4, 0.5% BSA] at 25 °C for 30 min before transfer to ice and washing with cold selection buffer. Bound antigen was detected by using streptavidin-Rphycoerythrin (Invitrogen) at a 1:200 dilution, and binding was normalized to the amount of scFv expression by using Alexa Fluor 488 goat antimouse antibody (Invitrogen) at a 1:100 dilution. Clones of interest were enriched using fluorescence-assisted cell sorting on a FACSAria cell sorter (BD Biosciences). Individual clones from CDR libraries with improved binding were sequenced. A combinatorial library combining the CDR regions of clones from the individual libraries with improved binding was constructed by single overlap extension–PCR, and subjected to further rounds of se-

GIP1-42 Sandwich Immunoassay

lection with increased stringency. Individual combinatorial clones were sequenced, and the relative binding affinity determined after antigen titration (20, 21 ). The resulting VH and VL genes were ligated into separate murine expression vectors, respectively, to create a final optimized variant, which was purified by protein A after large-scale transient transfection using HEK (human embryonic kidney) 293 cells. BIACORE ANALYSIS

Kinetic analysis and binding specificity of the anti-GIP1-42 specific antibody were determined by use of a Biacore T100 instrument. A Series S Sensor Chip CM5 (Biacore) with approximately 6000 response units of goat antimouse ␬ capture antibody (Southern Biotech) covalently attached was used to capture approximately 1500 response units of IgG diluted to 10 mg/L in running buffer (HEPES buffered saline containing 3 mmol/L EDTA and 0.05% Tween, Biacore). Increasing concentrations of GIP1-42 were injected for 5 min at a flow of 30 ␮L/min followed by a 10-min dissociation phase. The surfaces were regenerated following each dissociation phase with two 15-␮L injections of glycine-HCl, pH 1.5, at a flow rate of 100 ␮L/min. Kinetic constants were determined using the T100 Evaluation software. Specificity for GIP1-42 was tested by injections of GIP3-42 peptide under the described assay conditions. GIP1-42 ELISA

After the final purification process, a dual monoclonal GIP1-42 MesoScale Discovery (MSD) ELISA was constructed using the N-terminal–specific anti GIP1-42 monoclonal antibody and the anti–total GIP monoclonal antibody. Standard MSD 96-well plates were incubated for 1 h with 50 ␮L of anti–total GIP antibody (5 mg/L). Afterward, wells were aspirated and washed 3 times with TBST (Tris buffered saline containing 10 mmol/L Tris pH 7.40, 150 mmol/L NaCl with 1 mL Tween 20/L). Wells were blocked with 200 ␮L of TBScasein (Thermo Fisher). Next, 50 ␮L of GIP1-42 calibrators (varying concentrations of GIP1-42 protein in assay buffer consisting of 50 mmol/L HEPES, pH 7.40, 150 mmol/L NaCl, 10 mL/L Triton X-100, 5 mmol/L EDTA, and 5 mmol/L EGTA) were added to the wells to generate a calibration curve. Plasma samples were diluted 1:4 in assay buffer and added to their respective wells, and the ELISA plate was incubated for 1.5 h at room temperature. Following aspiration, wells were washed 3 times with TBST, and 50 ␮L of conjugate antibody (biotin-labeled N-terminal anti-GIP1-42 specific antibody, 1 mg/L) were added to the wells for a 1-h incubation at room temperature. After aspiration, wells were washed 3 times with TBST, followed by the addition of 50 ␮L of 1:5000 diluted streptavidinruthenium conjugate (Sulfo-Tag, MSD) for a 1-h incu-

bation at room temperature. After the final aspiration, wells were washed 3 times with TBST, 150 ␮L of 2⫻ MSD read buffer were added to the wells, and plates were developed using an MSD reader, which recorded ruthenium electrochemiluminescence. DATA ANALYSIS

