Serum GRP78 as a potential biomarker for early liver injury and active fibrogenesis in alcohol-associated liver disease

Abstract

Alcohol-associated liver disease(ALD) is an important cause of liver-related lesions and death, mainly caused by long-term or excessive alcohol consumption.Reactive oxygen species (ROS) produced during alcohol metabolism induce an oxidative stress response, which GRP78 playing an important regulatory role. This study aims to investigate the relationship between serum GRP78 levels and liver fibrosis/cirrhosis in ALD patients. A case control study was conducted, involving ALD patients (n = 122) and healthy participants (n = 28). Liver steatosis and fibrosis/cirrhosis was assessed by transient elastography (TE) and biochemical parameters were collected. Serum GRP78 was measured by sandwich ELISA based on two monoclonal anti-GRP78 antibodies and calibrated with a standard of recombinant GRP78. Serum GRP78 levels increased in alcohol-associated steatohepatitis and alcoholic hepatitis patients correlated with early fibrosis severity (F1 fibrosis stage), but were significantly downregulated in patients with cirrhosis (F4 fibrosis stage). Furthermore, serum GRP78 levels showed a weak negative correlation with liver stiffness measurements (LSM; r = − 0.2064, p = 0.0365) and demonstrated a notable association with the severely advanced fibrotic stage in ALD patients, while showing no significant correlation with the controlled attenuation parameter (CAP) associated with liver steatosis. This study suggests that serum GRP78 is a potential noninvasive biomarker for early liver injury and active fibrogenesis in ALD patients. When combined with LSM, it improves diagnostic accuracy for advanced fibrosis and cirrhosis.

Introduction

Alcohol-associated liver disease (ALD) is one of the main causes of chronic liver disease. The spectrum of ALD severity encompasses steatosis, steatohepatitis, advancing to fibrosis, cirrhosis, and hepatocellular carcinoma (HCC) [1]. ALD patients exhibit accelerated disease progression and tend to present for medical care at later disease stages compared to individuals with alternative liver disease etiologies [2].

Alcohol-associated steatosis (AS) represents the earliest histological stage of alcohol-related liver disease (ALD), characterized by the abnormal accumulation of lipids, primarily triglycerides, in hepatocytes. Continuous alcohol consumption leads some patients to progress to alcohol-associated steatohepatitis (ASH). Typical histological changes in the liver include ballooning degeneration and necrosis of liver cells, along with lobular inflammation predominantly involving neutrophils. Alcoholic hepatitis (AH) emerges as a distinct phenotype of advanced disease, marked by the rapid onset or exacerbation of jaundice. In severe instances, chronic liver failure occurs, presenting a mortality risk of 20% to 50% within one month [2].Persistent inflammation and hepatocyte damage contribute to the development of progressive liver fibrosis, ultimately resulting in alcohol-associated cirrhosis (AC) and significantly increasing the risk of hepatocellular carcinoma (HCC).

The mechanisms of ALD include oxidative stress, disturbance of hepatocyte metabolism, liver inflammation, modifications in the regeneration process, and translocation of bacterial products from the gut microbiota into the portal blood stream. Most studies are derived from animal models, which typically show signs of mild ALD [3]. Recent translational research on human samples has identified several molecular drivers of ALD. Ethanol consumption triggers oxidative stress by generating reactive oxygen species (ROS) through its metabolites. This process leads to steatosis and liver inflammation, which are critical for the development of Alcohol-associated liver disease [4].

HSP70 family is a group of stress response proteins, mainly represented by HSP70-1 and GRP78. Serum levels of HSP70 and GRP78 were elevated in patients with non-alcoholic steatohepatitis, but decreased following metabolic surgery. Until recently, HSPs have been regarded exclusively as molecular chaperones playing a pivotal role in protein folding. However, literature data show detrimental effects of beneficial effects of reduced expression of GRP78. When placed on high-fat diet (HFD), GRP78 partial knockout (Grp78+/−) mice are more resistant to the development of hyperglycemia and hyperinsulinemia, as well as liver steatosis and inflammation of the white adipose tissue [5]. The protein and mRNA levels of GRP78 were upregulated in the liver tissues of the NASH rats. In addition, GRP78 expression levels were positively correlated with the levels of ALT, TNFα, CD68 and hepatocytic apoptosis [6]. Its potentially an important player in the development of nonalcoholic steatohepatitis.

However, what is the expression level of GRP78 in ALD patients and whether it is related to hepatic steatosis and cirrhosis. In this study, we aim to investigate the serum GPR78 levels and their correlation with biochemical indicators in ALD subgroups, in order to explore the potential application value and related mechanisms of GRP78 in ALD.

