- Email: [email protected]
Plasma soluble human leukocyte antigen-G expression is a potential clinical biomarker in patients with hepatitis B virus infection
Human Immunology 72 (2011) 1068-1073
Contents lists available at SciVerse ScienceDirect
Plasma soluble human leukocyte antigen-G expression is a potential clinical biomarker in patients with hepatitis B virus infection Wei-Wu Shi a, Aifen Lin b, Dan-Ping Xu a, Wei-Guang Bao b, Jian-Gang Zhang b, Shi-Yong Chen c, Jun Li d, Wei-Hua Yan a,* a
Medical Research Center, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China Human Tissue Bank, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China c Department of Laboratory Medicine, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China d Department of Blood Bank, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China b
A R T I C L E
I N F O
Article history: Received 18 December 2010 Accepted 21 June 2011 Available online 1 July 2011
Keywords: HLA-G Hepatitis B virus Infection
A B S T R A C T
The significance of upregulated soluble human leukocyte antigen-G (sHLA-G) expression under various pathologic conditions has been discussed. In this study, we evaluated the potential significance of plasma sHLA-G expression in patients with hepatitis B virus (HBV) infection. The study included 90 acute hepatitis B patients (AHB), 131 chronic hepatitis B patients (CHB), 152 resolved hepatitis B individuals (RHB), and 129 normal controls. sHLA-G were determined using enzyme-linked immunosorbent assay. A receiver operating characteristic (ROC) curve was used to evaluate the feasibility of plasma sHLA-G as a biomarker for distinguishing patients with HBV infection. sHLA-G levels in AHB (median, 193.1 U/mL; p ⬍ 0.001), CHB (median, 324.6 U/mL; p ⬍ 0.001), and RHB (median, 14.8 U/mL; p ⫽ 0.006) patients was much higher than that in normal controls (median, 9.0 U/mL). A significant difference for sHLA-G levels was also observed between patients with HBV infection (AHB vs CHB, AHB vs RHB, and CHB vs RHB; all p ⬍ 0.001). The area under the ROC curve for sHLA-G levels was 1.000 (p ⬍ 0.001) for AHB, 0.993 (p ⬍ 0.001) for CHB, and 0.604 (p ⫽ 0.003) for RHB patients versus normal controls, respectively. Data also indicated that the percentage of CD4⫹CD25⫹FoxP3⫹ T regulatory cells and HLA-G⫹CD14⫹ monocytes was significantly increased in AHB and CHB patients compared with normal controls (all p ⬍ 0.001). Our findings indicated that induction of HLA-G expression may play a role in HBV immune evasion and sHLA-G levels could be a useful biomarker in HBV infection. 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.
1. Introduction Hepatitis B virus (HBV), a small DNA virus, belongs to the Hepadnaviridae family [1]. HBV is one of the most common serious infectious agents for humans and could result in acute and chronic hepatitis, cirrhosis, end-stage liver disease, and hepatocellular carcinoma worldwide [2]. Extensive research on HBV has been carried out for more than half a century, resulting in remarkable advances in our understanding of the virus and in development of means for diagnosis, prevention, and control of the HBV-associated diseases. In addition to a medical history and physical examination, evaluation of HBV infection usually requires serologic, biochemical, and virologic tests to confirm diagnosis, treatment responses, and disease progression monitoring [3]. Virologic and biochemical markers such as HBV antigens and their corresponding antibodies, serum HBV DNA, viral genotype, and serum alanine aminotransferase (ALT) applied in the diagnosis and monitoring of HBV disease clearly are useful for the * Corresponding author. E-mail address: [email protected] (W.-H. Yan).
evaluation of patients with HBV infection in clinical practice [4]. Beyond this, other potential useful markers are definitely needed to expand the understanding of HBV pathogenesis and to predict outcome and diagnosis for virus infection. Human leukocyte antigen-G (HLA-G) is a nonclassic HLA class I molecule that was first identified as being selectively expressed by choriocarcinoma cells and later observed under a broad spectrum of pathologic conditions [5]. Unlike classic HLA class I antigens, 7 HLA-G isoforms, including 4 membrane-bound (HLA-G1–G4) and 3 soluble HLA-G (HLA-G5–G7) proteins, are generated, and another soluble form of HLA-G could be produced by shedding of the proteolytically cleaved cell surface expressed HLA-G1 (sHLA-G1) [6,7]. Both isoforms of HLA-G bear similar immune-suppressive properties via binding to receptors such as immunoglobulin-like transcript-2/CD85j and -4/CD85d and KIR2DL4/CD158d expressed on various immune-competent cells [8]. Apart from its initial relevance in fetal–maternal immunotolerance, HLA-G has now been reported to be involved in tumor cell immune escape, inflammatory, autoimmune, and infectious diseases, and allograft acceptance [9].
0198-8859/11/$32.00 - see front matter 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.humimm.2011.06.012
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
In viral infection, induction of sHLA-G expression by virusinfected cells was proposed as a mechanism that helps viruses to subvert host antivirus defenses. Roles of induced HLA-G expression after infections with certain influenza A virus strains, human cytomegalovirus virus (HCMV), human immunodeficiency virus (HIV), neurotropic viruses such as rabies virus, and hepatitis B virus (HBV) were discussed [10 –14]. Beside the aims of elucidating the roles of HLA-G in immune regulation and its clinical relevance under pathologic conditions, studies that focus on potential applications of HLA-G expression in clinical practice may represent one of the new perspectives in HLA-G research [15]. Previous studies indicated that sHLA-G could be a tumor marker for malignant versus benign ascites in breast and ovarian cancer, and HLA-G5 as a good predictor of outcome in septic shock was also reported [16,17]. In the present study, plasma sHLA-G levels in different HBV infection status, including acute (AHB), chronic (CHB), and resolved hepatitis B (RHB) infected patients, as well as in normal controls, were measured, and its potential feasibility in distinguishing the outcome of HBV-associated diseases was analyzed. Moreover, HLA-G expression on CD14⫹ monocytes and the percentage of CD4⫹CD25⫹FoxP3⫹ T regulatory cells (Tregs) was determined.
1069
and hepatitis B envelope (HBe) antigen seroconversion for at least 6 months before sample collection. No patients received anti-HBV agent or steroid 6 months before sampling. Samples from 129 healthy individuals without HBV infection history and with similar age and sex characteristics were taken as controls. Concurrence of hepatitis C virus or HIV infections was excluded for all enrolled individuals. The study protocol was approved by the ethics committee of the Taizhou Hospital of Zhejiang Province, Wenzhou Medical College. The baseline characteristics of subjects enrolled in the study are presented in Table 1. 2.2. Virological assessments
2. Materials and methods
HBV serum markers were determined using commercial enzyme immunoassay kits (AXSYM System, Abbott, Wiesbaden, Germany). HBV DNA was extracted from serum samples and quantified using a commercial polymerase chain reaction diagnostic kit with detection limit of 100 copies/mL (Sybio, Shanghai, China). Serum ALT, aspartate aminotransferase (AST), alkaline phosphatase, ␥-glutamine transpeptidase, albumin, and total, direct, or indirect bilirubin concentration levels were measured using an autobiochemical analyzer (Olympus AU 5400, Olympus Corp., Tokyo, Japan).