MSD software and SigmaPlot version 8.0 were used for fitting ELISA calibration curves. Data were plotted and analyzed with Fig. P (version 2.98, Biosoft). In each case, a P value ⱕ0.05 was considered to indicate statistical significance. Results We first obtained an anti-GIP1-42 monoclonal antibody by immunizing mice with a peptide corresponding to the first 6 residues of human GIP conjugated to keyhole limpet hemocyanin. Initial Biacore analysis of this parental antibody (Fig. 1A) indicated that although it was promising, its affinity was likely not suitable for our ultimate goal. The parental antibody did, however, demonstrate specificity for GIP1-42 vs GIP3-42 (Fig. 1B). To improve the affinity of the antibody, the VH and VL genes were obtained from the parental hybridoma cell line, and mutations were introduced into the CDRs. Three rounds of selection with decreasing concentrations of GIP1-42 were performed, and sequences of individual variants with improved binding were determined. A combinatorial library, in which the CDR regions of clones from the individual CDR libraries with improved binding were combined, was constructed and subjected to further rounds of selection with increased stringency to identify additive or synergistic mutational pairings. Individual combinatorial clones were sequenced, and binding characteristics were determined. Biacore analysis (Fig. 1C) indicated that the optimized antibody had much improved affinity for GIP1-42. Importantly, the optimized antibody was also specific for GIP1-42 and did not recognize GIP3-42 (Fig. 1D). Table 1 summarizes the characteristics of the parental and optimized antibodies and demonstrates that the optimized anti-GIP1-42–specific antibody resulted in a ⬎100-fold improvement in the dissociation rate and 10-fold improvement in overall affinity for GIP1-42 compared to the parental antibody. This optimized antibody, together with an anti–total GIP monoclonal antibody, was then investigated for pairing in a sandwich ELISA. The optimal pairing was found to be anti–total GIP antibody as the capture antibody and anti–N-terminal specific GIP1-42 antibody as the conjugate antibody. Fig. 2 shows a typical calibration curve obtained with this ELISA orientation, in which GIP1-42 peptide was prepared at a concentration of 1 ␮g/L and serially diluted to create a calibration curve. Based on a 3-SD evalClinical Chemistry 57:6 (2011) 851

Fig. 1. Development, optimization, and characterization of an optimized N-terminal–specific anti-GIP1-42 monoclonal antibody. (A), The parental anti-GIP1-42 hybridoma clone was analyzed by using Biacore technology, with serial 3-fold dilutions of GIP1-42 antigen, starting at a concentration of 90 nmol/L. (B), Biacore analysis of the specificity of the parental anti-GIP1-42 monoclonal antibody was performed by using either 100 nmol/L GIP1-42 or 100 nmol/L GIP3-42. (C), The optimized anti-GIP1-42 variant was analyzed by using Biacore technology, with serial 3-fold dilutions of GIP1-42 antigen, starting at a concentration of 90 nmol/L. (D), Biacore analysis of the specificity of the optimized anti-GIP1-42 monoclonal antibody was performed by using either 100 nmol/L GIP1-42 or 100 nmol/L GIP3-42.

by using glucagon, GLP-1, GLP-2, secretin, and vasoactive intestinal peptide, each at concentrations up to 10 ␮g/L. No cross-reactivity was observed with any of these other peptides, further suggesting that the ELISA was specific for GIP1-42. ELISA dilution curves for the synthetic standard and actual human plasma samples were determined to be parallel, and the ELISA demonstrated acceptable dilutional linearity. When observed vs expected re-

uation of the zero calibrator, the limit of detection of the ELISA was determined to be 1 ng/L. The lower limit of quantification was determined to be 5 ng/L, and the analyte measurement range of the assay was determined to be 5 ng/L to 4000 ng/L. The sandwich ELISA was specific for GIP1-42 and did not recognize GIP3-42 (Fig. 2). Specificity of the ELISA for other members of the peptide superfamily was also tested

Table 1. Affinities of parental and optimized N-terminal–specific anti-GIP1-42 monoclonal antibodies. Clone

ka [(mol/L)ⴚ1 䡠 s]a

kd (1/s)

⌬kd

⫺2

1⫻ 180⫻

Parental

2.11 ⫻ 10

1.19 ⫻ 10

Optimized

2.22 ⫻ 106

6.59 ⫻ 10⫺5

7

KD (M)

