Quantitative Analysis of Newly Synthesized Proteins

Abstract

Measuring how proteomes respond to perturbations is crucial to understanding the underlying mechanisms involved. Traditional quantitative proteomic methods are limited by the large numbers of proteins in the proteome and the dynamic range of the mass spectrometer . A previously developed method uses the biorthogonal reagent azidohomoalanine (AHA), an analog of methionine, for labeling, enrichment and detection of newly synthesized proteins (NSPs). Newly synthesized AHA-proteins can be coupled to biotin via CuAAC-mediated click-chemistry and enriched using avidin-based affinity purification. . The combination of AHA-mediated NSP labeling with metabolic stable isotope labeling has subsequently allowed the quantitation of low-abundant, newly secreted proteins by mass spectrometry (MS). However, the resulting multiplicity of labeling complicates the analysis of the newly synthesized proteins. We developed a new NSPs quantification strategy called HILAQ (Heavy Isotope Labeled Azidohomoalanine Quantification) which uses a heavy isotope labeled AHA molecule to enable NSP labeling, enrichment, identification, and quantification. In addition, AHA-peptide enrichment employed in HILAQ improves both the identification and quantification of NSPs over AHA-protein enrichment. Here, we provide a step-by-step description of the HILAQ method that includes procedures for (1) pulse-labeling NSPs and harvesting; (2) addition of biotin by click reaction; (3) protein precipitation; (4) protein digestion; (5) enrichment of AHA-biotin peptides by neutravidin beads and four-step elution; (6) MS analysis; and (7) data analysis for identification and quantification by ProLuCID and pQuant. We demonstrate our HILAQ approach using by identifying NSPs from cell cultures, but we anticipate that it can be adapted for applications in animal models. The whole protocol generally takes six days to complete.

INTRODUCTION

Cells respond to perturbations or disease by changing their protein expression levels¹ and measuring these changes is crucial to understanding the underlying mechanisms involved. Traditional approaches have used techniques such as stable isotope labeling to compare protein expression levels between a perturbed and non-perturbed state².. Metabolic stable-isotope labeling based mass spectrometry (MS) analysis allows quantitation of steady-state proteome changes by comparing two fully labeled conditions^{3, 4}, but it usually takes days to weeks to generate these fully-labeled proteomes, and proteins in low abundance can be missed due to the complex nature of the analyzed samples. In addition, labeling efficiency can be an issue in primary cells and fully differentiated cells; consequently, monitoring the immediate proteome response to stimuli or early changes remains a challenge⁵. To overcome this, short-pulse application of stable isotope labeling can be used to only label a small portion of a proteome, but it is difficult to distinguish labeled from unlabeled proteins using MS-based analysis⁶. In addition, it can be challenging to quantitate the changes because protein expression may be small and the detection of newly synthesized proteins (NSPs) is often obscured by the overwhelming static proteome.

To create a means for selective enrichment of NSPs, Dietrich et al ⁷ developed a labeling strategy called biorthogonal non-canonical amino acid tagging (BONCAT), which uses click chemistry for protein enrichment. BONCAT employs L-Azidohomoalanine (AHA), a methionine analogue that can be readily accepted by endogenous methionine tRNA and incorporated into proteins using the cell’s normal translational machinery. In order to pulse-label newly synthesized proteins, AHA can simply be added to the culture media (for cultured cells⁸) or to food pellets (for animals⁹) without measurable adverse effects on cellular functions^10,11. The azide moiety in AHA can specifically react with an alkyne-bearing biotinylated tag in the presence of CuAAC catalyst, allowing subsequent selective enrichment of the modified protein pool using avidin affinity purification. BONCAT has been widely used for NSP identification in cell lines. Most analyses using AHA have been qualitative, but Howden et al used AHA labeling combined with heavy and light stable isotope labeled amino acids to quantitate low-abundance, newly secreted proteins by mass spectrometry (MS)¹², as well as to quantitate new proteins synthesized after T-cell stimulation (QuaNCAT)¹³. In the QuaNCAT method, AHA is exploited to enrich NSPs using avidin beads, and heavy isotope labeled amino acids (AA) not only confirm proteins as newly synthesized but also allow quantification in the MS. Although effective, the resulting multiplicity of labeling makes analysis of newly synthesized proteomes challenging. In addition, the limitations in stable isotope labeling of primary cells and fully differentiated cells remain.

To address the need for a more straightforward method to quantify NSPs, we recently developed a new strategy using a heavy isotope labeled AHA molecule (heavy-AHA or hAHA) ¹⁴. This strategy, termed HILAQ (Heavy Isotope Labeled Azidohomoalanine Quantification), enables NSP enrichment, confirmation and quantification using a single stable isotope labeled molecule. Here, we provide a step-by-step HILAQ protocol for cell labeling, click chemistry reaction, protein precipitation, trypsin digestion, enrichment of AHA-biotin peptides, elution, MS analysis and data analysis (Figure 1). One of the key stages in this protocol is the enrichment of NSPs on the peptide level. We have previously demonstrated that AHA peptide enrichment is more effective than protein enrichment in biotin labeled protein identification ¹⁵. We successfully applied our HILAQ strategy to a HT22 oxytosis model to quantitate the NSPs, demonstrating the validity of HT22 cells as a tool to study the molecular details of cell death involved in neurodegenerative diseases ¹⁴.

Figure 1. Flow chart of the protocol, including references to the steps in the procedure.

AHA = L-Azidohomoalanine; hAHA = heavy-AHA.

Advantages over other methods

Traditional quantitative metabolic labeling strategies take days to weeks to produce a fully labeled proteome and to generate differentially expressed protein networks by comparing two fully labeled proteomes⁴. However, using HILAQ we only need to label cells for hours or feed/inject animal for 1–4 days before enriching and analyzing NSPs. In theory, quantitating the newly synthesized proteome using AHA is more sensitive than measuring NSPs in the static proteome, as the ability to enrich improves the detectable abundance range (Figure 2).

Figure 2. Schematic diagram of the differences between the previous traditional quantitative strategy and HILAQ.

Size of the shapes does not correlate with the protein expression level. Briefly, the standard quantitative proteomic strategy compares the whole proteome after taking weeks to finish labeling, while NSPs quantitation compares the enriched NSPs part after pulse labeling for hours. AHA = L-Azidohomoalanine; hAHA = heavy-AHA.

Compared with the previously published NSPs quantification strategy QuaNCAT, the HILAQ workflow offers several advantages: (1) HILAQ simplifies the cell labeling and data search using a single heavy isotope labeled AHA molecule. (2) HILAQ uses peptide enrichment, which has been demonstrated to be more sensitive than the protein level enrichment used in QuaNCAT, with the same accuracy ^{14, 15}. (3) HILAQ is a more informative strategy, and many more NSPs (>3.5 times) were quantified with HILAQ than with QuaNCAT.

