Human Herpesvirus 8 Infects and Replicates in Primary Cultures of Activated B Lymphocytes

JOURNAL OF VIROLOGY, May 2008, p. 4793–4806 Vol. 82, No. 100022-538X/08/$08.00�0 doi:10.1128/JVI.01587-07Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Human Herpesvirus 8 Infects and Replicates in Primary Cultures ofActivated B Lymphocytes through DC-SIGN�

Giovanna Rappocciolo,1* Heather R. Hensler,1 Mariel Jais,1 Todd A. Reinhart,1 Amarendra Pegu,1Frank J. Jenkins,1,2 and Charles R. Rinaldo1,2

Department of Infectious Diseases and Microbiology, Graduate School of Public Health,1 and Department of Pathology, School ofMedicine,2 University of Pittsburgh, Pittsburgh, Pennsylvania 15261

Received 20 July 2007/Accepted 5 March 2008

Human herpesvirus 8 (HHV-8) is the etiological agent of Kaposi’s sarcoma, primary effusion lymphoma, andsome forms of multicentric Castleman’s disease. Although latent HHV-8 DNA can be detected in B cells frompersons with these cancers, there is little information on the replication of HHV-8 in B cells. Indeed, B cellsare relatively resistant to HHV-8 infection in vitro. We have recently shown that DC-SIGN, a C-type lectin firstidentified on dendritic cells (DC), is an entry receptor for HHV-8 on DC and macrophages. We have alsodemonstrated previously that B lymphocytes from peripheral blood and tonsils express DC-SIGN and that thisexpression increases after B-cell activation. Here we show that activated blood and tonsillar B cells can beproductively infected with HHV-8, as measured by an increase in viral DNA, the expression of viral lytic andlatency proteins, and the production of infectious virus. The infection of B cells with HHV-8 was blocked by thepretreatment of the cells with antibody specific for DC-SIGN or with mannan but not antibody specific for xCT,a cystine/glutamate exchange transporter that has been implicated in HHV-8 fusion to cells. The infection ofB cells with HHV-8 resulted in increased expression of DC-SIGN and a decrease in the expression of CD20 andmajor histocompatibility complex class I. HHV-8 could also infect and replicate in B-cell lines transduced toexpress full-length DC-SIGN but not in B-cell lines transduced to express DC-SIGN lacking the transmem-brane domain, demonstrating that the entry of HHV-8 into B cells is related to DC-SIGN-mediated endocytosis.The role of endocytosis in viral entry into activated B cells was confirmed by blocking HHV-8 infection withendocytic pathway inhibitors. Thus, the expression of DC-SIGN is essential for productive HHV-8 infection ofand replication in B cells.

Human herpesvirus 8 (HHV-8), also known as Kaposi’s sar-coma (KS)-associated herpesvirus, is the etiological agent ofKS, primary effusion lymphoma (PEL), and some forms ofmulticentric Castleman’s disease (MCD). The virus is found inendothelial cells of KS lesions but is also detected in B cells ofPEL and MCD lesions and the peripheral blood of KS patients(5). However, B cells from normal individuals are relativelyresistant to in vitro infection with HHV-8 (8). Attempts toestablish productive infections by using B-lymphoblastoid-celllines have also met with limited success (8). On the other hand,B-cell lines established from B cells from PEL patients, whichharbor HHV-8, can be induced to replicate virus by treatmentwith phorbol esters (37). These PEL B-cell lines have greatlyhelped studies of lytic and latent HHV-8 infections but are oflimited use as models of natural viral infection.

