Stable isotope labeling tandem mass spectrometry (SILT) to quantify protein production and clearance rates (original) (raw)

J Am Soc Mass Spectrom. Author manuscript; available in PMC 2008 Jun 1.

Published in final edited form as:

PMCID: PMC2040126

NIHMSID: NIHMS24094

Randall J. Bateman

1Department of Neurology, Washington University School of Medicine, St. Louis, MO 63110

4Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, MO 63110

5Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, MO 63110

Ling Y. Munsell

2Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110

Xianghong Chen

2Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110

David M. Holtzman

1Department of Neurology, Washington University School of Medicine, St. Louis, MO 63110

3Department of Molecular Biology & Pharmacology, Washington University School of Medicine, St. Louis, MO 63110

4Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, MO 63110

5Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, MO 63110

Kevin E. Yarasheski

2Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110

1Department of Neurology, Washington University School of Medicine, St. Louis, MO 63110

2Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110

3Department of Molecular Biology & Pharmacology, Washington University School of Medicine, St. Louis, MO 63110

4Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, MO 63110

5Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, MO 63110

Address all correspondence to Randall J. Bateman, Dept. of Neurology, Washington University School of Medicine, 660 S. Euclid, Box 8111, St. Louis, MO 63110; Phone: 314-747-7066; Fax: 314-362-2244; e-mail: ude.ltsuw@rnametab

Abstract

In all biological systems, protein amount is a function of the rate of production and clearance. The speed of a response to a disturbance in protein homeostasis is determined by turnover rate. Quantifying alterations in protein synthesis and clearance rates is vital to understanding disease pathogenesis (e.g., aging, inflammation). No methods exist for quantifying production and clearance rates of low abundance (femtomole) proteins in vivo. We describe a novel, mass spectrometry-based method for quantitating low abundance protein synthesis and clearance rates in vitro and in vivo in animals and humans. The utility of this method is demonstrated with amyloid-beta (Aß), an important low abundance protein involved in Alzheimer's disease pathogenesis. We used in vivo stable isotope labeling, immunoprecipitation of Aß from cerebrospinal fluid, and quantitative liquid chromatography electrospray-ionization tandem mass spectrometry (LC-ESI-tandem MS) to quantify human Aß protein production and clearance rates. The method is sensitive and specific for stable isotope labeled amino acid incorporation into CNS (± 1% accuracy). This in vivo method can be used to identify pathophysiologic changes in protein metabolism; and may serve as a biomarker for monitoring disease risk, progression, or response to novel therapeutic agents. The technique is adaptable to other macromolecules, such as carbohydrates or lipids.

Introduction

Methods of protein analysis

Recent advances in mass spectrometry have led to significant improvements in the types and complexity of samples that can be analyzed, sensitivity of detection, and amount of structural information that can be obtained. These advances have been combined with chemical and isotope labeling techniques that allow for the comparison of protein amounts under different experimental conditions[1,2]. SILAC ( Stable Isotope Labeling by Amino acids in Cell culture), a technique used to label proteins expressed in vitro by incubating cells with stable isotope labeled amino acids, provides an excellent non-radioactive method for labeling proteins during production that does not alter chemical or biologic properties[3]. This in vitro labeling method has been used with liquid chromatography mass spectrometry (LC-ESI-tandem MS) to obtain measurements of relative and total amounts of low abundant proteins. In addition, stable isotope labeling has also been used in vivo to produce highly-enriched protein/peptide internal standards for proteomic analysis[4].

Techniques to measure stable isotope incorporation in vivo have utilized gas chromatography mass spectrometry (GC-MS), which has high mass resolution, but low sensitivity, and so requires a relatively large amount of protein for quantitative analysis. In addition, GC-MS quantifies stable isotope enrichment of amino acids and does not provide specific identification of the source proteins/peptides. Human plasma and skeletal muscle protein synthesis rates have been quantified using this method, however, only the most abundant proteins have been studied[5]. Most proteins in biologic samples are present in less than nanomole per ml quantities, and approach the limit of detection for most GC-MS systems.

We describe a novel method for quantifying production and clearance rates for low abundant (femtomole) proteins using tandem mass spectrometry. By using MS/MS ions for isotope ratio quantitation, rather than the parent MS ion, significantly higher signal to noise can be achieved. This allows for isotope ratio quantitation for low abundance proteins/peptides that would not be detectable by standard MS analysis. The technique utilizes in vivo rates of stable isotope labeled amino acid incorporation and tandem MS quantitation, so that protein production and clearance rates are calculated while peptide sequence information is obtained for positive protein identification.

Protein production and clearance rates are important parameters that are tightly regulated under normal physiologic and pathophysiologic conditions [6-11]. Certain disease states are characterized by disturbances in protein production, accumulation, or clearance. For example, CNS diseases such as Alzheimer's disease, Frontal Temporal Dementia, Huntington and Creutzfeldt - Jakob disease, are associated with disturbance in Aß, tau, huntingtin, and prion proteins respectively. The amyloid hypothesis proposes that disturbances in the metabolism of amyloid-beta[12] (Aß) causes Alzheimer disease (AD). Aß is normally made predominantly by neurons and is present in cerebrospinal fluid and blood, however, no methods were available to quantify Aß synthesis or clearance rates in humans. In order to address critical questions about underlying AD pathogenesis and Aß metabolism in humans, we developed a method to quantify Aß fractional synthesis rate (FSR) and fractional clearance rate (FCR) in vivo in the CNS of humans. The findings indicate that reproducible rates of Aß synthesis and clearance can be quantified in normal humans by administering stable isotope labeled amino acids (13C6-leucine), sampling CSF, and using LC-ESI-tandem MS to quantify the amount and kinetics of 13C6-leucine labeling in Aß [13].