2.1. Study setting and participants
2.3. sHLA-G enzymed-linked immunosorbent assay (ELISA)
Consecutive blood samples were collected from 90 patients with AHB and 131 patients with CHB. One hundred fifty-two individuals with RHB were screened from blood donors. Specimens were collected from the Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical College. The diagnosis was complied with the diagnostic criteria of the 2000 Xi’an Viral Hepatitis Management Scheme issued by the Chinese Society of Infectious Diseases and Parasitology and the Chinese Society of Hepatology of the Chinese Medical Association [18]. Briefly, standards for diagnoses of AHB include displaying hepatitis B surface Ag (HBsAg)-negative conversion within 6 months after the initial onset of symptoms caused by HBV infection and being anti-HBc immunoglobulin M (IgM) positive. An HBV carrier with chronic hepatitis exhibits a clinical course of hepatitis B for ⬎6 months and may have exhibited symptoms or signs of hepatitis and abnormal hepatic function on this occasion. An additional group of 152 individuals with resolved HBV infections was included. The inclusion criteria for resolved subjects included a history of HBV infection, normal liver biochemistry, undetectable clinical activity, and HBV DNA, as well as HBsAg
Plasma sHLA-G concentrations were determined with the sHLAG-specific ELISA kit (sHLA-G kit; Exbio, Prague, Czech Republic), which measures sHLA-G1 and HLA-G5. Each sample (100 L) was measured in triplicate. The optical densities were measured at 450 nm (Spectra Max 250, Molecular Devices, Sunnyvale, CA). The final concentration was determined by optical density according to the standard curves (range, 0 –125 U/mL). When the concentration exceeded 125 U/mL, diluted samples were used and dilution factors were considered to calculate the sHLA-G concentration. The detection limits were 1 U/mL. Details of the performance were according to the manufacturer’s instruction. 2.4. Flow cytometry analysis Flow cytometry analysis was performed with a standard protocol according to the previous study [10]. Briefly, peripheral blood mononuclear cells (PBMC) obtained from HBV-infected patients and normal controls were gated with allophycocyanin-labeledanti-CD14 (BD Biosciences, San Jose, CA) for monocytes, and HLA-G expression was analyzed. Briefly, 1 ⫻ 106 mononuclear cells were
Table 1 Baseline characteristics of the study populations Variables
Age (years)a Sex (male/female) HBsAg positive (%) HBsAb positive (%) HBeAg positive (%) HBeAb positive (%) HBcAb positive (total, %) HBcAb (IgM) positive (%) ALT (IU/L) AST (IU/L) HBV DNA (copies/mL, log10)a TBil (mol/L) DBil (mol/L) IBil (mol/L) Albumin (g/L)
Controls
AHB
n ⫽ 129
n ⫽ 90
39.6 ⫾ 11.2 75/54 0 0 0 0 0 0 ⬍40 — — — — — —
43.1 ⫾ 12.7 54/36 90 (100%) 0 90 (100%) 0 90 (100%) 90 (100%) 1,117.0 ⫾ 686.9 647.1 ⫾ 417.1 6.8 (2.1–8.9) 49.2 ⫾ 45.0 19.1 ⫾ 24.8 24.5 ⫾ 25.9 39.8 ⫾ 9.9
CHB Correlate to sHLA-G (p)
0.626 0.743 0.203 0.844 0.960 0.633 0.479
n ⫽ 131 44.8 ⫾ 14.3 71/60 131 (100%) 0 41 (30.1%) 76 (55.9%) 127 (97.1%) 0 163.9 ⫾ 79.1 54.1 ⫾ 78.0 3.8 (2.0–9.3) 19.7 ⫾ 19.3 5.7 ⫾ 8.5 13.6 ⫾ 11.0 41.6 ⫾ 6.6
RHB Correlate to sHLA-G (p)
0.860 0.773 0.648 0.469 0.589 0.779 0.763 0.828 0.746 0.203
n ⫽ 152 42.2 ⫾ 13.1 89/63 0 123 (80.9%) 0 0 128 (84.2%) 0 ⬍40 — — — — — —
Correlate to sHLA-G (p)
0.672
0.542
AHB ⫽ acute hepatitis B patients; CHB ⫽ chronic hepatitis B patients; RHB ⫽ resolved hepatitis B individuals; ALT ⫽ alanine aminotransferase; AST ⫽ aspartate aminotransferase; HBV ⫽ hepatitis B virus; TBil ⫽ total bilirubin; DBil ⫽ direct bilirubin; IBil ⫽ indirect bilirubin; ⫽, not detected. a Expressed as median (range).
1070
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
incubated in 200 L of 1.0% phosphate-buffered saline/bovine serum albumin in a cold room for 30 minutes in the dark with 10 L (1 mg/mL) PE–anti-HLA-G MEM-G/9 (IgG1, Serotec, Oxford, UK), which reacts with the native form of human HLA-G1 on the cell surface, and an isotype IgG was added as a negative control (BD Biosciences). Intracellular Foxp3 stainning was detected using the eBioscience Foxp3 staining kit (eBioscience, San Diego, CA). Briefly, PBMC were first surface labeled with PE–CD4 (BD Biosciences) and FITC–CD25 (BD Biosciences). Then, intracellular analysis of FoxP3 was performed after fixation and permeabilization according to the manufacturer’s recommendation (eBioscience). Flow cytometry was performed on a FACSCalibur and analyzed using the CellQuest (BD Biosciences) software package. 2.5. Statistical analysis Statistical analysis was performed with SPSS 13.0 software (SPSS, Inc., Chicago, IL). Normality of continuous numeric data was analyzed by one-sample Kolmogorov–Smirnov test. Comparison of the percentage of HLA-G-expressing monocytes, the percentage of Tregs, and sHLA-G levels between different groups was performed using the Mann–Whitney U test. The feasibility of using sHLA-G as a potential biomarker for discrimination of HBV-infected patients from healthy individuals was assessed using receiver operating characteristic (ROC) curve analysis. The areas under the ROC curve were calculated and subjected to statistical analysis. The 95% confidence intervals were calculated for specificity and sensitivity. A two-sided p value ⬍ 0.05 was considered statistically significant.
ical parameters such as age, sex, HBV DNA copies, ALT, AST, alkaline phosphatase, ␥-glutamine transpeptidase, albumin, and total, direct, or indirect bilirubin concentration (Table 1). 3.2. ROC curve analysis for plasma sHLA-G as a biomarker for patients with HBV infection ROC curves were used to evaluate the performance of sHLA-G in predicting the outcome of patients with HBV infection. The area under the ROC curve for sHLA-G levels was 1.000 (95% CI 0.999 – 1.000, p ⬍ 0.001) for AHB versus normal controls (Fig. 2A). The area under the ROC curve for sHLA-G levels was 0.993 for CHB patients versus normal controls (Fig. 2B). The area under the ROC curve was 0.604 (95% CI 0.538 – 0.671, p ⫽ 0.003) for RHB patients versus normal controls (Fig. 2C). Given 100% specificity at the cutoff of corresponding plasma sHLA-G levels, the detection sensitivity was 97.8, 91.6, and 3.3% for AHB, CHB, and RHB patients, respectively. 3.3. Cell surface HLA-G expression in monocytes
3. Results
Cell surface HLA-G expression was analyzed on the CD14⫹ monocytes. The percentage of HLA-G-positive CD14⫹ monocytes was a median of 7.72% (n ⫽ 26; range: 1.12– 48.62%) in AHB patients, 4.62% (n ⫽ 23; range: 0.82–24.97%) in CHB patients, 1.93% (n ⫽ 31; range: 0.19 –11.87%) in RHB patients, and 1.62% (n ⫽ 36; range: 0.30 –9.58%) in normal controls, respectively. When compared, the percentage of HLA-G⫹CD14⫹ monocytes was much higher in AHB and CHB than in normal controls (all p ⬍ 0.001), and the percentage of HLA-G⫹CD14⫹ monocytes in AHB was much higher than in CHB patients (p ⫽ 0.021, Fig. 3A).
3.1. Plasma sHLA-G expression in study populations
3.4. The frequency of Tregs in HBV-infected patients
Plasma sHLA-G levels in 90 AHB, 131 CHB patients, and 152 RHB individuals and 129 normal controls were determined by ELISA. The concentration of the plasma sHLA-G was a median of 193.1 U/mL (range, 69.4 –504.3 U/mL) for AHB patients, 324.6 U/mL (range, 23.6 –786.0 U/mL) for CHB patients, 14.8 U/mL (range, 3.5– 86.5 U/mL) for RHB individuals, and 9.0 U/mL (range, 3.6 –71.6 U/mL) for normal controls. sHLA-G levels were dramatically increased in AHB, CHB, and RHB individuals when compared with that in normal controls (all p ⬍ 0.001). sHLA-G levels were also markedly different among HBV-infected individuals (AHB vs CHB, AHB vs RHB, and CHB vs RHB, all p ⬍ 0.001; Fig. 1). Furthermore, no correlation was observed between sHLA-G levels and patient clin-
sHLA-G comparison: Normal vs RHB, p=0.006; Normal vs AHB, p<0.001; Normal vs CHB, p<0.001; AHB vs CHB, p<0.001; AHB vs RHB, p<0.001.
Fig. 1. Distribution and comparison of plasma soluble human leukocyte antigen-G (sHLA-G) between study groups. The bar in the box represents the median. Comparison of sHLA-G levels between groups was performed using the Mann–Whitney U test. AHB, acute hepatitis B patients; CHB, chronic hepatitis B patients; RHB, resolved hepatitis B individuals.
Moreover, our data indicated that the percentage of CD4⫹ CD25⫹FoxP3⫹ Tregs was a median of 5.08% (n ⫽ 21; range, 1.69 – 9.58%) in AHB, 3.84% (n ⫽ 23; range: 0.64 – 8.76%; p ⫽ 0.098) in CHB, 2.98% (n ⫽ 21; range: 0.66 – 6.89%) in RHB patients and 2.76% (n ⫽ 34; range, 1.39 –5.01%) in normal controls, respectively. The proportion of Tregs in AHB (p ⬍ 0.001) and CHB patients (p ⫽ 0.016) was significantly increased compared with that in normal controls; however, no significance was observed between the AHB and CHB patients (p ⫽ 0.098, Fig. 3B). 4. Discussion Our study provides evidence of a marked elevation of plasma sHLA-G concentration in patients with HBV infection. Most important, increased sHLA-G concentrations could be a biomarker for the outcome of HBV infection. Furthermore, this result may be important in understanding the immune-suppressive roles of HLA-G in HBV pathogenesis. HBV is one of the most common serious infectious viruses for humans. More than one third of the world’s population has been infected with HBV and it is estimated that there are 350 million persistent carriers of HBV worldwide [19]. Most adults infected with the virus recover, but 5–10% are unable to clear the virus and become chronically infected [20]. HBV produces several antigens that can be detected in the blood and disappear as the body produces antibodies against them. The patterns of these and other markers provide clues to the phase of infection. HBsAg and HBV DNA are often the first detectable markers of acute infection, appearing before the onset of symptoms or before elevation of ALT levels [4]. Chronic HBV infection was defined as surface antigen persisting longer than 6 months. HBeAg is a marker of HBV replication and infectivity. In chronic infection, HBeAg can persist for years or decades. HBcAg cannot be detected in the serum, but antibodies against it can, first IgM and later IgG [3]. These serologic and virologic markers of HBV infection have been well accepted as useful tools in the diagnosis and monitoring of HBV-associated diseases.
