Sie sind auf Seite 1von 10

Journal. .of . .General. .Virology. . .(2001),. . .82,. . 20612070.. . . . . Printed. .in. . Great. .Britain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................. .... ................. .................

............. ..... ....................... ............ .. ......... ..........

Efcient downregulation of major histocompatibility complex class I molecules in human epithelial cells infected with cytomegalovirus
Christine Benz, Uwe Reusch, Walter Muranyi, Wolfram Brune, Ramazan Atalay and Hartmut Hengel
Max von Pettenkofer-Institut, Lehrstuhl Virologie, Ludwig-Maximilians-Universitat Munchen, Pettenkoferstrasse 9a, 80336 Munchen, $ $ $ Germany

Liver and intestinal epithelial cells are a major target of infection by cytomegaloviruses (CMV), causing severe disease in affected organs of immunocompromised patients. CMV downregulates major histocompatibility complex class I (MHC-I) molecule expression in broblasts in order to avoid lysis by CD8M cytotoxic T lymphocytes. However, MHC-I expression in human cytomegalovirus (HCMV)-infected hepatic tissue was reported to be increased. As it is unclear at present whether HCMV affects MHC-I expression in epithelial cells, new cell culture models for HCMV infection of differentiated hepatobiliary cell lines were established. HCMV immediate early gene expression was achieved in 60 to 95 % of cells. Progression of the HCMV replication cycle differed from prototypic infection of broblasts, since structural early and late proteins were produced at low levels and HCMV progeny yielded much lower titres in hepatobiliary cells. In contrast, HCMV glycoproteins, gpUS2, gpUS3, gpUS6 and gpUS11, that downregulate MHC-I expression were synthesized with temporal kinetics and in a similar quantity to that seen in broblasts. As a result, HCMV infection led to a drastic and selective downregulation of MHC-I expression in epithelial cells and was uniformly observed irrespective of the hepatic or biliary origin of the cells. The new models document for the rst time a stealth function of HCMV in epithelial cells and indicate that the downregulation of MHC-I expression by HCMV can occur in the virtual absence of virus replication.

Introduction
Cytomegaloviruses (CMV) represent prototype viruses of the -subgroup of the family Herpesviridae. Human cytomegalovirus (HCMV) gene expression in infected organs has been documented in a broad array of cell types and tissues, in particular in epithelial and endothelial cells, but also in broblasts, smooth muscle cells and macrophages (Sinzger et al., 1995, 1999). HCMV infection in the immunocompetent
Author for correspondence : Hartmut Hengel. Present address : Robert Koch-Institut, Division of Viral Infections, Nordufer 20, 13353 Berlin, Germany. Fax j49 1888 754 2328. e-mail hengelh!rki.de Present address : Medizinische Universitatskliniken, Joseph$ Stelzmann-Strasse 9, 50931 Koln, Germany. $ Present address : Robert Koch-Institut, Division of Viral Infections, Nordufer 20, 13353 Berlin, Germany.

host does not in most cases cause a signicant illness, although a mononucleosis-like syndrome, including a benign cause of hepatitis, has been observed (Snover & Horwitz, 1984 ; Ten Napel et al., 1984). Immune control does not achieve complete clearance of the virus. Instead, the viral genome persists in a nonproductive form. From latency, CMV replication is frequently reactivated and can result in recurrent infection and virus shedding. Impairment of cellular immune control represents a prime risk factor for CMV recurrence and disease (Britt & Alford, 1996). In AIDS patients, persistent virus replication leads to tissue injury in some but not all organs, of which the colon represents the most common gastrointestinal site of disease (Burako & Eglow, 1995). In organ transplant recipients, CMV infection frequently complicates post-transplantation (Paya et al., 1989 ; Arnold et al., 1992 ; Lautenschlager et al., 1997). After liver transplantation, CMV may cause severe hepatitis and cholangitis, leading to allograft dysfunction and, if untreated, to death (Stratta et al., 1989 ; Lautenschlager et al., 1990).
CAGB

0001-7473 # 2001 SGM

C. Benz and others

Table 1. Efciency of HCMV infection of epithelial cell lines and HFF cells
Cells were infected with HCMV strain AD169 at an m.o.i. of 10, as described in Methods. Immunouorescence microscopy was performed 72 h p.i. after staining the cells with an IE1\pp72-specic MAb. Evans blue was used as the counterstain. A total of 100 individual cells was analysed in at least two independent experiments. IE1/pp72-positive cells (%) 95 90 80 75 60 20 20 10

Cell line HFF Huh7 MzChA2 Hep3B Caco2 MzChA1 PLC\PRF\5 HepG2

Description Primary broblast Hepatoma Biliary carcinoma Hepatoma Colon carcinoma Biliary carcinoma Hepatoma Hepatoma

Source Hengel et al. (1995) Nakabayashi et al. (1982) Knuth et al. (1985) Aden et al. (1979) ; HB-8064 (ATCC) HTB-37 (ATCC) Knuth et al. (1985) Gabay et al. (1995) HB-8065 (ATCC)

Previous work has demonstrated that CMV downregulates the surface expression of major histocompatibility complex class I (MHC-I) molecules (Del Val et al., 1992 ; Beersma et al., 1993 ; Yamashita et al., 1993). In productively infected broblasts, presentation of antigens to MHC-I-restricted CD8+ T cells is abolished (Del Val et al., 1992 ; Warren et al., 1994 ; Hengel et al., 1995). Analysis of HCMV deletion mutants led to the identication of genes within the short segment of the HCMV genome, i.e. US2, US3, US6 and US11, all of which can mediate this eect independently (Jones et al., 1995, 1996 ; Hengel et al., 1996, 1997 ; Wiertz et al., 1996 a, b ; Ahn et al., 1996, 1997 ; Lehner et al., 1997). These genes belong to two related gene families encoding type I transmembrane glycoproteins. The common phenotype of transfected cells expressing these genes is a drastic reduction of MHC-I molecules on the cell surface. HCMV proteins employ sophisticated molecular mechanisms to interrupt the MHC-I pathway at distinct checkpoints. gpUS2 and gpUS11 displace nascent MHC-I molecules from the endoplasmic reticulum (ER) back to the cytosol for rapid proteolytic degradation by proteasomes (Wiertz et al., 1996 a, b). gpUS3 was shown to retain assembled MHC-I complexes in the ER, thus preventing transport to the cell surface (Ahn et al., 1996 ; Jones et al., 1996). The US6-encoded glycoprotein does not directly interact with MHC-I molecules, but inhibits the transporter associated with antigen processing-mediated transport of cytosolic peptides into the ER, thereby blocking the assembly of the peptide containing ternary MHC-I complexes (Hengel et al., 1997 ; Ahn et al., 1997 ; Lehner et al., 1997). To cover all phases of the protracted HCMV replication cycle, the synthesis of genes that control MHC-I surface expression is tightly regulated. Whereas the immediate early (IE) gene US3 is transcribed from 3 to 6 h post-infection (p.i.) (Ahn et al., 1996), US2 and US11 expression peaks in the early phase, between 24 and 48 h p.i., of virus replication (Jones & Muzithras, 1991). gpUS6 is the
CAGC