Materials and methods

ALD patients consulting at the outpatient clinics of the Hepatology Department of Xixi Hospital in Hangzhou between June 2022 and Oct 2024 were recruited in this case control study. ALD diagnosis was assessed during clinical examination including assessment for the presence of risk factors (obesity, type 2 diabetes, or metabolic syndrome), level of liver enzymes, and ultrasonic abdominal examination with absence of secondary causes for steatosis such as significant alcohol consumption. All patients with viral hepatitis, Wilson’s disease, α1-antitrypsin deficiency, autoimmune hepatitis, genetic hemochromatosis, primary biliary cirrhosis and biliary obstruction, and use of steatogenic drugs were excluded. Other exclusion criteria were recent gastrointestinal surgery and suffering from malignancy. According to the latest classification of ALD [7], a total of 122 patients were included in the study, consisting of 23 cases of AS, 49 cases of ASH, 19 cases of AH, and 31 case of AC. All patients completed a questionnaire and provided blood samples, and 103 of them also enrolled patients’ ultrasound-based TE performed. Healthy individuals from the general population (n = 28) were recruited at Medical Center. Controls had normal clinical and laboratory findings, and none of them had any evidence of fatty liver or previous liver disease. The study was conducted in accordance with the declaration of Helsinki, and the protocol was approved by the Research Ethics Committee at Xixi hospital. Written informed consent was provided by all participants.

Transient elastography

Liver stiffness management (LSM) and Controlled Attenuation Parameter (CAP) was assessed in all ALD patients by Experienced operators using TE (FibroScan; EchoSens, France), according to the manufacturer’s instructions [8]. With the patient lying in the dorsal decubitus position, the tip of the transducer probe was placed on the skin between the ribs over the right lobe of the liver. At this site, the distance between the skin and liver capsule (skin-capsular distance) assisted by a sonographic image was measured and an attempt was made to collect 10 valid LSMs. The median liver stiffness value was considered representative of the elastic modulus of the liver. The CAP measures ultrasonic attenuation in the liver at 3.5 MHz using signals acquired by the FibroScan^® M probe based on vibration-controlled transient elastography (VCTETM). The CAP is measured on validated measurements according to the same criteria used for LSM and on the same signals, ensuring that one obtains a liver ultrasonic attenuation simultaneously and in the same volume of liver parenchyma as the LSM. The final CAP value, which ranges from 100 to 400 decibels per meter (dB/m), is the median of individual measurements. Clinical evaluation was performed by the physician and patients were divided into categories based on the liver fibrosis stage (F0-F4) assessed by LSM. Stratification of ALD patients was done according to their LSM measurements, using the following cutoffs of 6.0 kPa for F ≥ F1 (early fibrosis), 8.0 kPa for F ≥ F2 (intermediate fibrosis), 9.0 kPa for F ≥ F3 (advanced fibrosis), and 12.5 kPa for F ≥ F4 (liver cirrhosis), while a normal LSM (F0 < 6 kPa) is thought to exclude liver pathology and liver fibrosis [9]. The liver steatosis stage (S0-S3) assessed by CAP, which the following cutoffs are 220 dB/m for S ≥ S1, 290 dB/m for S ≥ S2 and 339 dB/m for S ≥ S3. However, CAP below 220 dB/m (S = S0) ruled out steatosis [10].

Sample collection

Blood sample was performed in the morning, following not less than 12 h fasting period. 4 ml of peripheral blood were collected from patients and healthy participants by venipuncture for serum collection. Samples were centrifuged at 2000 rpm for 10 min at room temperature. Serum was aliquoted and stored at -80℃. The biochemical parameters of patients and healthy participants were detected, such as serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT) and alkaline phosphatase (ALP).

Measurement of serum GRP78 by ELISA

GRP78 were determined using a commercial sandwich ELISA kit (Human GRP78 ELISA Kit, Cat. No. D711329, Sangon Biotech, China) according to the manufacturer’s instructions. All reagents and samples were brought to room temperature (RT) for 20 min before use. The kit detects GRP78 with a pre‑coated anti‑GRP78 antibody, a biotin‑conjugated detection antibody, and HRP‑conjugated streptavidin.

A standard curve was prepared by two‑fold serial dilution of the recombinant GRP78 standard (40 ng/mL) with the provided standard/sample diluent to obtain concentrations of 40, 20, 10, 5, 2.5, 1.25, 0.63, and 0 ng/mL (blank). Serum samples were diluted 1:2 with the same diluent. 100 µL of each standard or diluted sample were added to duplicate wells, sealed, and incubated for 90 min at 37 °C. After discarding the liquid, 100 µL of biotin‑conjugated anti‑GRP78 antibody working solution were added to each well and incubated for 60 min at 37 °C. Wells were washed five times with 350 µL/well of washing buffer (1×, prepared from the supplied 25× concentrate), each time allowing a 1‑2 min soak. Subsequently, 100 µL of HRP‑conjugated streptavidin working solution were added and incubated for 30 min at 37 °C, followed by another five washes (300 µL/well, 30 s soak). Then, 90 µL of TMB substrate solution were added and incubated for 15 min at 37 °C in the dark. The reaction was stopped with 50 µL of stop solution, and the OD was measured immediately at 450 nm using a microplate reader.