The current method was developed to measure the production and clearance rate of Aß in humans, but is applicable to proteins being produced in vitro or in vivo, as all proteins are labeled simultaneously. This manuscript describes the methodology to perform quantitative measurements of proteins by stable isotope labeling and tandem mass spectrometry (SILT) and expands the technique to employ two stable isotope labeled amino acids to measure the production rate sequentially in the same subject.

Methods

Materials

13C6-leucine (98% 13C6) and 13C6-phenylalanine (95% 13C6) were obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA). M266 antibody against Aβ was provided by Eli Lilly Inc. CNBr-activated sepharose 4B beads were obtained from Amersham Biosciences (Piscataway, NJ). Formic acid (98%) and ammonium bicarbonate (ultra >99.5%,) were obtained from Fluka (AG, Buchs, Switzerland). Fetal bovine serum (FBS) and Dubelco's modified essential media (DMEM), with and without leucine, were obtained from the Washington University Tissue Culture Support Center. Hygromycin B, Streptomyces sp. #400051 was obtained from EMD Biosciences, Inc. (San Diego, CA) and made in 100x stock of 426 mg/ml; Blasticidin S, Hydrochloride (Streptomyces griseochromogenes #203350) obtained from CalBioChem (San Diego, CA) and made in 100x stocks of 10 mg/ml. Sequence grade modified trypsin was obtained from Promega (Madison, WI).

Cell culture and In Vitro labeling

Human neuroglioma cells that produce Aß [14] were grown in the presence of 13C6- leucine or unlabeled leucine. The unlabeled media was DMEM + 10% FBS + antibiotics. The labeled media was leucine-free DMEM + 105 mg/L 13C6-leucine + 3 kDa dialyzed FBS + antibiotics. Media was collected every 24-48 hours of incubation to obtain highly labeled or unlabeled Aß. The media was pooled separately and stored at −20°C. Aß concentration in each pool was determined by ELISA [15]. Using serial, volumetric dilution of labeled and unlabeled media, 13C6-Aβ enrichment standards of 20%, 10%, 5%, 2.5%, 1.25% and 0% labeled media were made, aliquoted and stored at −80°C.

Amyloid-β immunoprecipitation

Antibody beads were prepared by covalently binding m266 antibody to CNBr sepharose beads per the manufacturer's protocol at a concentration of 10 mg/ml m266 antibody. The antibody beads were stored at 4°C in a 50% slurry of PBS 0.02% azide.

The immunoprecipitation mixture was 250 μl of 5x RIPA, 12.5 μl of 100x protease inhibitors, and 30 μl of antibody-bead slurry added to 1 ml of sample in an Eppendorf micro-centrifuge tube which was rotated overnight at 4°C. The beads were rinsed once with 1x RIPA and twice with 1x ammonium bicarbonate. They were aspirated dry after the final rinse and Aß was eluted off the antibody-bead complex using 30 μl of 100% formic acid. After centrifuging the beads again, the formic acid supernatant was transferred to a new Eppendorf tube, and evaporated in a Savant speed-vac (model AES2010) for 15 minutes at low rate (ambient temperature) with radiant cover and full vacuum, followed by 30 minutes at medium rate (43°C) with radiant cover and full vacuum. The sample was reconstituted in 5μl of acetonitrile and 20μl of 25 mM ammonium bicarbonate, pH 8.0. The sample was digested with 400 ng sequence grade trypsin and incubated at 37 °C for 16 hours.

Liquid Chromatography/Mass Spectrometry

LCQ ESI-tandem MS was performed on a Thermo-Finnigan (San Jose, CA) LCQ-DECA equipped with an electrospray ionization source. The LCQ-DECA was operated in the positive ion mode using a spray voltage of 5kV, a capillary voltage of 7V, and a capillary temperature at 200°C. The instrument settings were set for ion specific analysis at full MS/MS mode at 28% normal collisional energy of the precursor MS ions at m/z= 663.5 and 666.5 (doubly charged [M+2H]+2). Mass spectra were collected over a 38 min period after a 10 minute delay time. The liquid chromatography for LCQ analysis was carried out on a Waters (Milford, MA) micro-capillary liquid chromatography system with auto injector interfaced to the mass spectrometer. Samples were chilled at 5°C until injection. A 5μl aliquot of each sample was injected on-column to a Vydac C18 capillary column (0.3 × 150mm MS 5μm). A HPLC gradient of 0-90 %B (0-5min at 0% B, 5-25 min to 50%B, 25-30 min to 90%B 30-33 min at 90% B, 33-35 min to 0% B, 35-40 min at 0% B) over 40 minutes (A= 0.1% Formic acid in water, B=0.1% Formic acid in 80% acetonitrile/20% water) at flow rate of 6 μl/min was used.

