Measurement of Homocitrulline, A Carbamylation‐derived Product, in Serum and Tissues by LC‐MS/MS

Carbamylation corresponds to the nonenzymatic binding of isocyanic acid to protein amino groups and participates in protein molecular aging, characterized by the alteration of their structural and functional properties. Carbamylated proteins exert deleterious effects in vivo and are involved in the progression of various diseases, including atherosclerosis and chronic kidney disease. Therefore, there is a growing interest in evaluating the carbamylation rate of blood or tissue proteins, since carbamylation‐derived products (CDPs) constitute valuable biomarkers in these contexts. Homocitrulline, formed by isocyanic acid covalently attaching to the ε‐NH2 group of lysine residue side chain, is the most characteristic CDP. Sensitive and specific quantification of homocitrulline requires mass spectrometry–based methods. This article describes a liquid chromatography–tandem mass spectrometry (LC‐MS/MS) method for the quantification of homocitrulline, with special emphasis on preanalytical steps that allow quantification of total or protein‐bound homocitrulline in serum or tissue samples. © 2023 The Authors. Current Protocols published by Wiley Periodicals LLC.


INTRODUCTION
Protein molecular aging, which is characterized by progressive and irreversible alterations of protein structure and function, is mainly explained by the accumulation of nonenzymatic post-translational modifications the proteins are exposed to during their biological life. These modifications correspond to the binding of small metabolites to free reactive groups of amino acid residues, usually followed by subsequent molecular rearrangements (Jaisson & Gillery, 2010). In the case of carbamylation, isocyanic acid reacts in an irreversible manner with αand ε-amino groups of proteins, generating α-carbamylated proteins and homocitrulline (HCit, ε-carbamoyllysine) residues, respectively (Kraus & Kraus, 2001). In vivo, isocyanic acid is formed by spontaneous dissociation of urea into cyanate and ammonia (Kraus & Kraus, 2001), by myeloperoxidasemediated transformation of thiocyanate (Wang et al., 2007), and to a lesser extent by exogenous sources from biomass burning or tobacco smoke (Roberts et al., 2011).
It is now well established that carbamylated proteins constitute a molecular substratum for many dysfunctions described in chronic diseases, such as chronic kidney disease, atherosclerosis, or rheumatoid arthritis (Kalim et al., 2014;Mastrangelo et al., 2015;Verbrugge et al., 2015). Accordingly, carbamylation-derived products (CDPs) are considered relevant biomarkers of these diseases. For example, elevated serum concentrations of HCit are associated with mortality and adverse outcomes in end-stage kidney disease (Koeth et al., 2013) and in various cardiovascular diseases Wang et al., 2007).
The quantification of HCit has long been a challenge because most conventional methods were not sensitive enough to detect the limited extent of protein carbamylation. For example, colorimetric assays based on the reactivity of carbamoyl groups with diacetylmonoxime have been described (Balion et al., 1998) but suffered from analytical interference when performed in complex matrices such as plasma. Similar limitations were encountered with immunoassays (Apostolov et al., 2005) because of a lack of standardization and especially the use of home-made antibodies, which are less specific and poorly characterized. Mass spectrometry-based methods are now considered the gold standard for the quantification of protein molecular aging markers, such as CDPs or advanced glycation end-products, because they offer excellent specificity and sensitivity.
Herein, we describe a liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) method to quantify HCit in serum and tissue samples (Jaisson et al., 2012). This method is divided into three parts. Basic Protocol 1 and Alternate Protocol describe preanalytical steps to quantify HCit in its different forms (total or protein bound) and in different matrices (serum, plasma, or tissue extracts), including details about the acid hydrolysis step, which is compulsory for the cleavage of peptide bonds and the release of HCit residues. Basic Protocol 2 describes the chromatographic separation based on a hydrophilic interaction liquid chromatography (HILIC) and mass detection using a triple quadrupole (e.g., TSQ Quantis, Thermo Scientific). Basic Protocol 3 describes the quantification of lysine (Lys) residues in hydrolysates, since HCit results are generally expressed as a ratio to Lys concentrations (Koeth et al., 2013;Wang et al., 2007). NOTE: All protocols using live animals or animal samples must first be reviewed and approved by an Institutional Animal Care and Use Committee (IACUC) or conform to local guidelines regarding the care and use of laboratory animals.
NOTE: All protocols involving humans or human samples must first be reviewed and approved by an Institutional Review Board (IRB) or must follow local guidelines for the use of human samples. All participants must provide informed consent.
GeBAflex mini dialysis unit, 8-kDa cutoff, 10 to 250 μl (e.g., Dominique Dutscher) 12-L bucket 0.2-and 1.5-ml microcentrifuge tubes Vortex mixer Microcentrifuge 2-ml glass ampoules for cell freezing in liquid nitrogen (e.g., Dominique Dutscher) Bunsen burner 110°C oven Evaporator for microcentrifuge tubes connected to a nitrogen stream (e.g., Dominique Dutscher) Chemical hood Orbital shaker for microcentrifuge tubes 0.45-μm polytetrafluoroethylene (PTFE) filters, 4-mm diameter (e.g., Phenomenex) Vials for liquid chromatography NOTE: Total HCit in serum or in plasma samples obtained from EDTA-or lithium heparinate-containing tubes can be quantified. Serum and plasma samples should be stored at -80°C. Quantification of protein-bound HCit requires preanalytical dialysis or protein precipitation to remove free and peptide-bound HCit. Though it takes longer, we prefer dialysis because precipitated proteins can be difficult to resuspend resulting in lower yields.

