Fibulin‐4 as a potential extracellular vesicle marker of fibrosis in patients with cirrhosis

Chronic liver injury leads to decreased liver function and increased fibrosis. Fibrosis is not only associated with the development of portal hypertension and carcinogenesis, but with the occurrence of events and a poor prognosis, highlighting the importance of non‐invasive fibrosis assessment in patients. In the present study, we searched for markers related to liver fibrosis via proteomic analysis of small extracellular vesicles (sEVs). In the discovery cohort, proteomic analysis was carried out in the sEVs extracted from the sera of 5 patients with decompensated cirrhosis, 5 patients with compensated cirrhosis, and 5 controls without liver disease. Interestingly, in this cohort, fibulin‐4 was significantly associated with cirrhosis while in the validation cohort [formed by 191 patients: 7 patients without disease, 16 patients without liver disease (other diseases), 38 patients with chronic liver disease (CLD), 75 patients with cirrhosis of Child–Pugh class A (36 without hepatocellular carcinoma [HCC], 29 with HCC), and 65 patients with cirrhosis of Child–Pugh class B–C (39 without HCC, 26 with HCC)], fibulin‐4/CD9 levels increased with cirrhosis progression. Furthermore, the fibulin‐4/CD9 ratio was significantly higher in patients with varices. Immunostaining also revealed strong fibulin‐4 expression in cholangiocytes within the fibrous areas and mesothelial cells in liver tissue blood vessels. Taken together, our results suggest that fibulin‐4, essential for lysyl oxidase activation, might be a new liver fibrosis marker found in the sEVs of patients with cirrhosis.

Chronic liver injury leads to decreased liver function and increased fibrosis.Fibrosis is not only associated with the development of portal hypertension and carcinogenesis, but with the occurrence of events and a poor prognosis, highlighting the importance of non-invasive fibrosis assessment in patients.In the present study, we searched for markers related to liver fibrosis via proteomic analysis of small extracellular vesicles (sEVs).In the discovery cohort, proteomic analysis was carried out in the sEVs extracted from the sera of 5 patients with decompensated cirrhosis, 5 patients with compensated cirrhosis, and 5 controls without liver disease.Interestingly, in this cohort, fibulin-4 was significantly associated with cirrhosis while in the validation cohort [formed by 191 patients: 7 patients without disease, 16 patients without liver disease (other diseases), 38 patients with chronic liver disease (CLD), 75 patients with cirrhosis of Child-Pugh class A (36 without hepatocellular carcinoma [HCC], 29 with HCC), and 65 patients with cirrhosis of Child-Pugh class B-C (39 without HCC, 26 with HCC)], fibulin-4/CD9 levels increased with cirrhosis progression.Furthermore, the fibulin-4/CD9 ratio was significantly higher in patients with varices.Immunostaining also revealed strong fibulin-4 expression in cholangiocytes within the fibrous areas and mesothelial cells in liver tissue blood vessels.Taken together, our results suggest that fibulin-4, essential for lysyl oxidase activation, might be a new liver fibrosis marker found in the sEVs of patients with cirrhosis.
Chronic liver injury can lead to liver dysfunction and fibrosis.Progressive fibrosis leads to portal hypertension, which in turn alters circulatory dynamics, resulting in ascites, portosystemic shunts, and varices. 1,2Furthermore, intestinal edema caused by portal hypertension can give rise to a variety of Abbreviations ATX, autotaxin; AUC, Area Under the ROC Curve; CLD, chronic liver disease; ECM, extracellular matrix; EFEMP2, EGF-containing fibulin extracellular matrix protein 2; ELISA, enzyme-linked immunosorbent assay; FIB-4, Fibrosis-4 index; HCC, hepatocellular carcinoma; LOX, Lysyl oxidase; M2BPGi, macrophage galactose-specific lectin-2-binding protein glycosylation isomer; P-III-P, Type III procollagen N-terminal peptide; pro-C3, N-terminal type III collagen propeptide; ROC, receiver operating characteristic; sEVs, small extracellular vesicles.pathological conditions, including changes in the immune system and metabolic disorders.This triggers systemic inflammation and the invasion of bacteria through the intestinal barrier, eventually affecting multiple organs and resulting in a decreased quality of life and, in some cases, death. 3n the past, liver biopsy was the mainstay for diagnosing fibrosis.However, a wide variety of non-invasive methods are currently used, ranging from blood tests, such as type IV collagen 7S, 4 hyaluronic acid, 5 type IIII procollagen N-terminal peptide (P-III-P), 6 N-terminal type III collagen propeptide (pro-C3), 7 platelet count, 8 autotaxin (ATX), 9 and macrophage galactose-specific lectin-2-binding protein glycosylation isomer (M2BPGi), 10 to the FIB-4 index, 11 liver stiffness measurement using FibroScan 12 and ultrasound, 13 as well as magnetic resonance elastography (MRE). 14,15While all fibrosis evaluation methods currently employed are excellent, there are variations and errors, highlighting the importance of combined diagnostic approaches.
7][18] They contain various nucleic acids and proteins, and their potential as biomarkers has been increasingly recognized.Small extracellular vesicles (sEVs), the main subset of EVs, are marked by CD9, CD63, and CD81.To date, no EV-based fibrosis marker has been reported, 19 as collecting EVs with high accuracy and evaluating them on a case-by-case basis is difficult.
Therefore, the current study aimed to reveal markers related to liver fibrosis through proteomic analysis of sEVs.We identified fibulin-4 as a potential marker of liver fibrosis in sEVs.

