Analysis of Serum Paraoxonase 1 Using Mass Spectrometry and Lectin Immunoassay in Patients With Alpha-Fetoprotein Negative Hepatocellular Carcinoma

The diagnosis of AFP (alpha-fetoprotein)-negative HCC (hepatocellular carcinoma) mostly relies on imaging and pathological examinations, and it lacks valuable and practical markers. Protein N-glycosylation is a crucial post-translation modifying process related to many biological functions in an organism. Alteration of N-glycosylation correlates with inflammatory diseases and infectious diseases including hepatocellular carcinoma. Here, serum N-linked intact glycopeptides with molecular weight (MW) of 40–55 kDa were analyzed in a discovery set (n = 40) including AFP-negative HCC and liver cirrhosis (LC) patients using label-free quantification methodology. Quantitative lens culinaris agglutin (LCA) ELISA was further used to confirm the difference of glycosylation on serum PON1 in liver diseases (n = 56). Then, the alteration of site-specific intact N-glycopeptides of PON1 was comprehensively assessed by using Immunoprecipitation (IP) and mass spectrometry based 16O/18O C-terminal labeling quantification method to distinguish AFP-negative HCC from LC patients in a validation set (n = 64). Totally 195 glycopeptides were identified using a dedicated search engine pGlyco. Among them, glycopeptides from APOH, HPT/HPTR, and PON1 were significantly changed in AFP-negative HCC as compared to LC. In addition, the reactivity of PON1 with LCA in HCC patients with negative AFP was significantly elevated than that in cirrhosis patients. The two glycopeptides HAN253WTLTPLK (H5N4S2) and (H5N4S1) corresponding to PON1 were significantly increased in AFP-negative HCC patients, as compared with LC patients. Variations in PON1 glycosylation may be associated with AFP-negative HCC and might be helpful to serve as potential glycomic-based biomarkers to distinguish AFP-negative HCC from cirrhosis.


INTRODUCTION
As the most common human cancer, hepatocellular carcinoma (HCC) is a dominant reason for cancer-related death all around the world (1,2). China is among the countries with the most serious liver cancer epidemic (3). Most cases of liver cancer in clinical practice develop on the background of hepatic cirrhosis (4). Continuous screening for HCC progression could help to increase survival in cirrhosis patients (5). Therefore, monitoring the development of liver cancer is a key issue for the management of cirrhosis patients.
Alpha-fetoprotein (AFP) is controversial for its clinical application in the clinical detection of liver cancer. Serum AFP is a widespread diagnostic biomarker for HCC, however, with limited diagnostic power (6)(7)(8). The sensitivity of ultrasound combined with AFP is only approximately 60% (9). In addition, about one-third of HCC patients have low or normal AFP concentration in the serum, and the diagnosis of HCC in this condition is a tough challenge (10). Several studies are in progress to discover novel biomarkers in HCC, especially for AFP-negative patients.
As an important and ubiquitous posttranslational modification, protein glycosylation plays an important role in pathological and physiological processes (11,12). In humans, nearly 70% of proteins are glycosylated (13). Moreover, Nglycosylation regulates the correct folding, transport, and biological function of various proteins (14) and has been shown to be involved in cancer, inflammatory diseases, and autoimmune diseases by participating in proliferation, cellmatrix interactions, migration, and differentiation (15,16). Glycosylation has been reported to be associated with most cancer serum biomarkers (17). For example, a serum glycomicbased approach can be applied to evaluate the risk for the development of HCC in patients with compensated cirrhosis (18). In recent years, changes in the glycosylation of liver diseaserelated sera proteins have attracted wide attention. For instance, the researchers employed MALDI-TOF mass spectrometry systems to profile glycan of serum AGP (19). It has been reported that the trifucosylated tetra-antennary glycan structure of AGP could distinguish HCC from cirrhosis. There are increasing pieces of evidence that glycosylation changes of serum proteins are closely related to cancer progression (20)(21)(22). Unique alterations of tumor-associated N-glycosylation have several inherent advantages, such as providing distinct biomarkers indicating significant and amplified changes (23). Characterizing the heterogeneity and dynamism of glycosylation will help to better understand the molecular pathogenesis of cancer and shed light on new clues for diagnosis, prognosis, and treatment (24,25).
In the present study, we aimed to discover new diagnostic biomarkers to distinguish AFP-negative HCC from LC patients based on glycoproteomic technology and confirmed the differential expression of the identified glycopeptides in a validation set.