most abundant protein in the late phase of the HCMV replication cycle (Hengel et al., 1997). While several studies have demonstrated downregulation of MHC-I molecules in productively CMV-infected broblasts leading to the abrogation of T cell recognition (Del Val et al., 1992 ; Beersma et al., 1993 ; Yamashita et al., 1993), presentation of murine CMV (MCMV) peptides by MCMV-infected macrophages to CD8+ cytotoxic T lymphocytes (CTL) was highly ecient and MHC-I expression in infected macrophages was essentially unaected (Hengel et al., 2000). Apparently, the type of cell used to analyse CMV infection may profoundly aect many characteristics of the infection process, including viral gene expression and protein tracking. Previous reports have suggested an increase of MHC-I expression in HCMVinfected liver (Steinho et al., 1988 ; Arnold et al., 1991, 1992) and biliary epithelial cells in vitro (Scholz et al., 1997). We have established new cell culture models for HCMV infection of liver and biliary cells to analyse the fate of MHC-I molecules. The data document a strong and selective loss of MHC-I assembly and surface expression on infected epithelial cells. Despite a very low level of HCMV replication, all HCMV proteins that aect MHC-I expression were eciently synthesized with similar kinetics to those seen in broblasts, indicating that MHC-I modulation is not linked to productive infection.

Methods

Cells. MRC5 foetal human lung broblasts (passages 620) (BioWhittaker) were grown in Dulbeccos minimal essential medium (DMEM) (GIBCO BRL) supplemented with 10 % foetal calf serum (FCS), penicillin, streptomycin and 2 mM glutamine. Human foreskin broblast (HFF) cells were established following standard procedures by mincing the skin and subsequently stirring it in pre-warmed trypsin. Recovered broblasts were passaged in DMEM supplemented with 10 % FCS, penicillin, streptomycin and 2 mM glutamine and used between culture

MHC-I in CMV-infected epithelial cells


(i) (a) (ii)

(b)

(c)

Fig. 1. HCMV gene expression in HFF, Huh7 and MzChA2 cells. HFF (a), Huh7, (b) and MzChA2 (c) cells were infected with HCMV strain AD169 at an m.o.i. of 5. At 24 h p.i., monolayers were xed, permeabilized and labelled for the presence of nuclear IE1/pp72 antigen using a pp72specic MAb and an FITC-conjugated goat anti-mouse IgG. The distribution of pp72 was visualized by confocal laser scanning microscopy (i) and the cell morphology of the same cells visualized by counterstaining with Evans blue (ii) is shown.

passages 6 and 20. Human hepatomas HepG2 (HB-8065 ; ATCC) and PLC\PRF\5 (Gabay et al., 1995), dierentiated human hepatocellular carcinoma Huh7 (Nakabayashi et al., 1982), less-well dierentiated biliary adenocarcinoma MzChA2 and dierentiated bile duct carcinoma MzChA1 (Knuth et al., 1985) cells were passaged in DMEM. Dierentiated human hepatocellular carcinoma Hep3B (HB-8064 ; ATCC) (Aden et al., 1979) and colon carcinoma Caco2 (HTB-37 ; ATCC) cells were passaged in RPMI 1640 medium supplemented with 10 % FCS, penicillin, streptomycin, 2 mM glutamine, nonessential amino acids and 1 mM sodium pyruvate (GIBCO BRL).
Viruses, infection conditions and virus titration. Virus stocks were prepared by propagating HCMV strains AD169, Towne, Toledo and a low passage clinical isolate, UL1271, derived from a congenitally CMV-infected infant in MRC5 cells. Infectious supernatants were harvested when 100 % of cells showed cytopathic eects. Virus titres were determined by standard plaque assay ; the titre of each virus stock was between approximately 1i10( and 5i10( p.f.u.\ml. HCMV infection was enhanced by centrifugation at 800 g for 30 min. To achieve cell-cycle synchronization, subconuent monolayers of HFF, Huh7, Hep3B, MzChA2 and Caco2 cells were maintained in serum-free DMEM for 24 h before cells were infected with HCMV at an m.o.i. of 5 or 10. In all experiments, control settings were included to ensure IE1\pp72 expression, which was assessed by immunouorescence microscopy

