Method for the Simultaneous Quantitation of Apolipoprotein E Isoforms using Tandem Mass Spectrometry

Abstract

Using Apolipoprotein E (ApoE) as a model protein, we developed a protein isoform analysis method utilizing Stable Isotope Labeling Tandem Mass Spectrometry (SILT MS). ApoE isoforms are quantitated using the intensities of the b and y ions of the ¹³C-labeled tryptic isoform-specific peptides versus unlabeled tryptic isoform-specific peptides. The ApoE protein isoform analysis using SILT allows for the simultaneous detection and relative quantitation of different ApoE isoforms from the same sample. This method provides a less biased assessment of ApoE isoforms compared to antibody-dependent methods, and may lead to a better understanding of the biological differences between isoforms.

Currently, many protein isoform quantitation methods depend upon the availability of specific antibodies to each protein isoform. Development of antibodies includes many steps, which can take months to years to complete, and often fail to produce an antibody with the desired specificity. Due to the difficulty, time, and resources required for antibody development, an alternative and more generalized quantitation method is highly desirable. We present an antibody-independent LC-MS method which enables the quantitation of multiple protein isoforms within a single sample.

ApoE is the greatest genetic risk factor for Alzheimer’s disease (AD) due to risk associated with the three common isoforms. These ApoE isoforms vary by a single amino acid change: ApoE2 (cys112, cys158), ApoE3 (cys112, arg158), and ApoE4 (arg112, arg158). ApoE2 decreases the risk for developing AD [1] while ApoE4 increases the risk of developing AD.[2] The only commercially available isoform-specific ApoE antibody is for the ApoE4 isoform.[3; 4] Current assays are limited by antibody specificity and are unable to detect the protein isoforms simultaneously. A method that can measure the ApoE isoforms within the same sample, independent of isoform-specific antibodies, will be useful in addressing important biological questions.

In this report, we present an antibody-independent method of detecting and quantitating ApoE isoform-specific proteins by combining general lipoprotein purification, stable-isotope-labeled internal standard, and directed tandem mass spectrometry quantitation. We utilize PHM-Liposorb^™ (Calbiochem, San Diego, CA), an absorbent typically used to remove lipids and lipoproteins from serum or plasma, to capture ApoE from biological fluids (Supplemental Figure 1. After tryptic digestion, ApoE isoform-specific peptides are quantitated using the Stable Isotope Labeling Tandem (SILT) MS approach. SILT MS maximizes sensitivity and selectivity by using the intensities of the MS/MS b and y ions for quantitation in lower resolution mass spectrometers.[5] The proteomic testing of SILT MS has been verified in vitro[6], and SILT has been applied to in vivo human protein kinetics studies.[7; 8]

ApoE2 and ApoE4 isoforms were obtained from the media of immortalized astrocytes derived from knock-in mice expressing human ApoE2 or ApoE4.[9] When pooled together and digested with trypsin, ApoE2 and ApoE4 media yield all four potential ApoE2, ApoE3 and ApoE4’s isoform-specific peptides: LGADMEDVC₁₁₂GR, LGADMEDVR₁₁₂, LAVYQAGAR, and C₁₅₈LAVYQAGAR (Supplemental Table 1). Pooled ApoE2 and ApoE4 media were incubated with PHM-Liposorb^™ (10:1 media:liposorb, 30 min 4°C). The PHM-Liposorb^™ was prepared according to the manufacturer’s instructions (1 g/50 mL of PBS). Adsorbed ApoE was denatured in 40% trifluoroethanol (TFE)/100mM triethylammonium bicarbonate (TEABC) (1h, 37°C). BSA(0.1 μg) was added during the TFE denaturation step to serve as an internal control as well as a means to normalize the matrix. Samples were reduced with 5 mM dithiothreitol (DTT) (30 min, RT), alkylated with 20 mM iodoacetamide (IAM) (30 min RT in the dark), and quenched with an additional 5 mM DTT (15 min, RT). Samples were diluted to 10% TFE with 100 mM TEABC, and then digested with trypsin (0.5 μg, 18h, 37°C). Before analysis, samples were desalted using Carbon Nu-tips following the manufacturer’s instructions (Glygen, Columbia, MD). After centrifugal evaporation (30 min at 25°C), samples were resuspended in 15 μL 1% acetonitrile/1% formic acid. Separation of peptides was carried out using an Eksigent nanoLC flowing at 200 nL/min. A NewObjective picofrit column (75 μm) was packed with Michrom Magic C18aq (5μm, 10cm). The gradient was held at 2%B increasing to 10%B over 20 min to 15%B at 30 min, and then to 95%B at 40 min for 5 min (A= 0.1% formic acid in water, B= 0.1 formic acid in acetonitrile). Peptides were detected by a Thermo LTQ operated in positive ion mode using a spray voltage of 1.8kV, 200°C capillary temperature, and 25% collision energy for MS². The MS method monitored the doubly-charged species of each isoform-specific peptide. All four isoform-specific peptides were detected with abundant signal in a single LC-MS run (Figure 1). The matching b and y ions for each peptide were identified in their respective mass spectra. Isoform-specific peptides were also identified in 0.1 mL human CSF from heterozygous young normal control individuals (E3/2 and E3/4). The identified isoform-specific peptides were consistent with the respective genotype of each individual (Supplemental Figure 2).