Limitations

While there are many advantages to using HILAQ, there are some technical limitations. First, the HILAQ approach is based on a heavy isotope labeled AHA molecule. The current workflow only allows duplex quantification, which can be used to compare two different conditions with a mass difference of 6 Da. This limitation was overcome in 2016 by combining AHA with isobaric labeling (iTRAQ) to enrich and quantitate NSPs¹⁶. This study used protein enrichment of biotinylated proteins, while HILAQ employs peptide enrichment. Our lab and others have demonstrated that peptide enrichment is ~5 times superior to protein enrichment for biotinylated proteins ¹⁵. In addition, this study performed biotin-alkyne removal before MS analysis and resulted in high false positive rate, while HILAQ retained biotin-alkyne tag to specify the mass-shift added on the newly synthesized peptides. However, in unpublished experiments, we have observed that Tandem Mass Tag (TMT)¹⁷ labeled AHA-biotin peptides show poor enrichment, resulting in a greatly reduced number of biotinylated peptides identified. We suspect that the addition of the large TMT tag along with the large biotin-alkyne tag negatively affects the solubility and the elution profile of the labeled peptides.

Second, as mentioned in previous publications¹⁸, AHA is known to be incorporated into NSPs less efficiently than methionine. Although previous studies showed that AHA labeling has no substantial adverse effect on protein synthesis or degradation, it is possible that methionine starvation and AHA labeling have some effect on certain signaling pathways.

Furthermore, HILAQ labeling is not applicable to methionine-free proteins. Although methionine-free proteins comprise only 1.02% of the whole known proteome, an additional 5.08% of proteins have a single N-terminal methionine which can be readily removed post-translationally¹⁰. Nevertheless, HILAQ can be used to analyze at least 94% of the proteome. In the HILAQ workflow, the added AHA-biotin modification is retained after enrichment and during MS and data analysis. This increases the confidence of identified NSPs, but it presents a challenge for liquid chromatography separation, as the hydrophobicity and mass of peptides are increased by addition of AHA-biotin. The HPLC gradients have been optimized to address this issue.

Potential applications

The capacity of HILAQ to detect and quantify NSPs with high sensitivity after brief labeling of cells makes it potentially useful for a wide range of studies, including studies of protein degradation¹⁹, investigation of immune response to perturbations ²⁰, translational regulation ²¹ and iPSC differentiation ²² using cell line models or primary cells . We anticipate that this strategy and workflow will be a powerful tool for quantifying proteome dynamics and will pave the way for new types of studies to discover mechanisms of disease. HILAQ can also be applied to studies using animal models. Our group previously published a study in which AHA labeling was successfully employed in mice and quantitation was performed using heavy isotope labeled biotin²³. We expect that heavy isotope labeled AHA is also compatible with mouse studies, and we have ongoing projects to test the use of HILAQ in mice.

Overview of the Procedure

The procedure of HILAQ is summarized in Figure 1 and consists of the following key stages:

• Cell labeling and harvest (steps 1–5): Methionine depletion is performed by culturing cells in methionine-free medium, followed by treatment with two different conditions (e.g. control and drug treatment) and labeling with either AHA or hAHA.

• Equal amounts of proteins from each condition are mixed together, followed by the addition of biotin to AHA or hAHA labeled NSPs by click reaction (step 6–18). In this protocol, we use 0.5mg of total protein as starting material. Although we have not tested the limit of the minimum amount of starting materials required for the HILAQ, we believe lower amounts may be used.

• Protein precipitation by TCA precipitation to remove excess click reagents (steps 19–23).

• Protein reduction, alkylation and trypsin digestion (steps 24–30)..

• Enrichment of AHA-biotin peptides by Neutravidin beads, followed by stringent washes to get rid of nonspecific peptides and a four-step elution to collect AHA/hAHA peptides (steps 31–48).

• MS analysis (steps 49–56).

• Data analysis for identification using ProLuCID from the software platform Integrated Proteomics Pipeline (IP2) and quantification using pQuant. MS and MS/MS data are extracted from RAW files by using RawXtractor, identified using ProLuCID from the software platform IP2 and quantitated using pQuant (steps 57–58).

Experimental design

Cell culture

AHA labeling efficiency can be affected by the composition of the growth medium, the cell density and cell type. For the majority of cell types, about 80% confluency at the time of treatment will be sufficient for labeling. A typical NSPs quantitation study compares cells in two conditions (e.g. condition A—drug treatment, condition B—control), and includes at least three biological replicates (condition A is labeled by hAHA and condition B is labeled by AHA in all three replicates) or a pair of swap experiments (condition A is labeled by hAHA and condition B is labeled by AHA in one replicate and reversed labeling in other replicate). Each replicate is an equal mixture of AHA/hAHA labeled cells. After removing the complete growth media, cells are cultured in methionine-free labeling medium for 30 min to deplete intracellular methionine (step 2). Cells are then cultured for 1hours with AHA or hAHA in methionine-free labeling medium supplemented with 10% (v/v) dialyzed FBS (step 3). Use of different AHA/hAHA concentrations and longer or shorter incubation times should be determined according to the specific nature of the experiment and the cells used. We found 1h labeling with 1mM AHA/hAHA is sufficient for HEK293T and HT22 Cell lines. We recommend using dialyzed FBS to avoid the residual L-methionine present in the normal serum, which will compete with AHA and decrease the efficiency of AHA incorporation. To check the AHA labeling efficiency of cells at the click reaction step, AHA labeling in HEK293T cells can be performed in parallel as a positive control.

Click reaction

The click reaction between AHA and biotin is one of the most critical steps in this workflow. To minimize interference from detergents with the click reaction, we optimized the protocol by balancing conditions that maximize protein solubility with conditions that support an efficient click reaction. Cells are harvested, suspended in DPBS containing protease inhibitors and sonicated (step 6). Equal amounts of protein from hAHA labeled or AHA labeled samples are mixed together followed by another round of sonication (step 7–9). After centrifugation, supernatants are transferred to new eppendorf tubes and an appropriate number of aliquots are made based on the amount of starting material (steps 11–12). The pellets, which contain insoluble proteins, are suspended and sonicated in 0.5% (v/v) SDS in DPBS until the solution becomes clear (step 12). It has been reported that the click reaction is less efficient with starting reaction volumes greater than 500 μl ²⁴. In the HILAQ work flow, the click reaction is performed in a 400 μl volume to maximize efficiency, with a final concentration of SDS of no more than 0.05% to minimize the interference of SDS with the click reaction. If the amount of starting material is equal or less than 2.5 mg, the supernatant is divided into 5 aliquots, as is the corresponding 0.5% SDS dissolved pellet; if the amount of starting material is greater than 2.5mg, the number of aliquots is determined by dividing starting material (in mg) by 0.5 (step 11). The number of aliquots of 0.5% SDS dissolved pellets should equal the number of aliquots of supernatant. The click reaction is performed separately on each aliquot (step 13). After the click reaction is completed, all tubes are combined and proteins are precipitated by trichloroacetic acid solution (TCA) precipitation to remove excess click reagents (step 19). For optimization of some cell lines or primary cells, before adding TCA to combined click reactions, approximately 50 μl should be removed to check the AHA labeling efficiency. Using click reactions from HEK293T cells with 1h labeling of 1mM AHA/hAHA as control, western blots with HRP conjugated streptavidin or biotin quantitation commercial kits are performed to provide a relative measurement of biotin. With that, researchers can better figure out AHA/hAHA labeling time and concentration comparing with the HEK293T datasets provided in this study.