(Table 1). Cells were infected with HCMV strain AD169 at an m.o.i. of 5 to determine the growth kinetics of HCMV in HFF and epithelial cells. Input virus was removed by exchanging the culture medium. Supernatant and cells were collected at the indicated time-points p.i. After homogenizing the cells, HCMV infectivity was determined in p.f.u.\ml by endpoint titration in MRC5 broblasts.
Reagents and antibodies. The antibodies used in this study are : Monoclonal antibody (MAb) W6\32 (HB-95 ; ATCC) recognizing MHCI heavy chain\ dimers (Parham et al., 1979) ; polyclonal rabbit antisera # recognizing calnexin (StressGen Biotechnologies) ; anti-HCMV gpUS6 MAb (Hengel et al., 1997) ; anti-HCMV gpUS2 MAb (a kind gift from E. Wiertz, Bilthoven, The Netherlands ; Wiertz et al., 1996 b) ; anti-HCMV gpUS11 and gpUS3 MAbs (a kind gift from K. Fruh, San Diego, CA, $ USA ; Ahn et al., 1996) ; anti-HCMV IE antigen pp72\pp86 MAb (Du Pont NEN) ; anti-HCMV IE antigen pp72 MAb p63-27 (Andreoni et al., 1989) ; anti-HCMV IE antigen pp86 MAb 295 (Plachter et al., 1993) ; antiHCMV glycoprotein B (gB) MAb 27156 (Spaete et al., 1988) ; antiHCMV pp28 MAb (ABI) ; anti-HCMV major capsid protein (MCP) UL86 MAb 28-4 (Sanchez et al., 1998) ; uorescein-conjugated mouse anti-IE antigen pp72 MAb (Chemicon) ; anti-intercellular adhesion molecule-1 (ICAM-1) (CD54) MAb ; anti-CD44 MAb ; anti-CD49a MAb (Dianova) ; uorescein-conjugated goat anti-mouse IgG (Sigma) ; indigocarbocyanin (Cy5)-conjugated goat anti-mouse IgG (Dianova) ; and peroxidaseconjugated aniPure goat anti-mouse IgG (HjL) (Dianova).
Western blot analysis. Following serum starvation, cells were either infected or mock-infected with virus at an m.o.i. of 10. Cells were then detached with trypsin, washed in ice-cold PBS, pelleted and lysed in lysis buer containing 0n125 M Tris, pH 6n8, 2 % SDS, 20 % glycerol, 1 % mercaptoethanol and 0n5 % bromophenol blue. After heating the samples at 95 mC for 5 min, lysates were separated by 10 % SDSPAGE, transferred to nitrocellulose and probed with the IEpp72\pp86 and anticalnexin MAbs, respectively. Peroxidase-conjugated antibodies, used as the secondary reagent, were detected with the ECL plus detection kit (AmershamPharmacia).
Flow cytometry. To analyse MHC-I surface expression, HFF, Huh7, Hep3B and Caco2 cells were infected with HCMV strain AD169 at an m.o.i. of 10 and incubated for 3 days. After harvesting, cells were pre-incubated in 10 % goat serum (Dianova) and immunostained with MAb W6\32. As a control, cells were immunostained with an antiICAM-1 MAb. Bound antibodies were visualized by the addition of uorescein-conjugated goat anti-mouse IgG. As a negative control, cells were incubated with only the uorescein-conjugated goat anti-mouse IgG antibody. A total of 10% cells was analysed for each uorescence histogram on a FACScan IV (Becton Dickinson). Double-stain analysis of uorescein-conjugated MHC-I surface expression and HCMV IE1\pp72 nuclear expression was performed as follows : 3 days after infection at an m.o.i. of 10, cells were harvested, pre-incubated with 10 % goat serum and immunostained with MAb W6\32. Bound antibodies were visualized by the addition of Cy5-conjugated goat anti-mouse IgG antibody and xed in 3 % paraformaldehyde. After permeabilization in cold methanol, cells were centrifuged and resuspended in 0n2 ml nuclear extraction buer (1 M NaCl, 0n5 M MgSO , 0n6 % NP-40, 2 % FCS and 0n01 M PBS) % (Thornthwaite et al., 1980) for 2 min. The cells were then immunostained with the uorescein-conjugated anti-IE1\pp72 MAb and xed in 3 % paraformaldehyde. As a negative control, cells were incubated with only the Cy5-conjugated goat anti-mouse IgG antibody. A total of 10% cells was analysed for each uorescence prole. Discrimination between expression levels was performed with quadrant statistics using the Cell Quest software (Becton Dickinson). After 72 h of incubation, Huh7 and HFF cells were harvested and double-stained for MHC-I surface expression and IE1\pp72 nuclear expression. CAGD

C. Benz and others


Cloning procedures and transfectants. The HCMV US6 and US11 genes were cut from pcDNAI constructs (Hengel et al., 1996, 1997) and subcloned into the SalI site of the pB45-Neo vector (Ohe et al., 1995). Huh7 cells were transfected with plasmid DNA using the SuperFect transfection reagent (Qiagen). Cell clones were selected in the presence of 0n5 mg\ml G418 and tested for gpUS6 and gpUS11 protein expression, respectively, with specic antibodies.
Metabolic labelling and immunoprecipitation. Subconuent monolayers of HFF, Huh7 and MzChA2 cells were infected with HCMV strain AD169. At 24, 48 and 72 h p.i., MHC-I complexes and HCMV proteins gpUS2, gpUS3 and gpUS11 were immunoprecipitated. HCMVinfected cells were washed and starved in cysteine- and methionine-free RPMI 1640 before labelling with [$&S]methionine and [$&S]cysteine (1200 Ci\mmol) (AmershamPharmacia) at a concentration of 500 Ci\ml. After 1 h, cells were washed and lysed in lysis buer (140 mM NaCl, 20 mM Tris, pH 7n6, 5 mM MgCl and 0n2 mM PMSF) # with 1 % NP-40. As a control, mock-infected cell lysates were precipitated with each antibody. Detection of gpUS6 and gpUS11 was controlled by the precipitation of proteins from lysates of Huh7 cells stably transfected with pB45US6 and pB45US11. The incorporation of $&S into proteins was quantied in all experiments by liquid scintillation counting of the lysate. All lysates used were adjusted to ensure comparability in quantitative terms. After the removal of cell nuclei by centrifugation, lysates were pre-cleared with pre-immune normal mouse serum and protein ASepharose. Immune complexes were eluted with sample buer and analysed by 1113 % SDSPAGE. After drying, gels were exposed to Kodak BioMax MS lms at k70 mC for 17 days.

develop new cell culture models for HCMV infection of cells with obvious pathophysiological signicance, we systematically screened a panel of epithelial cell lines derived from the liver, gut, kidney, exocrine glands and placenta for their susceptibility to HCMV infection. The expression of nonstructural HCMV IE1\pp72 nuclear antigen was visualized by immunouorescence using an IE1\pp72-specic MAb. The proportion of pp72-positive cells varied widely between cell lines, exceeding 50 % in the case of dierentiated Huh7 and Hep3B hepatoma cells and less-well dierentiated biliary adenocarcinoma MzChA2 cells. To further improve eciency, selected parameters of infection were modied. Among these, serum starvation leading to the synchronization of the cell cycle and centrifugal enhancement of infection signicantly increased the number of cells infected with HCMV. Depending on the cell line, between 60 and 95 % of cells were found to be IE1\pp72-positive (Table 1 and Fig. 1).
Viral protein synthesis and productivity of HCMV-infected epithelial cells

Results
Establishment of cell culture models for HCMV infection in epithelial cells from the liver and colon

Tissue tropism of HCMV in vivo is broad and includes epithelial cells from dierent organs (Sinzger et al., 1995). To

Biosynthesis of HCMV marker proteins of the IE, early and late replication phase was analysed. Using Western blots, IE1\pp72 was detected in AD169-infected Huh7, MzChA2 and Hep3B cells at similar levels from 12 h p.i. to 168 h p.i., similar to that seen in permissive MRC5 broblasts (Fig. 2 a). IE2\pp86 expression in MRC5 broblasts continued to increase over time. In MzChA2 and Huh7 cells, the level of IE2\pp86 required for its detection in Western blots was not reached (Fig. 2 a). In AD169-infected Hep3B cells, IE2\pp86 was detected between 24 and 168 h p.i., albeit at signicantly

(a)

(b)

(c)

(d)

Fig. 2. Synthesis of HCMV proteins in infected epithelial cells. (a) MRC5 broblasts, HFF, Huh7, Hep3B and MzChA2 cells were infected with HCMV strain AD169 at a m.o.i. of 10. Cell lysates were harvested at the indicated time-points and adjusted for protein content. IE1/pp72 and IE2/pp86 were detected with IE-specic MAbs in Western blots. As a control, calnexin was detected using a specic antiserum. At 72 h p.i., cells were metabolically labelled for 60 min and lysates precipitated with antibodies specic for gB (b), pp65 (c) and pp28 (d) before being separated by SDSPAGE. Note the background band in (d) migrated more slowly than the band corresponding to pp28, which is only detectable in HCMV-infected HFF cells.