Figure 1.

The Separation and Detection of ApoE Isoform-specific Tryptic Peptides. Human ApoE2 and ApoE4 from pooled astrocyte media was captured with PHM-Liposorb, denatured, reduced, alkylated, and digested. All four isoform-specific peptides are separated and detected in a single nanoLC-MS run with a high MS² signal-to-noise ratio. A. Extracted Ion Chromatographs for the doubly-charged ion of each isoform-specific peptide. B. Corresponding MS² spectra for each peptide with the b and y ions labeled. (Lower case c represents alkylated cysteine residue.)

Stable-isotope-labeled internal standards provide an ideal control to account for processing steps as both the labeled and unlabeled peptides respond similarly to biological and chemical processes. The ratios of labeled to unlabeled peptides can be used for absolute quantitation with exogenous labeled standards, or to measure the relative incorporation of amino acids into proteins during translation. Labeling with stable-isotope-labeled amino acids allows for calculations of protein production and clearance rates in cell culture, animal models, and human subjects.[7; 10; 11; 12; 13; 14; 15; 16; 17]

To demonstrate stable-isotope-labeling and relative quantitation of ApoE isoforms, immortalized ApoE4 and ApoE2 expressing astrocytes were grown to near confluency. The media was changed from serum-containing to serum-replacement containing dialyzed and delipidated fetal bovine serum. After approximately 12 hours in serum-replacement media, ¹³C₆-leucine was added at an equal concentration to the unlabeled leucine present in the media (maximum ¹³C-leucine labeling = 50%, 1:1). Samples were collected at various intervals for up to 48 hours. The MS monitored the ¹³C₆-labeled and unlabeled doubly-charged ions for all four isoform-specific peptides. Spectra were quantitated with the SILT method using the ratio of the sum of the b and y ion intensities of the labeled to unlabeled isoform-specific peptides.[5] (MS Excel spreadsheets customized for quantitation are available from the authors). To test the reproducibility of the method, standard error of the mean (SEM) of the tracer to tracee ratio (TTR) (¹³C₆-ApoE divided by ¹²C₆-ApoE) was calculated for biological triplicates. SEM ranged from 0.5–6% for LGADMEDVcGR, 0.5–5% for LGADMEDVR, 0.1–3% for LAVYQAGR, and 0.7–8% for cLAVYQAGAR. The higher SEM values were observed at hours with higher labeling (24–48h). After 24 hours of labeling, the TTR approached steady-state (Figure 2). The fractional synthetic rates (FSR) for the ApoE2 and ApoE4 specific peptides were calculated from the slope of the initial incorporation of the ¹³C-leucine divided by the TTR achieved at plateau.[16] The FSR for ApoE4 (averaged result for LGADMEDVR and LAVYQAGAR) and ApoE2 (averaged result for LGADMEDVcGR and cLAVYQAGAR) was found to be 8.6% per hour and 8.5% per hour, respectively (Figure 2). The similar FSR observed for ApoE4 and ApoE2 may be due to the fact that expression of both isoforms in these astrocytes was under control of the same endogenous ApoE mouse promoter; the cell lines were immortalized.[9] Though it is interesting that the FSRs were found to be similar for the E2 and E4 cells, this likely does not provide physiologic information on ApoE production and regulation in vivo. The cell culture system is an artificial, model system used to demonstrate an application of the SILT quantitation method. The results of the cell culture kinetics were used to determine at which point the ¹³C-labeling reached steady-state. The 48 hour time point was chosen to collect the 0–20% ¹³C-ApoE2 and ¹³C-ApoE4 media used to generate Supplemental Figure 3, which demonstrates the reproducibility of the method. The residuals of the four peptides’ standard curves demonstrate strong linear correlations (R²>0.99) over the 0–20% ¹³C-leucine labeling range (Supplemental Figure 3).

Figure 2.