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
A 52&&XUYH
HLA-G+CD14+ monocytes (%)
A
6HQVLWLYLW\
60
50
40
30
20
10
AUC: 1.000
Normal controls
B
12
6SHFLILFLW\
CD4+CD25+FoxP3+ Treg (%)
52&&XUYH
6HQVLWLYLW\
HLA-G+monocyte comparison: Normal vs RHB, p=0.758; Normal vs AHB, p<0.001; Normal vs CHB, p=0.001; AHB vs CHB, p=0.021; AHB vs RHB, p<0.001.
0
B
1071
10
8
RHB
AHB
CHB
AHB
CHB
Treg comparison: Normal vs RHB, p=0.782; Normal vs AHB, p<0.001; Normal vs CHB, p=0.016; AHB vs CHB, p=0.098; AHB vs RHB, p=0.005.
6
4
2
0
AUC: 0.993
6SHFLILFLW\
C
52&&XUYH
Normal controls
RHB
Fig. 3. Flow cytometry anaylsis of human leukocyte antigen-G (HLA-G)⫹CD14⫹ monocytes and CD4⫹CD25⫹FoxP3⫹ regulatory T cells (Tregs). (A) Comparison of the proportion of HLA-G⫹ monocytes among normal controls, resolved hepatitis B individuals (RHB), acute hepatitis B patients (AHB), and chronic hepatitis B patients (CHB). (B) Comparison of the frequencies of CD4⫹CD25⫹FoxP3⫹ Treg among normal controls, RHB, AHB, and CHB patients. The solid line represents the median and comparison was analyzed using the Mann–Whitney U test.
6HQVLWLYLW\
AUC: 0.604
6SHFLILFLW\ Fig. 2. Receiver operating characteristic (ROC) curve analysis of plasma soluble human leukocyte antigen-G (sHLA-G) levels (U/mL) for predicting the outcome of patients with hepatitis B virus (HBV) infection. Area under the curve (AUC): (A) 1.000 (95% CI 0.999 –1.000, p ⬍ 0.001) for acute hepatitis B patients, (B) 0.993 (95% CI 0.988 – 0.999, p ⬍ 0.001) for chronic hepatitis B patients, and (C) 0.604 (95% CI 0.538 – 0.671, p ⫽ 0.003) for resolved hepatitis B patients versus normal controls, respectively.
In the present study, our data revealed that plasma sHLA-G was dramatically increased in acute and chronic hepatitis B patients as well as resolved hepatitis B individuals and suggested that plasma sHLA-G expression could be a useful biomarker in predicting outcome in patients with hepatitis B. Induction of HLA-G expression and its potential roles in certain types of virus infection were addressed in previous studies [9]. We recently reported that both the PBMC surface HLA-G and the plasma sHLA-G levels were dramatically induced in active HCMV patients [11]. Another study demonstrated that both cell surface HLA-G and soluble HLA-G could be produced during viral reactivation in macrophages and that bronchoalveolar macrophages from patients with acute HCMV pneumonitis also express HLA-G molecules [21]. Moreover, only soluble HLA-G could be induced in CMVinfected U-373 MG astrocytoma cells through the cooperative action of the 2 IE1 pp72 and IE2 pp86 products [22]. In HIV-infected patients, HLA-G expression was upregulated in CD8⫹ T cells and monocytes, and long-term progressor patients expressed higher levels of plasma sHLA-G than long-term nonprogressor patients [12,23]. In neurotropic viral infections, herpes simplex virus type 1 and rabies virus upregulate the neuronal expression of HLA-G isoforms, both in infected cells and in neighboring uninfected cells
1072
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
[13]. Furthermore, upregulation of HLA-G gene expression was observed in a human neuronal line (NT2-N cells) after in vitro infection by a highly pathogenic strain of rabies virus or a strain of herpes virus inducing neuron latency. Importantly, HLA-G expression was restricted to rabies virus–infected neurons [24]. These findings suggested that upregulation of immunotolerant HLA-G expression in virus infection might be a potential virus escape mechanism from host antiviral immune responses. HBV has evolved a multitude of strategies to subvert host immune surveillance and responses by both the innate and the adaptive arms of the immune system, where dendritic cells, natural killer (NK) and NKT cells, and virus-specific CD4⫹ and CD8⫹ T cells are crucial for restricting viral replication [25]. The biologic role of HLA-G in HBV infection remains to be elucidated. However, both membrane-bound and soluble HLA-G were well described as tolerogenic molecules exhibiting strong immunosubversive activities against both innate and adaptive immune responses; recent studies also indicated that HLA-G was involved in inducing suppressor cells such as Tregs, dendritic cells, and NK cells, providing a long-term immune modulatory function [26]. Consequently, a decreased immune function induced by HLA-G leads to an advantage for the virus to subvert host defenses. It should be noted that the immune response in chronic carriers is weak or undetectable, resulting in an inability to eliminate the virus efficiently. This finding was supported by evidence that peripheral cytotoxic T lymphocyte (CTL) responses in chronically infected patients are difficult to detect and HBV-specific CTLs are present at a low frequency in the livers of chronically infected patients [27,28]. In this study, our results indicated that plasma sHLA-G, cell surface HLA-G expression in monocytes, and the percentage of the CD4⫹CD25⫹Foxp3 Treg cells were dramatically increased in patients with AHB and CHB. Unexpectedly, our data revealed that sHLA-G among AHB patients was significantly lower than that among CHB patients, but the reason remains unknown. Because peripheral monocytes are the main source of sHLA-G, the mechanisms underlying the retention of sHLA-G production in monocytes during HBV infection require further investigation [7]. High sHLA-G levels among CHB patients may result in the inhibition of CTL cytolysis and in induction of a CD8⫹ regulatory T-cell population, as previously reported [29]. In addition, sHLA-G has the ability to induce apoptosis of CD8⫹ T cells via binding to CD8 and through a Fas/FasL pathway [30,31]. Furthermore, during chronic HBV infection, the peripheral blood HLA class II–restricted CD4⫹ T-cell response to all viral antigens such as HBcAg and HBeAg is much less vigorous than that in patients with acute hepatitis [32,33]. The impaired CD4⫹ T-cell responses in CHB patients may be supported by the current findings that the sHLA-G level was positively associated with CD4⫹CD25⫹ CD152⫹ natural Treg frequency, and functionally, sHLA-G could induce CD4⫹CD25highFoxP3⫹ Tregs [34,35]. A recent study demonstrated that both soluble and membrane-bound HLA-G is frequently and largely expressed in hepatocytes, biliary epithelial cells, or both of chronic HBV-infected individuals, indicating that HLA-G is expressed in most cases of chronic HBV infection in all stages and may play a role in the persistence of HBV infection [14]. Thus, a dominant cause of viral infection could be the existence of a weak antiviral immune response induced by sHLA-G; however, the underlying mechanisms associated with sHLA-G in viral infection, such as HBV infection, require further investigation. Another intriguing result of the present study is that measurement of plasma sHLA-G is a highly sensitive technique to monitor the outcome of HBV infection. Studies on malignancies indicated that HLA-G was a fair discriminator for metastatic versus nonmetastatic endometrial adenocarcinoma, that sHLA-G could be a tumor marker for malignant versus benign ascites in breast and ovarian cancer, and that HLA-G5 was also a good predictor of outcome in septic shock [16,17,36]. Furthermore, significantly enhanced