⌬KD

⫺10

1⫻

2.97 ⫻ 10⫺11

20⫻

5.63 ⫻ 10

ka, association rate constant; kd, dissociation rate constant; ⌬kd, change in dissociation rate constant relative to parental clone; KD, equilibrium dissociation constant; ⌬KD, change in equilibrium dissociation constant relative to parental clone.

a

852 Clinical Chemistry 57:6 (2011)


Fig. 2. GIP1-42 sandwich ELISA calibration curve.

sults were compared for the dilution studies performed, the r value was 0.99 and the regression equation was: y ⫽ 1.17x– 4.71. We next evaluated the ELISA using human plasma samples to assess overall robustness. Freeze–thaw stability was evaluated by testing 4 different plasma samples. Results showed acceptable freeze–thaw stability with 80%–120% recovery after 5 freeze–thaw cycles. Individual results for freeze–thaw cycles were as follows: sample A, 23, 25, 24, 24, and 23 ng/L, respectively; sample B, 173, 171, 168, 158, and 153 ng/L, respectively; sample C, 470, 486, 477, 459, and 451 ng/L, respectively; and sample D, 738, 790, 740, 737, and 747 ng/L, respectively. We assessed precision of the ELISA using human plasma samples containing 24, 360, and 754 ng/L of endogenous GIP1-42. Intraassay (n ⫽ 20) imprecision results (CVs) were 3.3%, 3.1%, and 3.5%, respectively. Interassay imprecision was also determined by having 2 different operators analyze the above samples in quadruplicate on each of 3 different days. Interassay (n ⫽ 24) imprecision results (CVs) were 11.5%, 8.7%, and 6.1%, respectively. To assess recovery, GIP1-42 peptide was added to pooled fasting human plasma (containing 10 ng/L of endogenous GIP1-42), at concentrations of 500, 250, and 125 ng/L, and samples were analyzed using the ELISA. Mean (SD) results were 592 (43) ng/L, 287 (28) ng/L, and 139 (8) ng/L, indicating 80%–120% recovery at all concentrations of GIP1-42 tested. We next used our assay to measure GIP1-42 concentrations following a mixed meal test in healthy individuals. To do this, 16 healthy volunteers had blood drawn at 0800 following an overnight fast. Study participants were fed a mixed meal breakfast and had their blood drawn 1 and 2 h afterward. GIP1-42 increased dramatically (Fig. 3A) in the postprandial state compared to the fasting state. At baseline in the fasting state, the median plasma GIP1-42

Fig. 3. Comparison of GIP1-42 and total GIP concentrations in healthy volunteers during fasting and postprandial states. (A), Total GIP and GIP1-42 concentrations in healthy volunteers before and after a mixed meal challenge. Total GIP concentrations, open circles; GIP1-42 concentrations, solid circles. (B), GIP1-42 and total GIP concentrations from Fig. 3A compared. Fasting concentrations, open circles; postprandial concentrations, solid circles.

concentration was 18 ng/L with a 25%–75% range of 14 –29 ng/L. At 1 and 2-h postprandial time points, GIP1-42 concentrations increased dramatically (median 552 ng/L, 25%–75% range 439 –922 ng/L; and median 470 ng/L, 25%–75% range 278 – 644 ng/L), respectively (P ⬍ 0.001 compared to fasting for both postprandial time points). Total GIP concentrations were also measured and were also observed to increase dramatically when the study participants went from a fasted to fed state. At baseline in the fasting state, the median total GIP concentration was 33 ng/L with a 25%–75% range of 20 – 62 ng/L. At 1 and 2-h postprandial time points, total GIP concentrations increased dramatically (median 766 ng/L, 25%–75% range 594 –1195 ng/L; and median 773 ng/L, 25%–75% range 501–965 ng/L, respectively (P ⬍ 0.001 compared to fasting for both postprandial time points). Clinical Chemistry 57:6 (2011) 853

