Quantitative & PTM Proteomics
Measuring protein abundance, stoichiometry, and post-translational modifications across biological conditions
11.1 Label-Free Quantification
Label-free quantification (LFQ) measures protein abundance without introducing isotopic or chemical labels. Its chief advantages are simplicity, unlimited multiplexing (any number of samples), and compatibility with any sample type. The two principal LFQ strategies — spectral counting and ion intensity — extract quantitative information from different features of the LC-MS/MS data.
Spectral Counting
Spectral counting is predicated on the observation that more abundant proteins generate more MS/MS spectra (peptide-spectrum matches, PSMs). The number of spectra assigned to a protein is used as a semi-quantitative proxy for abundance. While simple, raw spectral counts are biased by protein size (larger proteins produce more tryptic peptides) and must be normalized.
Normalized Spectral Abundance Factor (NSAF)
where $\text{SpC}_k$ is the spectral count for protein $k$, $L_k$ is its length (number of amino acids), and the denominator sums over all $N$ identified proteins. NSAF normalizes for protein length and total spectral output, enabling cross-sample comparison.
Spectral Index (SI$_N$)
The spectral index incorporates the total number of fragment ions observed for all peptides of a protein, providing a more information-rich spectral counting metric than simple PSM counts.
Ion Intensity-Based Quantification
Ion intensity methods extract chromatographic peak areas (extracted ion chromatograms, XICs) from MS1 survey scans. The area under the curve of a peptide's elution profile is proportional to its abundance. This approach is more accurate than spectral counting (CV typically 10-20% vs. 30-50%) but requires high reproducibility in LC retention time and careful alignment across runs. MaxLFQ, implemented in MaxQuant, uses delayed normalization and pairwise peptide ratio extraction to provide robust protein-level quantification without labels.
Intensity-Based Absolute Quantification (iBAQ)
where $\sum I_i$ is the sum of all peptide ion intensities for a protein and $N_{\text{theoretical}}$ is the number of theoretically observable tryptic peptides (typically defined as those within 6-30 amino acids in length). iBAQ provides an estimate of absolute protein abundance within a sample. When calibrated with spiked-in protein standards of known concentration (UPS2 or AQUA peptides), iBAQ yields absolute copy numbers per cell.
Coefficient of Variation (CV)
where $\sigma$ is the standard deviation and $\mu$ is the mean of replicate measurements. CV is the standard metric for assessing quantitative reproducibility. LFQ methods typically achieve median CVs of 10-25% for technical replicates and 20-40% for biological replicates.
11.2 Isotope Labeling Strategies
Stable isotope labeling introduces a defined mass shift into peptides from different experimental conditions, allowing them to be distinguished by the mass spectrometer while being processed identically throughout sample preparation. This minimizes technical variability and enables highly accurate relative quantification.
SILAC (Metabolic Labeling)
Stable Isotope Labeling by Amino acids in Cell culture (SILAC), developed by Ong et al. (2002), achieves near-complete incorporation of heavy isotope-labeled amino acids into the proteome through metabolic turnover. Cells are cultured for at least 5-6 doublings in media containing either normal ("light") amino acids or their heavy counterparts:$^{13}$C$_6$-Lys (+6.0201 Da) and $^{13}$C$_6$,$^{15}$N$_4$-Arg (+10.0083 Da). Because trypsin cleaves at Lys and Arg, every tryptic peptide (except the C-terminal peptide) contains exactly one labeled residue, producing a defined mass shift.
SILAC Ratio Calculation
The ratio is computed from the extracted ion chromatograms of co-eluting heavy and light peptide pairs at the MS1 level. Protein ratios are typically the median of all peptide ratios. A ratio of 1.0 indicates equal abundance; deviations represent up- or down-regulation. Triple SILAC (light, medium with$^{2}$H$_4$-Lys/$^{13}$C$_6$-Arg, and heavy) enables three-condition comparisons in a single experiment.
Dynamic SILAC & Protein Turnover
By pulsing cells from light to heavy media and measuring the incorporation rate of heavy label over time, dynamic SILAC measures protein synthesis and degradation rates. The fraction of newly synthesized protein at time $t$ follows an exponential model: $f_H(t) = 1 - e^{-k_{\text{deg}} \cdot t}$, where$k_{\text{deg}}$ is the first-order degradation rate constant. The protein half-life is then$t_{1/2} = \ln(2) / k_{\text{deg}}$. This approach has revealed that protein half-lives in mammalian cells span four orders of magnitude, from minutes (e.g., p53, c-Myc) to weeks (e.g., histones, nuclear pore complex components).