We hypothesized that the lack of permissive infection of Bcells and B-cell lines with HHV-8 in vitro is related to thedifferential expression of the appropriate virus entry receptors.Several proteins have been reported to serve as HHV-8 entryreceptors (3, 25, 33). We have shown previously that DC-SIGN, a C-type lectin first identified on dendritic cells (DC)(18), is an entry receptor for HHV-8 on DC and macrophages

in vitro (33). DC-SIGN and its isomer DC-SIGNR were ini-tially shown to be restricted in expression in vivo to dermal andlymphatic DC, activated macrophages, and vascular endothe-lial cells (38, 40, 44). Recent studies from our laboratory andothers have demonstrated that B lymphocytes from peripheralblood and tonsils express DC-SIGN and that this expressionsignificantly increases after B-cell activation mediated by CD40ligand (CD40L) and interleukin 4 (IL-4) (22, 34). These datasuggest that DC-SIGN may also serve as an entry receptor onactivated B (aB) cells and that its lack of expression on restingB (rB) cells may explain why previous attempts to infect B cellswith HHV-8 have been had limited success.

In the present study, we show that activated blood and ton-sillar B cells expressing DC-SIGN can be productively infectedwith HHV-8, as determined by an increase in the level of viralDNA, the expression of lytic and latency-associated viral pro-teins, and the production of infectious virus. HHV-8 infectioncould be blocked by the pretreatment of the B cells with anti-DC-SIGN monoclonal antibody (MAb). These results are thefirst evidence of a fully productive infection of B cells andconfirm the role of DC-SIGN as an entry receptor for thisvirus. This model can provide insight into the life cycle ofHHV-8 and a better understanding of its pathogenesis.

MATERIALS AND METHODS

Blood and tonsil donors. Blood from HHV-8-seronegative, healthy donors wasused in this study. The HHV-8 antibody status was determined by an immuno-fluorescence assay (IFA) (33). Freshly excised tonsils were obtained from per-

* Corresponding author. Mailing address: Department of InfectiousDiseases and Microbiology, Graduate School of Public Health, Uni-versity of Pittsburgh, 604 Parran Hall, 130 De Soto St., Pittsburgh, PA15261. Phone: (412) 383-9590. Fax: (412) 624-4953. E-mail: [email protected].

� Published ahead of print on 12 March 2008.

4793

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

sons undergoing tonsillectomies. Informed consent was obtained according toUniversity of Pittsburgh Institutional Review Board guidelines.

Cell lines. BCBL-1 cells (HHV-8-positive, Epstein-Barr virus [EBV]-negativehuman B cells) were obtained through the AIDS Research and ReferenceReagent Program, Division of AIDS, NIAID, NIH, and used as a source of virus.EBV-transformed Raji cells, Raji cells transfected to express DC-SIGN (Raji-DC-SIGN cells) (48), which do not express DC-SIGN, and Raji cells expressinga truncated form of DC-SIGN lacking 20 N-terminal amino acids (Raji-DC-SIGN-�20 cells) were gifts from D. Littman (New York University) and V.KewalRamani (NCI). K562 cells (American Type Culture Collection), an eryth-roleukemia cell line that does not constitutively express DC-SIGN, were stablytransfected with plasmid pcDNA-DC-SIGN (NIH AIDS Research and Refer-ence Reagent Program) by using Lipofectamine 2000 (Invitrogen) and wereselected and maintained in medium containing G418 (Invitrogen). Cells weregrown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calfserum (FCS) and L-glutamine and were negative for mycoplasma (as determinedby a mycoplasma PCR enzyme-linked immunosorbent assay; Roche).

Preparation of B cells from blood. Peripheral blood mononuclear cells wereisolated from heparinized blood by Ficoll-Hypaque density gradient separation.To obtain purified B cells, cell populations were depleted of monocytes by tworounds of incubation with anti-CD14 MAb-coated immunomagnetic microbeadsaccording to the instructions of the manufacturer (Miltenyi). B cells (CD19�

cells) were then isolated from the CD14� cell fraction by incubation with anti-CD19 MAb-coated magnetic microbeads (Miltenyi) (34). The purity of thefractionated B cells, as determined by staining with anti-CD20 MAb, was �95%,with �1% CD14� and CD3� cells. aB cells were generated by the culture ofCD19� cells in RPMI 1640 medium supplemented with 10% heat-inactivatedFCS, 1,000 U of recombinant human IL-4 (R&D Systems)/ml, and 1 ug ofsoluble trimeric CD40L (Amgen)/ml. We have previously shown that thismethod of positive selection does not alter the B-cell function or phenotype (34).