The average OD of the blank was subtracted from all standard and sample readings. A standard curve was constructed using a four‑parameter logistic fit (OD vs. concentration), and GRP78 concentrations in samples were calculated from the curve. The assay range was 0.63‑40 ng/mL with a sensitivity of 0.38 ng/mL. Intra‑assay and inter‑assay CVs were < 10% as reported by the manufacturer. All samples were measured in duplicate in three independent experiments to assess reproducibility.

Statistical analysis

Continuous variables were expressed as means ± standard error. Two group comparisons were made using unpaired t-test or the Mann-Whitney test for continuous variables, based on the results of Kolmogorov-Smirnov normality test. Analysis of variance (ANOVA), independent sample test and Mann-Whitney U test or Kruskal-Wallis H test were used appropriately for comparisons of continuous variables between groups. Categorical variables are reported as counts and percentages; Chi-square or Fisher’s exact tests were used for comparing categorical factors. Spearman correlation coefficient was used to estimate the association of Serum GRP78 and variables of interest. Logistic regression analysis was used to assess the association between Serum GRP78 and LSM/CAP of TB and the severity of fibrosis/cirrhosis in ALD. To assess the diagnosis accuracy of combined biomarkers, a logistic regression analysis using the GRP78 and TE as covariates was first performed. The obtained predicted values were used to calculate the AUROC for the corresponding combination. Two-tailed P-values < 0.05 were considered statistically significant.

Results

Characteristics of ALD patients and healthy participants

Statistical analysis of 32 indicators among four groups of 122 ALD patients reveals and elucidates the pathophysiological characteristics of alcoholic liver disease progression(Table 1). Firstly, liver injury and inflammatory indicators show significantly elevated ALT and AST levels in the ASH and AH groups, indicating active hepatitis. Moreover, the AC group exhibits a notably increased AST/ALT ratio, associated with heightened mitochondrial AST release in alcoholic liver cirrhosis, potentially serving as a diagnostic indicator. Secondly, cholestasis and excretory function markers such as TBIL, TBA, and ALP are markedly elevated in the AC group, suggesting severe impairment of liver excretory function with advancing liver fibrosis and cirrhosis. Thirdly, liver synthetic function indicators including ALB, CHOL, APOA1, and CHE are significantly reduced in the AC group, indicating a comprehensive decline in liver synthetic function during the decompensated stage of liver cirrhosis. Additionally, metabolic and nutritional indicators demonstrate variations in lipid metabolism markers like TG and APOE among groups, reflecting the intricate influence of alcohol and its metabolites on lipid metabolism. Lastly, iron metabolism analysis reveals the highest SFe levels in the ASH group, potentially linked to the inflammatory response and iron metabolism disorder in alcoholic hepatitis.

Table 1.