Application of ApoE Isoform SILT Quantitation in vitro Using Immortalized Human ApoE2 and ApoE4 Knock-in Murine Astrocytes. ApoE2 and ApoE4 expressing astrocytes were labeled to ratio of 1:1 ¹³C-leucine: ¹²C-leucine. Media was collected over 48 hours. The SILT method was used to determine the incorporation of ¹³C-leucine into ApoE (biological triplicates, error bars SEM). (Lower case c represents alkylated cysteine residue.)

Using Apolipoprotein E as a model protein, we detail a specific protocol for simultaneous ApoE2, ApoE3, and ApoE4 isoform quantitation. This mass spectrometry based approach to detect and quantitate isoform-specific tryptic peptides is useful to study the less common ApoE isoforms (in humans, the prevalence of the apoE2 allele is 7%, apoE3 is 78%, and apoE4 is 15%).[18] Thus, ApoE isoforms can be analyzed in ApoE heterozygous subjects. The described tandem MS quantitation method can be applied to metabolism or kinetics studies for the calculation of protein isoform production and clearance rates, and can also be adapted for absolute quantitation experiments.

Supplementary Material

01. Appendix A. Supplementary data.

Supplementary data can be found in the online version.

02

Supplemental Figure 1. Capture of Human ApoE Isoforms from CSF and Astrocyte Media with PHM-Liposorb. 0.1 mL of CSF from young normal control participants (genotype E3/4 and E3/2), 1 mL of ApoE4 media and 1 mL of ApoE2 media was incubated with 0.1 mL of liposorb slurry. 1% of the sample was loaded onto a 10% bis-tris gel, and the gel run under reducing conditions. Human ApoE purified from VLDL (BioDesign, Saco, Maine) was used as a positive control. Proteins were transferred to nitrocellulose, blocked with 2% milk in TBST and probed with antihuman ApoE goat polyclonal antibody (1:10000, Calbiochem). Blots were washed in TBST and incubated with peroxidase labeled anti-goat IgG (horse) (1:2000, Vector Labs, Burlingame, CA) and Supersignal Fempto (Pierce). A Kodak phosphoimager station was used for detection. Liposorb absorbed greater than 90% of the ApoE from CSF as demonstrated by comparing the post incubation lane with the 1/10 of pre incubation lane. The bead lanes illustrate excellent recovery of the absorbed ApoE. Tryptic digest samples have no detectable ApoE, suggesting its complete digestion. (pre= pre-incubation, post= supernatant post-incubation, beads = liposorb post-incubation, digest= supernatant post incubation with trypsin. TBST – tris buffered saline with 0.05% tween20, pH8)

03

Supplemental Figure 2. Detection of Human ApoE Isoform-Specific Tryptic Peptides in CSF by NanoLC Tandem MS. 0.1 mL of CSF from a young normal control participant (genotype E3/4 or E3/2) was prepared and analyzed by nanoLCMS. Extracted Ion Chromatographs are shown in the top panels for the doubly-charged ion of each isoform-specific peptide, and the corresponding MS² spectra are displayed in the bottom panels. A. ApoE3/4 (LAVYQAGAR, LGADMEDVR, and LGADMEDVcGR). Strong signals for the ApoE3 and ApoE4 specific peptides were detected but the ApoE2 peptide was not. B. ApoE3/2 (LAVYQAGAR, cLAVYQAGAR, and LGADMEDVcGR). Likewise, strong signals were detected for the ApoE3 and ApoE2 peptides were detected but the ApoE4 peptide was not. The mass spectra indicate clear identification of the ApoE isoform-specific peptides.(Lower case c represents alkylated cysteine residue).

04

Supplemental Figure 3. Quantitation of ApoE Isoform-Specific Tryptic Peptides from ¹³C-labeled ApoE4 and ApoE2 Astrocyte Media. 0, 1.25, 2.5, 5, 10, and 20% ¹³C-leucine labeled pooled ApoE4 and ApoE2 astrocyte media (1:2) was processed and analyzed by nanoLCMS. Labeled and unlabeled ApoE isoform-specific peptides were detected by tandem MS and the percent label calculated from the ratio of the labeled to unlabeled ions. Standard curves of predicted vs. observed ratios are plotted and indicate high linear precision (R²=0.99) and reproducibility. Standard error of the mean (SEM) ranged from 0.02–0.2% for LGADMEDVcGR, 0.03–0.7% for LGADMEDVR, 0.05–0.6% for LAVYQAGR, and 0.08–0.5% for cLAVYQAGAR. (Each replicate represented by circle, triangle or square). (Lower case c represents alkylated cysteine residue).