Subjects Samples
Samples were collected from Guangxi Medical University Cancer Hospital, including 80 cases of cirrhosis and 80 cases of liver cancer ( Table 1). Medical Ethics Committee of Guangxi Medical University Cancer Hospital approved the acquisition and use of these specimens. All patients gave informed consent. Cirrhotic patients were collected retrospectively, and those who had received local treatment or incomplete clinical data were excluded. Cirrhosis was confirmed by B-ultrasound examination in the presence of chronic liver disease. Hepatocellular carcinoma diagnosis was based on CT, MRI characteristics, or the results of pathology examination.

Sample Preparation
For the discovery phase, the serum specimens were obtained from 20 patients with liver cirrhosis (four biological repetitions, each pooled from five random individuals) and 20 patients with AFP-negative hepatocellular carcinoma (four biological repetitions, each pooled from five random individuals). Serum AFP level <20 ng/ml was considered to be AFP negative, using standard kits (Abbott Labs). In addition, traditional lectin ELISA was conducted for each patient of another 56 patients for further confirmation. For the validation phase, an independent cohort including 32 patients with LC and 32 with AFP-negative HCC were assessed individually by 16

Immunoprecipitation
Serum was diluted with lysis buffer (Beyotime, Shanghai, China). In this procedure, 20 µl of protein G agarose slurry (Roche, Basel, Switzerland) was added to the diluted serum at 4°C for 3-4 h incubation to remove IgG interference. After high speed centrifugation, 30 µl of protein G agarose slurry (Roche) and anti-PON1 antibody (ab92466, Abcam, UK) were sequentially added to the IgG-depleted serum for overnight incubation. Then the beads were washed with the same lysis buffer for five times. IP samples were handled with SDS-PAGE for in-gel digestion and glycopeptide enrichment. 16

O/ 18 O Labeling of Glycopeptides
Immobilized trypsin (Thermo Scientific, Rockford, IL) was added and dried with a vacuum centrifuge. H 2 16 O/H 2 18 O (97%, Cambridge Isotope Laboratories, Andover, MA) was added respectively for 24 h at the temperature of 37°C. 16 Oand 18 O-labeled digests were pooled by equal volumes before mass spectrometry analysis. The 16 O/ 18 O labeled Nglycopeptides were identified and quantified according to a previous study (22).

Protein ELISA and Lectin-ELISA Assay
Levels of PON1 in serum were determined with ELISA kits (R&D, Minneapolis, USA). Samples were diluted 1:50 in phosphate-buffered saline before lectin-ELISA. The 96-well The serum AFP <20 ng/ml was considered to be AFP negative. *Eighty patients in each group are composed of a discovery set (n = 20), samples for ELISA (n = 28) and a validation set (n = 32). a ALT and AST levels were determined by ARCHITECT c16000 (Abbott, Sligo, Ireland).
plate was coated with diluted capture antibody. After overnight incubation, each well was blocked with (3%) BSA in PBS for more than 2 h. Then, the coated antibody was reacted with the oxidation buffer (50 mM citric acid, 100 mM NaIO 4 ) for 1 h without light. After washing, each well of plate was applied with 100 µl diluted serum. After incubation, each well was washed with PBST (0.05% Tween20 in PBS), and 1 µg/ml biotinylated LCA (Vector, Burlingame, CA) was then added at room temperature for 2 h. HRP-streptavidin was added to the plate followed by TMB and stop solution. OD value at 450 nm was measured. Twenty-eight LC and 28 AFP-negative HCC patients were conducted individually. ELISA Index was defined to be OD value of PON1 bound to lectin LCA, divided by OD value of protein PON1.

Statistical Analysis
Figures were plotted with GraphPad Prism (version 8.0, GraphPad Software Inc.) and OriginPro 8 software. Statistical comparisons were calculated using t-test. P <0.05 was considered to be statistically significant. All tests for significance were two-tailed.