sHLA-G levels were proven to be useful in assisting the diagnosis of malignant versus healthy conditions. The area under the ROC for sHLA-G levels in plasma between patients and normal controls was 0.953 in bladder cancer [37], 0.739 in renal cell carcinoma [38], 0.95 in breast cancer [39], 0.84 in distinguishing colorectal cancer from benign colorectal diseases [40], and 0.992 in esophageal squamous cell carcinoma [41], respectively. In this study, sHLA-G levels among AHB, CHB, and RHB patients were found to be much higher than in normal controls. The area under the ROC curve was 1.000, 0.993, and 0.604 for AHB, CHB, and RHB patients versus normal controls, respectively. Given 100% specificity at the cutoff of corresponding plasma sHLA-G levels, the detection sensitivity was 97.8, 91.6, and 3.3% for AHB, CHB, and RHB patients, respectively. These data suggested that sHLA-G could be a useful biomarker for HBV infection, particularly for AHB and CHB patients. In summary, this study revealed that HLA-G expression in monocytes, Tregs, and sHLA-G was dramatically increased in AHB and CHB patients. Given its immune-suppressive properties, we hypothesized a critical role of sHLA-G in the pathogenesis of HBV infection. Moreover, plasma sHLA-G may provide a novel approach to supplement other serologic or virologic marker tests in discrimination for the outcome for HBV-infected patients. However, further investigations are certainly required to confirm and reinforce the findings in this study. Acknowledgments This work was sponsored by the Zhejiang Provincial Program for the cultivation of high-level innovative health talents and by grants from the Science and Technology Bureau of Zhejiang Province (2008C33013 and 2009C33147). References [1] Ganem D, Prince AM. Hepatitis B virus infection—natural history and clinical consequences. N Engl J Med 2004;350:1118 –29. [2] Kao JH, Chen DS. Global control of hepatitis B virus infection. Lancet Infect Dis 2002;2:395– 403. [3] Kao JH. Diagnosis of hepatitis B virus infection through serological and virological markers. Expert Rev Gastroenterol Hepatol 2008;2:553– 62. [4] Elgouhari HM, Abu-Rajab Tamimi TI, Carey WD. Hepatitis B virus infection: understanding its epidemiology, course, and diagnosis. Cleve Clin J Med 2008; 75:881–9. [5] Carosella ED, Moreau P, Le Maoult J, Le Discorde M, Dausset J, Rouas-Freiss N. HLA-G molecules: from maternal–fetal tolerance to tissue acceptance. Adv Immunol 2003;81:199 –252. [6] Paul P, Cabestre FA, Ibrahim EC, Lefebvre S, Khalil-Daher I, Vazeux G, et al. Identification of HLA-G7 as a new splice variant of the HLA-G mRNA and expression of soluble HLA-G5, -G6, and -G7 transcripts in human transfected cells. Hum Immunol 2000;61:1138 – 49. [7] Rebmann V, Busemann A, Lindemann M, Grosse-Wilde H. Detection of HLA-G5 secreting cells. Hum Immunol 2003;64:1017–24. [8] Hofmeister V, Weiss EH. HLA-G modulates immune responses by diverse receptor interactions. Semin Cancer Biol 2003;13:317–23. [9] Carosella ED, Moreau P, Lemaoult J, Rouas-Freiss N. HLA-G: from biology to clinical benefits. Trends Immunol 2008;29:125–32. [10] Chen HX, Chen BG, Shi WW, Zhen R, Xu DP, Lin A, et al. Induction of cell surface human leukocyte antigen-G expression in pandemic H1N1 2009 and seasonal H1N1 influenza virus-infected patients. Hum Immunol 2011;72:159 – 65. [11] Yan WH, Lin A, Chen BG, Chen SY. Induction of both membrane-bound and soluble HLA-G expression in active human cytomegalovirus infection. J Infect Dis 2009;200:820 – 6. [12] Donaghy L, Gros F, Amiot L, Mary C, Maillard A, Guiguen C, et al. Elevated levels of soluble non-classical major histocompatibility class I molecule human leucocyte antigen (HLA)-G in the blood of HIV-infected patients with or without visceral leishmaniasis. Clin Exp Immunol 2007;147:236 – 40. [13] Lafon M, Prehaud C, Megret F, Lafage M, Mouillot G, Roa M, et al. Modulation of HLA-G expression in human neural cells after neurotropic viral infections. J Virol 2005;79:15226 –37. [14] Souto FJ, Crispim JC, Ferreira SC, et al. Liver HLA-G expression is associated with multiple clinical and histopathological forms of chronic hepatitis B virus infection. J Viral Hepat 2011;18:102–5. [15] Shih IeM. Application of human leukocyte antigen-G expression in the diagnosis of human cancer. Hum Immunol 2007;68:272– 6. [16] Singer G, Rebmann V, Chen YC, Liu HT, Ali SZ, Reinsberg J, et al. HLA-G is a potential tumor marker in malignant ascites. Clin Cancer Res 2003;9:4460 – 4.
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
[17] Monneret G, Voirin N, Krawice-Radanne I, BohÊ J, Lepape A, Rouas-Freiss N, et al. Soluble human leukocyte antigen-G5 in septic shock: marked and persisting elevation as a predictor of survival. Crit Care Med 2007;35:1942–7. [18] Chinese Society of Infectious Diseases and Parasitology and Chinese Society of Hepatology of Chinese Medical Association. National program for prevention and treatment of viral hepatitis. Chin J Hepatol 2000;8:324 –9. [19] McMahon BJ. The natural history of chronic hepatitis B virus infection. Hepatology 2009;49:S45–55. [20] Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology 2007;45:507–39. [21] Onno M, Pangault C, Le Friec G, Guilloux V, AndrÊ P, Fauchet R. Modulation of HLA-G antigens expression by human cytomegalovirus: specific induction in activated macrophages harboring human cytomegalovirus infection. J Immunol 2000;164:6426 –34. [22] Barel MT, Ressing M, Pizzato N, van Leeuwen D, Le Bouteiller P, Lenfant F, et al. Human cytomegalovirus-encoded US2 differentially affects surface expression of MHC class I locus products and targets membrane-bound, but not soluble HLA-G1 for degradation. J Immunol 2003;171:6757– 65. [23] Lozano JM, GonzÂlez R, KindelÂn JM, et al. Monocytes and T lymphocytes in HIV-1-positive patients express HLA-G molecule. AIDS 2002;16:347–51. [24] MÊgret F, Prehaud C, Lafage M, et al. Modulation of HLA-G and HLA-E expression in human neuronal cells after rabies virus or herpes virus simplex type 1 infections. Hum Immunol 2007;68:294 –302. [25] Jung MC, Pape GR. Immunology of hepatitis B infection. Lancet Infect Dis 2002;2:43–50. [26] Carosella ED, HoWangYin KY, Favier B, LeMaoult J. HLA-G-dependent suppressor cells: diverse by nature, function, and significance. Hum Immunol 2008; 69:700 –7. [27] Chung MK, Yoon H, Min SS, et al. Induction of cytotoxic T lymphocytes with peptides in vitro: identification of candidate T-cell epitopes in hepatitis B virus X antigen. J Immunother 1999;22:279 – 87. [28] L×hr HF, Gerken G, Schlicht HJ, Meryer zum Bu¨schenfelde KH, Fleischer B. Low frequency of cytotoxic liver-infiltrating T lymphocytes specific for endogenous processed surface and core proteins in chronic hepatitis B. J Infect Dis 1993; 168:1133–9. [29] Fainardi E, Rizzo R, Castellazzi M, Stignani M, Granieri E, Baricordi OR. Potential role of soluble human leukocyte antigen-G molecules in multiple sclerosis. Hum Immunol 2009;70:981–7.
1073
[30] Contini P, Ghio M, Poggi A, Filaci G, Indiveri F, Ferrone S, et al. Soluble HLA-A,-B,-C and -G molecules induce apoptosis in T and NK CD8⫹ cells and inhibit cytotoxic T cell activity through CD8 ligation. Eur J Immunol 2003;33:125–34. [31] Fournel S, Aguerre-Girr M, Huc X, Lenfant F, Alam A, Toubert A, et al. Cutting edge: soluble HLA-G1 triggers CD95/CD95 ligand-mediated apoptosis in activated CD8⫹ cells by interacting with CD8. J Immunol 2000;164:6100 – 4. [32] Jung MC, Spengler U, Schraut W, Hoffmann R, Zachoval R, Eisenburg J, et al. Hepatitis B virus antigen-specific T-cell activation in patients with acute and chronic hepatitis B. J Hepatol 1991;13:310 –7. [33] Ferrari C, Penna A, Bertoletti A, Valli A, Antoni AD, Giuberti T, et al. Cellular immune response to hepatitis B virus-encoded antigens in acute and chronic hepatitis B virus infection. J Immunol 1990;145:3442–9. [34] Le Maux A, NoÌl G, Birebent B, Grosset JM, Vu N, De Guibert S, et al. Soluble human leucocyte antigen-G molecules in peripheral blood haematopoietic stem cell transplantation: a specific role to prevent acute graft-versus-host disease and a link with regulatory T cells. Clin Exp Immunol 2008;152:50 – 6. [35] Selmani Z, Naji A, Zidi I, Favier B, Gaiffe E, Obert L, et al. Human leukocyte antigen-G5 secretion by human mesenchymal stem cells is required to suppress T lymphocyte and natural killer function and to induce CD4⫹ CD25highFOXP3⫹ regulatory T cells. Stem Cells 2008;26:212–22. [36] Barrier BF, Kendall BS, Sharpe-Timms KL, Kost ER. Characterization of human leukocyte antigen-G (HLA-G) expression in endometrial adenocarcinoma. Gynecol Oncol 2006;103:25–30. [37] Gan LH, Huang LF, Zhang X, Lin A, Xu DP, Wang Q, et al. Tumor-specific upregulation of human leukocyte antigen-G expression in bladder transitional cell carcinoma. Hum Immunol 2010;71:899 –904. [38] Li BL, Lin A, Zhang XJ, Zhang X, Zhang JG, Wang Q, et al. Characterization of HLA-G expression in renal cell carcinoma. Tissue Antigens 2009;74:213–21. [39] Chen HX, Lin A, Shen CJ, Zhen R, Chen BG, Zhang X, et al. Upregulation of human leukocyte antigen-G expression and its clinical significance in ductal breast cancer. Hum Immunol 2010;71:892– 8. [40] Zhu CB, Wang CX, Zhang X, Zhang J, Li W. Serum sHLA-G levels: a useful indicator in distinguishing colorectal cancer from benign colorectal diseases. Int J Cancer 2011;128:617–22. [41] Lin A, Zhang X, Zhou WJ, Ruan YY, Xu DP, Wang Q, et al. HLA-G expression is associated with a poor prognosis in patients with esophageal squamous cell carcinoma. Int J Cancer.