We next examined the correlation between total GIP and GIP1-42 (Fig. 3B). During the fasted state, GIP1-42 values were modestly but not significantly correlated with total GIP values (r ⫽ 0.42; P ⫽ 0.09). In the postprandial state, GIP1-42 concentrations were more highly correlated with total GIP concentrations (r ⫽ 0.93; P ⬍ 0.001). However, the ratio of GIP1-42 to total GIP was not the same in all study participants (Fig. 3B), suggesting large interindividual variability in the relative concentrations of GIP1-42 to total GIP in the fasting and postprandial states. In the fasting state, the range of the ratios of total GIP/GIP1-42 was 0.51– 4.62. At the 1-h postprandial time point, the range of the ratios was 0.96 –1.90, and at the 2-h postprandial time point, the range of ratios was 0.76 –3.00. Discussion To develop a sandwich ELISA for measuring GIP1-42, we generated a first-in-class monoclonal antibody targeting the N-terminus of the incretin peptide. Although this parental antibody possessed the required specificity, it had a dissociation rate that would ultimately limit the assay limit of quantification. The recent advent of in vitro display technologies (22–24 ) provided us the flexibility to use CDR saturation mutagenesis to increase its affinity. The optimized version of the antibody had a 100-fold improvement in the dissociation rate for GIP1-42 peptide compared to the parental antibody. At the same time, specificity for GIP1-42 was maintained after optimization, and no binding was observed to GIP3-42. We were then able to use this antibody to build a dual monoclonal sandwich ELISA for GIP1-42 and show that the sandwich ELISA was capable of specifically measuring GIP1-42 concentrations. This work builds upon previously reported assays for GIP1-42 (8, 15, 16 ). Compared to previously reported assays, however, which are LC-MS-based or rely on the use of a polyclonal antibody in an RIA format (8, 15, 16 ), our approach uses a monoclonal antibody specific for the N-terminus of GIP1-42 in a sandwich format to measure only active GIP. Using this assay, we were able to demonstrate that healthy individuals show dramatic increases in GIP1-42 in the postprandial state compared to the fasting state. Although GIP1-42 concentrations were correlated overall with total GIP concentrations, there was large interindividual variation in the ratio of active GIP1-42 to total GIP. In addition, this correlation was higher in the fed state than the fasted state, possibly due to the lack of uniformity among the study participants during their overnight fasting. As a result of the 2 monoclonal antibodies used, this sandwich ELISA is highly specific for GIP1-42. From a practical standpoint, the advantage of a sand854 Clinical Chemistry 57:6 (2011)

wich ELISA method over an LC-MS type method for measuring GIP1-42 concentrations is that the ELISA can be implemented in most laboratories that may have neither the complex equipment nor the highly specialized operator expertise required to routinely perform LC-MS type assays. In addition, the ELISA has the potential for higher throughput than an LC-MS assay and therefore provides the basis for the first sandwich ELISA method that can specifically measure GIP1-42 concentrations. Another advantage of this ELISA is its lower limit of quantification of 5 ng/L. This may be especially important because we observed that 50% of healthy individuals had plasma GIP1-42 concentrations ⱕ20 ng/L in the fasted state. For many years, there has been great interest in the role of GIP in regulating ␤-cell function and postprandial insulin secretion (25, 26 ). It has long been observed that the concentration of circulating insulin measured after administration of an oral glucose load is significantly greater than the circulating insulin concentration measured after a comparable amount of glucose has been administered intravenously (9 ). Secretion of GIP from small intestine K cells following an oral glucose load is increasingly being recognized as contributing to these differences in insulin response (25, 26 ). Recent publications have further drawn attention to the relationship that GIP may have with the risk for developing impaired glucose tolerance and/or type 2 diabetes. Renner and coworkers created a transgenic pig with a dominant negative GIP receptor (GIPR) and demonstrated that this animal model showed a normal response to an IV glucose tolerance test, but an impaired response to an oral glucose tolerance test (27 ). These observations were particularly interesting in light of previous studies, suggesting that an impaired response to an oral glucose tolerance test (OGTT) is associated with increased risk of developing type 2 diabetes and increased cardiovascular events (28 ). Very recently, Saxena and colleagues demonstrated that mutations in the GIPR were associated with decreased insulin secretion and an increase in the 2-h post-OGTT glucose value of 0.15 mmol/L, and they concluded that GIPR mutations may increase the risk of developing type 2 diabetes (29 ). It has also become clear that GIP1-42 is the main GIP species to have activity through the GIPR. Deacon and colleagues demonstrated that GIP3-42 bound the receptor with much less affinity than did GIP1-42 and had no effect on cAMP accumulation (30 ). The same researchers also demonstrated that although GIP3-42 weakly antagonized GIP1-42-induced cAMP accumulation and insulin output in vitro, it did not act as a physiological antagonist in vivo (30 ).


