Search

Copyright of Obésité is the property of Lavoisier and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In humans, glucocorticoids (GCs) are commonly prescribed because of their anti-inflammatory and immunosuppressive properties. However, high doses of GCs often lead to side effects, including diabetes and lipodystrophy. We recently reported that adipocyte glucocorticoid receptor (GR)-deficient (AdipoGR-KO) mice under corticosterone (CORT) treatment exhibited a massive adipose tissue (AT) expansion associated with a paradoxical improvement of metabolic health compared with control mice. However, whether GR may control adipose development remains unclear. Here, we show a specific induction of hypoxia-inducible factor 1α (HIF-1α) and proangiogenic vascular endothelial growth factor A (VEGFA) expression in GR-deficient adipocytes of AdipoGR-KO mice compared with control mice, together with an increased adipose vascular network, as assessed by three-dimensional imaging. GR activation reduced HIF-1α recruitment to the Vegfa promoter resulting from Hif-1α downregulation at the transcriptional and posttranslational levels. Importantly, in CORT-treated AdipoGR-KO mice, the blockade of VEGFA by a soluble decoy receptor prevented AT expansion and the healthy metabolic phenotype. Finally, in subcutaneous AT from patients with Cushing syndrome, higher VEGFA expression was associated with a better metabolic profile. Collectively, these results highlight that adipocyte GR negatively controls AT expansion and metabolic health through the downregulation of the major angiogenic effector VEGFA and inhibition of vascular network development., (© 2024 by the American Diabetes Association.)

The microvascular endothelium of the renal transplant is the first site of graft interaction with the host immune system and is often injured in chronic Antibody Mediated Rejection (AMR). Microvascular inflammation is an independent determinant of AMR and heightens endothelial expression of HLA molecules thereby increasing the possibility of Donor Specific Antibody (DSA) binding. Endothelial cells produce IL-6 in the steady-state and this is increased by inflammation or by HLA-DR antibody binding in an allogeneic setting. Because IL-6 has been implicated in AMR, IL-6 blockade is currently under investigation as a therapeutic target. To further understand the role of IL-6 in endothelial cell immunogenicity, we have examined whether humanized antibody blockade of IL-6 altered endothelial cell interactions with allogeneic PBMC and after anti-HLA or DSA binding to endothelial cells in an in vitro human experimental model. Soluble factors, endothelial phenotype, Stat-3 activation, CD4 + -T differentiation, and C4d deposition were examined. Blockade of IL-6 reduced endothelial cell secretion of IL-6 and of the monocyte chemoattractant MCP-1. Pre-activation of endothelial cells by anti-HLA or DSA binding increased IL-6 secretion, that was further increased by concurrent binding of both antibodies and this was inhibited by IL-6 blockade. Activation of Stat-3 in CD4 + -T mediated by soluble factors produced in endothelial-PBMC interactions, and endothelial differentiation of CD4 + -T cell subsets (Th1, Th17, Treg), were impaired whereas activation of Complement by anti-HLA antibody binding remained unchanged by IL-6 blockade. Together, these data identify EC-mediated pro-inflammatory responses (T cell expansion, EC auto-activation, chemokine secretion) targeted by IL-6 blockade., (© 2022 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.)

The microvascular endothelium of the renal transplant is the first site of graft interaction with the host immune system and is often injured in chronic Antibody Mediated Rejection (AMR). Microvascular inflammation is an independent determinant of AMR and heightens endothelial expression of human leukocyte antigen (HLA) molecules thereby increasing the possibility of Donor Specific Antibody (DSA) binding. Endothelial cells (ECs) produce IL-6 in the steady-state that is increased by inflammation or by HLA-DR antibody binding in an allogeneic setting. Because IL-6 has been implicated in AMR, IL-6 blockade is currently under investigation as a therapeutic target. To further understand the role of IL-6 in EC immunogenicity, we have examined whether humanized antibody blockade of IL-6 altered EC interactions with allogeneic PBMC and after anti-HLA or DSA binding to ECs in an in vitro human experimental model. Soluble factors, endothelial phenotype, Stat-3 activation, CD4 + -T differentiation and C4d deposition were examined. Blockade of IL-6 reduced EC secretion of IL-6 and of the monocyte chemoattractant MCP-1. Pre-activation of ECs by anti-HLA or DSA binding increased IL-6 secretion, that was further increased by concurrent binding of both antibodies and this was inhibited by IL-6 blockade. Activation of Stat-3 in CD4 + -T mediated by soluble factors produced in endothelial-PBMC interactions, and endothelial differentiation of CD4 + -T cell subsets (Th1, Treg), were impaired whereas activation of Complement by anti-HLA antibody binding remained unchanged by IL-6 blockade. Together, these data identify EC-mediated pro-inflammatory responses (T cell expansion, EC auto-activation, chemokine secretion) targeted by IL-6 blockade., (© 2022 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.)