CAGE

MHC-I in CMV-infected epithelial cells


(a)

derived from a congenitally infected infant. As depicted in Fig. 3, productivity of infection was not increased compared with AD169. From these data, we concluded that infection of Huh7, Hep3B and MzChA2 cells with HCMV is ecient, although productivity is signicantly less (Hep3B, MzChA2) or virtually abortive (Huh7) compared with broblasts.
HCMV infection prevents surface expression and assembly of MHC-I complexes in hepatobiliary and colonic cells

(b)

Fig. 3. Single-step growth curves of HCMV strain AD169 in HFF, Huh7, Hep3B and MzChA2 cells. Subconuent monolayers were infected with HCMV strain AD169 (a) or the clinical isolate UL1271 (b) at an m.o.i. of 5 for 2 h before the culture medium was changed. Cells and supernatants of duplicate cultures were harvested at the indicated time-points. Cellassociated virus was released by homogenization before titration on MRC5 broblasts. Titres represent the mean of three replicates.

lower levels as compared to MRC5 cells. This type of result was also found during immunoprecipitation studies of metabolically labelled cells using IE1\pp72- and IE2\pp86-specic antibodies (data not shown). In metabolically labelled cells, expression of the matrix phosphoprotein UL83\pp65 (Fig. 2 c) and UL86\MCP (data not shown) was clearly lower in all epithelial cells as compared to broblasts, while synthesis of gB in epithelial cells reached almost similar amounts to those seen in broblasts (Fig. 2 b), but depended on the virus titre used for infection. To test the biosynthesis of a true late protein, UL99\pp28 was precipitated 4 days p.i. While readily found in broblasts, pp28 synthesis could not be detected in any of the epithelial cells (Fig. 2 d ). These ndings suggested that the HCMV replication cycle may be less frequently completed in epithelial cells compared with broblasts. Quantitative assessment of AD169 progeny in epithelial cells revealed a more delayed kinetics of virus production in liver and colonic cell lines and reached titres of varying magnitude, but were signicantly lower than levels seen in broblasts (Fig. 3). To test whether infection of epithelial cells with a low passage clinical isolate strain of HCMV might be more productive, we plotted single-step growth curves of UL1271, a clinical isolate

HCMV infection of broblasts results in a strong and selective downregulation of MHC-I surface expression (Beersma et al., 1993 ; Yamashita et al., 1993 ; Warren et al., 1994 ; Hengel et al., 1995). In contrast, immunocytochemical analysis of MHC-I expression in the liver of CMV-infected patients (Steinho et al., 1988 ; Arnold et al., 1991, 1992) and cytouorometric analysis of infected biliary epithelial cells (Scholz et al., 1997) suggested an upregulation of MHC-I expression, implying that MHC-I expression in these cells was resistant to the eects of HCMV. In view of the conicting reports, we determined MHC-I surface expression in Huh7, Hep3B and MzChA2 cells during the course of HCMV infection. Similar to HFF cells, all epithelial cells had a large decrease in MHC-I surface density compared with mockinfected cells following AD169 infection (Fig. 4 a). Loss of MHC-I surface expression in hepatobiliary and colonic cells was almost as ecient as that seen in HFF cells and occurred selectively, since other molecules expressed on the cell surface, e.g. ICAM-1, CD44 and CD49a, were essentially not aected by HCMV infection (Fig. 4 a ; data not shown). Likewise, MHC-I downregulation was observed in epithelial cells infected with the HCMV strains Towne and Toledo (data not shown). In HFF cells, ICAM-1 surface expression was upregulated following HCMV infection, as reported previously (Ito et al., 1995). To analyse the fate of newly synthesized MHC-I molecules, the formation of MHC-I complexes in the ER was compared in AD169-infected HFF, Huh7 and MzChA2 cells during the course of infection. Metabolic labelling and immunoprecipitation with the conformation-dependent MAb W6\32 revealed a higher level of constitutive MHC-I complex assembly in HFF cells compared with hepatobiliary cells (Fig. 4 b). As expected, MHC-I complexes are greatly diminished in HFF cells upon HCMV infection and remained at low levels throughout the complete replication cycle. This phenotype was perfectly reproduced in HCMV-infected Huh7 and MzChA2 cells (Fig. 4 b) as well as in Hep3B cells (data not shown) at all time-points p.i, suggesting that genes that downregulate MHC-I expression, other than IE US3, are synthesized. To prove that cells exhibiting a reduced MHC-I phenotype were indeed HCMV-infected, double-staining for HCMV IE1\pp72 and surface MHC-I molecules was performed. As seen in Fig. 5, pp72 antigen was detected in 97n9 % of Huh7 cells and in 92n2 % of HFF cells. Among Huh7 cells,
CAGF

C. Benz and others


(a) (b)

Fig. 4. HCMV infection downregulates MHC-I surface expression in Huh7, Hep3B and MzChA2 cells. (a) HFF, Huh7, Hep3B and MzChA2 cells were infected with HCMV strain AD169 at an m.o.i. of 10. At 72 h p.i., mock-infected (thin lines) and infected (bold lines) cells were harvested and stained with MAb W6/32, specic for MHC-I complexes. Bound antibodies were visualized by the addition of FITC-conjugated goat anti-mouse IgG. As a negative control, mock-infected (dotted lines) and HCMV-infected (dashed lines) cells were stained with the second antibody alone. Likewise, cells were stained with an ICAM-1specic MAb (CD54). (b) Inhibition of MHC-I complex formation by HCMV in hepatobiliary cells. HFF, Huh7 and MzChA2 cells were infected with HCMV strain AD169 at an m.o.i. of 10 for 24, 48 and 72 h or left untreated before metabolic labelling for 60 min. Cell lysates were precipitated with MAb W6/32 and immunoprecipitates were separated by 11n513n5 % gradient SDSPAGE. MHC-I heavy chains (HC) and 2m are indicated by arrowheads.

eciently controls MHC-I expression in infected liver epithelial cells, despite comparatively poor or absent HCMV replication.
Sequential expression of HCMV proteins that affect MHC-I expression

Fig. 5. Huh7 cells exhibiting a reduced MHC-I phenotype express HCMV IE1/pp72. Huh7 and HFF cells were infected with HCMV strain AD169 at an m.o.i. of 10 or mock-infected. After 72 h, cells were harvested and stained with MHC-I-specic MAb W6/32 and Cy5-conjugated goat antimouse IgG. Next, cells were xed, permeabilized and incubated with an FITC-conjugated IE1/pp72-specic MAb. Quadrant statistics were performed using the Cell Quest software. The percentage of cells is indicated in each rectangle.