LTQ experiments were performed on a Thermo-Finnigan (San Jose, CA) LTQ equipped with a New Objective (Woburn, MA ) nanoflow electrospray ionization source. Sample injection and LC gradients were controlled by an Eksigent 2D-LC nanoflow pump operating in a 1D mode. Samples were chilled at 5°C until injection. A 5μl aliquot of each sample was injected on-column to a New Objective biobasic picofrit column (75 micrometer × 150mm). The HPLC gradient used was: 5-95 %B (0-5min at 5% B, 5-45 min to 50%B, 45-50 min to 90%B, 50-60 min to 5% B), over 60 minutes (A= 0.1% Formic acid in water, B=0.1% Formic acid in 80% acetonitrile/20% water) at a flow rate of 200 nl/min. The LTQ was operated in the positive ion mode using a spray voltage of 1.7 kV and a capillary temperature of 220°C. The MS/MS scanning was performed as an ion specific experiment with data independent scans for full MS and MS/MS scans of unlabeled and labeled precursor ions. The instrument settings were 28% normal collisional energy of the precursor MS ions at m/z= 663.5 and 666.5 (doubly charged species [M+2H]+2) with an isolation width of 2.5 Daltons. Mass spectra were collected over a 80 min period.

Gas chromatography-mass spectrometry

As described, 13C6-enrichment in plasma and CSF leucine, ketoisocaproic acid (KIC), and phenylalanine were quantified using capillary gas chromatography-mass spectrometry (GC-MS; Agilent 6890N gas chromatograph and Agilent 5973N mass selective detector, Palo Alto, CA) [16]. Leucine and phenylalanine were converted to their heptafluorobutyric propyl ester derivatives; 13C6-leucine (m/z 349 and 355) and 13C6-phenylalanine (m/z 383 and 389) enrichments were quantified using GC-MS in negative chemical ionization mode. In the same plasma samples, KIC was isolated, the trimethylsilyl quinoxalinol derivative was formed, and 13C6-enrichment was quantified using GC-positive chemical ionization-MS and selected ion monitoring (m/z 275 and 281)[16]. The GC-MS instrument response was calibrated using gravimetric standards of known isotope enrichment.

Calculation of labeled protein ratio

Percent labeled Aß was calculated as the ratio of all b- and y-tandem MS ion intensities from Aß*17-28 (m/z=666.5) divided by all b- and y- tandem MS ion intensities from Aβ17-28 (m/z=663.5). A custom Microsoft Excel spreadsheet with macros (available upon request) was used to sum all b- and y-series ion intensities, calculate the isotope ratios, and plot the labeled Aß curves. Specifically, these calculations were performed by copying the unlabeled and labeled tandem MS spectra signal intensities averaged over the elution peak directly from Xcalibur into the custom Excel spreadsheet. The spreadsheet parsed the data to identify and sum the signal intensities from predicted unlabeled and labeled b and y ions, and divided labeled to unlabeled ion intensities to produce a ratio for that data set. The ratio was calculated after summing all ion intensities for labeled and unlabeled tandem MS ions. Only the M0 (the most abundant for this peptide) and M6 ions were used for calculations. The labeled ratio for each data set and a standard curve were plotted automatically.

Participants and Sampling

All human studies were approved by the Washington University Human Studies Committee and the General Clinical Research Center Advisory Board. Verbal and written informed consent was obtained from the participant.

Results

Quantitation of newly synthesized proteins using stable isotope labeled amino acids consisted of 5 basic steps

  1. label proteins by application of stable isotope labeled amino acid(s), 2) isolate and concentrate the protein of interest from other labeled proteins, by immunoprecipitation or another method, 3) proteolytically cleave the protein into smaller peptide fragments 4) use tandem mass spectrometry to quantitate the relative amounts of labeled and unlabeled peptide fragments, 5) calculate the amount of unlabeled and labeled tandem MS ions (Fig. 1).

An external file that holds a picture, illustration, etc. Object name is nihms24094f1.jpg

Quantitation by tandem-MS of amino acid stable isotope labeling of newly synthesized proteins

(a) Trypsin fragments of the protein are separated and concentrated using liquid chromatography before mass spectrometry for detection and quantification of both labeled and unlabeled fragments using MS/MS ions. (b) Diagram of quantitation of labeled and unlabeled Aβ17-28 and graph of labeling curve to calculate synthesis and clearance rates of proteins.

In Vitro labeling and quantitation

We labeled newly synthesized Aß in vitro by incubating neuroglioma cells[14] with 13C6-leucine added to leucine-free media. This stable isotope labeled amino acid was chosen because it equilibrates across the blood brain barrier quickly via active transport [17], is an essential amino acid, is present in Aß protein, and is safe and non-radioactive. Labeling a protein with carbon-13 does not significantly change its chemical or biologic properties, as entire organisms have been grown on pure carbon-13 substrates/media without any deleterious effect [18]. Stable isotope labeled leucine was incorporated into Aß at amino acid positions 17 and 34.

Under these in vitro incubation conditions, all leucine-containing proteins synthesized are labeled. The protein of interest, Aß, was isolated and concentrated, using immunoprecipitation techniques. In CSF and neuroglioma cell culture conditions, Aß is present in low picomole amounts per ml. Immunoprecipitation conditions for unlabeled Aß were tested and refined using archived human CSF and cell culture media. Aß was immunoprecipitated from samples of human CSF or cell culture media using an Aß specific monoclonal antibody (m266; provided by Eli Lilly) covalently linked to cyanogen bromide sepharose beads. Aß was eluted from the antibody-bead complex using formic acid, and directly identified and characterized using matrix assisted laser desorption ionization time of flight (MALDI –TOF) mass spectrometry. Mass spectra for Aß were similar to previously published findings[19] (Fig. 2a).