During allotransplantation, the endothelium acts as semi-professional antigen-presenting cells with the ability to activate proliferation and to promote differentiation of CD4 + -T subsets. These abilities are dependent on the luminal expression of HLA class II antigens by microvascular endothelial cells, which is regulated by inflammatory cytokines. The upregulation of HLA-DR and HLA-DQ during rejection implies significant intragraft inflammation. Furthermore, the microvascular inflammation is an independent determinant for renal allograft failure. In this study, the potential of inflammation to modify endothelial regulation of peripheral CD4 + Treg cells was examined. Microvascular endothelial cells were exposed to pro-inflammatory cytokines for varying durations before co-culture with PBMC from non-HLA matched donors. Proliferation and expansion of CD4 + Treg and soluble factor secretion was determined. Early interactions were detected by phosphorylation of Akt. Video microscopy was used to examine spatial and temporal endothelial-CD4 + T interactions. Highly inflammatory conditions led to increased endothelial expression of HLA-DR, the adhesion molecule ICAM-1, the costimulatory molecule PD-L1 and de novo expression of HLA-DQ. Treg differentiation was impaired by exposure of endothelial cells to a high level of inflammation. Neither IL-6, IL-2 nor TGFβ were implicated in reducing Treg numbers. High PD-L1 expression interfered with early endothelial cell interactions with CD4 + T lymphocytes and led to modified TCR signaling. Blocking endothelial PD-L1 resulted in a partial restoration of Treg. The allogenic endothelial cell-mediated expansion of Treg depends on a critical threshold of inflammation. Manipulation of the PD-L1/PD-1 pathway or endothelial activation post-transplantation may promote or interfere with this intrinsic mechanism of allospecific Treg expansion., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2021 Cross, Lion, Poussin, Glotz and Mooney.)

Development of donor-specific antibodies is associated with reduced allograft survival in renal transplantation. Recent clinical studies highlight the prevalence of human leukocyte antigen (HLA)-DQ antibodies amongst de novo donor-specific antibodies (DSAs), yet the specific contribution of these DSAs to rejection has not been examined. Antibody-mediated rejection primarily targets the microvasculature, so this study explored how patient HLA-DQ alloantibodies can modulate endothelial activation and so immunoregulation. HLA-DQ antibodies phosphorylated Akt and S6 kinase in microvascular endothelial cells. This activation prior to culture with alloreactive lymphocytes increased IL-6 and RANTES secretion. The antibody-mediated upregulation of IL-6 was indeed Akt-dependent. The binding of HLA-DQ antibodies to endothelial cells selectively reduced T cell alloproliferation and FoxP3 high Treg differentiation. In clinical studies, detection of HLA-DQ DSAs with other DSAs is associated with worse graft survival than either alone. Endothelial cells stimulated with HLA-DR and HLA-DQ antibodies showed a synergistic increase in pro-inflammatory cytokine secretion and a decrease in Treg expansion. HLA-DQ antibodies strongly promote pro-inflammatory responses in isolation and in combination with other HLA antibodies. Thus, our data give new insights into the pathogenicity of HLA-DQ DSAs., (Copyright © 2019 International Society of Nephrology. Published by Elsevier Inc. All rights reserved.)