Preparation of B cells from tonsils. After surgical removal, tonsils were im-mediately transferred to the laboratory in cold phosphate-buffered saline sup-plemented with penicillin (100 U/ml), streptomycin (100 �g/ml), gentamicin (5�g/ml), amphotericin B (0.5 �g/m), and 5% FCS and processed into single-cellsuspensions as previously described (47). Lymphocytes from the collected cellsuspensions were isolated by Ficoll-Hypaque density gradient centrifugation.The cells collected at the interface were stained with anti-CD20, anti-CD3, andanti-CD14 MAbs to determine the relative percentages of B and T lymphocytesand were routinely shown to comprise 60 to 70% B lymphocytes, 30 to 40% Tlymphocytes, and 1% or fewer monocytes. B cells were purified by immunomag-netic bead separation by using anti-CD19 MAb-coated magnetic beads as de-scribed above.

HHV-8 purification and infection of cells. Concentrated supernatants andcell-free lysates from 2 � 109 12-O-tetradecanoylphorbol-13-acetate (Sigma)-induced BCBL-1 cells were used as sources of HHV-8, and virus was purified bya method modified from a procedure described by Cerimele et al. (9) as previ-ously detailed (33). An average of 107 HHV-8 virions (defined by determiningthe amount of DNase-resistant HHV-8 DNA) was used to infect 106 cells for 2 hat 37°C, after which the cells were washed and incubated for up to 3 days inRPMI 1640 medium supplemented with 10% FCS at 37°C in 5% CO2. For virusblocking studies, cells were pretreated with a 20-�g/ml concentration of anti-human DC-SIGN MAb (clone 120507; R&D Systems) or mouse immunoglob-ulin G (IgG; Sigma) or a 100-�g/ml concentration of mannan (Sigma) or 200 �lof undiluted anti-xCT rabbit antiserum for 1 h at 4°C before exposure to HHV-8.Sera from preimmune rabbits were used as a control for the anti-xCT blockingstudies.

Quantitative PCR assay for HHV-8. HHV-8 DNA levels were measured by aquantitative, real-time TaqMan PCR assay with a primer set specific for theHHV-8 alkaline exonuclease gene (the open reading frame 37 [ORF37] gene),previously shown to be highly specific for HHV-8 DNA (42), as previouslydescribed (33).

Real-time PCR measurement of DC-SIGN and xCT mRNAs. Total RNA wasextracted from cells by using TRIzol according to the instructions of the manu-facturer (Invitrogen Life Sciences, Carlsbad, CA). TaqMan amplification anddetection of DC-SIGN mRNA were carried out using primers and probes asdescribed previously (34), whereas for xCT mRNA, a commercially availableTaqMan assay (Applied Biosystems, Foster City, CA) was used. cDNA synthesiswith 400 ng of input RNA was performed in duplicate with SuperScript II reversetranscriptase (Invitrogen) and random hexamers as described previously (19), inparallel with control reactions with mixtures lacking reverse transcriptase. Am-plification and detection for 45 cycles were performed on a Prism 7000 sequencedetection system (Applied Biosystems). The expression levels of DC-SIGN and

xCT mRNAs were calculated relative to the expression level of the endogenouscontrol mRNA for �-glucuronidase (�-GUS; Applied Biosystems).

Generation of anti-xCT polyclonal antibodies. Rabbit antiserum specific to thexCT peptide located in the extracellular domain (amino acids 218 to 232 [peptideP218-232]; TQNFKDAFSGRDSSI) (26) was generated. All rabbit immuniza-tions and serum collections were performed commercially by Genosys-Sigma.