80n7 % were pp72-positive and exhibited a reduced MHC-I phenotype, whereas 88n8 % of all HFF cells showed this pattern. Taken together, these data show that HCMV
CAGG

In highly permissive broblasts, HCMV genes that downregulate MHC-I expression are synthesized in a multi-step order (Jones & Muzithras, 1991 ; Ahn et al., 1996). Therefore, we analysed the synthesis and kinetics of glycoproteins aecting MHC-I expression in weakly permissive Huh7 and MzChA2 cells. Expression of gpUS3 shows IE kinetics and peaks a few hours after infection of broblasts (Ahn et al., 1996). Infection at an m.o.i. of 10 led to the synthesis of both US3 products of 23 and 17 kDa, respectively, at similar levels in Huh7 and HFF cells, but at a slightly lower level in MzChA2 cells. Likewise, the expression of gpUS11 and gpUS2 in Huh7, MzChA2 (Fig. 6) and Hep3B cells (data not shown) was comparable to that seen in HFF cells after 24 and 48 h of infection and decreased at later time-points in each cell type. Moreover, synthesis of gpUS11 in infected Huh7 cells exceeded that in a stable Huh7-US11 transfectant (Fig. 6) showing a strong reduction of MHC-I molecules on the cell surface (data not shown). gpUS6 is characterized by earlylate kinetics, reaching the highest expression levels 72 h p.i. (Hengel et al., 1997). Remarkably, the 21 kDa gpUS6 early late protein was detected at comparable amounts in HFF, Huh7 and MzChA2 cells, reaching levels similar to those found in a stable Huh7-US6 transfectant (Fig. 6). Overall, the rate of synthesis of

MHC-I in CMV-infected epithelial cells

(a)

(b)

(c)

Fig. 6. Synthesis of HCMV proteins that regulate MHC-1 expression in infected hepatobiliary epithelial cells. HFF, Huh7 and MzChA2 cells were infected with HCMV strain AD169 at an m.o.i. of 10 for the indicated time or left untreated before metabolic labelling for 60 min. Cell lysates were precipitated with specic antibodies and immunoprecipitates were separated by SDSPAGE. The HCMV proteins gpUS3 (a), gpUS11 and gpUS2 (b) and gpUS6 (c) are indicated. For comparison, stable Huh7-pB45US11 and Huh7-pB45US6 transfectants as well as vector controls were included. The asterisk indicates a 17 kDa protein encoded by the US3 gene transcription unit.

MHC-I downregulating proteins correlated with the virus titre used for infection. In conclusion, all HCMV proteins that aect MHC-I expression are abundantly synthesized in infected hepatobiliary epithelial cells irrespective of their temporal classication.

Discussion
In this study, we have demonstrated that HCMV infection inhibits MHC-I surface expression in epithelial cells of hepatic and biliary origin. The loss of MHC-I surface expression in HCMV-infected epithelial cells reproduces the phenotype observed in broblasts. Indeed, all CMV genes that encode MHC-I downregulating proteins, i.e. US3, US2, US11 and US6, were expressed in the same temporal order and at similar quantitative levels in epithelial cells to those seen in the prototypic infection of highly permissive broblasts. Since Huh7, Hep3B, MzChA2 and Caco2 cells are little, if at all, productive for HCMV replication, this nding indicates that, in principle, the loss of MHC-I expression does not correlate with CMV replication and is not necessarily restricted to productively infected cells. CMV exhibits broad cell tropism in vivo. In infected organs, CMV was primarily identied in epithelial cells, including hepatocytes, bile duct cells and colon mucosa as well as endothelial cells (Ten Napel et al., 1984 ; Paya et al., 1990 ; Theise et al., 1993 ; Barkholt et al., 1994 ; Sinzger et al., 1995, 1999 ;

Podlech et al., 1998). Currently, most in vitro studies concerning HCMVhost cell interactions are conducted with embryonic or dermal broblasts. Only a few attempts of HCMV infection of dierentiated epithelial tissues like retinal epithelial cells or syncytiotrophoblasts have been reported in the past (Tugizov et al., 1997 ; Hemmings et al., 1998 ; Halwachs-Baumann et al., 1998 ; Sinzger et al., 1999). A small number of primary epithelial cells has been evaluated for supporting the growth of HCMV, but the preparation of cells from organs is cumbersome and the purity and characterization of these cells is limited (Hemmings et al., 1998 ; Sinzger et al., 1999). This holds particularly true for the liver, which consists of hepatocytes and bile duct epithelial cells, but in addition, of stromal, Kuper and Ito cells. The complexity of primary liver cell preparations and the poor availability of organ material from healthy donors are major obstacles for experimental analysis of the eects of the virus on dened cell types. To generate epithelial cell culture models of HCMV infection, we systematically screened a panel of dierentiated cell lines for their susceptibility to HCMV infection and varied infection conditions to optimize infection eciency. In this way, we were readily able to select liver and intestinal cell lines that supported infection with all laboratory HCMV strains and clinical isolates. The cells produced HCMV proteins IE1\pp72, IE2\pp86, pp65, gB and UL86\MCP, which represent dierent temporal classes of proteins, but with variable rates of synthesis compared with those seen in broblasts. Notably, the liver cell lines
CAGH