In light of these findings, it is logical to expect that there will be an increased need for robust assays to measure GIP1-42. By providing a relatively straightforward and robust method to do so, our GIP1-42 dualmonoclonal sandwich ELISA should help increase our understanding of the role of GIP in regulating glucosedependent insulin secretion. Because our assay provides specificity for the active form of the hormone, a lower limit of quantification of 5 ng/L, and an extremely broad dynamic range, it is possible that this ELISA or one similar to it could eventually be adopted for measurement of GIP1-42. This may be particularly helpful in evaluating GIP-based therapies or therapeutic molecules that work through the GIP pathway.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 re-

quirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article. Authors’ Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest: Employment or Leadership: Eli Lilly and Company Consultant or Advisory Role: None declared. Stock Ownership: J.S. Troutt, Eli Lilly and Company; R.W. Siegel, Eli Lilly and Company; J. Chen, Eli Lilly and Company; J.H. Sloan, Eli Lilly and Company; M.A. Deeg, Eli Lilly and Company; G. Cao, Eli Lilly and Company; R.J. Konrad, Eli Lilly and Company. Honoraria: None declared. Research Funding: Eli Lilly and Company. Expert Testimony: None declared. Role of Sponsor: The sponsor played a direct role in the design of the study, the review and interpretation of data, the preparation of the manuscript, and the final approval of the manuscript.

References 1. Baggio LL, Drucker DJ. Biology of incretins: GLP-1 and GIP. Gastroenterology 2007;132:2131–7. 2. Nauck MA. Unraveling the science of incretin biology. Am J Med 2009;122:S3–10. 3. Drucker DJ. Enhancing incretin action for the treatment of type 2 diabetes. Diabetes Care 2003; 26:2929 – 40. 4. Drucker DJ. The role of gut hormones in glucose homeostasis. J Clin Invest 2007;117:24 –32. 5. Kreymann B, Williams G, Ghatei MA, Bloom SR. Glucagon-like peptide-17-36: a physiological incretin in man. Lancet 1987;330:1300 – 4. 6. Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 2006;368:1696 –705. 7. Kieffer TJ, McIntosh CH, Pederson RA. Degradation of glucose-dependent insulinotropic polypeptide and truncated glucagon-like peptide 1 in vitro and in vivo by dipeptidyl peptidase IV. Endocrinology 1995;136:3585–96. 8. Deacon CF, Nauck MA, Meier J, Hucking K, Holst JJ. Degradation of endogenous and exogenous gastric inhibitory polypeptide in healthy and in type 2 diabetic subjects as revealed using a new assay for the intact peptide. J Clin Endocrinol Metab 2000;85:3575– 81. 9. McIntyre N, Holdsworth CD, Turner DS.fr New interpretation of oral glucose tolerance. Lancet 1964;284:20 –1. 10. Pamir N, Lynn FC, Buchan AM, Ehses J, Hinke SA, Pospisilik JA, et al. Glucose-dependent insulinotropic polypeptide receptor null mice exhibit compensatory changes in the enteroinsular axis. Am J Physiol Endocrinol Metab 2003;284:E931–9. 11. Ehses JA, Casilla VR, Doty T, Pospisilik JA, Winter KD, Demuth HU, et al. Glucose-dependent insulinotropic polypeptide promotes beta-(INS-1) cell survival via cyclic adenosine monophosphatemediated caspase-3 inhibition and regulation of p38 mitogen-activated protein kinase. Endocrinology 2003;144:4433– 45. 12. Kim SJ, Winter K, Nian C, Tsuneoka M, Koda Y,

13.

14.

15.

16.

17.

18.

19.