IFAs. IFAs of HHV-8-infected and mock-infected cells were done usinganti-HHV-8 ORF73 protein rat MAb and mouse MAb directed against ORF-K8.1A/B or ORF59 protein (ABI) or polyclonal rabbit anti-viral IL-6 (anti-vIL-6) IgG (ABI). Cells were counted and spotted onto poly-L-lysine-coatedmicroscope slides, fixed in 4% paraformaldehyde for 20 min, and permeab-ilized with buffer (0.55% bovine serum albumin, 0.1% saponin, 0.1% NaN3)for 20 min at room temperature. Cells were incubated with the primaryantibody (Ab) for 30 min at 4°C, washed extensively in buffer, and thenstained with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouseor anti-rat IgG (Sigma). Controls were the omission of the primary Ab andthe staining of uninfected cells. In some experiments, cells were also stainedby direct immunofluorescence with FITC-conjugated anti-DC-SIGN MAb(clone 120507; R&D Systems) or with K8.1A/B or ORF59 MAb directlyconjugated with phycoerythrin or Texas Red dyes by using a Zenon labelingkit (Molecular Probes). To avoid nonspecific binding of IgG, a blocking stepusing SuperBlock blocking buffer (Pierce) was added. In some experiments,cells were also incubated with anti-xCT rabbit antiserum, washed, and stainedwith anti-rabbit IgG conjugated with Alexa Fluor 350 or FITC.

Blocking of HHV-8 binding. For receptor blocking studies, cells were incu-bated with anti-DC-SIGN MAb (20 �g/ml; clone 120507) or mannan (100 �g/ml)for 1 h at 4°C prior to the addition of purified, [3H]thymidine-labeled HHV-8 bythe method of Akula et al. (4).

Flow cytometric analysis of cell phenotypes. The expression of cell surfacemolecules was assessed by flow cytometry (with a Beckman Coulter XL instru-ment) using FITC-conjugated MAb specific for HLA-A, HLA-B, HLA-C, CD20,CD23, CD54, CD80, CD86, CD83, and DC-SIGN (CD209) for 30 min at 4°C,and then cells were fixed with 1% paraformaldehyde. Cells were stained withisotype antibody or goat anti-mouse IgG–FITC as controls.

Inhibition of endocytosis. Chlorpromazine HCl (10 �g/ml), NH4Cl (50 mM),and bafilomycin A1 (BFLA-1; 50 nM) were used to inhibit endocytosis, asdescribed previously (1). The stock solutions were prepared in accordance withthe manufacturer’s recommendations (Sigma), and working dilutions in com-plete medium were prepared. B cells were incubated at 37°C for 1 h with andwithout the inhibitors and then infected with HHV-8 at 37°C for 2 h, washed, andfurther incubated in complete medium for 24 h. After the incubation, infectionwas detected by an indirect IFA for K8.1 and vIL-6.

RESULTS

aB cells support productive HHV-8 lytic and latent infec-tion. We have previously demonstrated the expression of HHV-8lytic- and latency-cycle proteins in infected DC and macrophagesin the absence of productive virus infection (33). That is, after theinfection of DC or macrophages with HHV-8, we observed anincrease in the number of cells expressing viral lytic- and latency-cycle proteins without increases in viral DNA over several days ofculture. This pattern of viral gene expression was consistent withthat shown for HHV-8 infection of vascular endothelial cells (28,33). In the present study, we determined whether B cells, a majortarget of HHV-8 infection in vivo, can support productive HHV-8infection in vitro. B cells were isolated to �95% purity fromperipheral blood mononuclear cells by CD19-specific magneticbead separation and either left untreated (rB cells) or treatedwith CD40L and IL-4 for up to 36 h (aB cells). We first assessedHHV-8 infection of these B cells by analyses of viral lytic- andlatency-cycle protein expression. The infection of rB cells withHHV-8 did not result in detectable virus infection over 3 days ofculture, as shown by the lack of expression of the K8.1 andORF59 lytic proteins (Fig. 1A, panels a and b). In contrast, similarto our results with HHV-8 infection of DC and macrophages,parallel cultures of purified, infected aB cells expressed both of