Reproducibility of Label-Free Quantitation
Hp was utilized as standard glycoprotein to assess the stability of label-free quantitation methodology ( Figure 1A). After running for each serum experiment, the blank was run in order to ensure no carry-over. A representative gel image was shown in Figure 1B. After enrichment, N-glycopeptides of standard Hp were investigated by nano LC-HCD-MS/MS. We further selected the high-abundance glycosites N184 and N241 to evaluate the quantitative reproducibility through four independent replicates at four different times. In order to simplify the glycan annotation, a four-digit nomenclature in HNSF order was used (H, Hexose; N, N-acetylhexosamine; S, Sialic acid; F, Fucose). As described in the figure (Figure 1C), the Pearson Correlation Coefficient R 2 values of the four technical repeats ranged from 0.955 to 0.998, indicating fine reproducibility. Pie chart ( Figure 1D) showed the frequency of intact glycopeptide quantification in four replicates, in three, in two, or only in one. The results indicated that 88.57% of glycopeptides were identified with at least two repeats. The quantification reproducibility of N241 glycopeptides was also carried out, and the difference of abundance was five orders of magnitude ( Figure 1E)

Quantification for N-Glycopeptides of 40-kDa Band
The label-free glycopeptide quantification method was performed to obtain comparisons of AFP-negative HCC and LC. Equal amounts of AFP-negative HCC and LC sera (one pool of five individuals) were obtained to separate target bands ( Figure 2A). In total, 195 glycopeptides were identified (Supplementary Table S1) and corresponded to 33 kinds of glycan compositions. For instance, H5N4S2 was considered as the most common glycan composition in the experiment ( Figure 2B). According to the four biological repeats, 53 Nglycopeptides appeared at least three times ( Figure 2C). Among them, 26 N-glycopeptides passed the strict filtering criteria (0.8<similarity ≤ 1 and 0<score interference ≤ 0.3) (Supplementary Table S2). As the cutoff criterion was set to fold change (FC) >2 or <0.5, three differentially expressed glycopeptides were identified in AFP-negative HCC samples compared to LC samples, including APOH_N253 (H5N4S2), PON1_N253 (H5N4S2), and HPTR/HPT_N126/184 (H5N4S1) ( Figure 2D). Compared with LC patients, PON1_N253 (H5N4S2) and HPTR/HPT_N126/184 (H5N4S1) were significantly up-regulated in AFP-negative HCC (p = 0.0026 and p = 0.0366, respectively), while APOH_N253 (H5N4S2) was considerably decreased (p = 0.0127). Interestingly, the protein level of PON1 was shown to be down-regulated in AFP-negative HCC in our recent study (31), which is opposite to the variation in glycopeptide abundance. Moreover, the increase of fucosylation and sialylation in patients' sera could play an essential role in the development of HCC (32). Finally, we selected PON1 as target glycoprotein in our study for further confirmation, whose glycosylation levels in AFP-negative HCC had only a few studies (33). Representative pGlyco annotation of PON1_N253 (H5N4S2) was shown in Figure 3A.

Measurement of PON1 Glycosylation Using ELISA Index
It was reported that LCA could recognize core fucosylated biantennary or tri-antennary N-glycans (34 Figure 3B).

Relative Quantitation of Site-Specific N-Glycans on PON1
To further discover and confirm N-glycosylation of PON1 in AFPnegative HCC, 16  O and subjected to LC-MS/MS analysis ( Figure 4A). The predicted N-glycosylation sites in PON1 were also considered (http://www.cbs.dtu.dk/ services/NetNGlyc/). Considering N253 and N324 were the most informative sites in PON1 (36,37), the glycosylation heterogeneity of these two sites was further studied. 16 glycoforms of the two glycosites (N253 and N324, detailed information shown in Supplementary Table S3) on PON1 were identified by pGlyco (FDR < 1%, Figure 4B). Among the identified glycopeptides, the intensity of HAN 253 WTLTPLK (H5N4S2) and (H5N4S1) was significantly increased in AFP-negative HCC patients ( Figure   4C, detailed information shown in Supplementary Table S4)