Trypsin digestion and enrichment

An important improvement in the HILAC strategy over the QuaNCAT strategy¹³ is the enrichment of labeled proteins on the peptide level instead of the protein level. To this end, trypsin digestion is performed on samples immediately after TCA precipitation (step 24–30). The use of mass spectrometry compatible surfactant (such asProteaseMAX ) in proteolytic digestion protocols has been shown to dramatically increase peptide and protein identifications in complex protein mixtures ^{25, 26}. ProteaseMAX, a commercial MS compatible surfactant, is used to shorten the digestion time while maximizing the trypsin digestion efficiency. Urea is used to solubilize protein pellet and increase protease efficiency in protein digestion. It is important to note that urea can cause carbamylation, which can interfere with enzymatic digestion of proteins, resulting in incompletely digested peptides. To overcome this, trypsin digestion has previously been optimized in the presence of ammonium bicarbonate to inhibit peptide carbamylation in Urea ²⁷. Using the digestion protocol provided in HILAQ workflow, carbamylation is detected at very low levels and does not cause a loss of protein identification. Peptide level enrichment using neutravidin resin is performed after trypsin digestion, followed by stringent washes on the shaker to eliminate unmodified peptides (steps 39–43). AHA/hAHA-biotin peptides are subsequently eluted by two treatments with elution buffer (80% (v/v) acetonitrile, 0.2% (v/v) formic acid, 0.1% TFA) on a shaker at RT and twice more at 70°C. It has been previously described that biotin-streptavidin interactions can be disrupted using water at temperatures above 70°C ²⁸, and we found that incubating at 70 °C in elution buffer leads to better recovery of biotin-labeled peptides. It should be noted that freeze/thaw of samples after elution reduces the number of identified NSPs. Therefore, it is better to elute the AHA-biotin peptides on the same day they are analyzed on the mass spectrometer.

Sample fractionation and mass spectrometry analysis

We fractionate the eluted peptides using MudPIT (steps xx-yy), which is a known powerful on line LC/LC fractionation strategy ²⁹ and offers several advantages for HILAQ samples.: First, MudPIT fractionation avoids sample loss by coupling C18 and strong cation exchange in a single column. Since the HILAQ is a strategy based on enrichment of newly synthesized proteins and it has been reported that <<1% of total proteins were labeled after a 2 h AHA pulse³⁰, the amount of AHA-labeled protein is expected to be small. Sample loss could therefore be a critical issue when performing off-line fractionation and can result in the loss of information, especially for samples with a small amount of starting material or low-labeling efficiency³¹. Second, A MudPIT fractionation strategy provides excellent dynamic range for the analysis of complex peptide mixtures, which is important since NSPs may cover a range of highly abundant housekeeping proteins to low level proteins. The use of an LC/LC method provides the best dynamic range of analysis that can be used over a range of mass spectrometers. MudPIT allows the analysis of HILAQ samples on slower scanning mass spectrometers with retention of good dynamic range. One dimensional LC (single phase) also can be employed in HILAQ sample analysis, but it requires a faster scanning mass spectrometer to get comparable results. The choice of method (MudPIT or single phase) depends on the complexity of the prepared HILAQ samples and the available mass spectrometer. For complex samples, MudPIT or off-line LC/LC is a highly recommended separation method for HILAQ sample analysis as it can be used on slow or fast scanning mass spectrometers to get comparable results, but for simple samples a one-dimensional LC (single phase) could be adequate on a late model faster scanning mass spectrometer.

Data processing

The quantitation of HILAQ relies on the quantification of MS1 spectra. A typical MS1 spectrum, which indicates the mass difference between a hAHA labeled peptide and the corresponding AHA labeled peptide, is shown in Supplementary Figure 1. We suggest that only quantitated proteins that have at least two occurrences in all replicates are suitable for further analysis. Data search and analysis of HILAQ samples is not a routine MS1 quantification. Because the AHA-biotin label is retained on peptides when subject to mass spectrometry in HILAQ, we treat AHA/hAHA-biotin labeling as a differential modification in a database search (steps 57–58), while other strategies, such as SILAC, are treated as a static modification. The reason these labels are treated differently is that AHA labeling is performed for a short time period. It can therefore not be assumed that every methionine is labeled, while SILAC assumes complete labeling of the proteome. Most stable isotope quantitation software tools assume complete labeling, and are not efficient for quantitation after pulse labeling. Pulse labeling requires that identification and quantitation tools are able to customize the modification with designed element composition. This customized modification should be as flexible as possible to accommodate the variations in the elements of the heavy isotopes. We have worked with the developers of ProLuCID³², pQuant ^{23, 33} to modify the fundamental data structures and configuration models to accommodate pulse labeling quantification. Specifically, pQuant recognizes the customized modification on methionine with light and heavy isoforms reported from ProLuCID, and extracts the corresponding precursor chromatograms for quantitation as set by users. Currently we use ProLuCID for identification and pQuant for quantification for HILAQ data analysis. We have provided step-by-step instructions for each software and all parameters we used in Supplementary Manual 1 and 2. Other software programs can be used if they can perform differential searches and quantitation for heavy and light modifications.

MATERIALS

REAGENTS

!CAUTION

The use and disposal of the listed reagents should follow the instructions in their respective material safety data sheets. It is also important to use the proper personal protective equipment (PPE) while handling the hazardous reagents.

• Cells of interest. In this Protocol we use HEK293T cells (ATCC, cat. no. CRL-3216) and HT22 cells (a generous gift from Salk Institute (La Jolla, CA, USA)). Cells should be approximately 80% confluent when used for labeling !CAUTION Cell lines should be regularly checked to ensure that they are authentic and not infected with mycoplasma.

• 1×DPBS (Dulbecco’s phosphate-buffered saline) (Gibco, cat.no.14190–136)

• Pierce BCA protein assay kit (reagent A, cat.no.23228; reagent B, cat.no.23224)

• Water, molecular biology grade (Sigma-Adrich, cat. no. W4502) !CRITICAL Unless otherwise specified, molecular biology grade water is used for the preparation of solutions used in the experiments. Molecular biology grade water is also used for all experimental procedures.

• L-Azidohomoalanine-HCl, 13C4, 15N2 (hAHA, Cambridge isotope laboratories, cat.no.CNLM-9461), !CRITICAL Store the powder at −20°C, protected from light.

• L-Azidohomoalanine-HCl (AHA, Cambridge isotope laboratories, cat.no.ULM-9460) !CRITICAL Store the powder at −20°C, protected from light.

• DMEM, high glucose medium (Gibco, cat.no.10569–010)

• protease inhibitor cocktail (Roche, cat. no. 04693124001)

• Dimethyl sulfoxide (DMSO) (Sigma-Adrich, cat. no. 276855) !CAUTION DMSO is highly skin-permeable and combustible, and can cause serious eye irritation. It should be properly handled with appropriate PPE (protective gloves, eye protection and face protection) under a fume hood when a large volume is used.

• 0.5% (v/v) SDS in molecular biology grade water. !CAUTION SDS is flammable and harmful to respiratory system. Avoid breathing its dust when scaling the powder, always wear appropriate gloves when handling it and keep it away from ignition sources.

• tert-butanol (Sigma-Adrich, cat. no. 360538) !CAUTION tert-butanol is flammable liquid with acute toxicity which can cause serious skin and eye irritation. It should be properly handled with appropriate PPE. Store in a segregated and approved area. Keep container in a cool, well-ventilated area. Keep container tightly closed and sealed until ready for use. Avoid all possible sources of ignition (spark or flame).

• Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine (TBTA) (Sigma, cat. no. 678937)

• Biotin-Alkyne (Click Chemistry tools, TA105–25)

• Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) !CAUTION TCEP powder is corrosive. It should be handled carefully with protective gloves.

• Copper Sulfate (Sigma-Adrich, cat. no. 451657)

• Trichloroacetic acid solution (TCA) (Sigma-Adrich, cat. no. T0699) !CAUTION TCA is a corrosive acid. It should be handled in a hood with personal protective equipment, including lab coat, gloves and safety glasses.

• Acetone, HPLC grade (Fisher chemical, cat. no. A949SK-4) !CAUTION Acetone has modest toxicity in small doses and should be handled in a hood using gloves. Be aware, Acetone has quite high flammability. Avoid all possible sources of ignition (spark or flame).

• Urea (Sigma-Adrich, cat. no.U5378).

• 2-Chloroacetamide (Sigma-Aldrich, cat. no.22790). Store protected from Light.

• Ammonium bicarbonate (Sigma-Adrich, cat. no.09830). Trypsin (Promega, cat.no. V5117)

• Fluorescence biotin quantitation kit (Thermo scientifics, cat.no. 46610)

• NeutrAvidin resin (Thermo scientifics, cat. no.29200)

• Acetonitrile (Sigma-Adrich, cat. no.271004) !CAUTION Acetonitrile has modest toxicity in small doses and should be handled in a hood using gloves.

• Trifluoroacetic acid (TFA, Sigma-Adrich, cat. no. T6508) !CAUTION TFA is a corrosive acid. It should be handled in a hood with personal protective equipment, including lab coat, gloves and safety glasses.

• Jupiter 10 μm proteo 90 Å C18 resin (Phenomenex, cat. no.04A-4397)

• Luna 5 μm Strong cation exchange resin 100 Å (Phenomenex)

• Jupiter 4 μm proteo 90 Å C18 resin (Phenomenex, cat. no.04A-4396)

EQUIPMENT

• CO₂ incubator (Thermo scientific, series 8000 WJ)

• Biohit Sartorius mLINE mechanical pipettes (Sigma-Adrich)

• Microtubes (1.7ml, Olympus plastics, cat. no.22–282)

• Falcon tubes (15ml, Corning, cat. no.352196)

• Refrigerated microcentrifuge: Prism R (Labnet, cat. no.C2500-R)

• Microfuge 18 Centrifuge (Beckman coulter)

• Thermomixer R (Eppendorf).

• Soniprep 150 (MSE, cat. no. MSS150.CX2.1). Tip with 3 mm diameter and 7:1 transformation rate is used for all sonications in the protocol. Power 6 means the number indicated on amplitude meter is six. For more details, explanations and instructions, see manual at http://www.bio.vu.nl/~microb/Protocols/Manuals/Soniprep150.pdf (check amplitude meter in page 11) !CAUTION The tip sonicater should be sound-insulated, or ear protection should be worn when operating the equipment.

• ProteomeLab SP (Beckman Coulter, cat. no. R-134a)

• CentriVap concentrator (Labconco, cat. no.7810016)

• CentriVap Cold Trap (Labconco, cat. no.7811020)

• Dry pump: DryFastUltra (Welch, cat. no.2042B-01)

• VWR vortexer (VWR, cat. no.58816–121)

• Agilent 1100 quaternary HPLC (Agilent)

• Mass spectrometer: Velos Pro mass spectrometer (Thermo Fisher Scientific) or Orbitrp Elite mass spectrometer (Thermo Fisher Scientific).

Software

• Rawxtractor 1.9.9.2 ( can be downloaded from http://fields.scripps.edu/downloads.php)

• IP2 - Integrated Proteomics Pipeline (http://goldfish.scripps.edu/ip2/login.jsp)

• pQuant (freely available from http://pfind.ict.ac.cn/software/pQuant/index.html)

• Java ( freely download from https://www.java.com/en/download/win10.jsp)

REAGENT SETUP

!CAUTION

Relevant institutional rules regarding chemical use and waste disposal should be observed.

• L-Azidohomoalanine-HCl 1M stock solution. Dissolve 18.655mg hAHA in 100ul molecular biology grade water to make 1M stock. This is sufficient to label 100ml media with 1mM final concentration of hAHA, and is comparable in cost to labeling with SILAC. This can be stored at −20°C, protected from light until precipitate can be seen

• L-Azidohomoalanine-HCl 1M stock solution. Dissolve 18.095mg AHA in 100ul molecular biology grade water to make 1M stock. This is sufficient to label 100ml media with 1mM final concentration of AHA, and is comparable in cost to labeling with SILAC. This can be stored at −20°C, protected from light until precipitate can be seen.

• Complete grow medium: DMEM, high flucose medium with 10% (v/v) FBS (SIGMA, cat.no.F0392) and 1% (v/v) Pen Strep (Gibco, cat.no.15140–122). This can be stored at 4°C, protected from light until precipitate can be seen.

• TBTA 50x stock solution. Make a 50× stock by dissolving 8.85mg TBTA in 200 μl of DMSO. This is stable at room temperature (RT, 20–24 °C) for years. Prepare 1.7 mM TBTA by adding 20 μl of 50× stock to a glass vial containing 180ul DMSO. Mix thoroughly and add 800ul of tert-butanol. This 1.7 mM solution is stable for months at RT. !CRITICAL Discard if precipitation is observed in stock or solutions.

• Biotin-Alkyne 5 mM stock solution. Dissolve 25mg in 10.927ml DMSO to get 5mM stock. Aliquot it by 500ul volume and store in −20°C for up to 1 year. Protected from light. !CRITICAL The solution should not be subjected to repeated freeze–thaw cycles, as its quality may be adversely affected.

• TCEP 50 mM stock solution. Dissolve 14.3mg in 1ml molecular biology grade water to make 50mM stock. Make it fresh for click reaction. !CAUTION TCEP powder is corrosive. It should be handled carefully with protective gloves.

• Copper Sulfate 50mM stock solution. Dissolve 7.98mg in 1ml molecular biology grade water to make 50mM stock. Separate into 500ul aliquots and store at −20°C for up to 6 months. Discard after thawing. !CRITICAL Solutions required for click chemistry may not contain ion chelators such as EDTA and EGTA in order to prevent the inactivation of the copper (I) catalyst.

• Urea 8M stock solution. To make 8M Urea, add water to 1ml with 480mg urea dissolved.Make it fresh before use.!CRITICAL There is a volume effect when dissolving urea in water.

• Ammonium bicarbonate 50mM stock solution. To make 50mM ammonium bicarbonate, dissolve 39.53 mg in 10ml water. Aliquot it by 500ul volume and store at −20°C for up to 6 months.

• Labeling media. Add 10% (v/v) dialyzed FBS (Hyclone, cat.no. SH30079.01HI), 1% (v/v) pen/strep, 1% (v/v) Glutamax (Gibco, cat.no. 35050–061), 1% (v/v) sodium pyruvate (Gibco, cat.no. 11360–070), 4 mM MgCl₂ (Sigma, cat. no. M4880) and 4 mM CaCl2 (Sigma, cat. no. C5670) in 500ml 1×HBSS (Gibco, cat. no. 14175095). Store at 4 °C until precipitate can be seen.