McIntosh CH. Glucose dependent insulinotropic polypeptide (GIP) stimulation of pancreatic betacell survival is dependent upon phosphatidylinositol 3-kinase (PI3K)/protein kinase B (PKB) signaling, inactivation of the forkhead transcription factor Foxo1, and down-regulation of bax expression. J Biol Chem 2005;280:22297–307. Ehses JA, Lee SS, Pederson RA, McIntosh CH. A new pathway for glucose-dependent insulinotropic polypeptide (GIP) receptor signaling: evidence for the involvement of phospholipase A2 in GIP-stimulated insulin secretion. J Biol Chem 2001;276:23667–73. Kim SJ, Nian C, Widenmaier S, McIntosh CH. Glucose-dependent insulinotropic polypeptidemediated up-regulation of beta-cell antiapoptotic Bcl-2 gene expression is coordinated by cyclic AMP (cAMP) response element binding protein (CREB) and cAMP-responsive CREB coactivator 2. Mol Cell Biol 2008;28:1644 –56. Siskos AP, Katsila T, Balafas E, Kostomitsopoulos N, Tamvakopoulos C. Simultaneous absolute quantification of the glucose-dependent insulinotropic polypeptides GIP1-42 and GIP3-42 in mouse plasma by LC/ESI-MS/MS: preclinical evaluation of DPP-IV inhibitors. J Proteome Res 2009;8: 3487–96. Wolf R, Rosche F, Hoffmann T, Demuth HU. Immunoprecipitation and liquid chromatographicmass spectrometric determination of the peptide glucose-dependent insulinotropic polypeptides GIP1-42 and GIP3-42 from human plasma samples: new sensitive method to analyze physiological concentrations of peptide hormones. J Chromatogr A 2001;10:21–7. Siegel RW. Antibody affinity optimization using yeast cell surface display. Methods Mol Biol 2009;504:351– 83. Gietz RD, Woods RA. Yeast transformation by the LiAc/SS Carrier DNA/PEG method. Methods Mol Biol 2006;313:107–20. Orr-Weaver TL, Szostak JW, Rothstein RJ. Genetic applications of yeast transformation with linear

20.

21.

22.

23.

24.

25.

26. 27.

28.

29.

30.

and gapped plasmids. Methods Enzymol 1983; 101:228 – 45. Chao G, Lau WL, Hackel BJ, Sazinsky SL, Lippow SM, Wittrup KD. Isolating and engineering human antibodies using yeast surface display. Nat Protoc 2006;1:755– 68. Feldhaus MJ. Siegel RW. Yeast display of antibody fragments: a discovery and characterization platform. J Immunol Methods 2004;290:69 – 80. Boder ET, Wittrup KD. Yeast surface display for screening combinatorial polypeptide libraries. Nat Biotech 1997;15:553–7. Hanes J, Plu¨ckthun A. In vitro selection and evolution of functional proteins by using ribosome display. Proc Natl Acad Sci U S A 1997;94:4937– 42. Parmley SF, Smith GP. Antibody-selectable filamentous fd phage vectors: affinity purification of target genes. Gene 1988;73:305–18. Kim W, Egan JM. The role of incretins in glucose homeostasis and diabetes treatment. Pharmacol Rev 2008;60:470 –512. Kulkami RN. GIP: no longer the neglected incretin twin? Sci Transl Med 2010;15:2(49):49ps47. Renner S, Fehlings C, Herbach N, Hofmann A, von Waldthausen DC, Kessler B, et al. Glucose intolerance and reduced proliferation of pancreatic ␤-cells in transgenic pigs with impaired glucosedependent insulinotropic polypeptide function. Diabetes 2010;59:1228 –38. Qiao Q, Tuomilehto J, Borch-Johnsen K. Postchallenge hyperglycaemia is associated with premature death and macrovascular complications. Diabetologia 2003;46 Suppl 1:M17–21. Saxena R, Hivert MF, Langenberg C, and MAGIC investigators. Genetic variation in GIPR influences the glucose and insulin responses to an oral glucose challenge. Nat Genet 2010;42: 142– 8. Deacon CF, Plamboeck A, Rosenkilde MM, de Heer J, Holst JJ. GIP-(3-42) does not antagonize insulinotropic effects of GIP at physiological concentrations. Am J Physiol Endocrinol Metab 2006; 291:E468 –75.

Clinical Chemistry 57:6 (2011) 855