the characteristics of three alcohol related liver disease groups

Characteristics	AS (n = 23)	ASH (n = 49)	AH (n = 19)	AC (n = 31)	P value
Gender, Male n (%)	23(100)	47(95.9)	18(94.7)	30(96.8)	0.812
Age (years)	52.91 ± 13.27	57.22 ± 11.91	56.84 ± 12.73	59.61 ± 9.83	0.148
BMI (kg/m2)	24.73 ± 4.73	25.54 ± 5.69	21.73 ± 3.97	22.47 ± 3.74	0.014
TBIL (µmol/L)	14.89(11.20 ~ 25.04)	18.84(12.91 ~ 34.74)	17.45(13.90 ~ 45.21)	27.23(19.75 ~ 63.90)	< 0.001
AST/ALT	1.15(0.83–1.78)	1.09(0.66–1.86)	1.23(0.58–2.41)	1.67(1.24–2.83)	0.003
ALT (U/L)	32.00(20.00 ~ 62.00)	111.00(48.00 ~ 206.00)	74.00(42.00 ~ 168.00)	43.00(30.00 ~ 80.00)	< 0.001
AST (U/L)	33.00(21.00 ~ 60.00)	100.00(59.00 ~ 182.00)	87.00(54.00 ~ 159.00)	80.00(42.00 ~ 145.00)	< 0.001
ALP (U/L)	100.76 ± 33.62	137.17 ± 84.12	137.58 ± 104.60	148.69 ± 73.92	0.028
GGT (U/L)	330.92 ± 299.20	438.92 ± 458.75	545.74 ± 693.51	287.92 ± 302.85	0.172
TBA (µmol/L)	11.90(3.70–37.80)	12.45(6.80–27.50)	12.90(6.30–90.60)	27.70(15.70–49.10)	< 0.001
LDH (U/L)	224.16 ± 76.34	223.83 ± 138.17	269.84 ± 221.53	228.62 ± 68.46	0.42
CHE (U/L)	6685.64 ± 2190.18	5501.38 ± 1674.88	5536.32 ± 2147.03	3885.85 ± 1478.15	< 0.001
ADA (U/L)	15.84 ± 8.42	13.93 ± 6.42	15.53 ± 7.09	24.27 ± 5.50	< 0.001
GFR (mL/min)	117.48 ± 31.01	101.21 ± 23.49	108.84 ± 29.71	119.38 ± 28.34	0.021
CHOL (mmol/L)	4.66 ± 1.30	4.41 ± 1.12	4.54 ± 1.58	3.47 ± 1.51	0.003
TG (mmol/L)	1.30(1.01–2.05)	1.41(0.92–2.14)	1.20(1.05–1.73)	0.95(0.68–1.16)	0.001
APOA1 (g/L)	1.22 ± 0.47	1.39 ± 0.43	1.42 ± 0.43	0.93 ± 0.42	< 0.001
APOB (g/L)	0.95 ± 0.43	0.75 ± 0.26	0.85 ± 0.32	0.63 ± 0.22	0.001
APOE (mg/L)	51.53 ± 20.24	59.08 ± 27.71	54.56 ± 31.64	35.43 ± 11.80	< 0.001
ALB (g/L)	39.83 ± 5.46	39.07 ± 6.31	39.68 ± 6.59	33.94 ± 6.32	< 0.001
GLB (g/L)	26.57 ± 4.87	28.83 ± 5.82	27.51 ± 4.87	31.86 ± 6.44	0.002
HGB (g/L)	142.57 ± 19.01	141.63 ± 19.67	133.68 ± 22.71	126.90 ± 23.69	0.016
AFP (µg/L)	3.55(1.71–5.74)	5.74(3.16–9.16)	4.74(3.59–7.13)	6.37(3.47–11.49)	0.002
SF (µg/L)	371.80(135.10-904.20)	490.40(279.20–1500.00)	877.90(311.60–1500.00)	314.90(117.40-900.20)	0.006
TRF (g/L)	1.79(1.55–2.23)	1.95(1.70–2.51)	1.82(1.48–2.18)	1.76(1.34–2.07)	0.079
SFe (µmol/L)	18.20(13.30–26.40)	31.20(21.40–40.20)	26.10(17.60–37.70)	24.70(14.30–30.30)	< 0.001
TIBC (µmol/L)	48.50(44.60–56.40)	51.30(45.00-65.80)	47.90(43.10–54.90)	46.30(37.80–51.70)	0.039
TSH (µIU/ml)	1.63(0.97–3.08)	1.62(1.08–2.41)	2.10(1.37–3.22)	2.38(1.14–3.89)	0.191
T3 (nmol/L)	1.47(1.44–1.60)	1.54(1.23–1.99)	1.57(1.38–1.94)	1.50(1.24–1.65)	0.562
FT3 (pmol/L)	4.89(4.53–5.27)	4.96(4.58–5.35)	5.13(4.52–5.40)	4.46(3.84–4.82)	< 0.001
T4 (nmol/L)	82.04(78.97-112.77)	107.44(95.46-134.98)	108.07(86.17-125.09)	99.93(89.75-133.73)	0.004
FT4 (pmol/L)	12.49(11.38–12.73)	12.11(10.82–12.89)	11.40(10.56–12.83)	12.07(10.32–14.16)	0.456

Normally distributed continuous variables are presented as mean ± SD and compared using one-way ANOVA test, categorical variables are presented as n (%) and compared using chi-square test. Non-normally distributed continuous variables are presented as median (95% confidence interval) and compared using Kruskal-Wallis test. Kolmogorov-Smirnov test for degerming the normality of Continuous variables. BMI, body mass index; TBIL, total bilirubin; ALT, alanine transferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, glutamyl transpeptidase; TBA, total bile acids; LDH, lactate dehydrogenase; CHE, cholinesterase; ADA, adenosine dehydrogenase; GFR, glomerular filtration rate; CHOL, total cholesterol; TG, triglycerides; APOA1, apolipoprotein A1; APOB, apolipoprotein B; APOE, apolipoprotein E; ALB, albumin; GLB, globulin; HGB, hemoglobin; AFP, alpha fetoprotein; SF, ferritin; TRF, transferrin; SFe, serum iron; TIBC, total iron-binding capacity; TSH, thyroid-stimulating hormone; T3, triiodothyronine; FT3, Free Triiodothyronine; T4,thyroxine; FT4, free thyroxine; P < 0.05 was considered statistically significant

This statistical analysis systematically quantified the dynamic changes of multiple biochemical indicators during the evolution of ALD from simple fatty liver (AS), steatohepatitis (ASH), alcoholic hepatitis (AH) to liver cirrhosis (AC). The P-value results clearly indicate that many indicators have significant differences between the AC group and other groups, and these differences objectively reflect the pathophysiological transition of liver function from compensation to decompensation. This analysis provides detailed data support for the clinical classification, disease assessment and prognosis judgment of ALD.

Differential expression of GRP78 in alcohol-related liver disease

The expression levels of GRP78 were analyzed across various stages of alcohol-related liver disease (ALD) to evaluate its potential as a biomarker for disease progression. As shown in Fig. 1A, GRP78 levels were significantly elevated in patients with ALD (n = 122) compared to healthy controls (HC, n = 28) (p = 0.0177), suggesting its upregulation in response to alcohol-induced liver injury. This finding indicates that GRP78, a key marker of endoplasmic reticulum stress, may play a role in the early pathogenesis of ALD.