Materials and methods
Proteomics samples, clinical data, and samples for discovery and validation cohort, and varices analysis sEVs extracted from the sera of five patients with decompensated cirrhosis, five patients with compensated cirrhosis, and five controls without liver disease were used for proteomic analysis (    2).Fibrosis-related markers were examined in some patients [platelets, fibrosis-4 (FIB-4) index, Mac-2 binding protein glycosylation isomer (M2BPGi), and autotaxin (ATX) in 189, 189, 54, and 23 patients, respectively].To analyze the association between varices and marker expression, all patients were divided into two groups: those with endoscopic esophageal, gastric, or duodenal varices (87 patients) and those without varices (104 patients).
This retrospective study was approved by the Institutional Review Board of Niigata University (2021-0032).Written informed consent was obtained from all patients.The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki.

EV isolation and mass spectrometry analysis
EVs were isolated from 200 lL of serum samples using a MagCapture Exosome Isolation Kit PS (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) according to the manufacturer's instructions, except that samples were eluted using 20 lL of 0.5% sodium dodecyl sulfate (SDS) in 50 mM triethylammonium bicarbonate.
After reduction with 10 mM TCEP at 100 °C for 10 min and alkylation with 25 mM iodoacetamide at ambient temperature for 45 min, protein samples were subjected to digestion with Trypsin/Lys-C Mix (Promega Corporation, Madison, WI, USA) at 47 °C for 2 h on S-Trap columns (ProtiFi, Fairport, NY, USA).
The resulting peptides were extracted from the gel fragments and analyzed using an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) combined with an UltiMate 3000 RSLC nano-flow ー high-performance liquid chromatographer (Thermo Fisher Scientific).Peptides were enriched using a l-Precolumn (0.3 mm i.d. 9 5 mm, 5 lm, Thermo Fisher Scientific) and separated on an AURORA column (0.075 mm i.d. 9 250 mm, 1.6 lm, Ion Opticks Pty Ltd, Fitzroy, Vic., Australia) using a two-step gradient: 2-40% acetonitrile for 110 min, followed by 40-95% acetonitrile for 5 min in the presence of 0.1% formic acid.The compensation voltages for gas-phase fractionation through a FAIMS Pro interface (Thermo Fisher Scientific) were set at À40, À60, and À80 V.The analytical parameters of the Orbitrap Fusion Lumos mass spectrometer were set as follows: resolution of full scans = 50 000, scan range (m/z) = 350-1500, maxi- The obtained MS/MS spectra were searched against the Homo sapiens protein sequence database in SwissProt using Proteome Discoverer 3.0 software (Thermo Fisher Scientific), in which peptide identification filters were set at "false discovery rate <1%."Label-free relative quantification analysis of proteins was performed using the default parameters of the Minora Feature Detector node, Feature Mapper node, and Precursor Ions Quantifier node in Proteome Discoverer 3.0 (Thermo Fisher Scientific).