C. Benz and others

Hep3B and MzChA2 supported HCMV production at least as eciently as reported recently in primary hepatocytes (Sinzger et al., 1999). Reecting the high tropism of HCMV for these epithelial cells, these cell lines provide suitable models for studying HCMV pathogenesis at a molecular level in vitro. Our nding that Caco2 cells are readily infected and exhibit a reduced MHC-I phenotype after HCMV infection (unpublished observation) adds to recent reports that demonstrate that HCMV enters the basolateral surface of polarized Caco2 cells (Jarvis et al., 1999 ; Esclatine et al., 2000). Applying the newly established cell culture models, we assessed the impact of HCMV infection upon MHC-I expression in cells of hepatic, biliary and colonic origin. Previous attempts to analyse MHC-I expression on cultured primary biliary epithelial cells revealed conicting results to ours, suggesting that MHC-I surface expression is upregulated rather than downregulated (Scholz et al., 1997). However, HCMV gene expression had not been determined, which raises the question of whether ecient infection of cells was achieved or not. Our data reveal a clear reduction of MHC-I molecules upon HCMV infection at the level of both complex formation and surface expression. Most convincingly, we demonstrate the abundant expression of the HCMV proteins responsible for controlling the MHC-I pathway. Their molecular functions have been elucidated in great detail when expressed in isolation (Wiertz et al., 1996 a, b ; Ahn et al., 1996, 1997 ; Jones et al., 1996 ; Hengel et al., 1997 ; Lehner et al., 1997), but there is little knowledge of their biological signicance in infected tissues so far. Our data show that the factors controlling MHC-1 expression also operate in parenchymal cells of the liver and colon. It is plausible to assume that the HCMV genes that regulate MHC-I expression promote virus pathogenesis in epithelial tissues in vivo, since animal studies using MCMV mutants decient in these gene functions showed that the proteins that aect the MHC-I pathway facilitate virus replication and contribute to CMV pathogenesis and virulence in the infected organs, including the liver (Krmpotic et al., 1999). Sequential phases of productive and nonproductive infection characterize the natural life-cycle of CMV in vivo (Jordan, 1983). In light of our ndings, it is therefore tempting to speculate that MHC-I control by CMV is not only used as a mechanism for escaping the immune system during productive infection but may also occur in latently infected cells and in those initiating the replicative program. Future analysis of the transcriptional programs during latency and reactivation of productive infection should therefore include the genes that regulate MHC-I expression. Mice infected with MCMV mutants decient in MHC-I downregulating genes (Krmpotic et al., 1999) may provide a suitable model to test whether genes regulating MHC-I expression are benecial in maintaining latency and supporting productive reactivation. On the other hand, avoidance of CD8+ T cell recognition and MHC-I expression by CMV has remarkable exceptions.
CAGI

Macrophages, shown to be exempt from eective class I downregulation, were readily lysed by MCMV-specic CTL (Hengel et al., 2000). According to current concepts, nonhaematopoietic cells are unable to induce a primary CTL response but guide immune recognition of primed T cell responses (Sigal et al., 1999). Considering the fact that HCMV controls MHC-I expression in broblasts and epithelial cells, and MCMV controls expression in broblasts but not in macrophages, it appears that the eector rather than the induction phase of the antiviral CD8+ T cell response is counteracted by CMV. However, inhibition of antigen presentation of infected nonhaematopoietic cell types is also not perfect, but regulated by certain cytokines. Interferon (IFN)- possesses the capacity to restore MHC-I expression and peptide presentation of broblasts to CD8+ T cells (Hengel et al., 1994, 1995). In epithelial cells, this eect is particularly pronounced. While broblasts largely lost responsiveness to IFN- after infection (Miller et al., 1998), IFN- eciently induced MHC-I expression in a signicant number of preinfected epithelial cells without aecting the expression of the relevant HCMV proteins (unpublished observation). Thus cytokines, like IFN-, could explain the paradoxical nding that hepatocytes in HCMV-infected liver tissue express higher levels of MHC-I compared with healthy controls (Steinho et al., 1988 ; Arnold et al., 1991, 1992). In summary, this study shows for the rst time that CMV infection downregulates MHC-I expression in a multitude of epithelial cell types even in the absence of ecient virus replication. This panel of hepatic and intestinal cell lines represents well-characterized and permissive cellular models for analysis of HCMV pathogenesis in distinct epithelial tissues of clinical importance. They will oer the opportunity to study whether natural HCMV infection may also deal with nonclassical MHC gene products exclusively expressed in these tissues (Balk et al., 1994 ; Groh et al., 1996). A further application is to test HCMV mutants with targeted gene disruptions in order to dene the specic roles of CMV genes on tissue-specic pathogenesis.
The technical assistance of Katja Wichmann is gratefully acknowledged. We are grateful to Professor Ulrich Koszinowski for continuous support and scientic advice, and to H.-G. Burgert and L. Theilmann for reading the manuscript. We thank K. Fruh, E. Wiertz, $ W. J. Britt, B. Plachter, T. Stamminger and E. Podack for their generous gifts of antibodies and plasmids, A. Knuth for providing MzChA2 cells and T. Mertens for the clinical isolate UL1271. This study received nancial support from the Deutsche Forschungsgemeinschaft, grant no. He2526\3-2, and Sonderforschungsbereich 455, project A6.

References
Aden, D., Fogel, A., Plotkin, S., Damjanov, I. & Knowles, B. B. (1979).

Controlled synthesis of HBsAg in a dierentiated human liver carcinomaderived cell line. Nature 282, 615616.
Ahn, K., Angulo, A., Ghazal, P., Peterson, P. A., Yang, Y. & Fruh, K. $ (1996). Human cytomegalovirus inhibits antigen presentation by a

MHC-I in CMV-infected epithelial cells sequential multistep process. Proceedings of the National Academy of Sciences, USA 93, 1099010995.
Ahn, K., Gruhler, A., Galocha, B., Jones, T. R., Wiertz, E. J. H. J., Ploegh, H. L., Peterson, P. A. & Fruh, K. (1997). The ER-luminal $ Hengel, H., Lucin, P., Jonjic, S., Ruppert, T. & Koszinowski, U. H. (1994). Restoration of cytomegalovirus antigen presentation by gamma

interferon combats viral escape. Journal of Virology 68, 289297.


Hengel, H., Elinger, C., Pool, J., Goulmy, E. & Koszinowski, U. H. (1995). Cytokines restore MHC class I complex formation and control

domain of the HCMV glycoprotein US6 inhibits peptide translocation by TAP. Immunity 6, 613621. Andreoni, M., Faircloth, M., Vugler, L. & Britt, W. J. (1989). A rapid microneutralization assay for the measurement of neutralizing antibody reactive with human cytomegalovirus. Journal of Virological Methods 23, 157167.
Arnold, J. C., Gmelin, K., Otto, G., Alexander, G. J. M. & Williams, R. (1991). Eect of cytomegalovirus infection on expression of HLA-

antigen presentation in human cytomegalovirus-infected cells. Journal of General Virology 76, 29872997.
Hengel, H., Flohr, T., Hammerling, G. H., Koszinowski, U. H. & $ Momburg, F. (1996). Human cytomegalovirus inhibits peptide trans-

location into the endoplasmic reticulum for MHC class I assembly. Journal of General Virology 77, 22872296.
Hengel, H., Koopmann, J. O., Flohr, T., Muranyi, W., Goulmy, E., Hammerling, G. J., Koszinowski, U. H. & Momburg, F. (1997). A viral $

antigens in liver allografts. Transplantation Proceedings 23, 442443.


Arnold, J. C., Portmann, B. C., OGrady, J. G., Naoumov, N. V., Alexander, G. J. M. & Williams, R. (1992). Cytomegalovirus infection

ER-resident glycoprotein inactivates the MHC-encoded peptide transporter. Immunity 6, 623632.