• Lysis buffer. Dissolve one protease inhibitor cocktail tablet in 10ml DPBS. Aliquot can be stored at −20 °C for up to 6 months.

• ProteaseMAX solution A. Dissolve 1mg ProteaseMAX™ surfactant (Promega, cat. no.v2072) in 100ul 50mM ammonium bicarbonate to make 1% (w/v) solution. Freeze at −20 °C for up to 6 months.

• ProteaseMAX solution B. Solution A is diluted 5X in 50mM ammonium bicarbonate to make 0.2% (w/v) solution. Freeze at −20 °C for up to 6 months.

• Elution Buffer. 80% (v/v) acetonitrile, 0.2% (v/v) formic acid, and 0.1% (v/v) TFA. Make fresh prior to use.

• Buffer A. 5% (v/v) acetonitrile, 95% (v/v) water and 0.1% (v/v) formic acid. Store at RT for up to 1 month

• Buffer B. 80% (v/v) acetonitrile, 20% water (v/v) and 0.1% (v/v) formic acid. Store at RT for up to 1 month.

• Buffer C. 500 mM ammonium acetate, 5% acetonitrile (v/v) and 0.1% (v/v) formic acid. Store at RT for up to 6 months.

EQUIPMENT SETUP

Agilent 1100 quaternary HPLC setup:

The sample was analyzed using MudPIT, with a modified 11-step separation described previously ³⁴. Step 1 consisted of a 10 min gradient from 0–10% Buffer B, a 50 min gradient from 10–50% Buffer B, a 10 min gradient from 50–100% Buffer B, and 20 min 100% Buffer A. Step 2 consisted of 1 min of 100% Buffer A, 4 min of 20% Buffer C, a 5 min gradient from 0–10% Buffer B, a 80 min gradient from 10–45% Buffer B, a 10 min gradient from 45–100% Buffer B, and 10 min 100% Buffer A. Steps 3–9 had the following profile: 1 min of 100% Buffer A, 4 min of X% Buffer C, a 5 min gradient from 0–15% Buffer B, a 90 min gradient from 15–45% Buffer B, and 10 min 100% Buffer A. The Buffer C percentages (X) were 30, 40, 50, 60, 70, 80, 100% for the steps 3–9, respectively. In the final two steps, the gradient contained: 1 min of 100% buffer A, 4 min of 90% buffer C plus 10% B, a 5 min gradient from 0–10% buffer B, a 80 min gradient from 10–45% buffer B, a 10 min gradient from 45–100% buffer B, and 10 min 100% buffer A.

	Time (min)	Buffer A (%)	Buffer B (%)	Buffer C (%)
Step 1	0 – 10	100 – 90	0 −10	0
	10 – 60	90 – 50	10 – 50	0
	60 – 70	50 – 0	50 – 100	0
	70 – 90	100	0	0
Step 2	0 – 1	100	0	0
	1–5	80	0	20
	5 – 10	100 – 90	0 – 10	0
	10 – 90	90 – 55	10 – 45	0
	90 – 100	55 – 0	45 – 100	0
	100 – 110	100	0	0
Step 3–9	Same as step 2	…	…	…
	1 – 5	70/60/50/40/30/20/0	0	30/40/50/60/70/80/100
	Same as step 2	…	…	…
Step 10–11	0 – 1	100	0	0
	1 – 5	0	10	90
	5 −10	100 – 90	0 – 10	0
	10 – 90	90 – 55	10 – 45	0
	90 – 100	55 – 0	45 – 100	0
	100 – 110	100	0	0

A capillary is used to test the flow rate (approx. 400 nl/min), and the length of waste line should be adjusted to make sure the pressure is between 60–80 bar with buffer A flow through. As peptides elute from the micro-capillary column, they are electrosprayed directly into an Orbitrap Elite mass spectrometer.

Thermo Easy nLCII pump setup:

Since the Thermo Easy nLCII pump only allows two mobile phases to flow at the same time, we adjust the buffers to achieve the same gradient as with the Agilent 1100 quaternary HPLC. Instead of using a buffer C pulse in each step, we used different percentages (X) of buffer C in buffer A or buffer B (20, 30, 40, 50, 60, 70, 80, 100% for the steps 2–9, 90% buffer C with 10% buffer B for both step 10 and 11), which were placed in auto-sampler in order. “Sample” (X% Buffer C) injection volume was set as 10 μl and the flow rate was set as 400 nl/min.

Mass spectrometer setup:

	Orbitrap Elite
Ionization mode	Positive
Electrospray voltage	2.5kV
MS1 Resolution	240k
MS2 Resolution	Rapid
MS1 scan range	400–2000
MS1 maximum injection time	250ms
MS2 maximum injection time	50ms
MS1 AGC target	1e6
MS2 AGC target	1e4
Isolation	2m/z with ion trap
Fragmentation	CID at 35%
Dynamic Exclusion	60 seconds
Charge states	+1 and unassigned are rejected
Intensity Threshold	500
DDA method	Top 20

PROCEDURE

Cell labeling (Timing: 2–4 h)

1. Prepare 1×DPBS labeling media with 1mM hAHA or 1mM AHA (see REAGENTS and REAGENTS SETUP) and warm to 37 °C for 30 min. 7ml of each solution is needed for a round 100mm dish.

2. Rinse the cells carefully in 6ml 1×DPBS, add the same volume of pre-warmed labeling media and incubate for 30 min in a 37 °C, 5% CO₂ incubator. This incubation step allows for the depletion of methionine. .

   CRITICAL STEP We have successfully used this approach for HEK293T and HT22 cells. When testing new cell types, it is important to test the labeling efficiency in Step 18. Incubation time and AHA/hAHA concentration can be modified to optimize labeling efficiency.

   CRITICAL STEP Cells should be approximately 80% confluent and grown in complete growth medium when starting the procedures.

3. Remove the labeling media and add 6ml fresh labeling media with 1mM hAHA or 1mM AHA to the cells. Move the cells back to the 37 °C, 5% CO₂ incubator for desired incubation time. In our experiment, we found an incubation of 1 h to be sufficient. However, longer and shorter incubation times should be selected according to the specific nature of the experiment and the cells used.

4. Remove the labeling media, rinse the cells carefully in 6 ml 1×DPBS.

5. Harvest the cells in 2 ml 1xDPBS using a cell scraper. Centrifuge at 1,000g for 5 min at 4°C and carefully remove all supernatant.

   ▪ PAUSE POINT: cell pellets can be stored at −80°C for several months.

   Click chemistry reaction (Timing: 2.5–3 h)

6. Resuspend the cells harvested from one 10mm dish in 200 μl lysis buffer (REAGENTS SETUP) in 1.5 ml eppendorf tubes and sonicate using the tip sonicator for 10 sec on ice, followed by 1min incubation on ice. Repeat the sonication cycle twice.

7. Measure the protein concentration using Pierce BCA protein assay kit.

8. Mix equal amounts of protein from the AHA labeled sample and the hAHA labeled sample.

9. Add lysis buffer to the appropriate volume and sonicate the protein mixture using the tip sonicator at power 6 for 10 secs on ice, followed by 1 min. incubation on ice. Repeat the sonication cycle twice.

