Quadrupole-Time-of-Flight Mass Spectrometric Identification of Hemoglobin Subunits α, β, γ and δ in Unknown Peaks of High Performance Liquid Chromatography of Hemoglobin in β-Thalassemias.

This is the first report of quadrupole time-of-flight (Q-TOF) mass spectrometric identification of the hemoglobin (Hb) subunits, α, β, δ and γ peptides, derived from enzymatic-digestion of proteins in the early unknown peaks of the cation exchange chromatography of Hb. The objectives were to identify the unknown high performance liquid chromatography (HPLC) peaks in healthy subjects and in patients with β-thalassemia (β-thal). The results demonstrate the existence of pools of free globin chains in red blood cells (RBCs). The α-, β-, δ- and γ-globin peptides were identified in the unknown HPLC peaks. The quantification and role of the free globin pool in patients with β-thal requires further investigation. Identification of all types of Hb subunits in the retention time (RT) before 1 min. suggests that altered Hbs is the nature of these fast-eluting peaks. Relevancy of thalassemias to the protein-aggregation disorders will require review of the role of free globin in the pathology of the disease.

Keywords: High performance liquid chromatography (HPLC); mass spectrometry (MS); peaks; quadrupole time-of-flight (Q-TOF); thalassemia

The free α-globin or hemoglobin (Hb) α subunit was previously called unbound, unpaired or unmatched globin [[1]]. The free α-globin was established in erythrocytes in patients with β-thalassemia (β-thal) [[4]]. The free α-globin pool measurement in the blood may help in discriminating different types of β-thal, scoring of disease severity and providing screening programs for hemoglobinopathies in diverse populations. Moreover, it may be applied to all diseases in which globin synthesis imbalance occurs and allows monitoring of response to various emerging therapeutics [[5]]. Mass spectrometry (MS) is the most powerful technique for biomarker identification, characterization and elucidation of their biological role [[7]]. Different MS approaches have been described for the identification of peptides and proteins, such as peptide de novo sequencing and sequence tag-based searching [[8]]. Enzymatic digestion of intact proteins into smaller peptides using proteolytic agents such as trypsin before introduction into an MS is the preferable approach [[9]]. Alternatively, intact proteins are ionized by electrospray ionization (ESI) or matrix-assisted laser desorption ionization (MALDI) prior to introduction into MS. Characterization of these peptide mixtures using liquid chromatography (LC)/MS, is called shotgun proteomics. The masses of the proteolytic peptides are used as input to search a database for predicted masses that result from digestion of known proteins. Evidence that a target protein is present in the original sample is confirmed when the protein sequences in the reference list match the experimental values [[10]]. Proteomic approaches could detect all the Hb subunits (α, β, γ and δ) in the plasma of patients with thalassemia [[11]].

Cation exchange high performance liquid chromatography (HPLC) is increasingly used for Hb analysis aiming for Hb variant determination. The analysis depends on the different charge states of the different Hb variants. Separation of the variants is based on the affinity of the cationic molecules of intact Hb to the polyaspartic acid-coated stationary phase. After mixing of the blood sample in a sampling station, it is loaded on the separation column and the separated Hb variants are monitored at 415 nm. The background signal is reduced by monitoring with a secondary wavelength at 690 nm [[13]]. The current software for Hb variants analysis (VARIANT II™, β-Thalassemia Short Program; Bio-Rad Laboratories, Hercules, CA, USA) does not integrate any peaks that elute before 0.75 min. The Hb variant determination is based on their retention times (RT) and their amount is calculated based on the peak areas of Hb variants [[14]]. Peaks that elute at undefined RTs are labeled as unknown peaks. However, the total percentage of the Hb measured by HPLC is not always 100.0% [[6]].