Dialyze serum/plasma samples to remove free and peptide-bound HCit
Dialysis removes free HCit and peptide-bound HCit, yielding samples containing only macromolecules (including proteins), so that only protein-bound HCit will be quantified.
1a. Transfer 150 μl serum or plasma into a GeBAflex mini dialysis unit with 8-kDa cutoff.
2a. Using the foam support provided with the kit, place GeBAflex dialysis unit into a bucket containing 10 L of 0.15 M NaCl. 4a. Remove dialyzed serum from the dialysis unit, and transfer to a 1.5-ml microcentrifuge tube. Tilt the ampoule to 45°above the Bunsen burner (focusing on the cone formed by the flame), and wait several seconds until the glass turns orange in color. Then gently pull with pliers.
2b. Shake tubes for 30 s using a vortex mixer. Place tubes on ice for 10 min, and then pellet proteins by centrifugation for 10 min at 10,000 × g, 4°C.
3b. Gently discard supernatant, and resuspend pellet in 1.2 ml of 6 M HCl by shaking for 10 s in a vortex mixer.
4b. Transfer solution to a 2-ml glass ampoule, and seal glass ampoule using a Bunsen burner (Fig. 1).
Follow the next steps described in the acid hydrolysis section, beginning at step 12. 10. Transfer 0.6 ml solutions into 2-ml glass ampoules, and add 0.6 ml of 12 M HCl to each ampoule.

Prepare calibrators
11. Seal glass ampoules with a Bunsen burner (Fig. 1 13. Cool ampoules on the bench for 30 min. Then break glass ampoules, and transfer 1 ml of each hydrolysate to new 1.5-ml microcentrifuge tubes.
14. Transfer tubes to an evaporator connected to a nitrogen gas inlet and located under a chemical hood.
15. Evaporate samples until dry (∼8 hr). Add 1 ml Milli-Q water to each sample, and evaporate overnight to ensure complete removal of HCl.
After evaporation, the tube walls will look brown.
The dried samples may be stored at -80°C before the next steps.
17. Shake tubes for 10 s on a vortex mixer. Transfer tubes to an orbital shaker, and agitate for 10 min at 1400 rpm.
Filtration should be performed in two steps (two filtrations of 50 μl). Transfer 50 μl into the filter, and then attach a 1-ml syringe to the filter. Push the plunger to force the liquid through the filter. Use a new filter for each sample, and wash the syringe with distilled water between each sample (taking care to dry the syringe with a paper towel to avoid any dilution of the sample).
The diluted hydrolysates may be stored at -80°C until ready for LC-MS/MS analysis.

PREANALYTICAL STEPS FOR THE QUANTIFICATION OF HOMOCITRULLINE IN TISSUE SAMPLES
Quantification of HCit may be performed in all types of tissues (fresh or frozen), including bones. In the case of frozen tissues, the samples should be cut into small pieces (100 to 150 mg each) and stored at -80°C until analysis. Using this protocol, only total HCit can be quantified.
Store the diluted homogenate at -80°C until ready for acid hydrolysis.

Prepare calibrators
8. Prepare calibrators as described in Basic Protocol 1, steps 5 to 8.