Immunosuppressive treatment is a prerequisite for both organ transplantation and tolerance of the allograft. However, long-term immunosuppression has been associated with a higher incidence of malignancies and infections. Immunosuppressors mainly target circulating immune cells and little is known of their "off-target" effects, such as their impact on endothelial cells (ECs). In chronic antibody-mediated rejection (AMR), the allograft endothelium is a target of damage, histologically detected as transplant glomerulopathy, and which correlates with poor graft survival. Under inflammatory conditions, EC expression of HLA class II antigens can lead to CD4 + -T lymphocyte alloactivation and selective expansion of pro-inflammatory Th17 and pro-tolerance Treg subsets. This response can be modified and preactivation of the EC by HLA-DR antibody binding promoted a proinflammatory Th17 response. However, whether or not immunosuppressors alter EC immunogenicity has not been examined. In alloimmunized patients with AMR, cyclosporine A (CsA) and mycophenolic acid (MPA) are often combined with intravenous immunoglobulins (IVIgs). This study reports changes in the microvascular EC phenotype and function after treatment with CsA, MPA, or IVIg. Both CsA and MPA decreased HLA-DR and increased CD54 expression, whereas IVIg increased HLA-DR expression. Interleukin 6 secretion was reduced by all three immunomodulators. Preincubation of ECs with CsA or MPA limited, while IVIg amplified, Treg expansion. Because CsA, MPA, and IVIg are known for their ability to act upon leukocytes, we confirmed that ECs maintained their immunoregulatory role when allogeneic leukocytes were pretreated with CsA, MPA, or IVIg. The results reveal that individual immunosuppressors, used in the induction and maintenance of renal allograft tolerance, had direct and distinct effects on ECs. Results of experiments associating IVIg with either CsA or MPA underlined the differences observed using individual immunosuppressors. Paradoxically, CsA or MPA may increase EC mediated inflammatory responses and long-term exposure may contribute to limitation of allograft tolerance. In contrast, IVIg interaction with the endothelium may mediate some of its immunosuppressive effects through promotion of Treg expansion, contributing to the maintenance of allograft tolerance.

Hepatocellular adenomas (HCA) are benign liver tumors predominantly developed in women using oral contraceptives. Here, exome sequencing identified recurrent somatic FRK mutations that induce constitutive kinase activity, STAT3 activation, and cell proliferation sensitive to Src inhibitors. We also found uncommon recurrent mutations activating JAK1, gp130, or β-catenin. Chromosome copy number and methylation profiling revealed patterns that correlated with specific gene mutations and tumor phenotypes. Finally, integrative analysis of HCAs transformed to hepatocellular carcinoma revealed β-catenin mutation as an early alteration and TERT promoter mutations as associated with the last step of the adenoma-carcinoma transition. In conclusion, we identified the genomic diversity in benign hepatocyte proliferation, several therapeutic targets, and the key genomic determinants of the adenoma-carcinoma transformation sequence., (Copyright © 2014 Elsevier Inc. All rights reserved.)

Inflammatory hepatocellular adenomas (IHCAs) are benign liver lesions that can be characterized histologically by the presence of an inflammatory infiltrate and at the molecular level by the overexpression of acute phase inflammatory response genes. Recurrent somatic mutations of the interleukin-6 (IL-6) signal transducer ( IL6ST) locus, encoding the critical component of the IL-6 signal transduction machinery gp130, are present in 60% of IHCAs and in a subset (2%) of hepatocellular carcinoma (HCCs). By screening of 256 human hepatic adenoma specimens (the largest genetic analysis of IL6ST performed to date in this setting), we identified 24 distinct somatic IL6ST mutations among 66 mutant adenomas. The functional analysis of nine different gp130 mutants expressed in hepatic cancer cell lines consistently revealed the constitutive and IL-6-independent activation of the JAK/STAT signaling pathway. We further demonstrated that the signaling activity of mutant gp130 in IHCA remains responsive to suppressor of cytokine signaling 3 (SOCS3), a physiological gp130 inhibitor. Specifically, cells expressing a double mutant variant of gp130 with a disrupted SOCS3-binding site at residue 759 (Y186/Y759F) displayed a hyperactivation of signal transducer and activator of transcription 3 (STAT3) as compared with cells expressing the endogenous IHCA-associated Y186 gp130 mutant. Notably, we identified that constitutive signaling via gp130 in IHCA requires the Janus kinase family member JAK1, but not JAK2 or tyrosine kinase 2. In support of this notion, AG490, a tyrosine kinase inhibitor that selectively blocks JAK2, had no effect on gp130 activity. In stark contrast, we showed that ruxolitinib, a JAK1/JAK2-selective tyrosine kinase inhibitor used to treat patients with myelofibrosis, dramatically impaired JAK1-STAT signaling downstream of all IHCA-associated gp130 mutants. In conclusion, our findings provide a rationale for the use of JAK1 inhibitors for the treatment of HCAs expressing mutant gp130 as well as a subset of HCCs that bear similar mutations.