An external file that holds a picture, illustration, etc. Object name is nihms24094f2.jpg

MALDI-TOF qualitative analysis of Aß

(a) Aß peptides were immunoprecipitated from human CSF with the central domain anti-Aß antibody, m266, directed against amino acids 13-28. Following immunoprecipitation, Aß was eluted with 100% formic acid and analyzed on a MALDI-TOF mass spectrometer. Mass spectral peaks are noted with their corresponding peptide variants; Aß38, Aß39, Aß40, and Aß42. The level of Aβ42 was approximately 10% of Aβ40 levels, as has been reported previously in human CSF.

(b) Unlabeled media from a human neuroglioma cell line producing Aß in vitro was collected and immunoprecipitated. Aß peptides were then cleaved with trypsin at sites 5, 16, and 28 producing the two fragment envelopes shown at masses 1325 and 1336. Note the two mass envelopes of Aß fragments Aβ17-28 (1325) and Aβ6-16 (1336) showing the statistical distribution of natural isotopes in unlabeled Aß.

(c) Human neuroglioma cells were cultured for 24 hours in the presence of 13C6-leucine. Media was collected and Aß was immunoprecipitated. Aß peptides were then cleaved with trypsin at sites 5, 16, and 28 producing the fragment envelopes shown at masses 1325, 1331, and 1336. Note the shift of mass (arrow) of Aβ17-28 from 1325 to 1331 that demonstrates the 13C6-leucine label. Aβ6-16 does not contain a leucine, and so is not labeled or mass shifted. A minor amount of Aβ17-28 remains unlabeled.

Naturally occurring isotopes, including 13C (1.1% of all carbon), cause a distribution of masses of larger molecules, including proteins. Due to the size of Aß and the presence of these naturally occurring isotopes, Aß was enzymatically cleaved into fragments (Aβ1-5, Aβ6-16, Aβ17-28, and Aβ29-40/42) using trypsin, prior to mass spectrometry analysis. The Aβ17-28 fragment has a nominal mass of 1324.6 Daltons (D), contains one leucine residue, and was detected using, MALDI-TOF-MS, as a singly-charged species at m/z 1325.6 [M+H]+1. The Aβ6-16 fragment has a nominal mass of 1335.7 D, does not contain a leucine residue, and was detected as a singly-charged species at m/z 1336.7 D [M+H]+1. MALDI-TOF-MS analysis identified the Aβ17-28 peptide fragment in tryptic digests of Aß isolated from cells grown in the presence of naturally abundant 13C-leucine (Fig. 2b), and from cells grown in the presence of 13C6-leucine (Fig. 2c). This confirmed that neuroglioma cells incorporated 13C6-leucine into the predicted Aß fragment, resulting in the expected 6 Dalton shift in Aβ17-28 peptide, and that this high level of 13C6-leucine incorporation could be easily detected using MALDI-TOF-MS. This confirmed the specificity of the immunoprecipitation techniques in CSF and cell media and the ability to qualitatively identify labeled and unlabeled Aß peptides.

After the isolation and detection of labeled and unlabeled Aß, the amount of labeled and unlabeled Aβ17-28 peptides was quantified using LC-ESI-tandem MS with quantitation of the full scan tandem MS ions. The product ions detected by tandem mass spectrometry specifically identified the amino acid sequence for the labeled peptides based on the signature mass plus 6 Daltons. This provided a highly specific “fingerprint” of the Aβ17-28 in both labeled and unlabeled forms. The tandem MS confirmed the expected amino acid sequence for Aβ17-28, and the mass spectra signal intensities were used to quantify the amount of labeled/unlabeled leucine in Aβ17-28.(Fig. 3 a).

An external file that holds a picture, illustration, etc. Object name is nihms24094f3.jpg

LCQ quantitation of tandem MS spectra of in vitro unlabeled and labeled Aβ17-28

(a) Neuroglioma cell media that was unlabeled (top) or labeled (bottom) with 13C6-leucine. The spectra were obtained using tandem MS analysis of unlabeled parent ion Aβ17-28 (m=1325) or labeled parent ion Aβ17-28(m=1331) by LCQ-ESI-MS. Note the tandem MS ions containing leucine at Aβ17 (see masses 213, 360, 921, and 978) are mass shifted by 6 Daltons demonstrating the labeled leucine (arrows). The Aß ions without leucine are not labeled and are not mass shifted by 6 Daltons (see mass 348.2 and 405.3 in both spectra).

(b) Base peak chromatograms of MS1 versus MS2 quantitation. 2.5% leucine labeled culture was analyzed in the same MS run after immunoprecipitation and trypsin digestion of Aβ. Reconstructed base peak chromatograms demonstrate the signal to noise ratio improvements of MS2 versus MS1 quantitation.