4794 RAPPOCCIOLO ET AL. J. VIROL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

FIG. 1. HHV-8 infects and replicates in B cells. (A and B) rB cells and aB cells were infected with HHV-8 and stained with MAb againstDC-SIGN (green fluorescence) and anti-K8.1 (24 h) or anti-ORF59 protein (48 h) (red fluorescence) or with MAb against ORF73 protein (72 h;red fluorescence) (panels a, b, and c, respectively). Uninfected B cells were negative for each of the three viral proteins (data not shown).(C) Expression of viral lytic and latency proteins in HHV-8-infected aB cells over time (bars represent means standard errors [SE]; K8.1, n 11; ORF59 protein, n 4; ORF73 protein, n 5). For each slide, at least four independent fields were counted and the results were averaged.(D) Quantitative HHV-8 DNA results from seven separate experiments with B cells from seven different donors. B cells were assayed by real-timePCR analyses of extracts of rB-cell or aB-cell pellets and the corresponding DNase-resistant, concentrated cell culture supernatants (sups). Bothuninfected cell pellets and supernatants were negative for HHV-8 DNA (data not shown). (E) Cell counts (bars) and viability (lines) for uninfected(blue) and HHV-8-infected (gray) aB cells (means SE; n 13).

VOL. 82, 2008 INFECTION OF B CELLS WITH HHV-8 4795

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

these proteins after 24 and 48 h of infection (Fig. 1B, panels a andb). The infected aB cell cultures exhibited peak mean expressionof the ORF59 protein at 24 h (36% 5% of the cells werepositive for the protein; n 4) and K8.1 at 48 h (22% 4% ofthe cells were positive; n 11) (Fig. 1C). The latency proteinencoded by the ORF73 gene (LANA-1) was first evident at 48 hpostinfection, with 10% 3% (n 4) of the aB cells beingpositive (Fig. 1C). By 72 h postinfection, 30% 3% of the cellswere positive for the ORF73 protein (Fig. 1B, panel c, and C). Itshould be noted that the data in Fig. 1C reflect the percentages ofimmunofluorescent-antibody-positive cells among all of the Bcells in the cultures, including both DC-SIGN-positive and -neg-ative cells. However, our IFA data indicated that only DC-SIGN-expressing B cells also expressed viral proteins, confirming thehypothesis that DC-SIGN expression is required for the infectionof B cells in culture. Finally, we have found that this effect is notunique for the Amgen trimeric CD40L used in this study, as othercommercial preparations of CD40L, i.e., those from Alexis andR&D Systems, together with IL-4, also upregulated the expres-sion of DC-SIGN on B cells and made the cells permissive forHHV-8 infection (data not shown).

We next assayed the cells and culture media for levels ofHHV-8 DNA. In contrast to the nonproductive HHV-8 infection

of DC and macrophages, in which there is no increase in HHV-8DNA over time and no release of virus into the cell culturesupernatant (33), we observed a significant increase in the num-bers of copies of HHV-8 DNA in cultures of purified aB cellsderived from healthy, HHV-8-seronegative donors (Fig. 1D). In-deed, compared to the amount at time zero, there was an average10-fold increase in the amount of HHV-8 DNA in aB-cell culturesupernatants at 24 h (P � 0.05; paired t test). There was somevariation in levels of HHV-8 DNA in the supernatants of culturesof cells from the seven donors at 48 and 72 h. There was a sixfoldincrease in the HHV-8 DNA levels in the cell pellets at 24 h (P �0.05), and these levels decreased at 48 and 72 h. There was noincrease in HHV-8 DNA in the infected rB-cell cultures. Thispattern of viral DNA replication was consistently found in B cellsobtained from seven different donors in multiple, replicate exper-iments (Fig. 1D). The pattern was accompanied by a relativelysmall but significant decrease in cell viability among the infectedaB cells compared to that among the uninfected aB cells, asdetermined by trypan blue exclusion analysis and cell recoveryfrom the HHV-8-infected cultures (e.g., viability levels were84.7% 6% for uninfected aB cells and 92.8% 1.4% forinfected aB cells at 48 h; P 0.05) (Fig. 1E).