   • CRITICAL STEP: The volume of lysis buffer added depends on the amount of starting material. For 5mg or less starting material, we add up to 200 μl of lysis buffer to make sure the sonication can be well performed. When using more than 5mg starting material, add 100 μl lysis buffer for every 2.5mg.

10. Centrifuge the samples at 21,000 g for 10 min at 4 °C.

11. Transfer the supernatant to a new tube and divide the supernatant into aliquots, according to the table below:

    Amount of starting material	Number of aliquots
    ≤ 2.5 mg	5
    > 2.5 mg	mg starting material / 5

12. Resuspend the pellets in 100 μl 0.5% (w/v) SDS in DPBS per 2.5mg starting material (with a minimum of 100 μl) Sonicate the samples to make sure there are no big chunks. Boil the samples for 10min at 100°C. After cooling to RT, divide into aliquots according to the table below:

    Amount of starting material	Number of aliquots
    ≤ 2.5 mg	5
    > 2.5 mg	mg starting material / 5

    • CRITICAL STEP: making appropriate aliquots is important for click reactions. Examples are shown here. For 2 mg starting materials, split the supernatants were into 5 aliquots after centrifugation; dissolve the pellets were dissolved in up to 100 μl 0.5% (w/v) SDS in DPBS and divide them into 5 tubes equally. For 5mg starting material, split the supernatants into 10 aliquots after centrifugation; dissolve the pellets in up to 200 μl 0.5% (w/v) SDS in DPBS and divide them into 10 tubes equally. Make sure that the final concentration of SDS is no more than 0.05% in 400 μl click reactions in step 13.

13. Mix the reagents in the order shown below (see REAGENTS and REAGENTS SETUP) in one eppendorf tube. After the addition of each reagent, vortex the tube for 30sec. This table shows the quantity of reagents needed for one click reaction. For multiple reactions, multiply the number of “μl added” (see form below) by the number of click reactions (add an additional two just in case) to be performed.

    Order	Reagent	[Stock]	ul added	[Final]
    1	TBTA	1.7mM	30	100uM
    2	Copper Sulfate	50mM	8	1mM
    3	Biotin-Alkyne	5mM	8	100uM
    4	TCEP	50mM	8	1mM

    • CRITICAL STEP: Fresh TCEP must be used to maximize click reaction efficiency, otherwise it will cause the copper to precipitate.

14. Add 54 μl to each aliquot from step 11&12.

15. Bring volume of each reaction up to 400 μl with 1×DPBS. Vortex each click reaction.

    • CRITICAL STEP: The reactions will require different amount of DPBS depending on the volume of samples.

16. Incubate for 30min at RT.

17. Vortex each click reaction and incubate for another 30 min at RT.

18. Combine all aliquots derived from the same sample in a 15 ml falcon tube. …. For new cell types, before adding TCA to combined click reactions, approximately 50 μl should be removed to check the AHA labeling efficiency for optimizing AHA/hAHA labeling time and concentration. Using click reactions from HEK293T cells with 1h labeling of 1mM AHA/hAHA as control, western blots with HRP conjugated streptavidin or fluorescence biotin quantitation kit are performed to provide a relative measurement of biotin.

    • PAUSE POINT : samples can be stored at −80°C at this point for short period of time (1 week)

    TCA precipitation (Timing: >8 h)

19. Add TCA to combined reactions as 1:4 (v/v) and incubate at 4°C on rotator overnight. Centrifuge at 3,000g for 30 min at 4 °C.

20. Discard the supernatant. Suspend the pellet with 1 ml ice cold acetone and transfer to a 2ml tube. Vortex to see if pellet can be dispersed without chunks. If not, sonicate using the tip sonicator at power 10 for 10 sec, then add another 1ml ice cold acetone. Centrifuge at 17,000g for 10 min at 4 °C.

21. Remove the supernatant quickly and carefully. Add 2ml ice cold acetone and vortex. Centrifuge at 17,000g for 10min at 4 °C. Repeat twice.

    • CRITICAL STEP: The pellet is very easily disturbed. Make sure to remove the supernatant as soon as centrifugation is finished.

22. Remove the supernatant completely.

    • PAUSE POINT: samples can be stored at −80°C at this point for several weeks.

23. Dry the pellet at RT for at least 30 min.

    • PAUSE POINT: samples can be stored at −80°C at this point for several weeks.

    Trypsin digestion (Timing: 4.5 h)

    !CRITICAL The volume of reagents used in steps 24–30 is based on 5 mg starting material or less. For larger amount of starting material, all reagents volumes in these steps should be proportionally increased according to the size of sample pellet from step 23.

24. Add 50 μl ProteaseMAX solution B (see REAGENTS SETUP) to each sample

25. Add 50 μl of 8M urea in water and pipette up and down to break pellets, if any.

26. Add 5 mM TCEP (final concentration) for 20 min at 55°C while shaking.

27. Add 10 mM 2-chloroacetamide (final concentration) and incubate for 20 min at room temperature in the dark.

28. Add 150 μl of 50 mM ammonium bicarbonate.

    • CRITICAL STEP: check to make sure there are no chunks. If chunks are present, sonicate it using tip sonicator at power 10 for 10 seconds.

29. Add 2.5 μl ProteaseMAX solution A.

30. Add trypsin at 1:50 (wt:wt, based on the amount of starting material in step 8) and then incubate on shaker at 37 °C for 3 h.

    • PAUSE POINT: samples can be stored at −80°C at this point for several months.

    NeutrAvidin Enrichment (Timing: 3.5 h)

31. Centrifuge the samples from step 30 at 17,000g for 10 min at RT.

32. Transfer the supernatants to a new 1.5 ml eppendorf tube.

33. Resuspend the pellets in 500 μl 1×DPBS, sonicate using tip sonicator for 10 seconds and centrifuge at 17,000g for 10 min at RT.

34. Combine the supernatants from step 32 and 33.

35. Prepare suitable volume of NeutrAvidin resin.

    • CRITICAL STEP: Too much resin can result in binding of nonspecific peptides and can affect the identification of modified peptides by the mass spectrometer. For HEK293 or HT22 cells, we typically use 100 μl of resin for each sample of 2mg or less, 150ul of resin for up to 10mg of starting material and 200ul of resin for up to 20mg of starting material (resin is provided as a 50% slurry, i.e., 100 μl of the slurry corresponds to approximately 50 μl resin or “bed” volume). There are many commercial biotin quantitation kits (see REAGENTS) available that can be used for measuring the amount of AHA-biotin peptides in the samples to determine the volume of resin required for enrichment.

36. Wash the resin from step 35 using 1 ml 1×DPBS in 1.5 ml eppendorf tubes and spin down at 500g for 2 min at RT. Discard the supernatant. Repeat this washing step twice.

37. Transfer the supernatants from step 34 to the washed resin and incubate the samples on a rotator for 2h at RT.

38. Remove the supernatant completely.

39. Wash the resins using 1 ml 1×DPBS on shaker for 5min at RT and spin down at 500g for 2 min at RT. Discard the supernatant and make sure not to disturb the resin. Repeat this washing step twice.