In this protocol, the IS d 3 -HCit is added after acid hydrolysis.
Store each calibrator and MB solution at -80°C until ready for acid hydrolysis.
Perform acid hydrolysis and evaporae hydrolysates 9. Transfer 0.6 ml diluted tissue homogenates into 2-ml glass ampoules, and add 0.6 ml of 12 M HCl to each ampoule.
After hydrolysis of the tissue samples, some black sediment may appear in the ampoule. These sediments correspond to tissue fragments that have not been hydrolyzed because there is too much material in the ampoule. If there is excessive sediment, the experiment must be repeated using a higher dilution of sample prior to hydrolysis in order to place less material in the ampoule.
13. Break glass ampoules, and transfer 1 ml of each hydrolysate to new 1.5-ml microcentrifuge tubes.
14. Transfer microcentrifuge tubes to an evaporator connected to a nitrogen gas inlet and located under a chemical hood.
15. Evaporate samples until dry (∼8 hr). Add 1 ml Milli-Q water to each sample, and evaporate overnight to ensure complete removal of HCl.
After evaporation, the tube walls will look brown.
Dried samples may be stored at -80°C until ready to perform the next steps.
17. Shake tubes for 10 s on a vortex mixer. Transfer tubes to an orbital shaker, and agitate for 10 min at 1400 rpm.

LC-MS/MS QUANTIFICATION OF HOMOCITRULLINE
This protocol describes LC-MS/MS analysis of protein hydrolysates to quantify HCit.
Ion source and mass detection parameters are detailed in Tables 2 and 3, respectively. The multiple reaction monitoring (MRM) transition 190.2 → 127.1 is used as the quantification transition, whereas 190.2 → 173.1 is used to confirm peak specificity. MRM transition 193.2 → 176.1 is used for IS (i.e., d 3 -HCit) quantification. Note that these mass spectrometry parameters have been optimized for this specific system and should be redefined if using a different mass spectrometer. For that purpose, directly infuse the standard solution of each analyte (HCit and IS) into the mass spectrometer source using a syringe pump.
Before each series, inject two blank samples (5 mM ammonium formate, pH 2.9) to condition the column. Insert a blank sample after the highest concentration calibrator and after samples with very high HCit concentration to avoid any cross-over contamination. An example of a chromatogram obtained by analyzing a serum sample is shown in Figure 2. HCit and its IS coelute at a retention time of 2.97 min.   Quantification of peak area is performed using Tracefinder 4.1 software. Ratios of HCit quantification peak area to IS peak area are used to derive the calibration curve and determine HCit concentrations in the samples. An example calibration curve is shown in Figure 3. When establishing the calibration curve, the HCit-to-IS ratio of the matrix blank sample must be subtracted from the ratios of all the calibrators before calculation.
To ensure specificity of quantification, the ratio of areas of HCit quantification peak (190.2 → 127.1) to HCit confirmation peak (190.2 → 173.1) must be between 0.6 and 0.8. Each sample should be injected twice, and mean peak areas should be used for calculations.

LC-MS/MS QUANTIFICATION OF LYSINE IN HYDROLYSATES
In order to evaluate the rate of protein carbamylation, it is important to normalize results toward the protein content in the serum sample or in the tissue extract. Thus, it is possible to perform usual protein quantification assays and to report HCit results expressed as mmol/g protein. However, most studies dealing with protein carbamylation express HCit results as a ratio to Lys content. Accordingly, here we describe an LC-MS/MS method for the quantification of Lys in hydrolysates.

Dilute hydrolysates
For serum and tissue samples, use the filtered hydrolysates obtained after evaporation (i.e., hydrolysates before dilution from Basic Protocol 1, step 19). The second dilution may be done directly in chromatography vials or in 96-well plates; both vials and plates may be used for injection.
The second dilution is performed with 5 mM ammonium formate with a pH at 2.9 to keep the pH close to the pH of mobile phase A.
The dilution factor for tissue will depend on the type of tissue and should be determined empirically.

Prepare calibrators for lysine quantification
The calibration curve for Lys quantification is calculated from calibrators prepared by spiking Lys standard into an MB sample prepared using diluted serum (or plasma) hydrolysate. However, it is not compulsory to submit calibrators to acid hydrolysis for establishing Lys calibration curve.

