MSACL 2018 US Abstract

Topic: Proteomics

Comparison of Protein Extraction, Detection, and Quantitation Methods for Therapeutic Monoclonal Antibodies; LC-MS/MS Tryptic Peptide vs Intact LC-MS

Paula Ladwig (Presenter)
Mayo Clinic

Bio: Paula M. Ladwig, M.S., MT (ASCP), is a Development Technologist Coordinator with the Clinical Mass Spectrometry Development Laboratory, Department of Laboratory Medicine and Pathology at Mayo Clinic in Rochester, MN. She has over 10 years of experience in the development and validation of new mass spectrometry tests. Her interests include the implementation of therapeutic drug monitoring of monoclonal antibody therapeutics by mass spectrometry.

Authorship: Paula M. Ladwig; Ann Rivard; Maria A. V. Willrich, Ph.D.
Department of Laboratory Medicine and Pathology, Mayo Clinic

Short Abstract

The established method of choice for protein quantitation has been tryptic peptide by liquid chromatography-tandem mass spectrometry. While some monoclonal antibody therapeutics (t-mAb) may fit this paradigm, developing assays to monitor more humanized t-mAbs has become challenging given the high homology with human immunoglobulin sequences. An alternative is to pursue t-mAb intact light chain quantitation by liquid chromatography-mass spectrometry. Here we present an assessment of extraction, detection and quantitation methodologies for t-mAbs, using eculizumab as an example. The methods discussed are readily transferrable to different mass spectrometry platforms as well as to new t-mAbs that may require therapeutic drug monitoring.

Long Abstract

Introduction

The established method of choice for protein quantitation has been tryptic peptide by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Triple-quadrupole mass spectrometers (MS) have been the instrument of choice in the clinical laboratories for many years due to their proven track record of providing precise and accurate quantitative results. These instruments are known to be extremely sensitive, specific and robust. Quadrupoles have the greatest linear dynamic range and are ideally suited for quantifying peptides produced from the tryptic digestion of proteins.

While proteins and peptides have traditionally been quantitated by immunological methods (RIA or ELISA), the tryptic peptide method by LC-MS/MS has now become well-established with the advantage of not requiring antibody reagents. Highly specific quantitation from a complex matrix is afforded by selected reaction monitoring (SRM) to allow for the monitoring of a single peptide transition, or by multiple transitions (MRM) to allow for even more specificity. Tryptic peptide LC-MS/MS offers cost effective, high throughput, multiplexed assays.

While this approach may work well for chimeric therapeutic monoclonal antibodies (t-mAbs), for example infliximab [1], the peptide method is limited when developing assays for humanized/human t-mAbs containing less than 5% of animal sequences. Occasionally the t-mAb signature peptides can also be found in a patient’s polyclonal, endogenous immunoglobulin background; reducing the sensitivity and specificity of the assay. Therefore, alternative approaches for t-mAb quantitation needed to be developed along with extraction techniques.

A recently published approach for rituximab monitors the unique molecular mass of the intact light chain of monoclonal immunoglobulins; miRAMM (monoclonal immunoglobulin Rapid Accurate Mass Measurement) utilizing microLC-ESI-Q-TOF MS [2]. A time-of-flight (TOF) MS has superior mass measurement accuracy, resolution and scanning speed over a triple quadrupole. Quantitation from full scan data can be performed from single or multiple charge states or from a reconstruct to the intact mass of the light chain. This approach is also advantageous at it allows for the quantitation of multiple t-mAbs while also monitoring the patient’s polyclonal immunoglobulin G (IgG) repertoire.

Monitoring of the intact light chain can also be accomplished utilizing an Orbitrap. The Q Exactive hybrid quadrupole-Orbitrap MS is a high-performance quadrupole that allows for precursor selection along with high-resolution, accurate mass (HRAM) detection. The Q Exactive has the ability for multiplexing of scans with LC multiplexing for implementation into a high volume clinical mass spectrometry laboratory (CMSL).

In this report we show highlights of development, encompassing different extraction, detection and quantitation approaches for measurement of t-mAbs using a humanized hybrid IgG2/IgG4 monoclonal antibody, eculizumab, as an example.