Inflammatory hepatocellular adenomas (IHCAs) are benign liver tumors. 60% of these tumors have IL-6 signal transducer (IL6ST; gp130) mutations that activate interleukin 6 (IL-6) signaling. Here, we report that 12% of IHCA subsets lacking IL6ST mutations harbor somatic signal transducer and activator of transcription 3 (STAT3) mutations (6/49). Most of these mutations are amino acid substitutions in the SH2 domain that directs STAT3 dimerization. In contrast to wild-type STAT3, IHCA STAT3 mutants constitutively activated the IL-6 signaling pathway independent of ligand in hepatocellular cells. Indeed, the IHCA STAT3 Y640 mutant homodimerized independent of IL-6 and was hypersensitive to IL-6 stimulation. This was associated with phosphorylation of tyrosine 705, a residue required for IL-6-induced STAT3 activation. Silencing or inhibiting the tyrosine kinases JAK1 or Src, which phosphorylate STAT3, impaired constitutive activity of IHCA STAT3 mutants in hepatocellular cells. Thus, we identified for the first time somatic STAT3 mutations in human tumors, revealing a new mechanism of recurrent STAT3 activation and underscoring the role of the IL-6-STAT3 pathway in benign hepatocellular tumorigenesis.

Inflammatory hepatocellular adenomas are benign liver tumours defined by the presence of inflammatory infiltrates and by the increased expression of inflammatory proteins in tumour hepatocytes. Here we show a marked activation of the interleukin (IL)-6 signalling pathway in this tumour type; sequencing candidate genes pinpointed this response to somatic gain-of-function mutations in the IL6ST gene, which encodes the signalling co-receptor gp130. Indeed, 60% of inflammatory hepatocellular adenomas harbour small in-frame deletions that target the binding site of gp130 for IL-6, and expression of four different gp130 mutants in hepatocellular cells activates signal transducer and activator of transcription 3 (STAT3) in the absence of ligand. Furthermore, analysis of hepatocellular carcinomas revealed that rare gp130 alterations are always accompanied by beta-catenin-activating mutations, suggesting a cooperative effect of these signalling pathways in the malignant conversion of hepatocytes. The recurrent gain-of-function gp130 mutations in these human hepatocellular adenomas fully explains activation of the acute inflammatory phase observed in tumourous hepatocytes, and suggests that similar alterations may occur in other inflammatory epithelial tumours with STAT3 activation.

Biallelic somatic mutations of TCF1 coding for hepatocyte nuclear factor 1alpha (HNF1alpha) are found in 50% of the hepatocellular adenoma (HCA) cases usually associated with oral contraception. In rare cases, HNF1alpha germ line mutations could also predispose to familial adenomatosis. In order to identify new genetic factors predisposing to HNF1alpha-mutated HCA, we searched for mutations in genes involved in the metabolism of estrogen. For 10 genes (CYP1A1, CYP1A2, CYP3A4, CYP3A5, COMT, UGT2B7, NQO1, GSTM1, GSTP1, and GSTT1), we did not find mutations nor differences in the allele distribution among 32 women presenting HNF1alpha-mutated adenomas compared with 58 controls. In contrast, we identified a CYP1B1 germ line heterozygous mutation in 4 of 32 women presenting HNF1alpha-mutated adenomas compared with none in 58 controls. We confirmed these results with the identification of four additional CYP1B1 mutations in a second series of 26 cases. No mutations were found in the control group, which was extended to 98 individuals, and only a known rare genetic variant was observed in two controls (P = 0.0003). We did an ethoxyresorufin O-deethylase assay to evaluate the functional consequence of the CYP1B1 mutations. We found reduced enzymatic activity in each CYP1B1 variant. In addition, an E229K CYP1B1 mutation was found in a woman with a germ line HNF1alpha mutation in a familial adenomatosis context. In this large family, all three patients with adenomatosis bore both HNF1 and CYP1B1 germ line mutations. In conclusion, our data suggested that CYP1B1 germ line-inactivating mutations might increase the incidence of HCA in women with HNF1alpha mutations.