We next examined whether there was production of infec-

FIG. 2. HHV-8 particles derived from infected B-cell primary cultures are infectious. (A) Supernatants from aB cells infected with HHV-8 for24 h were harvested, centrifuged, and used to infect freshly aB cells. At the indicated time points, cell pellets and supernatants were harvested andthe amount of HHV-8 DNA was determined by real-time PCR (line). Bars indicate the percentages of cells expressing lytic (K8.1) or latency(ORF73) viral proteins. The results shown are representative of two independent experiments. (B) Cells were stained for the expression ofDC-SIGN and of K8.1 and ORF73 viral proteins at 72 h postinfection.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

tious HHV-8 in the aB-cell cultures by determining if we couldtransfer infectious virus to uninfected B cells by using super-natants from our HHV-8-infected B-cell cultures. For this ex-periment, supernatants from the 24- and 48-h primary culturesof HHV-8-infected aB cells were centrifuged at 400 � g for 10min and mixed undiluted with a fresh culture of uninfected aBcells. Supernatants and pellets from the secondary cultures ofaB cells were examined for HHV-8 DNA and protein at dailyintervals, as were the primary aB-cell cultures. The resultsshow that infectious virus was produced in the secondary aB-cell cultures, as evidenced by increases in HHV-8 DNA levelsin the supernatants (Fig. 2A) and pellets (data not shown) andthe percentages of aB cells expressing lytic and latent proteins(Fig. 2) from 24 to 72 h postinfection.

Taken together, these results indicate that HHV-8 estab-

lishes an infection in CD40L- and IL-4-treated aB cells, withthe production of lytic and latency proteins, viral DNA, andinfectious virus and a moderate decrease in the number ofviable cells.

HHV-8 requires DC-SIGN for the infection of B cells. We haverecently shown that an average of 8% of peripheral blood B cellsconstitutively express DC-SIGN (CD209) on their surfaces andthat this level increases to an average of 29% after treatment invitro for 1 day with CD40L and IL-4 (34). We have also reportedthat DC-SIGN is a receptor for HHV-8 entry into monocyte-derived DC and macrophages (33). To address the question ofwhether DC-SIGN is a surface receptor for HHV-8 on B cells, wefirst assessed the binding of radioactively labeled virus to rB andaB cells that were left untreated or pretreated with anti-DC-SIGN MAb or mannan, which is a natural ligand for DC-SIGN

FIG. 3. HHV-8 binds and infects B cells through DC-SIGN. (A) Radioactively labeled HHV-8 was incubated with rB or aB cells that were leftuntreated or were pretreated with anti-DC-SIGN MAb or mannan. Each bar represents the mean of results from duplicate experiments ( SE). (B) aBcells were left untreated or treated with anti-DC-SIGN MAb or mannan prior to infection with HHV-8. Cells were stained for DC-SIGN (green) andK8.1 (red) or ORF59 protein (red) and counterstained with DAPI (4�,6�-diamidino-2-phenylindole; blue) at 24 h postinfection. (C) rB or aB cells wereleft untreated or incubated with anti-DC-SIGN MAb prior to infection with HHV-8. HHV-8 DNA in both cell pellets and cell culture supernatants (sups)was measured by real-time PCR as described in the legend to Fig. 1. No HHV-8 DNA was detected in uninfected B-cell cultures used as controls (datanot shown). (D) aB cells were left untreated or incubated with anti-DC-SIGN MAb at different concentrations prior to infection with HHV-8. MouseIgGs were used as controls at the highest concentration used for the anti-DC-SIGN MAb treatment (20 �g/ml). HHV-8 DNA in the cell pellets andsupernatants at 24 h postinfection was measured. The results shown are representative of two independent experiments.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

(14). As shown in Fig. 3A, the level of binding of HHV-8 to aBcells was greater than that of binding to rB cells and this bindingwas inhibited by pretreatment with anti-DC-SIGN MAb or man-nan (P � 0.05). The low level of binding to the rB cells that wemeasured may be due to the low levels of DC-SIGN or to bindingto heparan sulfate, since HHV-8 is known to attach to this surfacemolecule (4).

