Patients in this study were from our prospective registration study, in which patients with severe calcified coronary lesions on angiographic images who underwent percutaneous coronary intervention (PCI) with orbital atherectomy were recruited between February 2020 and May 2021 (registered on the Chinese Clinical Trial Registry [ChiCTR] 2000029775). Severe calcified coronary lesions were defined as the presence of apparent radiative opacities within the coronary lumen [10]. Based on qualitative visual assessment of calcified lesions using fluoroscopy, they were classified as severe (radio-opacities located on both sides of the coronary arterial wall before contrast injection and without cardiac movement), moderate (densities were acquired during the cardiac cycle before contrast injection), or none/mild calcified lesions according to the published literature [11]. An example of severe calcified coronary lesions on angiographic images is given in Fig. 1. Two cardiologists who were unaware of the patients included in this study assessed the angiographic calcification. Some cases of disagreement were adjudicated by a third independent cardiologist and given the final diagnosis. The recruited control subjects were matched by age, sex, and conventional risk factors. The control subjects were selected during the same period in the same hospital with non-calcific coronary angiography findings. Subjects with a known history of malignant tumors, serious system diseases, severe respiratory disease, renal or hepatic insufficiency, and acute infections were not included. The clinical and demographic information of the participants was recorded. The study protocol was approved by our institutional ethics committee, and written informed consent was obtained from all patients. The study was designed in accordance with the ethical guidelines of the Declaration of Helsinki.

Fig. 1.

Representative image of severe calcified coronary lesions detected by coronary angiography. (a) Controls without coronary artery calcification (non-CAC). (b) Patients with coronary artery calcification (CAC). Red marks indicate severe calcified left anterior descending artery (LAD) lesions detected by x-ray and the corresponding image from coronary computed tomography angiography (CTA).

Thirty individuals were randomly selected from each group and subjected to proteomics analysis in a discovery setting. Fasting venous blood samples were collected preoperatively from CAC and control patients, centrifuged at 2500 g for 10 min to obtain serum, collected in EDTA-coated tubes, and stored at –80 °C until subsequent analysis.

High-abundant proteins from 20 $\mu$L of each serum sample were removed using the Human 14 Multiple Affinity Removal Column (4.6 $\times$ 50 mm, Agilent Technologies, USA) according to the manufacturer’s protocol. The high- and low-abundance proteins were collected, and a 5 kDa ultrafiltration tube (Sartorius, Göttingen, Germany) was used for desalination and concentration of the high- and low-abundance components. SDT buffer (4% SDS, 100 mM DTT, 150 mM Tris-HCl pH 8.0) was added, boiled for 15 min, and centrifuged at 14,000 g for 20 min. The supernatant was quantified with the BCA Protein Assay Kit (Bio-Rad, Hercules, CA, USA). The sample was stored at –80 °C.

Both high- and low-abundant proteins were subjected to a digestion procedure modified from the filter-aided sample preparation (FASP) protocol as previously described [12]. Briefly, 200 $\mu$g protein was mixed with 30 $\mu$L SDT buffer (4% SDS, 100 mM DTT, 150 mM Tris-HCl pH 8.0). UA buffer (8M urea, 150 mM Tris HCl pH 8.0) accompanied by repeated ultrafiltration (Microcon-10 kDa Centrifugal Filter Unit; Millipore, Burlington, MA, USA) removed the detergent, DTT, and other low-molecular-weight components. Then 100 $\mu$L iodoacetamide (IAA) (100 mM IAA in UA buffer) was added to the samples and stored in the dark for 30 min at 4 °C. Subsequently, 100 $\mu$L UA buffer washed the filters three times and then twice with 100 $\mu$L 25 mM NH${}_4$HCO${}_3$. Next, 40 $\mu$L trypsin buffer (2 $\mu$g trypsin in 40 $\mu$L of 25 mM NH4HCO3 buffer) was added to the samples for trypsin digestion. Then the peptides of each sample were desalted on the C18 cartridge column (Empore™ SPE Cartridges C18 (standard density), bed I.D. 7 mm, volume 3 mL; Sigma, St. Louis, MO, USA) and dried under a vacuum. The peptide concentration was estimated by ultraviolet light spectral density at 280 nm using an extinction coefficient of 1.1 of 0.1% (g/L) solution, which was calculated based on the frequency of tryptophan and tyrosine in vertebrate proteins.