Contents lists available at SciVerse ScienceDirect
Plasma soluble human leukocyte antigen-G expression is a potential clinical biomarker in patients with hepatitis B virus infection Wei-Wu Shi a, Aifen Lin b, Dan-Ping Xu a, Wei-Guang Bao b, Jian-Gang Zhang b, Shi-Yong Chen c, Jun Li d, Wei-Hua Yan a,* a
Medical Research Center, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China Human Tissue Bank, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China c Department of Laboratory Medicine, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China d Department of Blood Bank, Taizhou Hospital of Zhejiang Province, Wenzhou Medical College, Linhai, Zhejiang, China b
A R T I C L E
I N F O
Article history: Received 18 December 2010 Accepted 21 June 2011 Available online 1 July 2011
Keywords: HLA-G Hepatitis B virus Infection
A B S T R A C T
The significance of upregulated soluble human leukocyte antigen-G (sHLA-G) expression under various pathologic conditions has been discussed. In this study, we evaluated the potential significance of plasma sHLA-G expression in patients with hepatitis B virus (HBV) infection. The study included 90 acute hepatitis B patients (AHB), 131 chronic hepatitis B patients (CHB), 152 resolved hepatitis B individuals (RHB), and 129 normal controls. sHLA-G were determined using enzyme-linked immunosorbent assay. A receiver operating characteristic (ROC) curve was used to evaluate the feasibility of plasma sHLA-G as a biomarker for distinguishing patients with HBV infection. sHLA-G levels in AHB (median, 193.1 U/mL; p ⬍ 0.001), CHB (median, 324.6 U/mL; p ⬍ 0.001), and RHB (median, 14.8 U/mL; p ⫽ 0.006) patients was much higher than that in normal controls (median, 9.0 U/mL). A significant difference for sHLA-G levels was also observed between patients with HBV infection (AHB vs CHB, AHB vs RHB, and CHB vs RHB; all p ⬍ 0.001). The area under the ROC curve for sHLA-G levels was 1.000 (p ⬍ 0.001) for AHB, 0.993 (p ⬍ 0.001) for CHB, and 0.604 (p ⫽ 0.003) for RHB patients versus normal controls, respectively. Data also indicated that the percentage of CD4⫹CD25⫹FoxP3⫹ T regulatory cells and HLA-G⫹CD14⫹ monocytes was significantly increased in AHB and CHB patients compared with normal controls (all p ⬍ 0.001). Our findings indicated that induction of HLA-G expression may play a role in HBV immune evasion and sHLA-G levels could be a useful biomarker in HBV infection. 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.
1. Introduction Hepatitis B virus (HBV), a small DNA virus, belongs to the Hepadnaviridae family [1]. HBV is one of the most common serious infectious agents for humans and could result in acute and chronic hepatitis, cirrhosis, end-stage liver disease, and hepatocellular carcinoma worldwide [2]. Extensive research on HBV has been carried out for more than half a century, resulting in remarkable advances in our understanding of the virus and in development of means for diagnosis, prevention, and control of the HBV-associated diseases. In addition to a medical history and physical examination, evaluation of HBV infection usually requires serologic, biochemical, and virologic tests to confirm diagnosis, treatment responses, and disease progression monitoring [3]. Virologic and biochemical markers such as HBV antigens and their corresponding antibodies, serum HBV DNA, viral genotype, and serum alanine aminotransferase (ALT) applied in the diagnosis and monitoring of HBV disease clearly are useful for the * Corresponding author. E-mail address: [email protected] (W.-H. Yan).
evaluation of patients with HBV infection in clinical practice [4]. Beyond this, other potential useful markers are definitely needed to expand the understanding of HBV pathogenesis and to predict outcome and diagnosis for virus infection. Human leukocyte antigen-G (HLA-G) is a nonclassic HLA class I molecule that was first identified as being selectively expressed by choriocarcinoma cells and later observed under a broad spectrum of pathologic conditions [5]. Unlike classic HLA class I antigens, 7 HLA-G isoforms, including 4 membrane-bound (HLA-G1–G4) and 3 soluble HLA-G (HLA-G5–G7) proteins, are generated, and another soluble form of HLA-G could be produced by shedding of the proteolytically cleaved cell surface expressed HLA-G1 (sHLA-G1) [6,7]. Both isoforms of HLA-G bear similar immune-suppressive properties via binding to receptors such as immunoglobulin-like transcript-2/CD85j and -4/CD85d and KIR2DL4/CD158d expressed on various immune-competent cells [8]. Apart from its initial relevance in fetal–maternal immunotolerance, HLA-G has now been reported to be involved in tumor cell immune escape, inflammatory, autoimmune, and infectious diseases, and allograft acceptance [9].
0198-8859/11/$32.00 - see front matter 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.humimm.2011.06.012
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
In viral infection, induction of sHLA-G expression by virusinfected cells was proposed as a mechanism that helps viruses to subvert host antivirus defenses. Roles of induced HLA-G expression after infections with certain influenza A virus strains, human cytomegalovirus virus (HCMV), human immunodeficiency virus (HIV), neurotropic viruses such as rabies virus, and hepatitis B virus (HBV) were discussed [10 –14]. Beside the aims of elucidating the roles of HLA-G in immune regulation and its clinical relevance under pathologic conditions, studies that focus on potential applications of HLA-G expression in clinical practice may represent one of the new perspectives in HLA-G research [15]. Previous studies indicated that sHLA-G could be a tumor marker for malignant versus benign ascites in breast and ovarian cancer, and HLA-G5 as a good predictor of outcome in septic shock was also reported [16,17]. In the present study, plasma sHLA-G levels in different HBV infection status, including acute (AHB), chronic (CHB), and resolved hepatitis B (RHB) infected patients, as well as in normal controls, were measured, and its potential feasibility in distinguishing the outcome of HBV-associated diseases was analyzed. Moreover, HLA-G expression on CD14⫹ monocytes and the percentage of CD4⫹CD25⫹FoxP3⫹ T regulatory cells (Tregs) was determined.
1069
and hepatitis B envelope (HBe) antigen seroconversion for at least 6 months before sample collection. No patients received anti-HBV agent or steroid 6 months before sampling. Samples from 129 healthy individuals without HBV infection history and with similar age and sex characteristics were taken as controls. Concurrence of hepatitis C virus or HIV infections was excluded for all enrolled individuals. The study protocol was approved by the ethics committee of the Taizhou Hospital of Zhejiang Province, Wenzhou Medical College. The baseline characteristics of subjects enrolled in the study are presented in Table 1. 2.2. Virological assessments
2. Materials and methods
HBV serum markers were determined using commercial enzyme immunoassay kits (AXSYM System, Abbott, Wiesbaden, Germany). HBV DNA was extracted from serum samples and quantified using a commercial polymerase chain reaction diagnostic kit with detection limit of 100 copies/mL (Sybio, Shanghai, China). Serum ALT, aspartate aminotransferase (AST), alkaline phosphatase, ␥-glutamine transpeptidase, albumin, and total, direct, or indirect bilirubin concentration levels were measured using an autobiochemical analyzer (Olympus AU 5400, Olympus Corp., Tokyo, Japan).