ELISA
EVs in patient sera were measured using the PS Capture TM Exosome ELISA Kit (FUJIFILM Wako Pure Chemical Corporation #298-80 601) and biotinylated anti-CD9 antibody (FUJIFILM Wako Pure Chemical Corporation #017-28 211) as per the manufacturer's instructions, except for the use of biotinylated antibodies.An anti-fibulin-4 antibody (polyclonal mouse anti-EFEMP2, purified immunoglobulin, Sigma-Aldrich #SAB1400526-50UG; We performed a fundamental primary investigation into the quantitative analysis of Fibulin-4 using this antibody with EVs derived from cell lines expressing Fibulin-4) was used.A Biotin Labeling Kit-NH2 (#LK03, Dojindo Molecular Technologies, Inc., Kumamoto, Japan) was used to conjugate biotin to the anti-fibulin-4 antibody.This antibody was confirmed to extract sEVs from the culture supernatant of TIG3, a cell line expressing fibulin-4, and to quantitatively measure sEVs.ELISA calculations were performed using an Infinite M Plex (Tecan Japan Co., Ltd., Kawasaki, Japan).

Statistical analyses
Data were processed, and volcano plots were created using GRAPHPAD PRISM V. 9.3.1 (GraphPad Software Inc., La Jolla, CA, USA).Data were assessed using the Mann-Whitney U test, and differences between groups were analyzed using one-way analysis of variance.Differences were considered statistically significant at P < 0.05.

Results
Fibulin-4 is strongly expressed in serum EVs from patients with decompensated cirrhosis in the discovery cohort We identified markers associated with fibrosis in the sEVs of patients with liver cirrhosis (Fig. 1A-G).First, we extracted sEVs from the sera of five patients with decompensated liver cirrhosis, five patients with compensated liver cirrhosis, and five controls without liver disease.These EVs were then subjected to proteomic analysis.We detected 2182, 2198, and 2068 EV-derived proteins in the sera from patients with decompensated cirrhosis, compensated cirrhosis, and controls without liver disease, respectively.
We then compared protein expression between the groups, observing these via volcano plots (Fig. 2A-C).We noted significant changes in certain protein levels between patients with cirrhosis and controls.We focused on fibulin-4 in particular, as it could not be detected at all in controls without liver disease.Meanwhile, fibulin-4 was detected in four of the five patients with decompensated cirrhosis and one of five patients with compensated cirrhosis (Fig. 2D).