(c) Standard curve of labeled Aß to unlabeled Aß. Labeled cultured media was serially diluted with unlabeled media to generate samples for a standard curve. Aß was immuno-precipitated from the media, trypsin digested, and Aβ17-28 fragments were analyzed on a LCQ-ESI-MS and the tandem mass spectra ions were quantitated using custom written software. The predicted percent labeled Aß versus the measured value is shown with a linear regression line.

Calculation of the ratio of labeled to unlabeled peptide

In order to assess the ratio of labeled to unlabeled Aß, the signal intensities from the product ions obtained in the tandem MS experiments were used to quantify the ratio of labeled to unlabeled Aβ17-28. The tandem MS spectra were averaged over the time that Aß eluted from the LC-column (number of tandem MS scans= 5) for both labeled and unlabeled tandem MS spectra. Each average intensity measurement was exported to an Excel spreadsheet where the masses of the expected Aß tandem MS ions were matched and the tandem MS ion signal intensities were summed. The labeled/unlabeled ratio was derived by dividing the total tandem MS ion intensities, detected in the labeled spectra, by the total tandem MS ion intensities, in the unlabeled spectra, using a custom written Microsoft Excel spreadsheet. Accuracy and precision were tested by generating a standard curve from serial dilutions of 13C6-leucine labeled and unlabeled Aß formed in vitro (Fig. 3c). The linear fit, from a range of 0% to 20% labeled Aß serial dilution standard curve, had an R2 value of 0.98 and slope of 1.08. Alternative quantitation techniques were evaluated, and included, using the parent ion-signal intensities and the MALDI-TOF-MS spectra signal intensities. However, the sensitivity and specificity of these quantitative methods were inferior to that achieved using the signal intensities of tandem MS product ions. Specifically, labeled precursor ions were difficult to detect in the TIC or base peak MS scans above baseline noise (Fig. 3b) and produce erroneous values (10% for a 2.5% labeled standard). However, they were easily detected and their signal intensities were quantified by tandem MS analysis (Fig. 3). We concluded that Aß can be labeled in vitro, and the amount of labeling can be quantitated accurately and precisely using LC-ESI-tandem MS.

In Vivo labeling and quantitation

Ten young, healthy human volunteers were enrolled in a metabolism study of Aß kinetics to determine the production and clearance rate of Aß[13]. The results indicate that this novel technique provided plausible in vivo Aß protein production and clearance rates. We further validate the in vivo method by quantifying two Aß production rates and a clearance rate in a middle aged (40-60 years old) healthy research participant during a 13C6-leucine infusion, followed by the infusion of a second stable isotope labeled amino acid (13C6-phenylalanine) in the same participant. We hypothesized that both tracers would provide similar estimates for in vivo Aß production rates. We intravenously administered a primed (2 mg/kg bolus over 10 minutes), constant 13C6-leucine (2mg/kg/hr) infusion for the first 9-hrs of the study, stopped the 13C6-leucine infusion at 9hrs, and intravenously administered a primed (3 mg/kg bolus over 10 minutes), constant 13C6-phenylalanine (3mg/kg/hr) infusion from 16-25hrs. CSF samples (6 ml) were obtained from an indwelling lumbar catheter every hour from 0 to 36 hours.

We utilized GC/MS to quantify 13C6 precursor amino acid enrichment in several accessible pools; CSF 13C6-leucine and 13C6-phenylalanine, plasma 13C6-leucine, 13C6-KIC, and 13C6-phenylalanine. The labeled amino acid ratio was calculated as the average of the CSF labeled to unlabeled ratio amino acid as measured by GC-MS over the hours during infusion (hours 1 to 9 for leucine and hours 17 to 25 for phenylalanine). The average labeled CSF leucine = 11.5%, plasma leucine = 16.4%, plasma KIC = 12.7%, CSF phenylalanine = 22.3%, and plasma phenylalanine = 26.9%.

Aß was immunoprecipitated from CSF, digested with trypsin, and 13C6-leucine and 13C6-phenylalanine abundance in AB17-28 were detected and quantified without significant interference (Fig 4 a, b). The calculated Aß FSRs were similar, accounting for the two-fold higher 13C6-phenylalanine enrichment in the CSF precursor during the infusion and the two-fold greater slope of the 13C6-phenylalanine incorporation into Aβ17-28 versus time curves compared to leucine. The Aß FSR was 4.8%/hr as measured by 13C6-leucine labeled Aß and 5.8%/hr from 13C6-phenylalanine labeled Aß (Fig 5). The Aß FCR was 3.9%/hr as measured by 13C6-leucine labeled Aß and could not be measured from 13C6-phenylalanine labeled Aß due to only 36 hours of sampling.

An external file that holds a picture, illustration, etc. Object name is nihms24094f4.jpg

LTQ quantitation of tandem mass spectra of Aß from human in vivo labeling with two amino acids

(a) Tandem MS spectra of Aβ17-28 from human CSF thirteen hours after _in vivo_13C6-leucine labeling demonstrate unlabeled Aβ17-28 tandem MS ions (top) and labeled Aβ17-28 tandem MS ions (bottom). The leucine containing tandem MS ions demonstrate the additional six Dalton label, as shown by arrows to leucine containing ions.

(b) Tandem MS spectra of Aβ17-28 from human CSF twelve hours after in vivo labeling with 13C6-phenylalanine demonstrate unlabeled ions (top), singly labeled ions (middle) and doubly labeled ions (bottom), as shown by arrows to phenylalanine containing ions.