Fig. 1.

GRP78 expression levels in healthy controls and liver disease subgroups. (A) Comparison of GRP78 levels between healthy controls (HC) and alcohol-related liver disease (ALD) patients. (B) GRP78 levels in HC, alcoholic non-cirrhosis (ANC), and alcoholic cirrhosis (AC). (C) GRP78 levels in HC, alcoholic steatosis (AS), alcoholic steatohepatitis (ASH), acute hepatitis (AH), and AC. Data are presented as mean ± SEM. Statistical significance is indicated by p-values (*p < 0.05, **p < 0.01, ***p < 0.001)

Further stratification of the ALD cohort revealed distinct patterns of GRP78 expression across disease stages (Fig. 1B). While no significant difference was observed between healthy controls (HC, n = 28) and patients with alcoholic cirrhosis (AC, n = 31) (p = 0.8713), the alcoholic non-cirrhosis group (ANC, n = 91) exhibited significantly higher GRP78 levels compared to both HC (p = 0.0024) and AC groups (p = 0.0163). These results suggest that GRP78 expression is particularly elevated during the non-cirrhotic phase of ALD, highlighting its potential as a marker for early-stage liver steatosis. The lack of significant difference between HC and AC groups further implies that GRP78 expression may not correlate with liver fibrosis or changes in liver hardness.

A detailed stage-specific analysis of GRP78 expression was conducted to further elucidate its role in ALD progression (Fig. 1C). Patients with alcoholic steatosis (AS, n = 23) showed significantly higher GRP78 levels compared to healthy controls (HC, n = 28) (p < 0.0001), as well as those with alcoholic steatohepatitis (ASH, n = 49) (p = 0.0040) and alcoholic cirrhosis (AC, n = 31) (p = 0.0007). Similarly, alcoholic hepatitis (AH, n = 19) demonstrated significantly elevated GRP78 expression relative to HC (p = 0.0013) and AC groups (p = 0.0336). In contrast, no significant differences in GRP78 levels were observed between HC and either ASH (p = 0.2218) or AC groups (p = 0.8713). These findings indicate that GRP78 upregulation is most pronounced during the early stages of ALD, particularly in alcoholic steatosis and alcoholic hepatitis, but diminishes in later stages such as steatohepatitis and cirrhosis.

Taken together, these results suggest that GRP78 may serve as a valuable biomarker for identifying patients at risk of early alcohol-related liver injury, particularly in the stages of alcoholic steatosis and alcoholic hepatitis. Its stage-specific expression pattern underscores its potential utility in the early diagnosis and monitoring of ALD progression.

Relationship between liver stiffness, steatosis, and GRP78 expression in ALD progression

Liver Stiffness Measurement (LSM) across ALD stages and fibrosis severity

Liver stiffness measurements (LSM) demonstrated a progressive increase across various stages of alcohol-related liver disease (ALD) (Fig. 2A). Patients with alcoholic steatosis (AS, n = 19) exhibited the lowest LSM values, significantly differing from those with alcoholic steatohepatitis (ASH, n = 44; p = 0.0068) and alcoholic cirrhosis (AC, n = 25; p < 0.0001). These findings align with the early-stage, minimal fibrosis seen in AS. In contrast, patients with alcoholic hepatitis (AH, n = 15) showed intermediate LSM values, which were significantly lower than those observed in AC (p = 0.0002) but did not differ significantly from AS (p = 0.1988) or ASH (p = 0.3809). This reflects the acute inflammatory injury and early fibrotic changes characteristic of AH. The highest LSM values were observed in AC patients (n = 25), with marked differences compared to AS (p < 0.0001) and ASH (p = 0.0005), indicating advanced fibrotic remodeling and structural disruption in cirrhosis. These results confirm that LSM serves as a reliable, non-invasive marker for disease severity in ALD, effectively distinguishing early-stage steatosis from advanced cirrhotic stages.

Fig. 2.

Correlation analysis between transient elastography (TE) indicators of the liver and serum GRP78 levels in patients with alcoholic liver disease (ALD). (A) Compare the liver stiffness (LSM) of patients at different stages of ALD.(B) Compare the LSM levels of patients with different stages of liver fibrosis (F0-F4).(C) Compare the serum GRP78 levels of the healthy control group (HC) with those of patients at different stages of liver fibrosis (F0-F4).(D) Analyze the correlation between the serum GRP78 level of patients and LSM.(E) Compare the hepatic fat content (CAP) of patients at different stages of ALD.(F) Compare the CAP levels of patients with different steatosis grades (S0-S3).(G) Compare the serum GRP78 levels of the healthy control group (HC) with those of patients with different steatosis grades (S0-S3).(H) Analyze the correlation between serum GRP78 levels and CAP. Data were expressed as mean ± SEM, and statistical significance was indicated as p < 0.05

Fibrosis stratification and LSM as a tool for disease staging

Further stratification by fibrosis stage (Fig. 2B) showed a stepwise increase in LSM correlating with the progression of liver fibrosis. Patients without fibrosis (F0, n = 10) had the lowest LSM values, while those with moderate fibrosis (F2, n = 9) and severe fibrosis (F3, n = 18) demonstrated progressively elevated LSM values (p < 0.0001 for all comparisons). Cirrhotic patients (F4, n = 43) exhibited the highest LSM values (p < 0.0001 vs. F0-F3), which are consistent with the structural alterations associated with end-stage disease. These findings underscore the role of LSM as a robust, non-invasive tool for fibrosis staging in ALD, offering valuable insight into the severity of liver damage and enabling accurate differentiation across fibrosis stages.