Fibulin-4/CD9 levels increase with cirrhosis progression in the validation cohort
Next, we performed ELISA using a biotinylated antibody confirmed to quantify EVs in a cell line expressing fibulin-4.The kit we used could quantitatively measure common EV markers CD9 and CD63 in all serum samples analyzed in this study.We confirmed a strong positive correlation between CD9 and CD63 levels (r = 0.8426) (Fig. 3A).CD9 also showed a positive correlation (r = 0.5697) (Fig. 3B) with platelets and a weak negative correlation (r = À0.2577) (Fig. 3C) with age.This finding indicates that the amount of EVs may decrease with the progression of fibrosis and aging.Since the number of EVs differed across patients, we decided to normalize the fibulin-4 level to that of CD9 per patient.We then measured serum samples from 191 patients: seven without disease, 16 without liver disease (other diseases), 38 with chronic liver disease (CLD), 65 with cirrhosis of Child-Pugh class A [36 without hepatocellular carcinoma (HCC), 29 with HCC], and 65 with cirrhosis of Child-Pugh class B-C (39 without HCC, 26 with HCC) (Table 2).The other (non-liver) disease group included cases of acute pancreatitis, autoimmune pancreatitis, intraductal papillary mucinous neoplasms, interstitial pneumonia, focal nodular hyperplasia, enteritis, early gastric cancer, early esophageal cancer, infected pancreatic cysts, cholangitis, sarcoidosis, and lung infection.
First, we evaluated the association of the fibulin-4/ CD9 ratio with liver fibrosis in the absence of HCC in seven healthy patients, 16 patients without liver disease (other diseases), 38 patients with CLD, 36 patients with Child-Pugh class A cirrhosis (without HCC), and 39 patients with Child-Pugh class B-C cirrhosis (without HCC).The median fibulin-4/CD9 ratio was found to increase as liver fibrosis progressed.A significant difference in the ratio value was found between CLD and Child-Pugh class A cirrhosis (P < 0.01), with an area under the receiver operating characteristic (ROC) curve (AUC) of 0.8000, as well as between CLD and Child-Pugh class B-C cirrhosis (P < 0.01), with an AUC of 0.9079.A significant difference was also noted between the Child-Pugh class A and B-C (P < 0.05) groups, with an AUC of 0.7283 (Fig. 4A,B).
Next, all patients, including those with HCC, were analyzed.The median fibulin-4/CD9 values increased as liver lesions progressed, with a significant difference between CLD and Child-Pugh class A (P < 0.01), with an AUC of 0.8121, as well as between CLD and Child-Pugh class B-C cirrhosis (P < 0.01), with an AUC of 0.8995.Child-Pugh class A and Child-Pugh class B-C (P < 0.01) also differed significantly, with an AUC of 0.7031 (Fig. 5A,B).Taken together, the fibulin-4/CD9 ratio increased with the degree of fibrosis, irrespective of the presence or absence of HCC.
Varices are associated with a higher fibulin-4/CD9 ratio Next, all patients were divided into two groups: those with endoscopic esophageal, gastric, or duodenal varices (87 patients) and those without varices (104 patients).Patients with varices had a significantly higher fibulin-4/CD9 ratio than those without (P < 0.001).A ROC curve was also generated, and the AUC of 0.8077 indicated that the ratio could discriminate well between patients with and without varices (Fig. 7A,B).

Fibulin-4 is upregulated in cholangiocytes within fibrotic areas and mesothelial cells in blood vessels
According to single-cell analysis of the Protein Atlas, 20 fibulin-4 is expressed in fibroblasts, smooth muscle cells, endothelial cells, and cholangiocytes.To verify whether fibulin-4 is indeed expressed in the liver and to determine which cell types express it, we performed r = 0.8426 r = 0.5697 r = -0.2577staining of tissues from patients with liver cirrhosis.Fibulin-4 expression was noted in bile duct cells near fibrotic regions in the liver and in the mesothelial cells of blood vessels (Fig. 8A-C).