We next addressed whether the blocking of DC-SIGN byanti-DC-SIGN MAb or mannan could inhibit the infection ofB cells by HHV-8. As shown in Fig. 3B, the pretreatment of aBcells with either anti-DC-SIGN MAb or mannan inhibitedHHV-8 infection, as determined by the expression of viralproteins at 24 h. The pretreatment of cells with control, mouseIgG did not inhibit viral infection (data not shown). The pre-treatment of aB cells with anti-DC-SIGN MAb also reducedthe amounts of viral DNA replication and encapsidated viralDNA in cell culture supernatants by 87% and those in cellpellets by 63% at 24 to 72 h (Fig. 3C). Moreover, the inhibitoryeffect of the anti-DC-SIGN MAb on HHV-8 infection wasdose dependent (Fig. 3D).

We next confirmed our previous results (27) that conventionalRaji cells and K562 cells, which do not express DC-SIGN, cannotbe infected with HHV-8 but that Raji and K562 cells stably trans-

fected to express DC-SIGN (Raji-DC-SIGN and K562-DC-SIGNcells) are susceptible to HHV-8 infection, as demonstrated by theproduction of lytic (Fig. 4A and C) and latency (data not shown)proteins. We extended these results by showing that HHV-8 es-tablished a productive infection in the DC-SIGN-expressing RajiB-cell and K562 cell lines, as determined by an increase in viralDNA in both cell pellets and cell culture supernatants within 24 hpostinfection (Fig. 4B and D).

Taken together, these results support the conclusion that theexpression of DC-SIGN is essential for HHV-8 infection ofperipheral blood B cells, B-cell lines, and non-B-cell lines.

Role of xCT in HHV-8 infection of B cells. Recently, xCT,the 12-transmembrane light chain of the human cystine/gluta-mate exchange transporter system x�

c, has been reported to beinvolved in cell fusion and the entry of HHV-8 into severaltypes of continuous cell lines (26). We therefore examined therole of xCT in the infection of B cells by the procedures ofKaleeba and Berger (26) using antisera from rabbits immu-nized with a peptide derived from the extracellular portion ofthe xCT protein (peptide P218-232) to inhibit HHV-8-cell fu-sion and entry into target cells. We first confirmed the obser-vations of Kaleeba and Berger (26) that rB cells do not expressxCT protein or mRNA (Fig. 5). However, we found that aB

FIG. 4. HHV-8 infects and replicates in DC-SIGN-transfected Raji and K562 cells. (A) Raji cells or Raji-DC-SIGN cells were infected withHHV-8 for 24 h as described in Materials and Methods and stained for DC-SIGN (green) and lytic protein K8.1 (red) and counterstained withDAPI (blue). (B) Quantitative viral DNA results from samples that were extracted from Raji-DC-SIGN cells and the corresponding DNase-resistant, concentrated cell culture supernatants (sups). Both uninfected and wild-type cell pellets and supernatants were negative for HHV-8 DNA(data not shown). (C) K562 cells or K562-DC-SIGN cells were infected with HHV-8 for 24 h as described in Materials and Methods and stainedfor DC-SIGN (green) and lytic protein K8.1 (red) and counterstained with DAPI (blue). (D) Quantitative viral DNA results from samples thatwere extracted from K562-DC-SIGN cells and the corresponding DNase-resistant, concentrated cell culture supernatants. Both uninfected andwild-type cell pellets and supernatants were negative for HHV-8 DNA (data not shown).The results shown are representative of two independentexperiments.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

cells derived from rB cells by treatment with CD40L and IL-4for 24 h expressed xCT protein (Fig. 5A) and mRNA (Fig. 5B).Moreover, in contrast to Kaleeba and Berger (26), we observedthat the K562 erythroleukemia cell line constitutively ex-pressed xCT, as determined by both protein and mRNA anal-yses (Fig. 5A). The expression of xCT in the K562 cells wasconfirmed by Western blotting (data not shown). Thus, bothK562 and K562-DC-SIGN cells express xCT, while only K562-DC-SIGN cells express DC-SIGN. We have previously re-ported that K562-DC-SIGN cells but not the parental K562cells are susceptible to HHV-8 infection (33). In the presentstudy, we confirmed that K562-DC-SIGN cells could be in-fected with HHV-8 (Fig. 6A), comparably to aB cells (Fig. 6B),and that HHV-8 could not infect either K562 cells or rB cells.