The expression of casein genes is specific to the mammary gland and maximal during lactation. However, among the numerous mammary cell lines described so far, only a few express some casein genes. The regulatory regions of casein genes have been largely described but the mechanisms explaining the mammary specific expression of these genes, and their silencing in most mammary cell lines, have not yet been fully elucidated. To test the hypothesis that the nuclear location of the casein genes may affect their expression, we transfected HC11 mouse mammary cell line with a 100 kb DNA fragment surrounding the rabbit alpha S1 casein gene. We derived stable clones which express or not the transfected rabbit casein gene, in the same cellular context, independently of the number of transgene copies. Metaphase spreads were prepared from the different clones and the transfected genes were localized. Unexpectedly, we observed that in the original HC11 cell line the number of chromosomes per metaphase spread is close to 80, suggesting that HC11 cells have undergone a duplication event, since the mouse karyotype is 2n = 40. In alpha S1 casein expressing cells, the expression level does not clearly correlate with a localization of the transfected DNA proximal to the centromeres or the telomeres. Analysis of the localization of the transfected DNA in nuclear halos allows us to conclude that when expressed, transfected DNA is more closely linked to the nuclear matrix. The next step will be to study the attachment of the endogenous casein gene in mammary nuclei during lactation., ((c) 2005 Wiley-Liss, Inc.)

We have examined whether the effects of singlet oxygen (1O2) produced by photodynamic action on the mitochondrial permeability transition (PT) can be modulated by the localization of photosensitizers in irradiated mitochondria. We have previously shown that oxidation due to 1O2 photogenerated in hematoporphyrin (HP)-loaded mitochondria can prevent opening of the PT pores, likely after degradation of some critical histidines (Salet et al, 1997, J. Biol. Chem. 272, 21938-21943). Equally, in the present study we have irradiated mitochondria in the presence of a structurally different photosensitizer producing 1O2, namely 4,5',8-trimethylpsoralen (TMP). Fluorescence studies show that TMP binds to protein sites which differ from those of HP. In sharp contrast with HP, TMP-driven photodynamic action triggers per se pore opening. Interestingly, this inducing effect is inhibited when TMP-treated mitochondria are irradiated after addition of mersalyl, a specific reagent protecting thiol groups of the inner mitochondrial membrane that are oriented toward the external hydrophilic phase. This fact suggests that 1O2-mediated thiol oxidation is responsible for TMP-photoinduced pore opening. Taken together, these findings suggest that 1O2 can activate or inactivate a cellular function such as mitochondrial PT depending on the site where it is produced in the mitochondrial membrane.

Chronic infection by HBV is the leading cause of hepatocellular carcinoma in man. Several lines of evidence suggest that the viral transactivator HBx plays a critical role in the molecular pathogenesis of HBV-related HCC. To study the actual impact of HBx and the mechanism of its action, we have recently cloned and characterized a set of X-sequences from HCC in patients with chronic infection by HBV. In the present study, we have compared the effects of HBx and its naturally arising mutants on cell growth and viability. We report that HBx inhibits clonal outgrowth of cells and induces apoptosis by a p53-independent pathway. Furthermore, HBx expression induced a late G1 cell cycle block prior to their counterselection by apoptosis. Importantly, mutations in the HBx-gene evolving in hepatocellular carcinoma abolished both HBx-induced growth arrest and apoptosis. Using a panel of engineered mutants we have mapped the growth suppressive effect of HBx to domains shown to be required for its transactivating function. Based on these results, we propose that abrogation of the anti-proliferative and apoptotic effects of HBx by naturally occurring mutations might render the hepatocytes susceptible to uncontrolled growth and contribute to multistep hepatocarcinogenesis associated with HBV-infection.

We compared the efficacy of 4 methods for isolating circulating tumor cells: immunocapture with Ber-EP4-coated magnetic beads, density gradient separation, ammonium chloride, and distilled water-mediated erythrocyte lysis. Human blood from healthy volunteers was mixed with serial dilutions of prostate (LNCaP) and liver (HepG2) derived tumor cells. Isolation of circulating tumor cells was followed by reverse transcriptase-polymerase chain reaction with primers specific for prostate-specific antigen and alpha-fetoprotein. Ber-EP4 antigen expression was evaluated by immunohistochemistry in 27 hepatocellular carcinomas and 34 prostate adenocarcinomas. Peripheral blood samples from 12 patients with hepatocellular carcinoma and 10 with prostate adenocarcinoma also were tested. Density gradient separation and Ber-EP4 immunocapture were the most sensitive techniques for isolating circulating tumor cells in in vitro tests. Isolation by density gradient separation was significantly more sensitive than Ber-EP4 immunocapture when applied to peripheral blood samples of patients with cancer, a result consistent with the variable expression of Ber-EP4 antigen that we found by immunohistochemistry in prostate adenocarcinomas and hepatocellular carcinomas.