2.1. Study setting and participants
2.3. sHLA-G enzymed-linked immunosorbent assay (ELISA)
Consecutive blood samples were collected from 90 patients with AHB and 131 patients with CHB. One hundred fifty-two individuals with RHB were screened from blood donors. Specimens were collected from the Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical College. The diagnosis was complied with the diagnostic criteria of the 2000 Xi’an Viral Hepatitis Management Scheme issued by the Chinese Society of Infectious Diseases and Parasitology and the Chinese Society of Hepatology of the Chinese Medical Association [18]. Briefly, standards for diagnoses of AHB include displaying hepatitis B surface Ag (HBsAg)-negative conversion within 6 months after the initial onset of symptoms caused by HBV infection and being anti-HBc immunoglobulin M (IgM) positive. An HBV carrier with chronic hepatitis exhibits a clinical course of hepatitis B for ⬎6 months and may have exhibited symptoms or signs of hepatitis and abnormal hepatic function on this occasion. An additional group of 152 individuals with resolved HBV infections was included. The inclusion criteria for resolved subjects included a history of HBV infection, normal liver biochemistry, undetectable clinical activity, and HBV DNA, as well as HBsAg
Plasma sHLA-G concentrations were determined with the sHLAG-specific ELISA kit (sHLA-G kit; Exbio, Prague, Czech Republic), which measures sHLA-G1 and HLA-G5. Each sample (100 L) was measured in triplicate. The optical densities were measured at 450 nm (Spectra Max 250, Molecular Devices, Sunnyvale, CA). The final concentration was determined by optical density according to the standard curves (range, 0 –125 U/mL). When the concentration exceeded 125 U/mL, diluted samples were used and dilution factors were considered to calculate the sHLA-G concentration. The detection limits were 1 U/mL. Details of the performance were according to the manufacturer’s instruction. 2.4. Flow cytometry analysis Flow cytometry analysis was performed with a standard protocol according to the previous study [10]. Briefly, peripheral blood mononuclear cells (PBMC) obtained from HBV-infected patients and normal controls were gated with allophycocyanin-labeledanti-CD14 (BD Biosciences, San Jose, CA) for monocytes, and HLA-G expression was analyzed. Briefly, 1 ⫻ 106 mononuclear cells were
Table 1 Baseline characteristics of the study populations Variables
Age (years)a Sex (male/female) HBsAg positive (%) HBsAb positive (%) HBeAg positive (%) HBeAb positive (%) HBcAb positive (total, %) HBcAb (IgM) positive (%) ALT (IU/L) AST (IU/L) HBV DNA (copies/mL, log10)a TBil (mol/L) DBil (mol/L) IBil (mol/L) Albumin (g/L)
Controls
AHB
n ⫽ 129
n ⫽ 90
39.6 ⫾ 11.2 75/54 0 0 0 0 0 0 ⬍40 — — — — — —
43.1 ⫾ 12.7 54/36 90 (100%) 0 90 (100%) 0 90 (100%) 90 (100%) 1,117.0 ⫾ 686.9 647.1 ⫾ 417.1 6.8 (2.1–8.9) 49.2 ⫾ 45.0 19.1 ⫾ 24.8 24.5 ⫾ 25.9 39.8 ⫾ 9.9
CHB Correlate to sHLA-G (p)
0.626 0.743 0.203 0.844 0.960 0.633 0.479
n ⫽ 131 44.8 ⫾ 14.3 71/60 131 (100%) 0 41 (30.1%) 76 (55.9%) 127 (97.1%) 0 163.9 ⫾ 79.1 54.1 ⫾ 78.0 3.8 (2.0–9.3) 19.7 ⫾ 19.3 5.7 ⫾ 8.5 13.6 ⫾ 11.0 41.6 ⫾ 6.6
RHB Correlate to sHLA-G (p)
0.860 0.773 0.648 0.469 0.589 0.779 0.763 0.828 0.746 0.203
n ⫽ 152 42.2 ⫾ 13.1 89/63 0 123 (80.9%) 0 0 128 (84.2%) 0 ⬍40 — — — — — —
Correlate to sHLA-G (p)
0.672
0.542
AHB ⫽ acute hepatitis B patients; CHB ⫽ chronic hepatitis B patients; RHB ⫽ resolved hepatitis B individuals; ALT ⫽ alanine aminotransferase; AST ⫽ aspartate aminotransferase; HBV ⫽ hepatitis B virus; TBil ⫽ total bilirubin; DBil ⫽ direct bilirubin; IBil ⫽ indirect bilirubin; ⫽, not detected. a Expressed as median (range).
1070
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
incubated in 200 L of 1.0% phosphate-buffered saline/bovine serum albumin in a cold room for 30 minutes in the dark with 10 L (1 mg/mL) PE–anti-HLA-G MEM-G/9 (IgG1, Serotec, Oxford, UK), which reacts with the native form of human HLA-G1 on the cell surface, and an isotype IgG was added as a negative control (BD Biosciences). Intracellular Foxp3 stainning was detected using the eBioscience Foxp3 staining kit (eBioscience, San Diego, CA). Briefly, PBMC were first surface labeled with PE–CD4 (BD Biosciences) and FITC–CD25 (BD Biosciences). Then, intracellular analysis of FoxP3 was performed after fixation and permeabilization according to the manufacturer’s recommendation (eBioscience). Flow cytometry was performed on a FACSCalibur and analyzed using the CellQuest (BD Biosciences) software package. 2.5. Statistical analysis Statistical analysis was performed with SPSS 13.0 software (SPSS, Inc., Chicago, IL). Normality of continuous numeric data was analyzed by one-sample Kolmogorov–Smirnov test. Comparison of the percentage of HLA-G-expressing monocytes, the percentage of Tregs, and sHLA-G levels between different groups was performed using the Mann–Whitney U test. The feasibility of using sHLA-G as a potential biomarker for discrimination of HBV-infected patients from healthy individuals was assessed using receiver operating characteristic (ROC) curve analysis. The areas under the ROC curve were calculated and subjected to statistical analysis. The 95% confidence intervals were calculated for specificity and sensitivity. A two-sided p value ⬍ 0.05 was considered statistically significant.
ical parameters such as age, sex, HBV DNA copies, ALT, AST, alkaline phosphatase, ␥-glutamine transpeptidase, albumin, and total, direct, or indirect bilirubin concentration (Table 1). 3.2. ROC curve analysis for plasma sHLA-G as a biomarker for patients with HBV infection ROC curves were used to evaluate the performance of sHLA-G in predicting the outcome of patients with HBV infection. The area under the ROC curve for sHLA-G levels was 1.000 (95% CI 0.999 – 1.000, p ⬍ 0.001) for AHB versus normal controls (Fig. 2A). The area under the ROC curve for sHLA-G levels was 0.993 for CHB patients versus normal controls (Fig. 2B). The area under the ROC curve was 0.604 (95% CI 0.538 – 0.671, p ⫽ 0.003) for RHB patients versus normal controls (Fig. 2C). Given 100% specificity at the cutoff of corresponding plasma sHLA-G levels, the detection sensitivity was 97.8, 91.6, and 3.3% for AHB, CHB, and RHB patients, respectively. 3.3. Cell surface HLA-G expression in monocytes
3. Results
Cell surface HLA-G expression was analyzed on the CD14⫹ monocytes. The percentage of HLA-G-positive CD14⫹ monocytes was a median of 7.72% (n ⫽ 26; range: 1.12– 48.62%) in AHB patients, 4.62% (n ⫽ 23; range: 0.82–24.97%) in CHB patients, 1.93% (n ⫽ 31; range: 0.19 –11.87%) in RHB patients, and 1.62% (n ⫽ 36; range: 0.30 –9.58%) in normal controls, respectively. When compared, the percentage of HLA-G⫹CD14⫹ monocytes was much higher in AHB and CHB than in normal controls (all p ⬍ 0.001), and the percentage of HLA-G⫹CD14⫹ monocytes in AHB was much higher than in CHB patients (p ⫽ 0.021, Fig. 3A).
3.1. Plasma sHLA-G expression in study populations
3.4. The frequency of Tregs in HBV-infected patients
Plasma sHLA-G levels in 90 AHB, 131 CHB patients, and 152 RHB individuals and 129 normal controls were determined by ELISA. The concentration of the plasma sHLA-G was a median of 193.1 U/mL (range, 69.4 –504.3 U/mL) for AHB patients, 324.6 U/mL (range, 23.6 –786.0 U/mL) for CHB patients, 14.8 U/mL (range, 3.5– 86.5 U/mL) for RHB individuals, and 9.0 U/mL (range, 3.6 –71.6 U/mL) for normal controls. sHLA-G levels were dramatically increased in AHB, CHB, and RHB individuals when compared with that in normal controls (all p ⬍ 0.001). sHLA-G levels were also markedly different among HBV-infected individuals (AHB vs CHB, AHB vs RHB, and CHB vs RHB, all p ⬍ 0.001; Fig. 1). Furthermore, no correlation was observed between sHLA-G levels and patient clin-
sHLA-G comparison: Normal vs RHB, p=0.006; Normal vs AHB, p<0.001; Normal vs CHB, p<0.001; AHB vs CHB, p<0.001; AHB vs RHB, p<0.001.
Fig. 1. Distribution and comparison of plasma soluble human leukocyte antigen-G (sHLA-G) between study groups. The bar in the box represents the median. Comparison of sHLA-G levels between groups was performed using the Mann–Whitney U test. AHB, acute hepatitis B patients; CHB, chronic hepatitis B patients; RHB, resolved hepatitis B individuals.
Moreover, our data indicated that the percentage of CD4⫹ CD25⫹FoxP3⫹ Tregs was a median of 5.08% (n ⫽ 21; range, 1.69 – 9.58%) in AHB, 3.84% (n ⫽ 23; range: 0.64 – 8.76%; p ⫽ 0.098) in CHB, 2.98% (n ⫽ 21; range: 0.66 – 6.89%) in RHB patients and 2.76% (n ⫽ 34; range, 1.39 –5.01%) in normal controls, respectively. The proportion of Tregs in AHB (p ⬍ 0.001) and CHB patients (p ⫽ 0.016) was significantly increased compared with that in normal controls; however, no significance was observed between the AHB and CHB patients (p ⫽ 0.098, Fig. 3B). 4. Discussion Our study provides evidence of a marked elevation of plasma sHLA-G concentration in patients with HBV infection. Most important, increased sHLA-G concentrations could be a biomarker for the outcome of HBV infection. Furthermore, this result may be important in understanding the immune-suppressive roles of HLA-G in HBV pathogenesis. HBV is one of the most common serious infectious viruses for humans. More than one third of the world’s population has been infected with HBV and it is estimated that there are 350 million persistent carriers of HBV worldwide [19]. Most adults infected with the virus recover, but 5–10% are unable to clear the virus and become chronically infected [20]. HBV produces several antigens that can be detected in the blood and disappear as the body produces antibodies against them. The patterns of these and other markers provide clues to the phase of infection. HBsAg and HBV DNA are often the first detectable markers of acute infection, appearing before the onset of symptoms or before elevation of ALT levels [4]. Chronic HBV infection was defined as surface antigen persisting longer than 6 months. HBeAg is a marker of HBV replication and infectivity. In chronic infection, HBeAg can persist for years or decades. HBcAg cannot be detected in the serum, but antibodies against it can, first IgM and later IgG [3]. These serologic and virologic markers of HBV infection have been well accepted as useful tools in the diagnosis and monitoring of HBV-associated diseases.