An external file that holds a picture, illustration, etc. Object name is nihms24094f5.jpg

Two sequential measurements of fractional synthesis rate of Aß in the same participant

The ratio of labeled Aß to unlabeled Aß over 36 hours is shown for a single participant who was intravenously given 13C6-leucine for 0 to 9 hours, followed by phenylalanine given from 16 to 25 hours of the study. Cerebral-spinal fluid was collected hourly during and after labeling for a total of 36 hours. Each hourly sample was immunoprecipitated for Aß, trypsin digested, and analyzed for percent leucine and phenylalanine labeled Aß. Note the rapid increase in leucine labeled Aß to plateau at 12 hours and a subsequent decline in labeled Aß after 24 hours. There is no detectable phenylalanine labeling until hour 20, followed by a rapid rise to plateau.

(a) Aβ17-28 labeled leucine tandem MS ions were quantified excluding 13C6-phenylalanine containing ions and plotted over 36 hours. Leucine labeled ions are detected 5 hours after onset of labeling. FSR was calculated using the slope of the linear regression shown divided by the 13C6-leucine enrichment in CSF.

**(b)**13C6-phenylalanine labeled Aβ17-28 tandem MS ions were quantified excluding 13C6-leucine containing ions and plotted over 36 hours from the same tandem MS data files as in (a). There are not detectable phenylalanine labeled ions during peak leucine labeled Aß times (5-20 hours). Phenylalanine labeled ions are detected 5 hours after onset of 13C6-phenylalanine labeling. FSR was calculated using the slope of the linear regression shown divided by the 13C6-phenylalanine enrichment in CSF.

Discussion

We describe a method to quantify low level stable isotope labeled amino acid incorporation into a low abundance (fmol) cerebral-spinal fluid protein (Aß), based on immunoprecipitation of the protein and tandem MS to quantify the abundance of 13C-labeled amino acid residues in tryptic peptides. This novel technique has been used to quantify human Aß production and clearance rates. The technique is reproducible as two different stable isotope labeled amino acids provided similar estimates of human Aß production rates. The technique can be adapted to simultaneously quantify any/several proteins produced and degraded in vitro or in vivo systems. The technique is robust as it can be used to quantify production and clearance for any protein that can be isolated (not necessarily to purity) from a biological fluid, and provides the advantage of confirming the protein/peptide identity because the amino acid sequence is derived from the tandem MS ions. Perhaps most importantly, this in vivo approach can be used to quantify alterations in low abundance protein production and clearance rates that may underlie the pathogenesis of human disease, to identify and quantify potential biomarkers for disease, and to evaluate the efficacy of proposed disease-modifying therapies. Potentially, this general approach can be applied to quantify the kinetics of other macromolecules produced in the CNS, including lipids, carbohydrates, and inflammatory markers, but care must be taken to select an appropriate stable isotope labeled precursor that crosses the blood-brain-barrier and does not enter rapidly recycling metabolic processes. In addition, the method may provide quantitative and comparative kinetic information about proteins that exist in multiple accessible compartments (blood, CSF or urine).

Established stable isotope tracer methods and mass spectrometric analytical approaches exist (GC/MS and gas isotope ratio MS) for quantifying the in vivo synthesis and clearance rates for abundant (microgram to milligram quantities) slowly synthesized proteins, e.g., albumin, apolipoprotein B, myosin, and surfactant [11,20-24]. Newer methods using SDS-PAGE and MALDI-TOF can quantify multiple in vivo protein synthesis and clearance rates [25], or in vitro prokaryotic synthesis to degradation ratios [26]. These advances have helped automate and extend the capabilities for quantitative proteomics. Although accurate, these methods were not suitable for a low abundance (fmol) protein like Aß. We have exploited the recent advances in LC-ESI-tandem MS to further refine these methods so that sensitive, specific, accurate (±1%), and reliable quantitation of stable isotope abundance in isolates of very low abundance proteins/peptides (fmol) can be achieved. The SILT approach demonstrates significantly higher sensitivity for total amount of protein (fmol), percent labeling (accurate to within 1% on a single analysis), and automation by utilizing LC-MS/MS without the need for protein separation by gel electrophoresis. We demonstrate this proof-of-principle by quantifying the in vivo kinetic rates for a biologically relevant protein that is involved in the pathogenesis of AD. Another advantage of this approach is that during the stable isotope labeling experiment, any/all proteins that are being produced are labeled simultaneously. This approach and quantitative mass spectrometry provide the potential to simultaneously quantify multiple protein production and clearance rates in the same samples by measuring percent labeling of many peptides using SILT analysis. The only requirements are that at least one peptide with a potential labeled amino acid is quantified. For the Aβ peptide detected in this study, the labeled leucine was located on the n-terminus, and thus all b-ions were labeled and y-ions were unlabeled. As the phenylalanine residues were located in central positions, some b- and some y-ions were labeled, while others were unlabeled. The SILT algorithm separately summed labeled MS2 ions (those containing the amino acid which could potentially be labeled) and unlabeled MS2 ions before calculating the ratio. The labeling can occur at any position, and after the peptide sequence is determined from the tandem MS spectra, the MS2 ions that are expected to be labeled, and thus mass shifted, are predicted from the sequence and can be used in the calculation of labeled to unlabeled ratio. The measurement and calculation are not dependent on amino acid location in the peptide sequence.