GRP78 expression and its relationship with fibrosis and steatosis

GRP78 expression exhibited limited association with fibrosis progression (Fig. 2C). Healthy controls (HC, n = 28) had the lowest GRP78 levels, while patients with early fibrosis (F0-F2, n = 42) did not show significant differences in GRP78 expression compared to HC (p = 0.5701–0.8692). Notably, a transient elevation in GRP78 expression was observed in patients with severe fibrosis (F3, n = 18; p = 0.0003 vs. HC), although levels returned to baseline in cirrhotic patients (F4, n = 43; p = 0.2629 vs. HC). This suggests that GRP78 upregulation may be associated with transient endoplasmic reticulum (ER) stress during active fibrogenesis, rather than being a sustained feature of advanced fibrosis. Moreover, a weak negative correlation between LSM and GRP78 expression was observed (r=-0.2064, p = 0.0365), reinforcing the dissociation between GRP78 expression and advanced fibrosis (Fig. 2D). These data indicate that GRP78 is more closely linked to the pathophysiology of early ALD stages, such as steatosis and inflammatory injury, than to late-stage fibrotic remodeling.

CAP, GRP78, and steatosis progression

The controlled attenuation parameter (CAP), which reflects hepatic fat content, exhibited significant variations across different stages of ALD. Patients with AS (n = 19) showed the highest CAP values, consistent with significant steatosis, while CAP values decreased progressively in patients with ASH (n = 44; p = 0.0058 vs. AS) and AH (n = 15; p = 0.0568 vs. ASH). The lowest CAP values were seen in AC patients (n = 25; p = 0.0046 vs. AS), likely due to the loss of fat in cirrhotic livers (Fig. 2E). Stratification by steatosis severity confirmed the sensitivity of CAP to fat accumulation, with significant increases in CAP values from S0 (no steatosis, n = 25) to S3 (severe steatosis, n = 4; p < 0.0001). GRP78 expression, however, showed a distinct pattern (Fig. 2F). While HC (n = 28) exhibited minimal GRP78 levels, patients with steatosis (S1-S3, n = 78) displayed significantly elevated GRP78 levels (p < 0.0001 vs. HC), peaking in severe steatosis (S3) (Fig. 2G). Despite this, no direct correlation between CAP and GRP78 was observed (r=-0.0092, p = 0.9268), suggesting that GRP78 upregulation is not directly associated with hepatic fat content (Fig. 2H). Instead, GRP78 may reflect ER stress triggered by lipid toxicity or oxidative damage in the liver, independent of fat accumulation.

Synthesis of findings

Collectively, the data highlight GRP78 as a dynamic biomarker in ALD, with elevated expression in early disease stages such as steatosis and alcoholic hepatitis. In contrast, GRP78’s relevance diminishes in advanced fibrosis and cirrhosis. While LSM and CAP are reliable markers for assessing structural damage and fat content, respectively, GRP78 offers unique insights into the early pathophysiological processes of ALD, such as ER stress and inflammatory injury. These findings underscore the potential of GRP78 as a diagnostic tool for identifying early alcohol-induced liver injury, emphasizing the need for multimodal biomarker approaches to improve ALD staging and monitor therapeutic interventions.

Serum GRP78 is associated with cirrhosis in adjusted analysis

Serum GRP78 levels exhibit significant variations at different stages of alcohol-related liver disease (ALD) (Fig. 3A). The serum GRP78 level in the fibrosis group (n = 50) is significantly higher than that in the healthy control group (Normal, n = 28, P = 0.0005) and the cirrhosis group (Cirrhosis, n = 43, P = 0.0113). However, no significant difference exists between the healthy control group and the cirrhosis group (P = 0.5926). These results indicate that serum GRP78 is upregulated in the early stage of alcohol-related liver fibrosis and decreases in the later stage of liver cirrhosis, suggesting its potential as an early diagnostic marker.

Fig. 3.