Methods

Eculizumab (Soliris; Alexion Pharmaceuticals) was obtained from our pharmacy and the drug was spiked into purchased normal human pooled serum at various concentrations for standards and quality control. Pools were aliquoted and frozen at -20°C for daily use.

Tryptic Peptide Method

Using a previously published tryptic method for infliximab by HPLC-tandem MS [1] as a guide, a peptide assay for eculizumab was developed; ammonium sulfate extraction, tryptic digest, followed by reverse-phase C8 (Waters XBridge C8, 3-30mm, 3 micron) separation at 400 mcL/min. Two transitions from a peptide, LLIYGATNLADGVPSR, from the c-terminal variable region of the eculizumab light chain were monitored in the MRM MS method along with 2 transitions from a surrogate internal standard (IS); horse IgG1. The peptide HPLC-MS/MS assay utilized a Thermo Cohesive TLX2 connected to an AB Sciex API 5000 triple quadrupole; the workhorse instrument of choice for our CMSL. HPLC gradient and MS conditions were as previously published.

Serum Immunoenrichment

There are many methods to extract t-mAbs from patient serum for before tryptic digest or for intact light chain approaches. Saturated ammonium sulfate can be used for protein fractionation before digestion to relieve some of the protein load; as shown in our published infliximab method [1]. Melon™ Gel allows for the extraction of immunoglobulins from serum by binding and removing non-immunoglobulins serum proteins. The procedure is quick, easy and gentle to antibodies. The elution contains the mAb of interest along with the patient’s immunoglobulin background, which is advantageous for monitoring/quantitating mAbs or monitoring the patient’s immunoglobulin repertoire [2].

Taking advantage of eculizumab being an IgG2/IgG4 t-mAb, we utilized an extraction method to specifically target IgG subclass 4 (~ 4%) and remove it from the larger immunoglobulin background. Life Technologies CaptureSelect™ IgG4 (Hu) Affinity Matrix was used for the extraction of eculizumab from serum followed by reduction [3]. Briefly, 25 mcL serum, 25 mcL surrogate IS (150 mcg/mL pembrolizumab), 100 mcL washed IgG4 Affinity Matrix resin, and 250 mcL PBS were combined in a well of a 96-well 0.2 mM PVDF filter plate, and incubated rotating for 1 hour at room temperature. Positive pressure was used to wash wells 2 times with 1 mL water. Next, 150 mcL of 5% acetic acid was added to each well for elution into a 96-well collection plate. Eluate was then denatured by adding 75 mcL of 100 mM DTT in 1 M ammonium bicarbonate, and incubated rocking for 30 minutes at 55°C. The reduction step makes the enriched sample now suitable for analysis of the Ig intact light chains.

Immunoglobulins Intact Light Chains Methods

A previously published method for t-mAb light chain measurement by microLC-ESI-Q-TOF MS [2] developed for rituximab was adapted to eculizumab; LC and MS conditions were used as published. Briefly, 2 mcL of the extract were separated by reverse phase using a Poroshell 300SB-C3 (1x75mm; 5 micron) column heated to 60°C, run at 25 mcL/minute on an Eksigent Ekspert 200 microLC. Full scan data was acquired from 600 – 2500 m/z on a Sciex TripleTOF 5600 quadrupole time-of-flight mass spectrometer.

We have recently developed and validated a novel intact light chain approach utilizing a Thermo Scientific™ Q Exactive™ Plus Orbitrap MS connected to a Transcend TX4 HPLC system. Extracts were injected (10 mcL) across 2 LC channels on an Agilent Poroshell 300SB C3 (2.1x75mm, 5 micron) column heated to 60°C running at 300 mcL/min with a total run time of 17 minutes. The gradient started at 90% mobile phase A (MPA; water + 0.1 % formic acid), 10% mobile phase B (MPB; 90% acetonitrile + 9% isopropanol + 1% formic acid). After 90 sec, the system ramped to 27% MPB over 1 minute, to 36% MPB over 6 minutes, to 50% MPB over 1 min, and to 98% MPB over 1 minute. Flow rate was increased to 500 mcL/min and column washed through 2 cycles of a 45 second ramp to 20% MPB followed by 45 second ramp to 80% MPB. Column flow rate was returned to 300 mcL/min and 10% MPB over 30 seconds and then equilibrated for 3 minutes before the next injection. The Q Exactive HESI source settings were as follows; sheath gas flow rate 50, aux gas flow rate 12, spray voltage 3.5 kV, capillary temp 300°C, s-lens RF level 65, and aux gas heater temp 300°C. Two scan types were multiplexed. A t-SIM method was run at 140k resolution (25 AGC, 125 ms maximum IT) for the inclusion list to include the +10, +11 and +12 charge states for eculizumab and the +11 charge state for the surrogate IS; pembrolizumab. This was multiplexed with a full scan method run at 140k resolution (16 AGC, 50 ms maximum IT) over 1000 to 2500 m/z.