To date, proposed reasons for the unknown peaks that occur at <0.75 min. are Hb Bart's (γ4), Hb H (β4), composed of tetramer of one globin chain type, Hb F1 (acetylated or altered Hb F in neonates), bilirubin and injection artifacts [[15]]. We proposed that the free Hb subunits may also be eluted at the very beginning in the cation exchange HPLC analysis of Hb. The objective was to identify the content of early-eluting unknown peaks that occur during the first minute of the cation exchange HPLC analysis of Hb in blood samples collected from healthy blood donors and patients with different types of β-thalassemias.

Approval of the ethics committees of the National Medical Research Register (NMRR) [NMRR-3-1471-15105], Medical Research and Ethics Committee (MREC) [(8) dlm.KKM/NIHSEC/P14-462] and Jawatankuasa Etika Penyelidikan (Manusia), Universiiti Sains Malaysia (USM) (JEPeM) [JEPeM Code: USM/JEPeM/15] and Universiti Sultan Zainal Abidin (UniSZA) [UniSZA/C/2/CRIM/431-2 (24)] were obtained at the initial stage of the study.

Venous blood samples (5 mL) were collected in EDTA vacutainers from selected healthy blood donors with normal red blood cell (RBC) parameters including: normal Hb level, mean corpuscular volume (MCV) of more than 87.0 fL, mean corpuscular Hb (MCH) of more than 27.0 pg and mean corpuscular Hb concentration (MCHC) of more than 34.0 g/dL, as well as from patients with different types of β-thal [β-thal trait, β-thal major (β-TM) and Hb E (HBB: c.79G > A)̸β-thal]. A total of 20 respondents (five respondents representing each group) were enrolled. All respondents were less than 20 years old (ranging from 5 to 20). The β-thal types were defined according to the standard clinical and hematological criteria. The enrolled transfusion-dependent patients with β-thal were treated at the Paediatric Department of Hospital Kuala Lumpur-Kuala Lumpur (HKL-KL), Malaysia. The venous blood sample from each respondent was processed for full blood cell count using the XE-5000™ Automated Hematology System (Sysmex Corp., Chuo-ku, Kobe, Japan) at HKL-KL, Malaysia, The samples were stored at 4 °C until HPLC analysis, generally less than 2 weeks after blood collection.

The HPLC analysis was done using the VARIANT II™, β-Thalassemia Short Program (Bio-Rad Laboratories) as recommended by the manufacturer's instructions. Each 6.5 min. assay cycle of this HPLC provides quantitative results for Hb A2 and Hb F percentages as well as detecting Hb variants [[13]]. The eluted fraction of each blood sample was manually collected in a sterile and capped plastic tube from time of injection (0 min.) to 1.0 min. for a total volume of 2.0 mL. This fraction represented the fast-eluting unknown HPLC peaks occurring during the first minute of the chromatography analysis of Hb.

Total protein concentration of each sample was measured using the NanoDrop™ 2000/2000 Spectrophotometer (Thermo Fisher Scientific, Wilmington DE, USA). Then 100 μg of protein in each eluted fraction was reconstituted and mixed with 100 μL of 6 M urea in 50 mM Tris-HCl, pH 8.0. Next, a 5 μL of 200 mM dithiotreitol (DTT) in 50 mM ammonium bicarbonate, pH 8.0, was added into the mixture and incubated at room temperature for 1 hour. A 20 μL of 200 mM iodoacetamide (IAA) in 50 mM ammonium bicarbonate, pH 8.0, was added into the mixture, followed by incubation at room temperature in the dark. The excess IAA was chelated by the addition of 20 μL of 50 mM DTT in 50 mM ammonium bicarbonate, pH 8.0, and incubated in the dark for another hour at room temperature. A 775 μL of 50 mM ammonium bicarbonate was added into the mixture to reduce the concentration of urea to 0.6 M prior to the addition of trypsin. A 2 μg MS grade trypsin (Thermo Fisher Scientific) was added into the mixture giving a final ratio of 1:50 (w/w) trypsin:protein. The mixture was vortexed gently and incubated at 37 °C for at least 18 hours. Finally, 2 μL of neat formic acid was added to the digested protein mixture to stop the trypsin activity. Following the proteolytic digestion, the samples were placed in a receiver tube and stored at –20 °C until use. Prior to injection into the LC with tandem MS (MS/MS) system, all tryptic digests were desalted using Pierce C18 Spin Column (Cat. #89873; Thermo Fisher Scientific) following equilibration in 50.0% acetonitrile (ACN) in 0.1% formic acid. Samples were then desalted using Pierce C18 Spin Column (Thermo Fisher Scientific) following the manufacturer's instructions prior to LC/MS/MS analysis.