All fractions for data-dependent acquisition (DDA) library generation were analyzed by the Thermo Scientific Q Exactive HF X Mass Spectrometer connected to the Easy-nLC 1200 Chromatography System (Thermo Fisher Scientific, Waltham, MA, USA). The peptide (1.5 $\mu$g) was first loaded onto the EASY-SprayTM C18 Trap Column (P/N 164946, 3 $\mu$m, 75 $\mu$m $\times$ 2 cm; Thermo Fisher Scientific) and then separated on the EASY-SprayTM C18 LC Analytical Column (ES802, 2 $\mu$m, 75 $\mu$m $\times$ 25 cm; Thermo Fisher Scientific) with a linear gradient of buffer B (84% acetonitrile and 0.1% formic acid) at a flow rate of 250 nL/min over 120 min. The mass spectrometry (MS) detection method was positive ion, the scan range was 300–1800 m/z, the resolution for the MS1 scan was 60,000 at 200 m/z, automatic gain control (AGC) was 3e6, maximum IT was 25 ms, and dynamic exclusion was 30.0 s. Each full MS-SIM scan followed 20 ddMS2 scans. The resolution for the MS2 scan was 15,000, the AGC target was 5e4, the maximum IT was 25 ms, and the normalized collision energy was 30 eV.

The peptides from each sample were analyzed by liquid chromatography-tandem MS operating in the data-independent acquisition (DIA) mode by Shanghai Applied Protein Technology Co., Ltd. (Shanghai, China). Each DIA cycle contained one full MS–selected ion monitoring (SIM) scan, and 30 DIA scans covered a mass range of 350–1800 m/z with the following settings: SIM full scan resolution was 120,000 at 200 m/z, AGC 3e6, maximum IT 50 ms, and profile mode. The DIA scans were set at a resolution of 15,000, AGC target 3e6, maximum IT auto, and normalized collision energy 30 eV. The run time was 120 min with a linear gradient of buffer B (84% acetonitrile and 0.1% formic acid) at a flow rate of 250 nL/min. Quality control samples (pooled sample from an equal aliquot of each sample in the experiment) were injected in DIA mode at the beginning of the MS study and after every six injections throughout the experiment, which were used to monitor the MS performance.

The raw MS data were analyzed using MaxQuant software (Max Planck Institute of Biochemistry in Martinsried, Germany, version 1.5.3.17). The database was Uniprot_human database, and iRT peptide sequence was added (iRT-Kit; Biognosys, Schlieren, Switzerland). The parameters were set as follows: enzyme was trypsin, max missed cleavages was 2, fixed modification was carbamidomethyl (C), and the dynamic modification was oxidation (M) and acetyl (protein N-terminus). All reported data were based on 99% confidence for protein identification as determined by the false discovery rate (FDR = N(decoy) $\times$ 2/(N(decoy) + N(target))) $\le$1%). A spectral library was constructed by importing the original raw files and DDA search results into Spectronaut Pulsar X TM_12.0.20491.4 (Biognosys). DIA data were analyzed with SpectronautTM 14.4.200727.47784 search of the above constructed spectral library. Main software parameters were set as follows: retention time prediction type was dynamic iRT, interference on MS2 level correction was enabled, and cross run normalization was enabled. All results were filtered based on a Q value cutoff of 0.01 (equivalent to FDR 1%).

The DEPs were further analyzed by Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) using DAVID 6.8 database (https://david.ncifcrf.gov/).

The expression of the selected proteins from proteomics was verified by specific human ELISA kits (Abcam, Cambridge, MA, USA), and the experimental steps were performed in duplicate according to the manufacturer’s instructions.

Statistical analyses were performed using SPSS Statistics version 24.0 (SPSS Inc, Chicago, IL, USA). The clinical characteristics of the participants were analyzed using the Student’s t-test, chi-square test, or Fisher’s exact test as appropriate. Continuous variables with normal distribution are presented as the mean $\pm$ standard deviation (SD) and nominal data are presented as counts and percentages. Receiver-operating characteristic (ROC) curve analysis and calculation of the area under the curves (AUC) were conducted. p 0.05 was considered statistically significant.

A total of 80 CAC patients and 80 matched-controls were included in this study. Among them, 60 serum samples (30 from CAC patients vs 30 from controls) were collected for proteomic analysis. For the validation set, serum samples from 100 participants, including 50 CAC patients and 50 matched controls, were utilized for ELISA. The clinical characteristics of the participants in the discovery and validation sets are presented in Table 1. There were no significant difference were observed in age, sex distribution, or traditional risk factors between the cases and controls. Consequently, the spectrum of clinical presentations was also well matched between the two groups in either the discovery or validation phase.

A total of 1590 serum proteins were detected between the CAC and non-CAC groups, and the quantitative values for all proteins are shown in Supplementary Table 1. Proteins with fold change $\ge$1.5 or $\le$0.66 and p 0.05, and a selected peptide chain $>$1 were considered as differentially expressed. In total, 110 proteins were found to be differentially expressed between the two groups, of which 81 were upregulated and 29 were downregulated, as illustrated in the volcano plot and by clustering analysis (Fig. 2).

Fig. 2.