Serum GRP78 levels in the healthy control group, liver fibrosis group and liver cirrhosis group and their diagnostic value for liver cirrhosis. (A) Comparison of serum GRP78 levels between the healthy control group (Normal) and the liver fibrosis group as well as the liver cirrhosis group. Data were expressed as mean ± SEM, and statistical significance was marked as p value (p < 0.05). (B) ROC curves based on the diagnostic capabilities of GRP78, CAP, LSM and their combinations for liver cirrhosis

Further ROC curve analysis (Fig. 3B) reveals that the diagnostic efficacy of serum GRP78 alone for liver cirrhosis (AUC = 0.7965) surpasses that of CAP (AUC = 0.5070) but remains lower than that of LSM (AUC = 0.9301). Notably, when GRP78 combines with LSM, the diagnostic efficiency significantly improves (AUC = 0.9441). The AUC for LSM alone was 0.9301 (95% CI: 0.882–0.978), while the combined GRP78 + LSM model yielded an AUC of 0.9441 (95% CI: 0.901–0.987). DeLong’s test showed that the improvement was statistically significant (p = 0.041). Although the absolute increase is modest, the combination provides complementary information by capturing both structural (LSM) and cellular stress (GRP78) aspects of ALD progression. This finding indicates that molecular markers and imaging parameters possess complementary diagnostic value, and their combined application enhances the accuracy of risk stratification and clinical decision-making for liver fibrosis and cirrhosis in ALD patients.

Conclusion

This study systematically analyzes the expression characteristics of serum GRP78 at various stages of ALD and its relationship with LSM and CAP. Serum GRP78 shows significant upregulation in AS and AH, while it decreases in AC, indicating its primary involvement in the pathological processes of early liver injury. The combination of GRP78 and LSM significantly enhances the diagnostic efficiency for liver cirrhosis, suggesting that molecular markers and imaging parameters serve as complementary tools.

Serum GRP78 serves as a potential biomarker for early liver injury in ALD, facilitating early screening and intervention. Serum GRP78 levels showed a weak negative correlation with liver stiffness measurements. When combined with TE, it enhances the accuracy of risk stratification and clinical management for advanced liver fibrosis and cirrhosis. This study offers new insights into the early diagnosis, staging assessment, and application of multimodal markers for ALD.

Discussion

The pathogenesis of ALD is complex and multifactorial, mainly including acetaldehyde mediated toxicity, oxidative and endoplasmic reticulum (ER) stress, gut derived mediators, pro-inflammatory responses, and cell death [11].Exposure of cells to different stressors such as heat shock, toxic chemicals, infections, and alcohol leads to the induction of stress-mediated proteostasis chaperones, also known as heat shock proteins (HSPs) [12, 13]. GRP78 is an important member of the HSP70 family among heat shock proteins (HSPs). Previous studies have shown that alcohol-mediated ER stress induces another crucial ER chaperone, GRP78, in the live [14, 15]. Our research corroborates this finding.

The occurrence and progression of ALD and endoplasmic reticulum stress exhibit dynamic characteristics (Fig. 4). During alcohol metabolism, the CYP2E1 pathway generates a substantial amount of reactive oxygen species (ROS), which induces protein misfolding and disrupts endoplasmic reticulum homeostasis, subsequently activating the unfolded protein response (UPR) [16].GRP78, a crucial molecular chaperone and central regulatory protein in endoplasmic reticulum stress, plays a vital role in maintaining endoplasmic reticulum homeostasis by detecting the accumulation of unfolded proteins and activating the UPR. Mild endoplasmic reticulum stress activates the PERK, IRE1, and ATF6α pathways, initiating the adaptive UPR signaling pathway. Severe endoplasmic reticulum stress can induce maladaptive UPR, leading to cell death [17].

Fig. 4.

Nonlinear serum GRP78 dynamics reflect ER stress adaptation in Alcohol-Related Liver Disease. Serum GRP78 levels exhibit stage-specific changes during ALD progression. GRP78 is upregulated during early disease (AS/ASH), peaks during active fibrosis (F3), and declines in cirrhosis (F4/AC), reflecting ER stress activation and eventual adaptation failure. Liver stiffness (LSM) increases monotonically, while GRP78 shows a nonlinear pattern. Mechanistic pathway illustrates alcohol-induced ROS triggering unfolded protein accumulation, GRP78 release, and UPR activation via PERK/IRE1 signaling, with failed adaptation in cirrhosis

This study observes elevated serum GRP78 levels in AS and AH. In the early stages of the disease, ER stress activates rapidly, leading to significant upregulation of GRP78 expression, which may increase release into the serum. As the disease progresses to the fibrotic stage, particularly in F3, the study notes a further increase in GRP78 expression, suggesting a link to ongoing ER stress during active fibrogenesis. In the cirrhotic stage, GRP78 levels may gradually decline due to a persistently excessive protein folding load, intensified liver cell damage, and reduced protein synthesis capacity. The results demonstrate that serum GRP78 exhibits a downward trend during liver cirrhosis and shows a weak negative correlation with LSM, further supporting the proposed mechanistic hypothesis.