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
A 52&&XUYH
HLA-G+CD14+ monocytes (%)
A
6HQVLWLYLW\
60
50
40
30
20
10
AUC: 1.000
Normal controls
B
12
6SHFLILFLW\
CD4+CD25+FoxP3+ Treg (%)
52&&XUYH
6HQVLWLYLW\
HLA-G+monocyte comparison: Normal vs RHB, p=0.758; Normal vs AHB, p<0.001; Normal vs CHB, p=0.001; AHB vs CHB, p=0.021; AHB vs RHB, p<0.001.
0
B
1071
10
8
RHB
AHB
CHB
AHB
CHB
Treg comparison: Normal vs RHB, p=0.782; Normal vs AHB, p<0.001; Normal vs CHB, p=0.016; AHB vs CHB, p=0.098; AHB vs RHB, p=0.005.
6
4
2
0
AUC: 0.993
6SHFLILFLW\
C
52&&XUYH
Normal controls
RHB
Fig. 3. Flow cytometry anaylsis of human leukocyte antigen-G (HLA-G)⫹CD14⫹ monocytes and CD4⫹CD25⫹FoxP3⫹ regulatory T cells (Tregs). (A) Comparison of the proportion of HLA-G⫹ monocytes among normal controls, resolved hepatitis B individuals (RHB), acute hepatitis B patients (AHB), and chronic hepatitis B patients (CHB). (B) Comparison of the frequencies of CD4⫹CD25⫹FoxP3⫹ Treg among normal controls, RHB, AHB, and CHB patients. The solid line represents the median and comparison was analyzed using the Mann–Whitney U test.
6HQVLWLYLW\
AUC: 0.604
6SHFLILFLW\ Fig. 2. Receiver operating characteristic (ROC) curve analysis of plasma soluble human leukocyte antigen-G (sHLA-G) levels (U/mL) for predicting the outcome of patients with hepatitis B virus (HBV) infection. Area under the curve (AUC): (A) 1.000 (95% CI 0.999 –1.000, p ⬍ 0.001) for acute hepatitis B patients, (B) 0.993 (95% CI 0.988 – 0.999, p ⬍ 0.001) for chronic hepatitis B patients, and (C) 0.604 (95% CI 0.538 – 0.671, p ⫽ 0.003) for resolved hepatitis B patients versus normal controls, respectively.
In the present study, our data revealed that plasma sHLA-G was dramatically increased in acute and chronic hepatitis B patients as well as resolved hepatitis B individuals and suggested that plasma sHLA-G expression could be a useful biomarker in predicting outcome in patients with hepatitis B. Induction of HLA-G expression and its potential roles in certain types of virus infection were addressed in previous studies [9]. We recently reported that both the PBMC surface HLA-G and the plasma sHLA-G levels were dramatically induced in active HCMV patients [11]. Another study demonstrated that both cell surface HLA-G and soluble HLA-G could be produced during viral reactivation in macrophages and that bronchoalveolar macrophages from patients with acute HCMV pneumonitis also express HLA-G molecules [21]. Moreover, only soluble HLA-G could be induced in CMVinfected U-373 MG astrocytoma cells through the cooperative action of the 2 IE1 pp72 and IE2 pp86 products [22]. In HIV-infected patients, HLA-G expression was upregulated in CD8⫹ T cells and monocytes, and long-term progressor patients expressed higher levels of plasma sHLA-G than long-term nonprogressor patients [12,23]. In neurotropic viral infections, herpes simplex virus type 1 and rabies virus upregulate the neuronal expression of HLA-G isoforms, both in infected cells and in neighboring uninfected cells
1072
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
[13]. Furthermore, upregulation of HLA-G gene expression was observed in a human neuronal line (NT2-N cells) after in vitro infection by a highly pathogenic strain of rabies virus or a strain of herpes virus inducing neuron latency. Importantly, HLA-G expression was restricted to rabies virus–infected neurons [24]. These findings suggested that upregulation of immunotolerant HLA-G expression in virus infection might be a potential virus escape mechanism from host antiviral immune responses. HBV has evolved a multitude of strategies to subvert host immune surveillance and responses by both the innate and the adaptive arms of the immune system, where dendritic cells, natural killer (NK) and NKT cells, and virus-specific CD4⫹ and CD8⫹ T cells are crucial for restricting viral replication [25]. The biologic role of HLA-G in HBV infection remains to be elucidated. However, both membrane-bound and soluble HLA-G were well described as tolerogenic molecules exhibiting strong immunosubversive activities against both innate and adaptive immune responses; recent studies also indicated that HLA-G was involved in inducing suppressor cells such as Tregs, dendritic cells, and NK cells, providing a long-term immune modulatory function [26]. Consequently, a decreased immune function induced by HLA-G leads to an advantage for the virus to subvert host defenses. It should be noted that the immune response in chronic carriers is weak or undetectable, resulting in an inability to eliminate the virus efficiently. This finding was supported by evidence that peripheral cytotoxic T lymphocyte (CTL) responses in chronically infected patients are difficult to detect and HBV-specific CTLs are present at a low frequency in the livers of chronically infected patients [27,28]. In this study, our results indicated that plasma sHLA-G, cell surface HLA-G expression in monocytes, and the percentage of the CD4⫹CD25⫹Foxp3 Treg cells were dramatically increased in patients with AHB and CHB. Unexpectedly, our data revealed that sHLA-G among AHB patients was significantly lower than that among CHB patients, but the reason remains unknown. Because peripheral monocytes are the main source of sHLA-G, the mechanisms underlying the retention of sHLA-G production in monocytes during HBV infection require further investigation [7]. High sHLA-G levels among CHB patients may result in the inhibition of CTL cytolysis and in induction of a CD8⫹ regulatory T-cell population, as previously reported [29]. In addition, sHLA-G has the ability to induce apoptosis of CD8⫹ T cells via binding to CD8 and through a Fas/FasL pathway [30,31]. Furthermore, during chronic HBV infection, the peripheral blood HLA class II–restricted CD4⫹ T-cell response to all viral antigens such as HBcAg and HBeAg is much less vigorous than that in patients with acute hepatitis [32,33]. The impaired CD4⫹ T-cell responses in CHB patients may be supported by the current findings that the sHLA-G level was positively associated with CD4⫹CD25⫹ CD152⫹ natural Treg frequency, and functionally, sHLA-G could induce CD4⫹CD25highFoxP3⫹ Tregs [34,35]. A recent study demonstrated that both soluble and membrane-bound HLA-G is frequently and largely expressed in hepatocytes, biliary epithelial cells, or both of chronic HBV-infected individuals, indicating that HLA-G is expressed in most cases of chronic HBV infection in all stages and may play a role in the persistence of HBV infection [14]. Thus, a dominant cause of viral infection could be the existence of a weak antiviral immune response induced by sHLA-G; however, the underlying mechanisms associated with sHLA-G in viral infection, such as HBV infection, require further investigation. Another intriguing result of the present study is that measurement of plasma sHLA-G is a highly sensitive technique to monitor the outcome of HBV infection. Studies on malignancies indicated that HLA-G was a fair discriminator for metastatic versus nonmetastatic endometrial adenocarcinoma, that sHLA-G could be a tumor marker for malignant versus benign ascites in breast and ovarian cancer, and that HLA-G5 was also a good predictor of outcome in septic shock [16,17,36]. Furthermore, significantly enhanced