We chose to use the CSF 13C6-leucine and 13C6-phenylalanine enrichment to represent the precursor pool enrichment for Aβ synthesis in our analysis based on several facts. CSF is within the central nervous system, and should reflect brain tissue extracellular enrichment better than plasma. Also, the CSF 13C6-leucine enrichment was similar to plasma 13C6-KIC enrichment and KIC likely reflects the intracellular pool enrichment better than plasma leucine. Because there are two phenylalanine residues in the Aβ peptide fragment analyzed, an elegant precursor enrichment measurement, mass isotopomer distribution analysis (MIDA), may theoretically be used to measure the precursor enrichment directly [27]. However, because there is only one leucine present in this peptide fragment, MIDA could not be used with the leucine tracer, and would not be suitable for comparison of leucine and phenylalanine precursor enrichments.

Protein quantitation using tandem MS is well established based on extracted ion chromatograms, which quantitate the MS parent ions, as opposed to the MS2 ions used in this study[28]. Using MS precursor ions for quantitation, a change of 30% to 50% in protein amounts can be detected. The tandem MS analysis employed here is different because it strongly suggests that LC-ESI-tandem MS can provide very sensitive (∼1%) and accurate (+/−1% SD) quantitation of stable isotope enrichment (isotope ratios) in very low abundant proteins/peptides. The minimum level of protein we have analyzed by SILT has been 33 fmol Aβ before immunoprecipitation and trypsin digest with a sensitivity of 1% labeling and a linear fit of the standard curve with an R2 = 0.99. This level of sensitivity was not possible utilizing precursor MS1 quantitation. For example, the labeled Aβ could not be detected above baseline in the MS1 scan (Fig. 3b). Among other reasons, this is possible because we quantified the tandem MS signal intensities for b- and y-series ions, for both the labeled and unlabeled Aβ17-28 peptide. This ‘summing’ of the tandem MS ion signal intensities effectively increased both the sensitivity and specificity of the isotope ratio quantitation.

SILT offers a way to measure protein isotope ratios in vitro and in vivo for a large number of proteins that were previously too low in abundance for quantitative isotope abundance analysis. The technique may be applied for total protein quantitation using stable isotope labeled standards or for isotope ratios in determining protein production and clearance rates. This may lead to new information about the physiology and pathophysiology of biologically and medically relevant proteins or other macromolecules.

Acknowledgments

This work was supported by grants from the American Academy of Neurology Foundation, NIH grants K08 AG 027091-01, ADRC (P50 AG05681), Blanchette Hooker Rockefeller Foundation, GCRC (MO1 RR00036), Mass Spectrometry Resource (RR000954), the Clinical Nutrition Research Unit (DK056341), and the Diabetes Research and Training Center (DK020579). We are grateful to the participant for his time, to Cambridge Isotope Laboratories (Ronald Trolard) for providing the 13C6-phenylalanine, to Eli Lilly for providing m266 antibody, and to Dr. Robert Swarm for providing support for lumbar catheterization.

Footnotes

COMPETING INTERESTS STATEMENT

The authors declare that there is a pending application for patent of some of the methods described in this article.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Literature Cited

1. Andersen JS, Lam YW, Leung AK, Ong SE, Lyon CE, Lamond AI, Mann M. Nucleolar proteome dynamics. Nature. 2005;433:77–83. [PubMed] [Google Scholar]

2. Zhang G, Neubert TA. Automated comparative proteomics based on multiplex tandem mass spectrometry and stable isotope labeling. Mol Cell Proteomics. 2006;5:401–11. [PubMed] [Google Scholar]

3. Ong SE, Blagoev B, Kratchmarova I, Kristensen DB, Steen H, Pandey A, Mann M. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics. 2002;1:376–86. [PubMed] [Google Scholar]

4. Wu CC, MacCoss MJ, Howell KE, Matthews DE, Yates JR. 3rd Metabolic labeling of mammalian organisms with stable isotopes for quantitative proteomic analysis. Anal Chem. 2004;76:4951–9. [PubMed] [Google Scholar]

5. Jaleel A, Nehra V, Persson XM, Boirie Y, Bigelow M, Nair KS. In Vivo Measurement Of Synthesis Rate Of Multiple Plasma Proteins In Humans. Am J Physiol Endocrinol Metab. 2006 [PubMed] [Google Scholar]

6. Wolfe RR. Regulation of skeletal muscle protein metabolism in catabolic states. Curr Opin Clin Nutr Metab Care. 2005;8:61–5. [PubMed] [Google Scholar]

7. Wolfe RR. Regulation of muscle protein by amino acids. J Nutr. 2002;132:3219S–24S. [PubMed] [Google Scholar]

8. San Pietro A, Rittenberg D. A study of the rate of protein synthesis in humans. II. Measurement of the metabolic pool and the rate of protein synthesis. J Biol Chem. 1953;201:457–73. [PubMed] [Google Scholar]

9. Balagopal P, Rooyackers OE, Adey DB, Ades PA, Nair KS. Effects of aging on in vivo synthesis of skeletal muscle myosin heavy-chain and sarcoplasmic protein in humans. Am J Physiol. 1997;273:E790–800. [PubMed] [Google Scholar]