Hengel, H., Reusch, U., Geginat, G., Holtappels, R., Ruppert, T., Hellebrand, E. & Koszinowski, U. H. (2000). Macrophages escape

persists in the liver graft in the vanishing bile duct syndrome. Hepatology 16, 285292.
Balk, S. P., Burke, S., Polischuk, J. E., Frantz, M. E., Yang, L., Porcelli, S., Colgan, S. P. & Blumberg, R. S. (1994). 2-microglobulin-in-

inhibition of major histocompatibility complex class I-dependent antigen presentation by cytomegalovirus. Journal of Virology 74, 78617868.
Ito, M., Watanabe, M., Ihara, T., Kamiya, H. & Sakurai, M. (1995).

dependent MHC class Ib molecule expressed by human intestinal epithelium. Science 265, 259262. Barkholt, L. M., Ehrnst, A. & Veress, B. (1994). Clinical use of immunohistopathologic methods for the diagnosis of cytomegalovirus hepatitis in human liver allograft biopsy specimens. Scandinavian Journal of Gastroenterology 29, 553560. Beersma, M. F. C., Bijlmakers, M. J. E. & Ploegh, H. L. (1993). Human cytomegalovirus down-regulates HLA class I expression by reducing the stability of class I H chains. Journal of Immunology 151, 44554464. Britt, W. J. & Alford, C. A. (1996). Cytomegalovirus. In Fields Virology, 3rd edn, pp. 24932523. Edited by B. N. Fields, D. M. Knipe & P. M. Howley. Philadelphia : LippincottRaven. Burakoff, R. & Eglow, R. (1995). Cytomegalovirus infection. In Bockus Gastroenterology, 5th edn, vol. 1, pp. 503517. Edited by W. S. Haubrich, F. Schaner & J. E. Berk. Philadelphia : W. B. Saunders.
Del Val, M., Hengel, H., Hacker, H., Hartlaub, U., Ruppert, T., Lucin, P. $ & Koszinowski, U. H. (1992). Cytomegalovirus prevents antigen

Increased expression of adhesion molecules (CD54, CD29 and CD44) on broblasts infected with cytomegalovirus. Microbiology and Immunology 39, 129133.
Jarvis, M. A., Wang, C. E., Meyers, H. L., Smith, P. P., Corless, C. L., Henderson, G. J., Vieira, J., Britt, W. J. & Nelson, J. A. (1999). Human

cytomegalovirus infection of Caco-2 cells occurs at the basolateral membrane and is dierentiation state dependent. Journal of Virology 73, 45524560. Jones, T. R. & Muzithras, V. P. (1991). Fine mapping of transcripts expressed from the US6 gene family of human cytomegalovirus strain AD169. Journal of Virology 65, 20242036.
Jones, T. R., Hanson, L. K., Sun, L., Slater, J. S., Stenberg, R. M. & Campbell, A. E. (1995). Multiple independent loci within the human

cytomegalovirus unique short region down-regulate expression of major histocompatibility complex class I heavy chains. Journal of Virology 69, 48304841.
Jones, T. R., Wiertz, E. J. H. J., Sun, L., Fish, K. N., Nelson, J. A. & Ploegh, H. L. (1996). Human cytomegalovirus US3 impairs transport

presentation by blocking the transport of peptide-loaded major histocompatibility complex class I molecules into the medial-Golgi compartment. Journal of Experimental Medicine 172, 729738.
Esclatine, A., Lemullois, M., Servin, A. L., Quero, A.-M. & GeniteauLegendre, M. (2000). Human cytomegalovirus infects Caco-2 intestinal

and maturation of major histocompatibility complex class I heavy chains. Proceedings of the National Academy of Sciences, USA 93, 1132711333. Jordan, M. C. (1983). Latent infection and the elusive cytomegalovirus. Reviews of Infectious Diseases 5, 205215.
Knuth, A., Gabbert, H., Dippold, W., Klein, O., Sachsse, W., BitterSuermann, D., Prellwitz, W. & Meyer zum Buschenfelde, K. H. (1985). $

epithelial cells basolaterally regardless of the dierentiation state. Journal of Virology 74, 513517.
Gabay, C., Silacci, P., Genin, B., Mentha, G., Le Coultre, C. & Guerne, P. A. (1995). Soluble interleukin-6 receptor strongly increases the

Biliary adenocarcinoma. Characterisation of three new human tumor cell lines. Journal of Hepatology 1, 579596.
Krmpotic, A., Messerle, M., Crnkovic-Mertens, I., Polic, B., Jonjic, S. & Koszinowski, U. H. (1999). The immunoevasive function encoded by

production of acute-phase protein by hepatoma cells but exerts minimal changes on human primary hepatocytes. European Journal of Immunology 25, 23782383.
Groh, V., Bahram, S., Bauer, S., Herman, A., Beauchamp, M. & Spies, T. (1996). Cell stress-regulated human major histocompatibility complex

the mouse cytomegalovirus gene m152 protects the virus against T cell control in vivo. Journal of Experimental Medicine 190, 12851295.
Lautenschlager, I., Hockerstedt, K., Salmela, K., Isoniemi, H., $ Holmberg, C., Jalanko, H. & Hayry, P. (1990). Fine-needle aspiration

class I gene expressed in gastrointestinal epithelium. Proceedings of the National Academy of Sciences, USA 93, 1244512450.
Halwachs-Baumann, G., Wilders-Truschnig, M., Desoye, G., Hahn, T., Kiesel, L., Klingel, K., Rieger, P., Jahn, G. & Sinzger, C. (1998). Human

biopsy in the monitoring of liver allografts. Dierent cellular ndings during rejection and cytomegalovirus infection. Transplantation 50, 798803.
Lautenschlager, I., Hockerstedt, K., Jalanko, H., Loginov, R., Salmela, $ K., Taskinen, E. & Ahonen, J. (1997). Persistent cytomegalovirus in liver

trophoblast cells are permissive to the complete replicative cycle of human cytomegalovirus. Journal of Virology 72, 75987602.
Hemmings, D. G., Kilani, R., Nykiforuk, C., Preiksaitis, J. & Guilbert, L. J. (1998). Permissive cytomegalovirus infection of primary villous

allografts with chronic rejection. Hepatology 25, 190194.


Lehner, P. J., Karttunen, J. T., Wilkinson, G. W. G. & Cresswell, P. (1997). The human cytomegalovirus US6 glycoprotein inhibits trans-

term and rst trimester trophoblasts. Journal of Virology 72, 49704979.