Chemical Labeling: TMT & iTRAQ
Chemical isobaric tagging allows multiplexed comparison of up to 18 samples (TMTpro 18-plex) in a single LC-MS/MS run. Tandem Mass Tags (TMT) and isobaric Tags for Relative and Absolute Quantification (iTRAQ) share the same principle: each tag is an isobaric label (same total mass) composed of a reactive NHS-ester group (attaches to peptide primary amines), a mass normalizer (balance group), and a reporter ion of unique mass. At the MS1 level, identically labeled peptides from all conditions appear as a single peak. Upon MS/MS fragmentation (HCD at higher normalized collision energy), each tag releases its reporter ion in the low$m/z$ region (e.g., 126-134 for TMT-10plex), and the reporter ion intensities provide the relative quantification.
| Feature | SILAC | TMT/iTRAQ | Label-Free |
|---|---|---|---|
| Labeling point | In vivo (culture) | In vitro (post-digestion) | None |
| Multiplexing | 2-3 conditions | Up to 18 (TMTpro) | Unlimited |
| Quantification level | MS1 | MS2 (or MS3) | MS1 or spectral count |
| Accuracy | Excellent (CV ~10%) | Good to excellent | Moderate (CV ~20%) |
| Applicability | Cell culture only | Any sample type | Any sample type |
| Known issue | Incomplete incorporation | Ratio compression | Missing values |
TMT Ratio Compression & MS3 Methods
A well-known artifact of isobaric tagging is ratio compression: co-isolated contaminating peptides contribute reporter ions that bias ratios toward 1:1. SPS-MS3 (synchronous precursor selection MS3), implemented on Orbitrap tribrid instruments, addresses this by isolating the top fragment ions from MS2 in an ion trap, transferring them to the Orbitrap for HCD fragmentation and clean reporter ion quantification. This substantially improves quantitative accuracy at the cost of reduced scan speed and sensitivity.
11.3 Data Acquisition Strategies: DDA vs. DIA
The acquisition strategy — how the mass spectrometer selects precursor ions for MS/MS — fundamentally shapes the depth, reproducibility, and quantitative quality of proteomic data. Two paradigms dominate modern proteomics: data-dependent acquisition (DDA) and data-independent acquisition (DIA).
Data-Dependent Acquisition (DDA)
In DDA (also called "shotgun"), a survey MS1 scan identifies the most intense precursor ions, and the instrument sequentially isolates and fragments the top N (typically 10-20) ions in real time. This produces high-quality, single-precursor MS/MS spectra amenable to standard database searching.
Strengths: Clean spectra, well-established search tools, deep coverage with fractionation.
Weaknesses: Stochastic sampling (under-sampling of low-abundance peptides), poor run-to-run reproducibility, missing values in large sample sets. Dynamic exclusion (15-30 s) reduces redundant sampling of dominant peptides.
Data-Independent Acquisition (DIA/SWATH)
In DIA, the mass spectrometer systematically fragments all precursor ions within predefined$m/z$ windows (typically 20-25 Da wide), cycling through the entire $m/z$ range in ~3 seconds. Every detectable peptide is fragmented in every run, eliminating the stochastic sampling problem of DDA and dramatically improving quantitative reproducibility.
Strengths: Complete, reproducible sampling; fewer missing values; excellent quantification.
Weaknesses: Highly multiplexed spectra require specialized analysis (spectral library matching, DIA-NN, Spectronaut, or library-free tools). Initial spectral library generation from DDA is often needed.
The Quantitative Advantage of DIA
In large clinical cohorts (hundreds to thousands of samples), DIA achieves >90% data completeness (quantified in >90% of runs) compared with <50% for DDA without match-between-runs. Modern DIA analysis tools such as DIA-NN use deep learning to predict spectral libraries in silico from the proteome FASTA file, eliminating the need for a pre-acquired DDA library and enabling fully library-free DIA workflows with >8,000 protein groups identified from single-shot analyses of human cell lines.
11.4 Post-Translational Modification (PTM) Analysis
Post-translational modifications (PTMs) vastly expand the functional repertoire of the proteome beyond what is encoded by the genome. Over 400 distinct PTM types have been cataloged, regulating protein activity, localization, interactions, and turnover. Mass spectrometry is the premier technology for global, unbiased PTM mapping, as each modification imparts a characteristic mass shift that is detectable at both MS1 (intact peptide mass) and MS2 (fragment ion level for site localization).
