Obese Patients With Nonalcoholic Fatty Liver Disease Have an Increase in Soluble Plasma CD163 and a Concurrent Decrease in Hepatic Expression of CD163

Abstract

Background and aims: Macrophages play an important role in the development of nonalcoholic fatty liver disease (NAFLD) and its progression to nonalcoholic steatohepatitis (NASH). In this study, we investigated the hepatic expression of the macrophage scavenger receptor CD163 and the plasma level of its shed soluble form (sCD163) in patients with obesity and NASH, non-NASH NAFLD (NAFL), or healthy livers (no NAFLD).

Methods: Paired liver biopsies and plasma samples were collected from 61 patients with obesity (body mass index ≥35). Hepatic expression of CD163 was analyzed by immunohistochemistry and data-independent acquisition mass spectrometry, whilst plasma levels of sCD163 were determined by enzyme-linked immunosorbent assay and data-independent acquisition mass spectrometry. NAFLD stage and activity were assessed using the Kleiner fibrosis and NASH Clinical Research Network (NAS-CRN) scoring system.

Results: sCD163 turned out as a promising predictor of NASH with an area under the receiver-operating characteristic curve of 0.78 [0.65;0.92] (P = .0008). sCD163 increased with more severe NAFLD both in univariate (odds ratio [OR] = 3.31[1.80;6.11], P < .001) and multivariable ordinal logistic regression adjusting for NAFLD risk factors (OR = 2.02 [1.03;3.97], P = .042). On the other hand, hepatic expression of CD163 was negatively associated with more severe NAFLD in univariate ordinal logistic regression determined by immunohistochemistry (OR = 0.91[0.84;0.98], P = .015) and proteomics (OR = 0.13[0.02;0.80], P = .028). Taking NAFLD risk factors into account, hepatic expression of CD163 was only associated with the fibrosis stage (OR = 0.01 [0.0003;0.21], P = .004). Accordingly, hepatic CD163 surface expression and sCD163 were negatively correlated (rho = -0.478, P = .0001).

Conclusion: An increased plasma sCD163 and a concurrent decreased hepatic expression of CD163 are strongly associated with NAFLD in obese patients.

Keywords: CD163; Immunohistochemistry; NAFLD; NASH; Proteomics; Translational.

Figure 1

Soluble CD163 (sCD163) increases with NAFLD severity and predicts NASH. (A) Plasma levels of sCD163 increase with NAFLD severity (No NAFLD = 12, NAFL = 33, NASH n = 16). (B) Correlation between sCD163 determined by enzyme-linked immunosorbent assay and proteomics. (C) Receiver-operator characteristic (ROC) curve with sCD163 as a predictor for NASH (No NASH n = 45, NASH n = 16).

Figure 2

Association between hepatic CD163 and soluble (sCD163) with primary and secondary outcomes in univariate and 2 multivariable ordinal logistic regressions. Outcome variables were diagnostic scores according to the fatty liver inhibition of progression (FLIP) algorithm or NASH activity score (NAS) as well as histological features of NASH individually assessed according to NAS-CRN. Model 1: Univariate logistic regression with hepatic expressed CD163 and sCD163 as the explanatory variables. Model 2: Multivariable logistic regression with hepatic expressed CD163 or sCD163 as explanatory variable and sex as covariate. Model 3: Multivariable logistic regression with hepatic expressed CD163 or sCD163 as explanatory variable and NAFLD risk factors (sex, age, body mass index (BMI), insulin resistance (HOMA-IR), triglycerides and hypertension) as covariates. Odds ratios are represented for the association of hepatic expressed CD163 or sCD163 to the respective outcome variables. Empty symbols indicate violated proportional odds assumption and therefore not significant.

Figure 3

Hepatic expression of CD163 decreases with increased NAFLD severity. (A) Fold change of CD163 measured in liver biopsies using proteomics (No NAFLD n = 12, NAFL n = 31, and NASH n = 15). (B) Relative positively stained area of CD163 in liver biopsies using IHC quantified by QuPath (No NAFLD n = 12, NAFL n = 33, and NASH n = 16). (C) Normalized expression of CD163 at mRNA level in liver biopsies using bulk RNA sequencing from a publicly available library (GSE207310, No NAFLD n = 5, NAFL n = 15, and NASH n = 10). (D) Representative IHC staining of CD163 in liver specimens obtained by whole slide scanner NanoZoomer 2.0HT. NASH activity score (NAS) was determined by an expert pathologist and the positive area of CD163 staining in the total region of interest was quantified using QuPath. Scale bars represent 250μm. (E) Correlation between hepatic expression of CD68 and CD163 determined by IHC. (F) Correlation between CD163 as per IHC compared to proteomics.

Figure 4

(A) Correlation between hepatic expressed CD163 and sCD163 levels in plasma in patients with obesity. (B) Correlation between ADAM17-shed soluble TREM2 (sTREM2) and sCD163 levels in plasma (No NAFLD n = 11, NAFL n = 26, and NASH n = 15). Correlations were assessed by Spearman's rank correlation.

Figure A1

Gender difference in hepatic expression of CD163 determined by immunohistochemistry (A) and data-independent acquisition mass spectrometry (B). The difference between genders was tested by t test and the difference between NAFLD diagnosis stratified by gender was tested by ordinary two-way analysis of variance followed by Tukey multiple comparisons test.