Discussion
In this study, we identified fibulin-4 [also known as epidermal growth factor-containing fibulin extracellular Fibulin-4; extracellular vesicle marker of cirrhosis matrix protein 2 (EFEMP2)] as a novel fibrosis marker in serum sEVs.It was particularly upregulated in patients with decompensated liver cirrhosis in both the discovery and validation cohorts.In addition to its elevation with the progression of cirrhosis, the levels of this marker were significantly higher in patients with endoscopic varices than in those without varices, indicating portal hypertension.Within the liver, fibulin-4 expression was observed close to regions of fibrosis, namely, the bile ducts and mesothelial cells of the arteries.
The matricellular protein fibulin-4 is essential for lysyl oxidase (LOX) activation. 21This LOX family member cross-links structural extracellular matrix (ECM) components, particularly fibrous collagen and elastin, being strongly implicated in fibrosis progression and resistance to fibrosis reversal.LOX proteins (LOX, LOXL1, LOXL2, LOXL3, and LOXL4) are extracellular copper-dependent enzymes and have been identified as potential therapeutic targets in liver fibrosis.Further, fibulin-4 has been reported to play a specific role in promoting the proper maturation of elastic fibers by linking LOX to tropoelastin through interactions with the LOX propeptide, which catalyzes the cross-linking of elastin molecules. 22Fibulin-4, a LOX-related protein, differs from conventional indicators of fibrosis and does not correlate well with conventional indicators.Moreover, its significance may deepen in many clinical specimens in the future.
Fibulin-4 is a member of the fibulin protein family composed of long (fibulin-1, -2, -6) and short fibulins (fibulin-3, -4, -5, -7) involved in protein-protein interactions between components of the basement membrane and ECM.Fibulin-1, -2, -3, -4, and -5 bind the  monomeric form of elastin (tropoelastin) in vitro.Further, fibulin-2, -3, -4, and -5 are involved in various aspects of elastic fiber development in vivo.In particular, fibulin-4 and -5 are non-dispensable for elastic fiber assembly. 23Fibulin-3, À4, and -5 are closely related paralogs with very similar domain structures and sequences, but have independent molecular functions in elastogenesis, as indicated by the distinct phenotypes noted in knockout mice and patients with mutations in the corresponding genes. 23ore recently, evidence suggesting that fibulin-4 is necessary for fibrillar collagen assembly has emerged. 23ibulin-4 À/À mice die immediately before birth because of arterial hemorrhage, while fibulin-4 +/À mice appear normal. 24Insufficient levels of fibulin-4 compromise the structural integrity of the aortic wall and can lead to aneurysm.Consistent with this, patients with mutations in fibulin-4 suffer from cardiovascular complications, including aortic aneurysms, arterial tortuosity, and elastin abnormalities as described in detail previously. 25Mutations in fibulin-4 also cause autosomal recessive cutis laxa 1B, characterized by loose skin, with vascular, lung, and skeletal abnormalities.
Although reports on the association of fibulin-4 with liver disease are rare, Pantano et al. have identified gene expression clusters that strongly correlate with the fibrosis stage, including four genes consistently reported across previously published transcriptomic studies on metabolic dysfunction-associated steatohepatitis: COL1A2, EFEMP2 (fibulin-4), FBLN5 (fibulin-5), and THBS2. 26This report supports our findings, suggesting that fibulin-4 and fibulin-5 are associated with liver fibrosis.
In this study, we detected fibulin-4 in sEVs.In general, EVs disappear from the blood in a relatively short period and thus dynamically reflect fibrosis at any given time.However, we did not identify fibulin-5 in sEVs.This suggests that members of the fibulin family may have different roles and kinetics, particularly in sEVs from patients with fibrosis.
Zhang et al. 27 reported that fibulin-4 promotes osteosarcoma invasion and metastasis by inducing the epithelial-to-mesenchymal transition (EMT) via the PI3K/Akt/mTOR pathway.Li et al. 28 reported that EFEMP2 increases the invasive ability of cervical cancer cells by promoting EMT via the Raf/MEK/ERK signaling pathway.In contrast, Kang et al. 29 reported that EFEMP2 inhibited breast cancer invasion and metastasis in vitro and in vivo.
Currently, various indices are available to evaluate fibrosis.Liver stiffness measurements using FibroScan, ultrasound, and MRE are widely used in clinical practice.One drawback of these existing methods is that they reflect inflammation.However, fibulin-4 is theoretically related to fiber maturation and may be less susceptible to the effects of inflammation.In addition, unlike fibulin-4, other markers are not related to fiber maturation.Thus, we believe that fibulin-4 could be a marker that reflects the progression of liver stiffness in a clinical setting.
In this study, fibulin-4 was expressed in cholangiocytes, particularly those undergoing ductular reactions, and in mesothelial cells of the surrounding blood vessels.The ductal cells undergoing ductular reactions are constantly present within fibrosis lesions.Fibulin-4 may contribute to the stabilization of growing cholangiocytes and may initially play an important role in the tissue repair process, with excessive growth potentially leading to fibrosis.
The study had some limitations.It did not assess the samples for acute liver and non-liver diseases.The clinical significance of this marker should be further explored.In future studies, samples from patients with acute liver and non-liver diseases should also be assessed in order to clarify the expression patterns of this marker.Moreover, the diagnostic value of fibulin-4 compared to other markers should be determined in patients with severe cirrhosis.The serum was not assayed in the present study owing to the unavailability of ELISA or other suitable assay systems.However, it may be more meaningful to evaluate whether the serum can be used to determine the difference in the future.The present study is novel in that it is the first study to use EVs to evaluate individual cases, which is technically difficult.The novelty of the present study lies in the fact that the ELISA method used herein was able to achieve this.
Despite these limitations, our identification of a new serum EV-based marker of liver fibrosis is significant.We believe that fibulin-4 has potential as a new biomarker of fibrosis in the clinic.
mum injection time of full scans = 50 ms, AGC target of full scans = 4 9 10 5 , dynamic exclusion duration = 30 s, cycle time of data-dependent mass spectrometry (MS)/MS acquisition = 2 s, activation type = HCD, detector of MS/MS = Ion trap, maximum injection time of MS/ MS = 35 ms, and AGC target of MS/MS = 1 9 10 4 .