We have investigated whether liver resection and needle liver biopsy cause dissemination of liver cells into peripheral blood circulation, using a reverse-transcription polymerase chain reaction (RT-PCR)-based assay targeted against alpha-fetoprotein (AFP) mRNA. Twelve patients with and 16 without primary liver cancer (PLC) undergoing liver resection were tested before skin incision, after liver mobilization, after hepatic parenchyma transection, after abdominal wall suture, and 4 days after surgery. Two patients with and 20 without PLC were tested before, 20 minutes after, and 24 hours after needle liver biopsy. Six of 14 patients with and 0 of 36 patients without PLC scored positive before intervention (P <.001). Liver cell spreading was induced at different times after surgery and liver biopsy in 14 of 14 patients with but also 23 of 36 without PLC (P <.05). We conclude that liver resection and needle liver biopsy induce release of cells from the liver, which are not necessarily liver tumor cells, into the peripheral blood circulation. This may be, however, an important mechanism of liver cancer cell dissemination deserving further investigations.

To explore the role of hepatitis B virus (HBV) X protein in liver carcinogenesis, independently from its role in viral replication, we have analyzed X gene structure and expression in tumorous and non-tumorous tissues obtained from 9 hepatitis B surface antigen (HBsAg)-negative, HBV DNA-positive patients. HBV replication was undetectable in tumorous tissues. HBV X gene was truncated at its 3' end in 5 of 9 tumorous tissues and 1 of 8 non-tumorous livers. Sequence analysis performed on uninterrupted X genes from 3 tumors and 3 surrounding non-tumorous tissues showed a high rate of mutations, selectively in the tumorous livers. In 1 of the 3 tumors, a frameshift mutation induced a new stop at codon 129. HBV RNAs were tested by reverse transcriptase-polymerase chain reaction (RT-PCR) with surface (S), core (C) and X specific primers. X, but not S and C, RNA expression was found in 6 of 8 tumors and in 6 of 7 non-tumorous tissues. This finding was consistent with immunohistochemical detection of X, but not S and C, antigens in all tumors also expressing X RNA. Our results provide evidence for selective expression of HBV X, but not S and C, RNA and protein in the tumorous and non-tumorous tissue of HBsAg-negative, HBV DNA-positive patients. It also shows that the structure of the X gene is modified (interrupted or highly mutated) in the majority of tumorous livers. Taken together, our findings are consistent with a potential role of mutated X proteins in HBV-related liver oncogenesis.

The rapid and reproducible identification of new cellular DNA sequences adjacent to known sequences is difficult to achieve with the currently available procedures. Here we describe a novel approach based on the polymerase chain reaction (PCR) using a primer specific to the known sequence and another directed to a human Alu repeat. To avoid undesirable amplifications between Alu sequences, primers are constructed with dUTPs and destroyed by uracil DNA glycosylase treatment after 10 initial cycles of amplification. Only desirable fragments are then further amplified with specific primers to the known region and to a tag sequence introduced in the Alu-specific primer. Using this protocol, we have successfully identified cellular sequences flanking integrated hepatitis B virus DNA from the human genome of three hepatoma tissues. The method enables a direct specific amplification without any ligation or nonspecific annealing steps as required by previous PCR-based protocols. This rapid and straightforward approach will be a powerful tool for the study of viral integration sites, but is also widely applicable to other studies of the human genome.

In HBsAg-negative patients with hepatocellular carcinoma (HCC), hepatitis B virus (HBV) genomes are present at a low copy number per cell, and the role of HBV in liver transformation is still unclear. We have mapped by polymerase chain reaction (PCR) the HBV genome in 19 HBsAg-negative tumorous and 9 corresponding nontumorous tissues and evaluated, by RT-PCR, the presence of HBV S, X, and C transcripts in the tumorous and nontumorous tissue of nine HBsAg-negative and, for comparison, six HBsAg-positive patients. Disrupted, presumably integrated, HBV genomes were detected by PCR in 10 of 19 tumorous tissues and in only one of nine nontumorous tissues. Significant accumulation of viral RNAs containing X but not C or S sequences was shown in 7/9 tumors and 7/8 nontumorous tissues from HBsAg-negative patients. In contrast, viral RNAs revealed by X-as well as by S- and C-specific primers were detected in five of six tumors and in six of six nontumorous tissues from HBsAg-positive patients. In conclusion, our results suggest the frequent integration of the HBV genome and the accumulation of X-related RNAs in HCCs developing in HBsAg-negative patients. This finding is consistent with a role, in these cases, for the potentially transforming X protein.