The samples were reconstituted with 50 μL of 0.1% formic acid in double distilled water and loaded on an Agilent Large Capacity Chip, 300 Å, C18, 160 nL (Cat. #G4240-62010; Agilent Technologies Deutschland GmbH, Waldbronn, Germany) using Agilent 1200 HPLC-Chip/MS Interface, coupled with Agilent 6520 Accurate-Mass quadrupole time-of-flight (Q-TOF) LC/MS column. Flow-rate was 4 μL/min. from Agilent 1200 Series Capillary pump and 0.3 μL/min. from Agilent 1200 Series Nano Pump. A 0.1% formic acid solution (solvent A) and 90.0% ACN with 0.1% formic acid (solvent B) were used as the mobile phases. The injection volume of the sample was 1-2 μL. A 60 min. gradient method was used for the LC separation. Sample loading onto the enrichment column was done at 5.0% B. The gradient used for the analytical column began at 5.0% B, increasing to 95.0% B in 60 min. and then returning to 5.0% B at 65 min. The column was equilibrated for 5 min. before subsequent injection.

Peptide identification using automated protein de novo sequencing software PEAKS studio: Identification of proteins was performed using PEAKS Studio (version 7.5) (http://www.bioinfor.com/peaks-studio/), and UniProtKB protein database (https://www.uniprot.org/). Data were searched against a publicly available human database containing 741,510 protein entries (https://www.ncbi.nlm.nih.gov/protein). The user-defined search parameters included: enzyme, trypsin; allowance of up to three missed cleavages, variable modification, methionine oxidation, fixed modification, carbamidomethyl cysteine, peptide mass tolerance, ±20 ppm, and fragment mass tolerance of 0.1 Da. The false discovery rate (FDR) was set at 1.0%. Proteins identified by two or more unique peptides were of high confidence. For each globin chain type, only the most consistently-detected peptide was selected as unique peptides for globin chain identification. The characteristics of the selected unique peptides used for the identification of each of the globin chains or Hb subunits α, β, δ and γ, are listed in Table 1.

Table 1. Properties of the unique peptides of the free globin chains or hemoglobin subunits identified by quadrupole time-of-flight mass spectrometry.

1 Criteria proposed by the World Anti-Doping Agency (WADA) (https://www.wada.ama.org/) indicate that at least two product ion transitions of the peptide are required to ensure unambiguous identification of the compound of interest.

In the cation exchange HPLC analysis of Hb, the eluted fraction from the separation column prior to the start of integration, which appears as fast-eluting peaks, is unknown. We observed the association of high-amplitude of the unknown early-eluting peaks in patients with different clinical phenotypes of β-thal. Also, these early-eluting peaks were absent or of low-amplitude during Hb analysis of blood from healthy individuals or even patients with qualitative defect of Hb such as Hb S (HBB: c.20A>T). The peak heights correlate with the levels of Hb F or Hb A2, particularly in patients with Hb E/β-thal and β-TM (Figure 1). Properties of the unique peptides used for identification of free globin chains and the MS/MS spectra of each peptide analysis are shown in Table 1 and Figure 2, respectively.

Graph: Figure 1. Examples of the fast-eluting unknown peaks during the first minute in the cation exchange HPLC analysis of blood from patients with different types of β-thal and from healthy blood donors. In the cation exchange HPLC analysis of Hb, the eluted fraction from the separation column prior to the start of integration, which appears as fast-eluting peaks, is unknown. The observed association of the higher-amplitude of the unknown early-eluting peaks in patients with different clinical phenotypes of β-thal compared to the low-amplitude or absence of these peaks in healthy individuals or other Hb variants such as Hb S. The peaks amplitudes correlate with the levels of Hb F or Hb A2, particularly in patients with Hb E/β-thal and β-TM.