| PTM | Mass Shift (Da) | Target Residues | Enrichment Method | Biological Role |
|---|---|---|---|---|
| Phosphorylation | +79.9663 | Ser, Thr, Tyr | TiO$_2$, IMAC, anti-pTyr Ab | Signaling, cell cycle, metabolism |
| Ubiquitination | +114.0429 (diGly remnant) | Lys | Anti-diGly (K-$\varepsilon$-GG) Ab | Protein degradation, signaling |
| Acetylation | +42.0106 | Lys, N-terminus | Anti-acetyl-Lys Ab | Chromatin regulation, metabolism |
| Methylation | +14.0157 (mono) | Lys, Arg | Anti-methyl Ab, HILIC | Epigenetics, RNA processing |
| Glycosylation | Variable (glycan mass) | Asn (N-linked), Ser/Thr (O-linked) | Lectin, HILIC, PNGase F | Protein folding, cell adhesion |
Phosphoproteomics Enrichment
Phosphopeptides are typically of low stoichiometry and are suppressed by abundant non-phosphorylated species in direct analysis. Enrichment is therefore essential. The two dominant methods are:
TiO$_2$ Enrichment
Titanium dioxide microspheres selectively bind phosphopeptides via bidentate coordination of the phosphate group with surface Ti(IV) atoms under acidic conditions (pH 2-3 in 80% acetonitrile with 6% TFA or lactic acid as a non-phosphopeptide excluder). Bound phosphopeptides are eluted at high pH (NH$_4$OH, pH 10.5). TiO$_2$ shows a preference for singly phosphorylated peptides and enriches pSer/pThr more efficiently than pTyr.
IMAC Enrichment
Immobilized Metal Affinity Chromatography uses Fe$^{3+}$, Ga$^{3+}$, or Ti$^{4+}$ ions chelated to a NTA or IDA resin. Phosphopeptides coordinate to the metal ion through their phosphate group. Elution is achieved with phosphate buffer or ammonium hydroxide. Fe$^{3+}$-NTA IMAC is widely used in large-scale phosphoproteomics and is particularly effective for multiply phosphorylated peptides. Modern high-throughput workflows process 96 samples in parallel using IMAC on magnetic beads.
PTM Site Localization & Stoichiometry
Identifying which residue within a peptide carries the modification requires careful analysis of site-determining fragment ions. Algorithms such as PhosphoRS, PTM-Score (in MaxQuant), and AScore calculate a probability score for each candidate site. A localization probability >0.75 is typically required for confident site assignment.
PTM Stoichiometry Estimation
PTM stoichiometry (site occupancy) is the fraction of protein molecules bearing a modification at a given site. It can be estimated from the ratio of modified to total (modified + unmodified) peptide intensities. However, this requires that modified and unmodified peptides have similar ionization efficiencies, which is often not the case. More robust approaches use chemical derivatization (e.g., phosphatase treatment followed by isotope labeling of the newly freed sites) or the "subtraction" method comparing phosphopeptide enrichment before and after phosphatase treatment.
11.5 Targeted Proteomics: SRM, MRM & PRM
While discovery proteomics casts a wide net, targeted proteomics provides the sensitivity, precision, and throughput needed for hypothesis-driven studies and clinical biomarker validation. Targeted methods monitor predefined peptides (proteotypic surrogates for proteins of interest) with exquisite selectivity and quantitative accuracy, analogous to an ELISA but multiplexable and antibody-free.
Selected/Multiple Reaction Monitoring (SRM/MRM)
SRM (also called MRM) is performed on triple quadrupole (QqQ) instruments. Q1 is set to transmit only the precursor ion of a target peptide at a specific $m/z$; the collision cell (q2) fragments the precursor; and Q3 is set to transmit only a specific fragment ion. Each precursor-to-fragment ion pair is called a "transition." Typically, 3-5 transitions are monitored per peptide for confident identification. The selectivity of two stages of mass filtering provides exceptional signal-to-noise, enabling quantification of targets present at low ng/mL concentrations in complex matrices (e.g., plasma).
SRM Transition Ion Ratio Criterion
where $R_{\text{sample}}$ is the ratio of transition ion areas in the biological sample and$R_{\text{reference}}$ is the expected ratio from a synthetic peptide standard. A deviation <20% across transitions confirms that the signal originates from the target peptide rather than an interfering species. Co-elution of all transitions at the expected retention time provides additional confidence.
Parallel Reaction Monitoring (PRM)
PRM, performed on Q-Orbitrap or Q-TOF instruments, isolates a target precursor in Q1 and records a full high-resolution MS/MS spectrum of all fragment ions simultaneously. Unlike SRM, which monitors selected transitions a priori, PRM captures all transitions in each scan, offering greater flexibility and the ability to retrospectively extract any fragment ion. The high resolution of the Orbitrap (>30,000) provides excellent selectivity even without the second mass filtering step, and PRM has been shown to match or exceed SRM sensitivity for most applications.
Chapter Summary: Key Concepts
- ◈Label-free quantification (spectral counting, ion intensity, iBAQ) is the simplest approach but requires careful normalization and suffers from missing values.
- ◈SILAC provides highly accurate metabolic labeling for cell culture; TMT/iTRAQ enable high-plex chemical labeling applicable to any sample type.
- ◈DIA/SWATH eliminates the stochastic sampling problem of DDA, achieving near-complete quantitative matrices in large-scale studies.
- ◈PTM analysis requires enrichment (TiO$_2$, IMAC for phospho; antibodies for ubiquitin, acetylation) and careful site localization scoring.
- ◈Targeted proteomics (SRM/MRM, PRM) provides the gold-standard quantification needed for biomarker validation and clinical translation.