Quantify Lys by LC-MS/MS
The same LC-MS/MS system, LC column, and mobile phases as for HCit quantification (Basic Protocol 2) are used for the quantification of Lys. Details about the buffer gradient, flow rate, and injection volume are shown in Table 4.
Ion source and mass detection parameters are detailed in Tables 5 and 6, respectively. The MRM transition 147.1 → 84.1 is used as the quantification transition, whereas 147.1 → 130.1 is used to confirm peak specificity. MRM transition 155.2 → 92.2 is used for IS (i.e., d 8 -Lys) quantification. As with HCit quantification, these parameters have been optimized for this mass spectrometry system and should be redefined if another mass spectrometer is used. For that purpose, directly infuse the standard solution of each analyte (Lys and IS) into the mass spectrometer source using a syringe pump.
Before each series, inject two blank samples (5 mM ammonium formate, pH 2.9) to condition the column. Insert a blank sample after the calibrator with the highest concentration and after samples with very high Lys concentration to avoid any cross-over contamination.  Nitrogen is used as curtain and nebulization gas, and argon is used as collision gas CID, collision-induced dissociation; FWHM, full width at half maximum; IS, internal standard; Lys, lysine; MRM, multiple reaction monitoring; MS/MS, tandem mass spectrometry. Other parameters: Q1 resolution (FWHM), 0.7; Q3 resolution (FWHM), 1.2; CID gas (mTorr), 1.5; source fragmentation, 0; chromatographic peak width, 6.
Quantification of peak area is performed using Tracefinder 4.1 software. Ratios of Lys quantification peak area to IS peak area are used to calculate the calibration curve and to determine Lys concentrations in the samples. When establishing the calibration curve, the Lys-to-IS ratio of the matrix blank sample has to be subtracted from the ratios of all the calibrators before calculation.
To ensure specificity of quantification, the ratio of areas of Lys quantification peak (147.1 → 84.1) to Lys confirmation peak (147.1 → 130.1) must be between 2.1 and 2.3. Each sample should be injected twice, and mean peak areas should be used for calculations. For the calculation of Lys concentrations used for the expression of HCit results, be sure to use the correct dilution factor. For serum (or plasma) samples, the dilution factor is 1:1200 (v/v): 1:40 dilution through acid hydrolysis, 10-fold concentration after evaporation (from 1 ml to 100 μl), and 1:300 dilution before LC-M/MS. For tissue samples, it is easier to use HCit results corresponding to HCit concentrations in the undiluted hydrolysates. In this case, dilution factors for Lys concentrations range from 100 to 1000.

Background Information
HCit is a general marker of protein carbamylation formed by the binding of isocyanic acid to ε-NH2 group of Lys side chains. As carbamylated proteins have been clearly identified as contributing factors in the development of long-term complications related to chronic diseases (Verbrugge et al., 2015), CDPs (e.g., HCit) are emerging biomarkers. Various studies have pointed out the added value of HCit quantification in different pathological contexts, including chronic kidney disease or cardiovascular diseases (Jaisson et al., 2015;Koeth et al., 2013).
Accordingly, it is essential to use robust analytical methods for the quantification of carbamylation biomarkers to boost their use in clinical studies. Few methods have been described so far, and most rely on colorimetry, immunoassays, or mass spectrometry. Colorimetric techniques lack the sensitivity and specificity needed to quantify carbamyltion in complex matrices such as serum. Immunoassays may constitute alternate methods but suffer from a lack of standardization (i.e., a lack of commercially available HCit antibodies). Indeed, the only immunoassays described use home-made antibodies, which are poorly characterized. As a consequence, LC-MS/MS methods have emerged as the gold standard for the quantification of CDPs, especially owing to their high sensitivity and specificity.
The disadvantage of LC-MS methods over colorimetric or immunological assays is that they only enable the quantification of HCit in its free amino acid form. Since it is necessary to quantify protein-bound HCit (and not free HCit) for evaluating protein carbamylation rate, HCit residues must be released from proteins by cleaving peptide bonds. This requires important preanalytical steps, descriptions of which are often only summarized in publications. Thus, the goal of this protocol is to detail these different steps.
The first step of the experimental design is to determine which HCit form has to be assayed (total or protein bound). The quantification of protein-bound HCit requires the removal of free HCit, either by sample dialysis or by protein precipitation. Both are described in this article. Though more time consuming, we prefer dialysis over protein precipitation. Indeed, the protein pellets obtained by protein precipitation are often difficult to suspend in 6 M HCl, and undissolved portions of the pellet stick to the walls of the tube and are often lost during transfer to the glass ampoules for hydrolysis.
The crucial preanalytical step is acid hydrolysis, which allows the cleavage of peptide bonds. The experimental protocol must be strictly controlled to ensure that all conditions (temperature, incubation time, HCl concentration) are identical for all samples. Indeed, HCit is partially degraded during this step (∼30%), but this degradation rate is constant under given experimental conditions (Jaisson et al., 2012). This problem can therefore be bypassed by a strict adhesion to the protocol. It is also compulsory to subject the calibrators to acid hydrolysis under the same conditions as the assayed samples. An alternative to acid hydrolysis is enzymatic hydrolysis. However, this step is more time consuming, and its main disadvantage is the difficulty of ensuring exhaustive peptide bond cleavage, especially because modified amino acids may limit the action of proteolytic enzymes.
As a conclusion, the LC-MS/MS method described herein includes many timeconsuming preanalytical steps that are crucial for robust and reliable HCit quantification. Therefore, particular attention must be paid to these steps.