TE shows that LSM increases gradually with ALD staging, providing a better reflection of fibrosis severity. CAP peaks during ASH, then decreases gradually, reaching its lowest point in liver cirrhosis, aligning with the pattern of decreasing steatosis as the disease progresses. This finding supports a recent imaging review’s conclusion that LSM effectively distinguishes significant fibrosis from cirrhosis but lacks sensitivity to early pathological changes [18]. Notably, GRP78 shows no correlation with CAP, indicating that it cannot simply reflect liver fat content but rather more prominently display inflammatory responses and endoplasmic reticulum stress.

ROC analysis demonstrates that serum GRP78’s diagnostic efficacy for liver cirrhosis (AUC 0.7965) surpasses CAP (0.5070) but lags behind LSM (0.9301). The combined diagnostic model of GRP78 and LSM achieves an AUC of 0.9441, significantly outperforming individual indicators, indicating the complementary advantages of molecular markers and imaging parameters. The observed improvement, though statistically significant, was modest; however, the primary clinical value lies in elucidating the complementary roles of molecular and imaging markers rather than achieving a dramatic AUC elevation. Recent research suggests that integrating biomarkers like PRO-C3 into models (e.g., ALPACA) enables early identification and risk stratification [19].This aligns with the concept of the joint model in our study, further endorsing the viability of the “GRP78 + LSM” joint model. Moreover, utilizing TE for opportunistic screening in alcoholic outpatients enhances disease detection rates and patient adherence [20, 21]. Incorporating serum GRP78 testing in this context could potentially evaluate structural risks and inflammatory stress loads concurrently in a single visit, thereby optimizing the follow-up and intervention window.

This study presents several limitations. First, the sample size is limited, and the number of cases in certain subgroups is relatively small. The cross-sectional design does not allow for the dynamic observation of GRP78 changes. Second, liver biopsy serves as the gold standard for diagnosing hepatitis. However, due to its invasive and traumatic nature, some ALD patients in this study receive diagnoses solely through ultrasound examination, which has inherent limitations and cannot entirely eliminate the possibility of misdiagnosis. Future research must involve large-scale prospective cohorts and integrate multi-omics data to validate the dynamic monitoring value of GRP78 and its potential mechanistic role. Third, liver fibrosis staging was based on transient elastography (TE) rather than the gold standard of liver biopsy. TE measurements can be influenced by hepatic inflammation, cholestasis, congestion, and recent alcohol intake. Although we excluded patients with acute cholestasis or reported heavy drinking within one week prior to TE, and all examinations were performed by experienced operators following standardized protocols, the lack of histological confirmation remains a significant limitation. Future studies with biopsy‑proven fibrosis stages are needed to validate our findings. Fourth, the sample size is modest, with particularly small numbers in certain subgroups (AS, n = 23; AH, n = 19; fibrosis stage F3, n = 18). This limits statistical power and may render the stage‑specific GRP78 pattern vulnerable to sampling variability. Large‑scale, multi‑center prospective cohorts are necessary to confirm the observed nonlinear GRP78 trajectory. Fifth, this study is exploratory in nature and lacks several key validations. We did not perform liver tissue staining for GRP78 to confirm that serum levels correlate with hepatic expression. There is no independent external validation cohort, and the cross‑sectional design precludes longitudinal assessment of GRP78 dynamics during disease progression or after alcohol cessation. Additionally, serum GRP78 may originate not only from stressed hepatocytes but also from other cell types or from passive release during cell death. Therefore, the mechanistic link between serum GRP78 levels and intrahepatic ER stress remains inferential. Future studies should include paired liver tissue and serum samples, longitudinal follow‑up, and functional experiments (e.g., GRP78 knockdown or overexpression in ALD models) to establish causality.

In conclusion, serum GRP78 significantly increases during the early stage of ALD and serves as a potential molecular marker for early inflammation and ER stress in ALD. Its combination with LSM enhances the diagnostic efficacy for staging and liver cirrhosis, providing evidence for the construction of a non-invasive diagnostic system based on the combination of molecular and imaging, and has a good prospect for clinical transformation.

Author contributions

Y.L. and Y.S. were responsible for the conceptualization of the study and overall research design. Y.S. performed the main data analysis and drafted the original manuscript. Y.Y., Y.M., and Y.Z. contributed to data collection, clinical information acquisition, and literature review. X.J., W.H., and K.Y. assisted in laboratory measurements, including serum GRP78 detection and transient elastography data organization. Y.C., X.W., and C.C. participated in data curation, statistical analysis, and preparation of Figs and tables. Y.L. provided methodological guidance and supervised the experimental design. Y.S. and Y.L. contributed to critical revision of the manuscript for important intellectual content and overall supervision of the study. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Construction Fund of Key Medical Disciplines of Hangzhou (Grant No. 2025HZGF09) and Zhejiang Province Medical and Health Science and Technology Plan Project (Grant No. 2024KY212).

Data availability

The data presented in this study are available on request from the corresponding author due to privacy or ethical restrictions.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

The study was conducted in accordance with the Declaration of Helsinki and approved by Ethics Committee of Hangzhou Xixi hospital (Approval No: 杭西医伦审2023研第060号, Date: 26 Sept 2023).

Consent to participate

Written informed consent was obtained from all subjects involved in the study.