10. Yarasheski KE. Exercise, aging, and muscle protein metabolism. J Gerontol A Biol Sci Med Sci. 2003;58:M918–22. [PubMed] [Google Scholar]

11. Elias N, Patterson BW, Schonfeld G. In vivo metabolism of ApoB, ApoA-I, and VLDL triglycerides in a form of hypobetalipoproteinemia not linked to the ApoB gene. Arterioscler Thromb Vasc Biol. 2000;20:1309–15. [PubMed] [Google Scholar]

12. Podlisny MB, Lee G, Selkoe DJ. Gene dosage of the amyloid beta precursor protein in Alzheimer's disease. Science. 1987;238:669–71. [PubMed] [Google Scholar]

13. Bateman RJ, Munsell LY, Morris JC, Swarm R, Yarasheski KE, Holtzman DM. Human amyloid-beta synthesis and clearance rates as measured in cerebrospinal fluid in vivo. Nat Med. 2006;12:856–61. [PMC free article] [PubMed] [Google Scholar]

14. Murphy MP, Uljon SN, Fraser PE, Fauq A, Lookingbill HA, Findlay KA, Smith TE, Lewis PA, McLendon DC, Wang R, Golde TE. Presenilin 1 regulates pharmacologically distinct gamma -secretase activities. Implications for the role of presenilin in gamma -secretase cleavage. J Biol Chem. 2000;275:26277–84. [PubMed] [Google Scholar]

15. DeMattos RB, Bales KR, Cummins DJ, Dodart JC, Paul SM, Holtzman DM. Peripheral anti-A beta antibody alters CNS and plasma A beta clearance and decreases brain A beta burden in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A. 2001;98:8850–5. [PMC free article] [PubMed] [Google Scholar]

16. Yarasheski KE, Smith SR, Powderly WG. Reducing plasma HIV RNA improves muscle amino acid metabolism. Am J Physiol Endocrinol Metab. 2005;288:E278–84. [PMC free article] [PubMed] [Google Scholar]

17. Smith QR, Momma S, Aoyagi M, Rapoport SI. Kinetics of neutral amino acid transport across the blood-brain barrier. J Neurochem. 1987;49:1651–8. [PubMed] [Google Scholar]

18. Krijgsveld J, Ketting RF, Mahmoudi T, Johansen J, Artal-Sanz M, Verrijzer CP, Plasterk RH, Heck AJ. Metabolic labeling of C. elegans and D. melanogaster for quantitative proteomics. Nat Biotechnol. 2003;21:927–31. [PubMed] [Google Scholar]

19. Wang R, Sweeney D, Gandy SE, Sisodia SS. The profile of soluble amyloid beta protein in cultured cell media. Detection and quantification of amyloid beta protein and variants by immunoprecipitation-mass spectrometry. J Biol Chem. 1996;271:31894–902. [PubMed] [Google Scholar]

20. Patterson BW. Use of stable isotopically labeled tracers for studies of metabolic kinetics: an overview. Metabolism. 1997;46:322–9. [PubMed] [Google Scholar]

21. Yarasheski KE, Smith K, Rennie MJ, Bier DM. Measurement of muscle protein fractional synthetic rate by capillary gas chromatography/combustion isotope ratio mass spectrometry. Biol Mass Spectrom. 1992;21:486–90. [PMC free article] [PubMed] [Google Scholar]

22. Hasten DL, Morris GS, Ramanadham S, Yarasheski KE. Isolation of human skeletal muscle myosin heavy chain and actin for measurement of fractional synthesis rates. Am J Physiol. 1998;275:E1092–9. [PMC free article] [PubMed] [Google Scholar]

23. Merchak A, Patterson BW, Yarasheski KE, Hamvas A. Use of stable isotope labeling technique and mass isotopomer distribution analysis of [(13)C]palmitate isolated from surfactant disaturated phospholipids to study surfactant in vivo kinetics in a premature infant. J Mass Spectrom. 2000;35:734–8. [PubMed] [Google Scholar]

24. Schulte JN, Yarasheski KE. Effects of resistance training on the rate of muscle protein synthesis in frail elderly people. Int J Sport Nutr Exerc Metab. 2001;11(Suppl):S111–8. [PubMed] [Google Scholar]

25. Doherty MK, Whitehead C, McCormack H, Gaskell SJ, Beynon RJ. Proteome dynamics in complex organisms: using stable isotopes to monitor individual protein turnover rates. Proteomics. 2005;5:522–33. [PubMed] [Google Scholar]

26. Cargile BJ, Bundy JL, Grunden AM, Stephenson JL., Jr Synthesis/degradation ratio mass spectrometry for measuring relative dynamic protein turnover. Anal Chem. 2004;76:86–97. [PubMed] [Google Scholar]

27. Papageorgopoulos C, Caldwell K, Shackleton C, Schweingrubber H, Hellerstein MK. Measuring protein synthesis by mass isotopomer distribution analysis (MIDA) Anal Biochem. 1999;267:1–16. [PubMed] [Google Scholar]

28. Li XJ, Zhang H, Ranish JA, Aebersold R. Automated statistical analysis of protein abundance ratios from data generated by stable-isotope dilution and tandem mass spectrometry. Anal Chem. 2003;75:6648–57. [PubMed] [Google Scholar]