sHLA-G levels were proven to be useful in assisting the diagnosis of malignant versus healthy conditions. The area under the ROC for sHLA-G levels in plasma between patients and normal controls was 0.953 in bladder cancer [37], 0.739 in renal cell carcinoma [38], 0.95 in breast cancer [39], 0.84 in distinguishing colorectal cancer from benign colorectal diseases [40], and 0.992 in esophageal squamous cell carcinoma [41], respectively. In this study, sHLA-G levels among AHB, CHB, and RHB patients were found to be much higher than in normal controls. The area under the ROC curve was 1.000, 0.993, and 0.604 for AHB, CHB, and RHB patients versus normal controls, respectively. Given 100% specificity at the cutoff of corresponding plasma sHLA-G levels, the detection sensitivity was 97.8, 91.6, and 3.3% for AHB, CHB, and RHB patients, respectively. These data suggested that sHLA-G could be a useful biomarker for HBV infection, particularly for AHB and CHB patients. In summary, this study revealed that HLA-G expression in monocytes, Tregs, and sHLA-G was dramatically increased in AHB and CHB patients. Given its immune-suppressive properties, we hypothesized a critical role of sHLA-G in the pathogenesis of HBV infection. Moreover, plasma sHLA-G may provide a novel approach to supplement other serologic or virologic marker tests in discrimination for the outcome for HBV-infected patients. However, further investigations are certainly required to confirm and reinforce the findings in this study. Acknowledgments This work was sponsored by the Zhejiang Provincial Program for the cultivation of high-level innovative health talents and by grants from the Science and Technology Bureau of Zhejiang Province (2008C33013 and 2009C33147). References [1] Ganem D, Prince AM. Hepatitis B virus infection—natural history and clinical consequences. N Engl J Med 2004;350:1118 –29. [2] Kao JH, Chen DS. Global control of hepatitis B virus infection. Lancet Infect Dis 2002;2:395– 403. [3] Kao JH. Diagnosis of hepatitis B virus infection through serological and virological markers. Expert Rev Gastroenterol Hepatol 2008;2:553– 62. [4] Elgouhari HM, Abu-Rajab Tamimi TI, Carey WD. Hepatitis B virus infection: understanding its epidemiology, course, and diagnosis. Cleve Clin J Med 2008; 75:881–9. [5] Carosella ED, Moreau P, Le Maoult J, Le Discorde M, Dausset J, Rouas-Freiss N. HLA-G molecules: from maternal–fetal tolerance to tissue acceptance. Adv Immunol 2003;81:199 –252. [6] Paul P, Cabestre FA, Ibrahim EC, Lefebvre S, Khalil-Daher I, Vazeux G, et al. Identification of HLA-G7 as a new splice variant of the HLA-G mRNA and expression of soluble HLA-G5, -G6, and -G7 transcripts in human transfected cells. Hum Immunol 2000;61:1138 – 49. [7] Rebmann V, Busemann A, Lindemann M, Grosse-Wilde H. Detection of HLA-G5 secreting cells. Hum Immunol 2003;64:1017–24. [8] Hofmeister V, Weiss EH. HLA-G modulates immune responses by diverse receptor interactions. Semin Cancer Biol 2003;13:317–23. [9] Carosella ED, Moreau P, Lemaoult J, Rouas-Freiss N. HLA-G: from biology to clinical benefits. Trends Immunol 2008;29:125–32. [10] Chen HX, Chen BG, Shi WW, Zhen R, Xu DP, Lin A, et al. Induction of cell surface human leukocyte antigen-G expression in pandemic H1N1 2009 and seasonal H1N1 influenza virus-infected patients. Hum Immunol 2011;72:159 – 65. [11] Yan WH, Lin A, Chen BG, Chen SY. Induction of both membrane-bound and soluble HLA-G expression in active human cytomegalovirus infection. J Infect Dis 2009;200:820 – 6. [12] Donaghy L, Gros F, Amiot L, Mary C, Maillard A, Guiguen C, et al. Elevated levels of soluble non-classical major histocompatibility class I molecule human leucocyte antigen (HLA)-G in the blood of HIV-infected patients with or without visceral leishmaniasis. Clin Exp Immunol 2007;147:236 – 40. [13] Lafon M, Prehaud C, Megret F, Lafage M, Mouillot G, Roa M, et al. Modulation of HLA-G expression in human neural cells after neurotropic viral infections. J Virol 2005;79:15226 –37. [14] Souto FJ, Crispim JC, Ferreira SC, et al. Liver HLA-G expression is associated with multiple clinical and histopathological forms of chronic hepatitis B virus infection. J Viral Hepat 2011;18:102–5. [15] Shih IeM. Application of human leukocyte antigen-G expression in the diagnosis of human cancer. Hum Immunol 2007;68:272– 6. [16] Singer G, Rebmann V, Chen YC, Liu HT, Ali SZ, Reinsberg J, et al. HLA-G is a potential tumor marker in malignant ascites. Clin Cancer Res 2003;9:4460 – 4.
W.-W. Shi et al. / Human Immunology 72 (2011) 1068-1073
[17] Monneret G, Voirin N, Krawice-Radanne I, BohÊ J, Lepape A, Rouas-Freiss N, et al. Soluble human leukocyte antigen-G5 in septic shock: marked and persisting elevation as a predictor of survival. Crit Care Med 2007;35:1942–7. [18] Chinese Society of Infectious Diseases and Parasitology and Chinese Society of Hepatology of Chinese Medical Association. National program for prevention and treatment of viral hepatitis. Chin J Hepatol 2000;8:324 –9. [19] McMahon BJ. The natural history of chronic hepatitis B virus infection. Hepatology 2009;49:S45–55. [20] Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology 2007;45:507–39. [21] Onno M, Pangault C, Le Friec G, Guilloux V, AndrÊ P, Fauchet R. Modulation of HLA-G antigens expression by human cytomegalovirus: specific induction in activated macrophages harboring human cytomegalovirus infection. J Immunol 2000;164:6426 –34. [22] Barel MT, Ressing M, Pizzato N, van Leeuwen D, Le Bouteiller P, Lenfant F, et al. Human cytomegalovirus-encoded US2 differentially affects surface expression of MHC class I locus products and targets membrane-bound, but not soluble HLA-G1 for degradation. J Immunol 2003;171:6757– 65. [23] Lozano JM, GonzÂlez R, KindelÂn JM, et al. Monocytes and T lymphocytes in HIV-1-positive patients express HLA-G molecule. AIDS 2002;16:347–51. [24] MÊgret F, Prehaud C, Lafage M, et al. Modulation of HLA-G and HLA-E expression in human neuronal cells after rabies virus or herpes virus simplex type 1 infections. Hum Immunol 2007;68:294 –302. [25] Jung MC, Pape GR. Immunology of hepatitis B infection. Lancet Infect Dis 2002;2:43–50. [26] Carosella ED, HoWangYin KY, Favier B, LeMaoult J. HLA-G-dependent suppressor cells: diverse by nature, function, and significance. Hum Immunol 2008; 69:700 –7. [27] Chung MK, Yoon H, Min SS, et al. Induction of cytotoxic T lymphocytes with peptides in vitro: identification of candidate T-cell epitopes in hepatitis B virus X antigen. J Immunother 1999;22:279 – 87. [28] L×hr HF, Gerken G, Schlicht HJ, Meryer zum Bu¨schenfelde KH, Fleischer B. Low frequency of cytotoxic liver-infiltrating T lymphocytes specific for endogenous processed surface and core proteins in chronic hepatitis B. J Infect Dis 1993; 168:1133–9. [29] Fainardi E, Rizzo R, Castellazzi M, Stignani M, Granieri E, Baricordi OR. Potential role of soluble human leukocyte antigen-G molecules in multiple sclerosis. Hum Immunol 2009;70:981–7.
1073
[30] Contini P, Ghio M, Poggi A, Filaci G, Indiveri F, Ferrone S, et al. Soluble HLA-A,-B,-C and -G molecules induce apoptosis in T and NK CD8⫹ cells and inhibit cytotoxic T cell activity through CD8 ligation. Eur J Immunol 2003;33:125–34. [31] Fournel S, Aguerre-Girr M, Huc X, Lenfant F, Alam A, Toubert A, et al. Cutting edge: soluble HLA-G1 triggers CD95/CD95 ligand-mediated apoptosis in activated CD8⫹ cells by interacting with CD8. J Immunol 2000;164:6100 – 4. [32] Jung MC, Spengler U, Schraut W, Hoffmann R, Zachoval R, Eisenburg J, et al. Hepatitis B virus antigen-specific T-cell activation in patients with acute and chronic hepatitis B. J Hepatol 1991;13:310 –7. [33] Ferrari C, Penna A, Bertoletti A, Valli A, Antoni AD, Giuberti T, et al. Cellular immune response to hepatitis B virus-encoded antigens in acute and chronic hepatitis B virus infection. J Immunol 1990;145:3442–9. [34] Le Maux A, NoÌl G, Birebent B, Grosset JM, Vu N, De Guibert S, et al. Soluble human leucocyte antigen-G molecules in peripheral blood haematopoietic stem cell transplantation: a specific role to prevent acute graft-versus-host disease and a link with regulatory T cells. Clin Exp Immunol 2008;152:50 – 6. [35] Selmani Z, Naji A, Zidi I, Favier B, Gaiffe E, Obert L, et al. Human leukocyte antigen-G5 secretion by human mesenchymal stem cells is required to suppress T lymphocyte and natural killer function and to induce CD4⫹ CD25highFOXP3⫹ regulatory T cells. Stem Cells 2008;26:212–22. [36] Barrier BF, Kendall BS, Sharpe-Timms KL, Kost ER. Characterization of human leukocyte antigen-G (HLA-G) expression in endometrial adenocarcinoma. Gynecol Oncol 2006;103:25–30. [37] Gan LH, Huang LF, Zhang X, Lin A, Xu DP, Wang Q, et al. Tumor-specific upregulation of human leukocyte antigen-G expression in bladder transitional cell carcinoma. Hum Immunol 2010;71:899 –904. [38] Li BL, Lin A, Zhang XJ, Zhang X, Zhang JG, Wang Q, et al. Characterization of HLA-G expression in renal cell carcinoma. Tissue Antigens 2009;74:213–21. [39] Chen HX, Lin A, Shen CJ, Zhen R, Chen BG, Zhang X, et al. Upregulation of human leukocyte antigen-G expression and its clinical significance in ductal breast cancer. Hum Immunol 2010;71:892– 8. [40] Zhu CB, Wang CX, Zhang X, Zhang J, Li W. Serum sHLA-G levels: a useful indicator in distinguishing colorectal cancer from benign colorectal diseases. Int J Cancer 2011;128:617–22. [41] Lin A, Zhang X, Zhou WJ, Ruan YY, Xu DP, Wang Q, et al. HLA-G expression is associated with a poor prognosis in patients with esophageal squamous cell carcinoma. Int J Cancer.