Differentially expressed proteins between patients with coronary artery calcification (CAC) and controls. (A) Volcano plot. X-axis represents log2 fold change and Y-axis shows –log10-transformed p values. Red dots indicate upregulated proteins, green dots indicate downregulated proteins, and gray dots represent proteins with no significant change. (B) Hierarchical clustering of the 110 dysregulated proteins. Red colour indicates the upregulated protein expression and blue indicates the downregulated protein expression.

GO and KEGG pathway analyses were performed to investigate the molecular mechanisms and functional activity of DEPs in coronary calcification. Detailed information on the biological process (BP), molecular function (MF), and cellular component (CC) by the GO enrichment analysis is shown in Fig. 3. In the BP category, DEPs were mainly involved in cytoskeleton organization, regulation of vasculature development, and actin cytoskeleton organization. Regarding MF classification, DEPs were associated with actin binding, cytoskeletal protein binding, signaling receptor binding, and cytokine receptor binding. Similarly, CC was mainly localized in the extracellular space, extracellular region, and extracellular vesicle. KEGG pathway results showed that the DEPs were mainly enriched in the complement and coagulation cascades, glycolysis/gluconeogenesis, regulation of the actin cytoskeleton, and platelet activation (Fig. 4).

Fig. 3.

Gene ontology (GO) analysis of the differentially expressed proteins between patients with coronary artery calcification (CAC) and controls. GO function consist of biological process (BP), molecular function (MF) and cellular component (CC). The Y-axis represents the number of different proteins under each functional classification. The color gradient reflects the size of the p value (p 0.05). The label at the top of the bar graph shows rich fator $\le$1, and the enrichment factor represents the ratio of the number of differentially expressed proteins (DEPs) over the total number of proteins in each category.

Fig. 4.

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis of the differentially expressed proteins between patients with coronary artery calcification (CAC) and controls. The bubble color represents significant p value, the bubble size represents the number of differentially expressed proteins (DEPs) in the pathway. The X-axis is rich factor representing the ratio of DEPs over the total number of proteins in each pathway.

The validation of candidate protein biomarkers was based on the following criteria: identified proteins containing more than one unique peptide, differential expression fold change $\ge$1.5, significance based on the standard Student’s t-test (p 0.05), quantified proteins corresponding to the gene names were found in the UniProt database, and potential functional or pathological significance in vascular calcification. In this study, KEGG pathway enrichment analysis suggested that complement and coagulation cascade, glycolysis/gluconeogenesis, regulation of actin cytoskeleton, and platelet activation signal pathways are significantly altered in the CAC patients, which are strongly associated with pathological calcification [13, 14, 15]. Because upregulated proteins may represent the disease process and have the potential to be a serum biomarker, the following six upregulated proteins, associated with the pathogenesis of cardiovascular calcification, were chosen for validation by ELISA in independent cohort samples: complement C5 (C5), C6, C7, fibrinogen gamma (FGG), pyruvate kinase isoform M2 (PKM2), and tropomyosin 4 (TPM4).The results demonstrated that serum C5 level in CAC patients was higher than that in non-CAC controls (137.83 $\pm$ 30.85 ng/mL vs 100.88 $\pm$ 22.95 ng/mL; p 0.001). Serum FGG had the same trend between the two groups (115.18 $\pm$ 16.06 $\mu$g/mL vs 95.88 $\pm$ 20.60 $\mu$g/mL; p 0.001). Compared to the non-CAC group (2.32 $\pm$ 0.88 mg/mL), the serum level of PKM2 was also significantly increased in the CAC patients group (3.48 $\pm$ 1.09 mg/mL; p 0.001). Additionally, patients with CAC showed elevated serum levels of TPM4 in serum compared with the non-CAC cohort (4.48 $\pm$ 1.14 mg/mL vs 2.93 $\pm$ 1.01 mg/mL; p 0.001). No significant difference in C6, and C7 expression was found between the two groups (Fig. 5A).

Fig. 5.

Validation of candidate serum proteins in an independent cohort. (A) ELISA validation of six candidate serum proteins (C5, C6, C7, FGG, PKM2, and TPM4) in 50 patients with severe coronary artery calcification and 50 matched controls. Data are presented as mean values $\pm$ standard deviation(SD). (B) The diagnostic performances of C5, FGG, PKM, and TPM4 by receiver operating characteristic (ROC) curve analysis.

ROC curves were calculated to further determine the diagnostic power of serum C5, FGG, PKM, and TPM4 in differentiating CAC patients from non-CAC controls (Fig. 5B). As shown in Table 2, serum C5, FGG, PKM and TPM4 had an optimal AUC $>$0.75, of which C5 had the largest AUC (0.821 [95% confidence interval: 0.732–0.891]; p 0.001). Moreover, the combination of C5, FGG, PKM and TPM4 had higher AUC, sensitivity, and specificity in discriminating coronary calcification than a single marker alone.