Critical Parameters and Troubleshooting
Assay calibration (1): As HCit is partially degraded during acid hydrolysis (about 30%), calibrators must be subjected to acid hydrolysis under the same conditions as the samples to avoid underestimation of values. The HCit degradation rate is constant in given experimental conditions, as demonstrated by linearity of the calibration curve (Fig. 3). Moreover, excellent recovery rates are obtained when hydrolyzed calibrators are used for the quantification of HCit (Jaisson et al., 2012).
Assay calibration (2): Calibrators should be prepared by spiking HCit in a diluted serum (or plasma) sample because the presence of matrix is very important for LC separation of HCit. HCit in buffer does not behave like samples (different retention time and peak shape). However, the initial HCit content in the matrix blank must be determined and subtracted from obtained values when calculating the calibration curve.
Acid hydrolysis: Take care to strictly respect the incubation time (18 hr) at 110°C. Due to HCit degradation during acid hydrolysis, any variation will affect reliability of the results.
Evaporation: It is very important to perform evaporation under a nitrogen stream (to avoid oxidation) until dryness and to perform two successive evaporations (i.e., evaporate HCl, add 1 ml water, and evaporate again) to remove all residual acid. Water must be added only when the samples are totally dry. If evaporation is incomplete or if the acid is not completely removed, the pH of the suspended hydrolysates will be too low and will affect chromatographic separation of HCit (i.e., retention time and peak shape).
Filtration: Hydrolysates must be filtered before injection of the sample to avoid clogging the LC system. Use a new filter for each sample, and be careful to correctly wash the syringe and to dry it between each sample to avoid diluting the next hydrolysate.
Homogenization of tissue samples: The homogenization time and the type of Lysing Matrix tubes to use may vary depending on the tissue. Preliminary tests should be used to determine the best conditions that will result in homogeneous tissue extracts.
LC-MS/MS analysis: In case of abnormal HCit peaks (asymmetric peak, modification of retention time), check the pH of mobile phase A and the samples (which should range from 4 to 5). If sensitivity is lost, check the ionization source, and clean it if necessary.
Calculations: Be very careful to account for the dilution of the hydrolysates and to use the correct concentrations of calibrators (which are different for serum/plasma and tissue samples).

Understanding Results
Serum HCit concentrations typically range from 100 to 150 μmol/mol Lys in control participants and may reach 1500 μmol/mol Lys in patients with chronic kidney disease (Desmons et al., 2016).
Tissue HCit concentrations may vary depending on the type of tissue. For instance, in mice HCit concentrations range from 200 to 500 μmol/mol Lys in bone, liver, kidney, aorta, and skin. In human skin extracts, HCit concentrations may reach 10,000 μmol/mol Lys in older adults (Gorisse et al., 2016;Pietrement et al., 2013).

Time Considerations
Dialysis of serum samples takes 1 day. Precipitation of protein (serum/plasma samples) takes 2 hr depending on the number of samples. Tissue homogenization requires 2 days, and acid hydrolysis (including preparation of tubes, sealing of glass ampoules, and incubation) takes 1 day. The evaporation steps require 1 day, and filtration of hydrolysates takes 3 hr depending on the number of samples. LC-MS/MS analysis takes 2 days, which includes 1 day for HCit quantification and 1 day for Lys quantification. Note that each step may be performed separately, except evaporation, which must be done immediately after acid hydrolysis.