Upregulation of hemoglobin expression by oxidative stress in hepatocytes and its implication in nonalcoholic steatohepatitis

Abstract

Recent studies revealed that hemoglobin is expressed in some non-erythrocytes and it suppresses oxidative stress when overexpressed. Oxidative stress plays a critical role in the pathogenesis of non-alcoholic steatohepatitis (NASH). This study was designed to investigate whether hemoglobin is expressed in hepatocytes and how it is related to oxidative stress in NASH patients. Analysis of microarray gene expression data revealed a significant increase in the expression of hemoglobin alpha (HBA1) and beta (HBB) in liver biopsies from NASH patients. Increased hemoglobin expression in NASH was validated by quantitative real time PCR. However, the expression of hematopoietic transcriptional factors and erythrocyte specific marker genes were not increased, indicating that increased hemoglobin expression in NASH was not from erythropoiesis, but could result from increased expression in hepatocytes. Immunofluorescence staining demonstrated positive HBA1 and HBB expression in the hepatocytes of NASH livers. Hemoglobin expression was also observed in human hepatocellular carcinoma HepG2 cell line. Furthermore, treatment with hydrogen peroxide, a known oxidative stress inducer, increased HBA1 and HBB expression in HepG2 and HEK293 cells. Importantly, hemoglobin overexpression suppressed oxidative stress in HepG2 cells. We concluded that hemoglobin is expressed by hepatocytes and oxidative stress upregulates its expression. Suppression of oxidative stress by hemoglobin could be a mechanism to protect hepatocytes from oxidative damage in NASH.

Figure 1. Increased expression of HBA1 and HBB in NASH livers.

Quantitative real time PCR was performed with the same NASH (n = 12) and control (n = 5) samples used for microarray analysis. GAPDH was used as an endogenous control. Hemoglobin levels in controls were normalized to 1. Data were expressed as Mean ± SD. * and ** represent P<.05 and P<.01, respectively.

Figure 2. HBA1 and HBB expression in the hepatocytes of NASH livers.

Immunofluorescence staining was performed on cryosections of NASH liver biopsies. Radixin is expressed in liver bile canaliculi. Radixin staining was used to identify hepatocytes. Images were recorded with an LSM510 beta laser scanning microscope (Zeiss). Cytoplasmic staining of HBA1 (A&B) and HBB (C&D) in hepatocytes of NASH livers were detected, respectively. Bars in A&C: 50 µm, bars in B&D: 10 µm.

Figure 3. Expression of HBA1 and HBB in HepG2 cells.

A, HBA1 and HBB mRNA were detected by reverse transcription PCR. PCR products were separated on 2% agrose gel and visualized with ethidium bromide. PCR products were further confirmed by sequencing. B, HBA1 protein was detected by Western blot analysis. Whole cell lysates from HepG2 cells transfected with control or HBA&HBB expression plasmid were analyzed for HBA1 protein. Blood cells were used as the positive control. Beta actin was probed as a loading control.

Figure 4. Upregulation of hemoglobin expression by oxidative stress.

A, Induction of hemoglobin expression by H₂O₂ in a dose dependent manner. HepG2 cells were treated with a series of concentrations (0, 0.25, 0.5, and 1 mM) of hydrogen proxide (H₂O₂) for 8, 16, or 24 hours. Relative HBA1 mRNA levels were determined by qRT-PCR. Data were expressed as Mean ± SD (n = 3). HBA1 level in control (0mM, 8hr) was normalized to 1. One way ANOVA and Tukey post hoc tests for multiple comparisons were used for analyzing the differences among treatments with different doses. **: p<0.01; ***: p<0.001, when compared to 0 mM treatment. B, HBA1 proteins were induced by H₂O₂ treatment. Whole cell lysates from HepG2 cells treated with or without H₂O₂ (1 mM, 48 hours) and blood cells were analyzed for HBA1 protein. Actin was probed as a loading control. C, HBB mRNA was upregulated by H₂O₂ treatment. HepG2 cells were treated with H₂O₂ (1 mM) for 48 hours. Reverse transcription PCR products were separated on 2% agrose gel and visualized with ethidium bromide. D, HBA1 and HBB mRNA in HEK293 cells were induced by H₂O₂ treatment. HEK293 cells were treated with H₂O₂ (1 mM) for 24 hours and hemoglobin mRNA levels were determined by qRT-PCR. Hemoglobin levels in controls were normalized to 1. Data were expressed as Mean ± SD (n = 3).

Figure 5. Reduction of oxidative stress by hemoglobin overexpression.

HepG2 cells were transfected with control or HBA&HBB expression plasmid for 48 hours and then treated with H₂O₂ (1 mM, 5 minutes) or mock treated. Intracellular H₂O₂ levels were measured by flow cytometry. A, Representative histogram plots of different treatments. Plots 1 and 2: control and HBA&HBB plasmid without H₂O₂ treatment, respectively. Plots 3 and 4: control and HBA&HBB plasmid with H₂O₂ treatment, respectively. B, Quantitative analysis of mean fluorescence intensity. Data were expressed as Mean ± SD (n = 4).