CAGJ

C. Benz and others porter associated with antigen processing-dependent peptide translocation. Proceedings of the National Academy of Sciences, USA 94, 69046909.
Miller, D. M., Rahill, B. M., Boss, J. M., Lairmore, M. D., Durbin, J. E., Waldman, J. W. & Sedmak, D. D. (1998). Human cytomegalovirus Sinzger, C., Bissinger, A. L., Viebahn, R., Oettle, H., Radke, C., Schmidt, C. A. & Jahn, G. (1999). Hepatocytes are permissive for human

inhibits major histocompatibility complex class II expression by disruption of the Jak\Stat pathway. Journal of Experimental Medicine 187, 675683.
Nakabayashi, H., Takeda, K., Miyano, K., Yamane, T. & Sato, J. (1982).

cytomegalovirus infection in human liver cell culture and in vivo. Journal of Infectious Diseases 180, 976986. Snover, D. C. & Horwitz, C. A. (1984). Liver disease in cytomegalovirus mononucleosis : a light microscopical and immunoperoxidase study of six cases. Hepatology 4, 408412.
Spaete, R. R., Thayer, R. M., Probert, W. S., Masiarz, F. R., Chamberlain, S. H., Rasmussen, L., Merigan, T. C. & Pachl, C. (1988). Human

Growth of human hepatoma cell lines with dierentiated functions in chemically dened medium. Cancer Research 42, 38583863. Ohe, Y., Zhao, D., Saijo, N. & Podack, E. R. (1995). Construction of a novel bovine papillomavirus vector without detectable transforming activity suitable for gene transfer. Human Gene Therapy 6, 325333. Parham, P., Barnstable, C. J. & Bodmer, W. F. (1979). Use of monoclonal antibody (W6\32) in structural studies of HLA-A, B, C antigens. Journal of Immunology 123, 342349.
Paya, C. V., Hermans, P. E., Wiesner, R. H., Ludwig, J., Smith, T. F., Rakela, J. & Krom, R. A. (1989). Cytomegalovirus hepatitis in liver

cytomegalovirus strain Towne glycoprotein B is processed by proteolytic cleavage. Virology 176, 207225. Steinhoff, G., Wonigeit, K. & Pichlmayr, R. (1988). Analysis of sequential changes in major histocompatibility complex expression in human liver grafts after transplantation. Transplantation 45, 394401.
Stratta, R. J., Shaefer, M. S., Markin, R. S., Wood, R. P., Kennedy, E. M., Langnas, A. N., Reed, E. C., Woods, G. L., Donovan, J. P., Pillen, T. J., Duckworth, R. M. & Shaw, B. W. (1989). Clinical patterns of

transplantation : prospective analysis of 93 consecutive orthotopic liver transplantations. Journal of Infectious Diseases 160, 752758.
Paya, C. V., Holley, K. E., Wiesner, R. H., Balasubramaniam, K., Smith, T. F., Espy, M. J., Ludwig, J., Batts, K. P., Hermans, P. E. & Krom, R. A. (1990). Early diagnosis of cytomegalovirus hepatitis in liver transplant

cytomegalovirus disease after liver transplantation. Archives of Surgery 124, 14431450. Ten Napel, H. H., Houthoff, H. J. & The, T. H. (1984). Cytomegalovirus hepatitis in normal and immune compromised hosts. Liver 4, 184194. Theise, N. D., Conn, M. & Thung, S. N. (1993). Localization of cytomegalovirus antigens in liver allografts over time. Human Pathology 24, 103108.
Thornthwaite, J. T., Sugarbaker, E. V. & Temple, W. J. (1980).

recipients : role of immunostaining, DNA hybridization and culture of hepatic tissue. Hepatology 12, 119126.
Plachter, B., Britt, W., Vornhagen, R., Stamminger, T. & Jahn, G. (1993). Analysis of proteins encoded by IE regions 1 and 2 of human

cytomegalovirus using monoclonal antibodies generated against recombinant antigens. Virology 193, 642652.
Podlech, J., Holtappels, R., Wirtz, N., Steffens, H.-P. & Reddehase, M. J. (1998). Reconstitution of CD8 T cells is essential for the prevention

Preparation of tissue for DNA ow cytometric analysis. Cytometry 1, 229234. Tugizov, S., Maidji, E. & Pereira, L. (1997). Role of apical and basolateral membranes in replication of human cytomegalovirus in polarized retinal pigment epithelial cells. Journal of General Virology 77, 6174.
Warren, A. P., Ducroq, D. H., Lehner, P. J. & Borysiewicz, L. K. (1994).

of multiple-organ cytomegalovirus histopathology after bone marrow transplantation. Journal of General Virology 79, 20992104.
Sanchez, V., Angeletti, P. C., Engler, J. A. & Britt, W. J. (1998).

Human cytomegalovirus-infected cells have unstable assembly of major histocompatibility complex class I complexes and are resistant to lysis by cytotoxic T lymphocytes. Journal of Virology 68, 28222829.
Wiertz, E. J. H. J., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J. & Ploegh, H. L. (1996 a). The human cytomegalovirus US11 gene product

Localization of human cytomegalovirus structural proteins to the nuclear matrix of infected human broblasts. Journal of Virology 72, 33213329.
Scholz, M., Cinatl, J., Blaheta, R. A., Kornhuber, B., Markus, B. H. & Doerr, H. W. (1997). Expression of human leukocyte antigens class I and

dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84, 769779.
Wiertz, E. J. H. J., Tortorella, D., Bogyo, M., Yu, J., Mothes, W., Jones, T. R., Rapoport, T. A. & Ploegh, H. L. (1996 b). Sec61-mediated transfer

class II on cultured biliary epithelial cells after cytomegalovirus infection. Tissue Antigens 49, 640643. Sigal, L. J., Crotty, S., Andino, R. & Rock, K. L. (1999). Cytotoxic T-cell immunity to virus-infected non-haematopoietic cells requires presentation of exogenous antigen. Nature 398, 7780.
Sinzger, C., Grefte, A., Plachter, B., Gouw, A. S. H., The, T. H. & Jahn, G. (1995). Fibroblasts, epithelial cells, endothelial cells and smooth

of a membrane protein from the endoplasmic reticulum to the proteasome for destruction. Nature 384, 432438.
Yamashita, Y., Shimokata, K., Mizuno, S., Yamaguchi, H. & Nishiyama, Y. (1993). Down-regulation of the surface expression of class I MHC

antigens by human cytomegalovirus. Virology 193, 727736.


Received 12 October 2000 ; Accepted 10 May 2001

muscle cells are major targets of human cytomegalovirus infection in lung and gastrointestinal tissues. Journal of General Virology 76, 741750.

CAHA

Das könnte Ihnen auch gefallen