PHOTO (COLOR): Figure 2. Mass spectrum of each peptide. Unique peptides representative of each globin chains were analyzed by the nano-HPLC-MS/MS Q-TOF system. All mass spectrums of peptides from the analysis were searched against the human proteome library downloaded from the National Center for Biotechnology Information (NCBI) human proteome database.

Complicated methods such as the use of radio-actively labeled globin are required to measure globin chain synthesis ratio to discriminate β-thal and to determine the prognosis [[17]]. Cation exchange HPLC (Bio-Rad Laboratories) is increasingly used to separate and measure various normal and abnormal Hb variants [[6]]. We observed the occurrence of visually-detected unknown peaks within the first minute in the HPLC analysis of blood from patients with different phenotypes of β-thal. The height of the peaks was lower during HPLC analysis of Hb from healthy subjects or even qualitatively defective Hb variants [e.g. Hb C (HBB: c.19G>A) and Hb S]. The heights of these peaks in the HPLC chromatogram are correlated with the levels of Hb F or Hb A2 (Figure 1) in healthy individuals and patients with different phenotypes of β-thal. The RT of these unknown peaks is close to where tetramers of β chains (Hb H) and tetramers of γ chains (Hb Bart's) are eluted and visually-detected in the cation exchange HPLC chromatogram. Hb Bart's and Hb H elute from the separation column during the first minute, prior to the start of integration, and appear as a fast-eluting fraction in cation exchange HPLC [[5], [13]]. The VARIANT II™ HPLC system (Bio-Rad Laboratories) does not integrate any peaks that elute before 0.75 min. or report them in the printout [[6], [15]].

This is the first report of Q-TOF mass spectrometric detection and identification of globin peptides (globin digests) identified in the fast-eluting unknown HPLC peaks. Acylated or altered Hb F is the only Hb type that was reported to be present in the fast-eluted fraction within the first minute during Hb analysis by cation exchange HPLC for patients with β-thal [[15]]. The altered Hb F might refer to its alteration from tetramers to homopolymers composed of free γ-globin [[18]]. Identification of all types of Hb subunits in the RT before 1 min. suggests that altered Hb A, Hb F and Hb A2 is the nature of these fast-eluting peaks. It may also indicate that the nature of these globin chains are aggregates of homopolymers (macroglobin molecules) consisting of identical free globin of α, β, δ or γ chains. However, the Hb E variant is known to cause weak union between α- and β-globin [[19]]. This weak union might be the reason for the elution of the free globin chains within the first minute causing high-amplitude of the unknown HPLC peaks, particularly in patients with Hb E/β-thal and those with β-TM. Relevancy of thalassemias to the protein-aggregation disorders will require a review of the role of the free globin pool in the pathology of this disease.

This study revealed the existence of free globin chains in the fast-eluting unknown HPLC peaks in normal human RBCs and in patients with different β-thal phenotypes. Further investigation and quantification of the free globin chains pool recommended. Identification of all types of Hb subunits in the RT time before 1 min. suggests that altered Hbs is the nature of these fast-eluting peaks. Early prognosis using pertinent biomarkers with diagnostic and prognostic utility allows more effective and early therapeutic intervention with more successful therapeutic outcome. The limitation of the study was in the recruitment of newly diagnosed untreated patients with β-TM and those with Hb E/β-thal.

The authors' sincere gratitude and thanks to Dr. Raudhawati Osman (Head of Unit) and Mrs. Norafzam Binti Muhamad at the Haematology Unit of the Pathology Department, HKL-KL, Kuala Lumpur, Malaysia, for their contribution to the data collection and sample processing and storage. We are grateful to all the respondents who participated in this study.

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.