Results

For the tryptic peptide HPLC-MS/MS method, a MRM method was used for detection, and a transition specific for the light chain of the eculizumab peptide was used for quantitation. Using Analyst, peptide peak areas were normalized to a transition from a surrogate IS. A linear curve with 1/x2 weighting gave AMR from 5 to 600 mcg/mL. The lower limit of quantitation (LLOQ) was 5 mcg/mL. Intra-assay imprecision at 3 levels was less than 10% CV; 15 mcg/mL (N=13, 6%CV), 75 mcg/mL (N=8, 5%CV) and 150 mcg/mL (N=9, 7%CV).

While we utilized a previously published miRAMM intact light chain method for detection of eculizumab, we had to find a different extraction method than published. While the published Melon™ Gel extraction method showed adequate intra-assay (4% at 93 mcg/mL, 7% at 139 mcg/mL, and 13% at 331 mcg/mL; N=12) and inter-assay imprecision (10% at 81 mcg/mL, 12% at 142 mcg/mL, and 14% at 291 mcg/mL; N=12), this published method was not as sensitive (LLOQ= 75 mcg/mL) for eculizumab as needed. In addition, the Melon Gel mechanism of action was found to remove eculizumab’s target (complement component C5) from the serum matrix, while also removing some of the eculizumab bound to it. This target interference specific for eculizumab proved challenging to overcome (data not shown), especially considering that C5 is present in high concentration in normal serum. Therefore, we utilized the IgG4 Affinity Matrix which besides giving us an increase in sensitivity by extracting only subclass 4 specific IgG, allowed for measuring total eculizumab [3].

The intact light chain method required a manual quantitation method at first, as the Sciex 5600 Q-TOF Analyst software did not have quantitation capabilities. Using either Analyst or PeakView, the m/z of the +11 charge state was used to produce an extracted ion chromatogram (XIC) for eculizumab and surrogate IS, the retention time window summed, and area under the curve (AUC) of the +11 charge state collected for quantitation. Excel was used for generating a standard curve and calculating unknowns from this curve. A linear curve with 1/x2 weighting gave an AMR from 5 to 600 mcg/mL. LLOQ was 5 mcg/mL. Intra-assay imprecision at 2 levels was less than 10%CV; 64 mcg/mL (N=10, 6%CV) and 155 mcg/mL (N=10, 9%CV). Residual waste serum samples from patients undergoing therapy with eculizumab (N=65) showed agreement when compared with the peptide HPLC-MS/MS; y=1.1368x+2.1878, R2=0.9412.

While the HPLC-Orbitrap method multiplexed scan modes, the t-SIM method was used for quantitation; allowing for targeting the +10, +11 and +12 charge states from the light chain of eculizumab. We utilized TraceFinder to automate data processing and quantitation. Multiple isotopes from each of the 3 charge states were added together to give one XIC used for quantitation. A quadratic curve with 1/x2 weighting gave an AMR from 5 to 600 mcg/mL. LLOQ was 5 mcg/mL. Intra- and inter-assay imprecision were accessed for 4 levels (15, 100, 400, and 550 mcg/mL) along with at LLOQ. Intra-assay precision of 20 replicates at each level analyzed in one run were all <10%CV with 15% at the LLOQ. Inter-assay precision of the 4 levels run over > 20 runs over 2 months with >50 data points gave ≤10%CV with 11% at the LLOQ. Residual waste serum samples from patients on eculizumab (N=24) showed agreement when compared to the microLC-ESI-Q-TOF method; y=0.988x+2.070, R2=0.975.