Fig. 1 .
Fig.1.Flowchart of this experiment.(A) From serum, (B) we extracted extracellular vesicles, (C) performed proteomics analysis, and (D) selected candidate proteins.(E) We verified whether antibodies could be used to detect the target protein and then quantitatively measured target protein expression via enzyme-linked immunoassay (ELISA).Finally, (F) a large number of samples were assayed using ELISA, and (G) associations with clinical data were determined.

Fig. 2 .
Fig. 2. Proteomics analysis.Volcano plot showing protein expression in (A) control vs. patients with compensated cirrhosis, (B) control vs. patients with decompensated cirrhosis, and (C) patients with compensated cirrhosis vs. those with decompensated cirrhosis.(D) Representative results of fibulin-4 expression as per proteomic analysis from five cases of each group.

Fig. 3 .
Fig. 3. Association of extracellular vesicle CD9 vs. CD63 levels, platelets, and age.(A) Correlation of CD9 and CD63 levels in extracellular vesicles, (B) correlation between extracellular vesicle CD9 levels and platelets, and (C) correlation between extracellular vesicle CD9 levels and age.

Fig. 4 .
Fig. 4. Fibulin-4/CD9 ratio in serum extracellular vesicles based on disease state (excluding patients with hepatocellular carcinoma).(A) Fibulin-4/CD9 ratio values for each disease state, excluding hepatocellular carcinoma.(B) Receiver operating characteristic (ROC) curves showing the comparison between disease states and the diagnostic accuracy of the fibulin-4/CD9 ratio for distinguishing each disease state.*P < 0,05, **P < 0.01.AUC, area under the ROC curve; C-P, Child-Pugh.[Correction added on 5 August 2024, after first online publication; Figure 4A has been updated to correct the alignment of the significance bars.]

Fig. 5 .
Fig. 5. Fibulin-4/CD9 ratio in serum extracellular vesicles by disease state (including patients with hepatocellular carcinoma).(A) Ratio values are measured for each disease state, including hepatocellular carcinoma.(B) Receiver operating characteristic (ROC) curves showing the comparison of each disease state and the diagnostic accuracy of the fibulin-4/CD9 ratio for distinguishing between each disease state.**P < 0.01.AUC, Area under the ROC curve; C-P, Child-Pugh.[Correction added on 5 August 2024, after first online publication; Figure 5A has been updated to correct the alignment of the significance bars.]

Fig. 8 .
Fig. 8. Immunohistochemical staining of fibulin-4 in cirrhotic liver tissues.(A) Immunohistochemical staining of cirrhotic tissue with an anti-fibulin-4 antibody.(B) A magnified view of the black frame in (A).(C) Magnified view of the red frame in (B) Scale bar = 50 lm.