40. Wash the resin using 1ml 1×DPBS with 5% (v/v) acetonitrile as described in step 39. Repeat this step once.

    • CRITICAL STEP: the number of repetitions in each wash step (steps 40–42) can be increased to three for better results.

41. Wash the resins using 1ml 1×DPBS as described in step 39. Repeat this step once.

42. Wash the resin using water as described in step 39. Repeat this step once.

    • CRITICAL STEP: this step is for desalting the samples.

43. Remove the supernatant completely.

    • CRITICAL STEP: leave approximate 20 μl water in the resin tubes if the samples are stored before elution.

    • PAUSE POINT: peptide-bound resin can be stored at −80°C for several weeks.

    Elution (Timing: 1 h)

    CRITICAL: Freeze/thaw of samples after elution reduces the number of identified NSPs. We therefore recommend eluting the AHA-biotin peptides on the same day they are analyzed on the mass spectrometer.

44. Add 200 μl elution buffer to the resin from step 43 and incubate on shaker for 5 min at RT. Spin down at 500g for 2 min at RT, and transfer the supernatant to a new 1.5ml eppendorf tube.

45. Add 200 μl elution buffer to the resin and incubate on 70 °C shaker for 5 min at RT. Spin down at 500g for 2 min at RT, and combine supernatant with in the supernatant from step 44.

46. Repeat step 44 and 45 and combine all supernatants from a single sample.

47. Centrifuge the combined supernatant at 500g for 2 min at RT, leave about 20ul liquid and transfer the supernatant to a new 1.7 ml eppendorf tube.

48. Vacuum dry the supernatants using a CentriVap system (at least 2 h at 37°C).

    • CRITICAL STEP: be careful and do not disturb the resin when transferring the supernatant. Make sure there is no resin left in the elution before vacuum drying it, otherwise the biotinylated peptides will be bound to the resin again.

    Sample loading (Timing: 6 h)

49. Prepare the following two columns:

    Loading column: a 250-μm i.d capillary with a kasil frit containing 1.9–2.1 cm of 10 μm Jupiter C18-A material (Phenomenex) followed by 1.9–2.1 cm 5 μm Partisphere strong cation exchanger (SCX, Whatman). (C18-SCX-Frit).

    Analytical column: A 100 μm i.d capillary with a 5 μm pulled tip packed with 15 cm 4 μm Jupiter C18 material.

50. Wash the loading column using 200 μl buffer B.

51. Wash the loading column using 200 μl buffer A.

52. Dissolve the dried pellet from step 48 in 600 μl buffer A, sonicate the sample in a water-bath sonicator for 10min and pressure-load it into the washed loading column.

53. Wash the loading column using 500 μl buffer A.

54. Attach the loading column to the analytical column with a union and place the entire split-column (loading column–union–analytical column) in line with an Agilent 1100 quaternary HPLC or a Thermo Easy nLCII pump.

    Mass spectrometry analysis (Timing: 1 d)

55. Set the LC settings as described in the Equipment Setup.

56. Set MS settings as described in the Equipment Setup and run the samples.

    Data analysis (Timing: 8–10 h)

    !CRITICAL Currently we use ProLuCID for identification and pQuant for quantification. ProLuCID are part of Intergrated Proteomics Pipeline (IP2). One replicate of a null experiment using HILAQ in anticipated results part including all raw data is provided at an external platform (ftp://massive.ucsd.edu/MSV000081962) for downloading and testing.

57. Identification using ProLuCID: Convert Xcalibur data (.RAW) into MS1 and MS2 (tandem mass spectra) formats using in house software (RAW_Xtractor)³⁵. Search the tandem mass spectra against a protein sequence database using ProLuCID. The following modifications are searched for HILAQ analysis: a static modification of 57.02146 on cysteine for all analyses, a differential modification of 452.2376 on methionine for AHA-biotin-alkyne or 458.2452 for HAHA-biotin-alkyne. Assemble and filtered the ProluCID results using the DTASelect (version 2.0) program^{36, 37}. In DTASelect, the modified peptides are required to be less than 5ppm deviation from peptide match, and a FDR at the spectra level of 0.01. Only modified peptides are retained for further analysis. The step by step instructions with screenshots using ProLuCID for AHA analysis are described in Supplementary Manual 1.

    CRITICAL STEP: For all datasets in anticipated results, protein FDR was <1% and peptide FDR<0.5%. However, the level of FDR in search settings can be varied depending on the aims of study. Researchers can try more or less stringent parameters only if both protein and peptide FDR of datasets are <1% in outputs. Both ProLuCID and DTASelect are modules of IP2.

58. Quantification using pQuant: Filtered DTASelect files with modified peptides only and MS1 files converted by RAW_Xtractor are used as the inputs for pQuant. pQuant then assigns a confidence score to each heavy/light ratio from zero to one. Zero, the highest confidence, means there is no interference signal, while a confidence score of one means the peptide signals are corrupted by interference signals (i.e. very noisy). Generally, protein ratios are calculated fromthe median of peptide ratios with sigma less than or equal 0.5 . Both peptide and protein ratios can be normalized as described before ³⁸ . pQuant is executed based on java script; all required parameter files are provided in Supplementary Data 1–4. Detailed instructions are shown in Supplementary Manual 2.

TIMING

Step 1–5, cell labeling: 2–4 h

Step 6–18, click chemistry reaction: 2.5–3 h

Step 19–23, TCA precipitation: >8 h (overnight)

Step 24–30, trypsin digestion: 4.5 h

Step 31–43, Neutravidin enrichment: 3.5 h

Step 44–48, elution: 1 h

Step 49–54, sample loading: 6 h

Step 55–56, mass spectrometry analysis: 1 d

Step 57–58, data analysis: 8–10 h

TROUBLESHOOTING

Trouble shooting advice can be found in Table 1.

Table 1.

Troubleshooting table

Step	Problem	Possible reason	Solution
12	Pellet cannot be dissolved in 0.5% SDS completely	Too much proteins	Increase the lysis buffer volume, as well as increase the volume of 0.5% SDS. Make more aliquots to ensure the final concentration of SDS in each reaction is no more than 0.05%.
22	Difficulty removing supernatant completely	Waiting too long after centrifugation before removing supernatant	Do not centrifuge too many samples at the same time and remove the supernatant as soon as centrifugation is finished.
25	Pellet is difficult to dissolve	Pellet was dried at RT too long	Stop drying once found the pellet becomes solid white.
30	Incomplete digestion	Big chunks of proteins in the digestion	Pipette or sonicate to diminish all chunks before step 29.
52	Slow or stuck sample loading	Bubbles in loading column	Use heat gun for 2–3 seconds to drive out the bubbles.
57	More than 50% unmodified peptides in the total identified peptides	Low labeling efficiency	Extend the labeling time.
		Too much resin used in step 35	Using commercial biotin quantitative kits to measure the amount of AHA-biotin peptides in samples. Then use the appropriate amount of resin.
		Wash steps (39–42) are not stringent enough	Increase washing time on shaker.

ANTICIPATED RESULTS

The workflow described in this Protocol can be used to measure how proteomes change in response to perturbations or disease. Multiple strategies utilizing AHA in combination with isobaric (in vitro) or metabolic (in vivo) labeling have been developed to achieve NSPs quantitation. The HILAQ workflow described here provides an alternative method which achieves higher sensitivity by reducing the labeling complexity and enables NSPs enrichment, confirmation and quantification using a single stable isotope labeled hAHA molecule.