Discussion
Currently, there are many advances in the treatment of liver cancer, including liver transplantation, surgical resection, and interventional therapy (38,39). However, numerous HCC patients are diagnosed only after the obvious clinical symptoms appear in the middle or late stage, which makes early diagnosis a top priority (40,41). At present, AFP is used as a serum biomarker to assist in the diagnosis or screening of high-risk liver cancer patients (42), but approximately 30-40% of overall hepatocellular carcinoma patients have very low serum AFP levels (<20 ng/ml), which is called AFP-negative HCC (43). an urgent need for new, economical, and effective biomarkers to predict the diagnosis of HCC. Glycosylation of serum glycoproteins is always associated with tumor progression. Changes in protein glycosylation are becoming increasingly popular to be used as diagnostic, prognostic, and therapeutic drug target indicators (49). Most of the N-glycans found in the whole serum are attached to the serum proteins produced in the liver. Changes in glycosylation of specific serum proteins secreted by the liver, such as haptoglobin, hemopexin, and vitronectin have been associated with HCC (50). In the present research, the N-glycopeptides of serum glycoprotein cluster with molecular weight of 40-55 kDa were quantified in LC versus AFP-negative HCC patients. Three differentially expressed intact N-glycopeptides belonging to APOH, PON1, and HPTR/HPT were screened. PON1_N253 (H5N4S2) and HPTR/HPT_N126/184 (H5N4S1) were up-regulated in AFPnegative HCC samples in this study. Whereas, APOH_N253 (H5N4S2) was reduced in HCC compared with LC patients. Then, the ELISA Indexes of protein ELISA and LCA-ELISA were used to validate glycan alterations of serum PON1 in AFPnegative patients with HCC. It was revealed that the changes of glycans on PON1 were significant.
PON1 is a high-density lipoprotein (HDL)-related serum enzyme that is mainly synthesized in the liver and released into the blood stream (51), which plays an important role in the prevention of vascular disease by metabolizing oxidized lipids (52). It has been found that changes in circulating PON1 levels are associated with various diseases involving oxidative stress reaction and inflammation (53). In our latest study, protein levels of PON1 in LC patients were elevated significantly than those in AFPnegative HCC patients (31), which is consistent with our PON1 ELISA results in this study. While this seemingly opposite trend may imply a dramatic change in PON1 glycosylation, which is independent of protein level. The fucosylation of PON1 was reported in HCC and lung cancer (33,54), and there were two informative N-glycosylation sites (N253 and N324) on PON1 (36,37,55). A variety of studies have confirmed changes in the N-glycosylation state of PON1, while only a few studies have proved the site-specific glycosylation of PON1 using glycoproteomic approach.
Intact glycopeptide analysis including glycosylation site and site-specific glycan is crucial for the understanding of the  biological significance of glycosylation (56). Based on labeling quantification, two N-glycopeptides HAN 253 WTLTPLK (H5N4S2) and (H5N4S1) of PON1 with high abundance have been realized to be increased significantly in AFP-negative HCC as compared with LC patients. It is the first reported study about site-specific PON1 glycosylation in AFP-negative liver disease conditions. However, our present study has limitations in terms of sample size and the absence of controls with benign liver diseases. This study has clarified the existence of potential biomarkers, such as the two glycopeptides of PON1 in AFP-negative liver cancer. Other candidate biomarkers should be further studied using other proteomic methodology.
To sum up, MS-based quantification methods combining labeling/label-free and lectin-based ELISA were used to screen and verify glycopeptide biomarkers in patients with LC and AFP-negative HCC. The results showed that two site-specific glycoforms on PON1 were significantly elevated in AFPnegative HCC and could be potential biomarkers to screen of AFP-negative HCC versus LC patients. For the clinical application of new-found biomarkers, preclinical validation is usually required with high-throughput quantitative methods and independent large sample sets (57). Multicenter studies on large samples are needed to evaluate the practicability of the results and are expected to translate them into clinical practice.

CONCLUSION
In conclusion, we found that N-glycosylation of PON1 played a key role in AFP-negative HCC tumorigenesis. Moreover, HAN 253 WTLTPLK (H5N4S2) and (H5N4S1) of PON1 could be served as potential diagnostic indicators for distinguishing AFP-negative HCC from cirrhosis.

DATA AVAILABILITY STATEMENT
The proteomics data and analyzed result datasets have been partially deposited to the ProteomeXchange Consortium (http:// proteomecentral.proteomexchange.org) via its official member iProX repository (58) with the dataset identifier PXD023500.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Research Ethics Committee of Cancer Hospital of Guangxi Medical University (no. LW2019043). The patients/ participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
HL and HS conceived and designed the research. All authors were responsible for acquisition and analysis of data. Furthermore, XC, YS, and XM performed the research. CL, GY, and LZ provided technical assistance. XC, ZC, CC, and GH performed the data analyses. HS and XC drafted the manuscript. HS and HL revised and commented on the draft. All authors contributed to the article and approved the submitted version.