Conclusions & Discussion

Our goal has been the development and validation of assays for t-mAbs into the CMSL [4]. While we developed a tryptic peptide method for the chimeric t-mAb infliximab by HPLC-MS/MS, we found that for the more humanized t-mAbs, such as eculizumab, this method suffered from non-specificity.

As for immunoenrichment strategies used prior to intact light chain methods, the Melon™ Gel works well for the majority of immunoglobulins; its disadvantage is the LLOQ is also dependent on the immunological background, which proved challenging for a IgG2/IgG4 t-mAb such as eculizumab, with a high concentration of target in circulation. A more specific immunoenrichment / extraction method utilizing CaptureSelect™ Affinity Matrix efficiently extracted only the IgG4 from the other IgG classes.

We developed an intact light chain MS method for eculizumab utilizing microLC-ESI-Q-TOF MS. This method showed great sensitivity and specificity, but lacked the high throughput needed for a high volume CMSL. We also found that while there are many automated data processing and quantitation products available for peptides that there were limited options for proteins. For the Sciex 5600 Q-TOF, we quantitated manually due to the main software not having quantitation capabilities. While we used the +11 charge state for quantitation, for better sensitivity and specificity this should be taken a step further and quantitation performed utilizing multiple charge states added together or utilizing a protein reconstruct to the intact mass and the AUC of the intact mass used for quantitation. MultiQuant was later purchased as companion software to streamline data processing and quantitation. MultiQuant will allow for quantitation using 1 charge state or multiple charge states, but does not allow for protein reconstruction to the intact light chain for quantitation. Multiple quantitation methods for the Sciex 5600 Q-TOF were compared and agreed within 80-120%.

Last, we developed and validated a dual t-SIM with full scan method on the Q Exactive Plus. The Q Exactive platform with the TLX Transcend gave us the advantage a HRAM MS with normal flow and the ability to multiplex. TraceFinder software was utilized for data processing and quantitation, but also had limitations in its fit for proteins. While the software quantitated from one or multiple charge states or even from one or more isotopes when taking advantage of the high resolution capability, the software also was limited by not having the ability to perform protein reconstruct and quantitating from the intact mass; which in theory should be the gold standard.

We have shown here multiple extraction, detection and quantitation options for t-mAbs. These methodologies for t-mAb quantitation are readily transferrable to different mass spectrometry platforms as well as to new t-mAbs that may require therapeutic drug monitoring.


References & Acknowledgements:

References:

[1] Willrich MA, Murray DL, Barnidge DR, Ladwig PM, Snyder MR., Quantitation of infliximab using clonotypic peptides and selective reaction monitoring by LC-MS/MS. Int Immunopharmacol. 2015 Sep;28(1):513-20.

[2] Mills JR, Cornec D, Dasari S, Ladwig PM, Hummel AM, Cheu M, Murray DL, Willrich MA, Snyder MR, Hoffman GS, Kallenberg CG, Langford CA, Merkel PA, Monach PA, Seo P, Spiera RF, St Clair EW, Stone JH, Specks U, Barnidge DR. Using Mass Spectrometry to Quantify Rituximab and Perform Individualized Immunoglobulin Phenotyping in ANCA-Associated Vasculitis. Anal Chem 2016 Jun 21;88(12):6317-25.

[3] Ladwig PM, Barnidge DR, Willrich MAV., Quantification of the IgG2/4 kappa Monoclonal Therapeutic Eculizumab from Serum Using Isotype Specific Affinity Purification and Microflow LC-ESI-Q-TOF Mass Spectrometry. J Am Soc Mass Spectrom. 2017 May;28(5):811-817.

[4] Ladwig PM, Barnidge DR, Willrich MAV., Mass Spectrometry Approaches for Identification and Quantitation of Therapeutic Monoclonal Antibodies in the Clinical Laboratory. Clin Vaccine Immunol. 2017 May 5;24(5).

Acknowledgements:

We would like to acknowledge David Barnidge for his invaluable work with miRAMM by microLC-ESI-Q-TOF MS over the years on several of these projects and method developments.


Financial Disclosure

DescriptionY/NSource
Grantsno
SalaryyesMayo Clinic
Board MemberyesCLSI
Stockno
Expensesno

IP Royalty: no

Planning to mention or discuss specific products or technology of the company(ies) listed above:

no