Experimental variations and accuracy of HILAQ

To validate the performance of HILAQ, we applied it to a null experiment and compared it with QuaNCAT. For HILAQ, HEK293T cells were pulse labeled with either AHA or hAHA without undergoing any additional treatment. For QuaNCAT, HEK293T cells were pulse labeled with AHA in combination with either two “heavy” or two “medium” SILAC amino acids. After 1h pulse labeling, three HILAQ or QuaNCAT biological replicates were prepared and analyzed. Boxplot analysis show similar variation among replicates in the measured protein ratios for HILAQ and QuaNCAT (Supplementary Figure 2), suggesting that the HILAQ and QuaNCAT strategies have comparable accuracy. The NSPs detection sensitivity (the number of NSP over the whole proteome) in QuaNCAT could be increased from 6.19% to 23.3% through the use of peptide level enrichment instead of protein level enrichment used. NSPs detection sensitivity for HILAQ was 35.2% (Supplementary Figure 3 and Supplementary Table S1), indicating that the improved performance of the HILAQ over the QuaNCAT approach is due to multiple factors, not just peptide level enrichment. Newer mass spectrometers may result in more identifications and quantitation with higher sensitivity, dynamic range and faster speed. We anticipate that this strategy and workflow will quickly become the standard for monitoring proteome dynamics and will pave the way for discovering the root cause of disease.

It is also noteworthy that unmodified peptides typically comprised 93.5% of the identified proteins in QuaNCAT, whereas this number decreased to 14.6% in HILAQ (Supplementary Table S2) due to the different enrichment strategies. A large number of unmodified peptides can negatively affect the detection of modified peptides by mass spectrometry. Even though HILAQ has multiple advantages due to its peptide enrichment strategy, researchers should still pay attention to the number of identified unmodified peptides when performing HILAQ. If data shows unmodified peptides comprise more than 50% of identified peptides using HILAQ, improving the sample preparation to reduce the number of unmodified peptides will allow the identification of more NSPs. Failure to keep the composition of unmodified peptides in HILAQ low may be the result of (1) low AHA labeling efficiency (2) too much neutravidin resin used in enrichment steps (3) weak washes after enrichment. The rate of protein synthesis and HILAQ labeling efficiency may vary among cell lines, primary cells and animal models (unpublished observations). The labeling strategy should be optimized according to the specific nature of the experiment and the models used. Nonspecific binding is a common problem for neutravidin beads. Use of an appropriate amount of resin helps to reduce the binding of unmodified peptides. Stringent washes after enrichment helps to reduce nonspecific bindings and clean up samples.

Application of HILAQ in a biological study

Oxytosis is a recently discovered form of programmed cell death ³⁹ and plays an important role in neurodegenerative diseases. Neuronal cell line HT22 cells has previously been demonstrated to be an in vitro oxytosis model, due to its high sensitivity to glutamate relative to other cell types⁴⁰. We used our approach to assess whether the proteome profile of HT22 cells could explain their increased vulnerability to oxytosis. In this experiment, HT22 cells were pulsed with hAHA for 1 hour, whereas control HEK293T cells were pulsed with AHA. In the second experiment, labeling of the cells was reversed. In total, 818 NSPs were quantified in the swap experiments and the ratios of proteins showed a Gaussian distribution pattern (Figure 3A). A comparison of the swap experiments between two biological replicates showed sufficient reproducibility with a Pearson correlation coefficient of 0.79 (Figure 3B). 226 proteins were observed to have at least a twofold change, with 136 up-regulated and 90 down-regulated in the HT22 cells (Supplementary Table S3). It is noteworthy that the most significantly (p value = 1.16 e −16) enriched pathway was cell death, with 108 proteins annotated (Figure 4 and Supplementary Table S4), and many of these altered proteins have previously been connected to cell death in brain tissue and disease models such as BIRC6, TRAP1, PINK1, NDUFV2 etc. These results help to illustrate the vulnerability features of HT22 to cell death, which is related to oxytosis, and verify the reliability of HILAQ in biological study. This HILAQ analysis also demonstrates the validity of HT22 cells as a tool to study the molecular details of cell death involved in neurodegenerative diseases.

Figure 3. HILAQ performed in HT22 oxytosis model.

A: protein ratios from one of the biological replicates were plotted and indicated a typical normal distribution. B: The peptides ratios from two independent HILAQ experiments were compared. The AHA and hAHA labels were swapped in the two experiments. A was made from published dataset of ref. 14. B is adapted from ref.14.

Fig. 4:

Analysis of 108 cell death pathway–enriched proteins with twofold change after 1 h of labeling in HT22 cells as compared with HEK293T cells.

Supplementary Material

Supplementary Figure 1.

A typical MS1 spectrum of HILAQ samples. Top: one MS1 spectrum with m/z range from 400 to 2000. Bottom: enlarged m/z range (560–660). The blue framed area indicated one pair of hAHA and AHA peptides (mass difference of 6 da).

Supplementary Figure 2.

Boxplot analysis on every replicate of HILAQ and QuaNCAT using Graphpad Prism (version 5.01). The boxplot was created with Turkey whiskers. This table is adapted from ref.14.

Supplementary Figure 3.

Comparison of NSPs identified by HILAQ, QuaNCAT or QuanNCAT-pep. Three independent biological replicates were performed using each strategy. Error-bar represents “Mean with SD”.

Supplementary Table 1.

NSP sensitivity of HILAQ, QuaNCAT and QuanNCAT-pep. HILAQ is the strategy with higher sensitivity due to multiple factors. Only the proteins quantified in all three replicates were considered as confident NSP and counted for the calculation of NSP sensitivity. “QuaNCAT-pep” is a modified QuaNCAT protocol with peptide level enrichment. This table is adapted from ref.14.

Supplementary Table 2.

Identification results for HILAQ and QuaNCAT. Enrichment efficiency (EE) is calculated from this formula: EE=100%*Modified peptide number/(modified peptide number + unmodified peptide number). This table is adapted from ref.14

Supplementary Table 3.

List of proteins with two-fold change.

Supplementary Table 4.

List of proteins enriched in cell death pathway.

Supplementary Manual 1.

Step-by-step instructions for identification using ProLuCID.

Supplementary Manual 2.

Step-by-step instructions for quantitation using pQuant.

Supplementary Data 1.

pQuant parameter file 1: elementE

Supplementary Data 2.

pQuant parameter file 2: aa… .

Supplementary Data 3.

pQuant parameter file 3: modification….

Supplementary Data 4.

Completed pQuant.cfg file

Figure 4.	The pellet are not dissolved completely	Pipette or sonicate the samples to remove all chunks.
52	The loading column gets clogged or sample loading occurs very slowly	Bubbles are present in the loading column	Use a heat gun for 2–3 seconds to drive out the bubbles.
57	More than 50% of the total identified peptides were found to be unmodified	Low labeling efficiency	Extend the labeling time.
		Too much resin was used in step 35	Use a commercial biotin quantification kit (such as …) to measure the amount of AHA-biotin peptide in samples. Then use the appropriate amount of resin.
		Wash steps (39–42) are not stringent enough	Increase the number of washes using 1×DPBS on shaker.