We next used MAb directed against DC-SIGN and antiseraspecific for xCT P218-232 to try to block the infection of aBcells and K562-DC-SIGN cells. The pretreatment of K562-DC-SIGN cells with anti-DC-SIGN MAb inhibited HHV-8 infec-tion, as shown by the blocking of the expression of K8.1, whileblocking with anti-xCT antisera had a minimal inhibitory effect(Fig. 6C). These data were confirmed using aB cells, for whichthe blocking of DC-SIGN inhibited HHV-8 infection by over80% at 24 h but the blocking of xCT resulted in a decreaseof only 15% in the level of aB cells expressing K8.1 (Fig.6D). The pretreatment of the aB cells with anti-DC-SIGNMAb also decreased levels of HHV-8 DNA in cells andsupernatants by 68 and 99%, respectively (Fig. 6E). Pre-treatment with xCT antisera, on the other hand, reducedHHV-8 DNA levels in cells and supernatants by 8 and 20%,respectively (Fig. 6E).

Taken together, these data indicate that the infection of

peripheral blood B cells and B-cell lines with HHV-8 requiresthe expression of DC-SIGN but is not dependent on xCTexpression.

HHV-8 infects tonsillar B cells via DC-SIGN. We have pre-viously shown that B cells derived from tonsils constitutively ex-press DC-SIGN at levels comparable to those on in vitro-acti-vated peripheral blood B cells, i.e., an average of 26% of the cellsare positive for DC-SIGN, and they coexpress B-cell activationmarkers (34) (a representative example is presented in Fig. 7A).Therefore, we examined whether this level of expression of DC-SIGN was sufficient to allow the infection of tonsillar B cells byHHV-8, without pretreatment with CD40L and IL-4. We foundthat the tonsillar B cells were susceptible to HHV-8 infection,with approximately 30% of the total B-cell population expressingviral K8.1 protein by 24 h postinfection (data not shown). Doublestaining of the infected cells with anti-DC-SIGN and K8.1 MAbsdemonstrated that only cells expressing DC-SIGN were infectedwith HHV-8, as evidenced by the coexpression of K8.1 (Fig. 7B).The infection of tonsillar B cells also resulted in a fivefold in-crease in viral DNA by 24 h and a peak level of virus in thesupernatant by 48 h postinfection (Fig. 7D). Furthermore, over95% of HHV-8 infection of the tonsillar B cells was blocked bythe pretreatment of the cells with anti-DC-SIGN MAb (Fig. 7B).

We next examined tonsillar B cells for the presence of xCT.We could not distinguish xCT protein by IFA of tonsillar Bcells derived from five individuals due to high levels of back-ground staining with the control rabbit preimmune serum(data not shown). However, we detected xCT mRNA in onlyone of the five tonsillar B-cell preparations (Fig. 7C). Weattempted to infect these tonsillar B cells in the presence ofanti-xCT antiserum and stained them for K8.1 expression 24 h

FIG. 5. Expression of xCT on B cells and K562 cells. (A) Freshly isolated rB or aB cells or K562 or K562-DC-SIGN cells were incubated withrabbit antisera against xCT peptide and goat anti-rabbit IgG–FITC and counterstained with DAPI. (B) DC-SIGN- and xCT-specific mRNA levelsin K562, K562-DC-SIGN, rB, and aB cells. The levels are expressed relative to mRNA levels for the �-GUS gene (an internal housekeeping gene)in these cells.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

postinfection. As shown in Fig. 7B, bottom right panel, treat-ment with anti-xCT antisera did not interfere with the infectionof these cells with HHV-8. These results indicate that DC-SIGN but not xCT expression is required for the infection oftonsillar tissue B cells.

HHV-8 infection of B cells increases the expression of DC-SIGN. We have previously reported that HHV-8 infection ofDC and macrophages induces the downregulation of DC-SIGN expression within 24 h, with a progressive loss through72 h postinfection (33). Therefore, we studied the expressionof DC-SIGN and other costimulatory molecules in infected Bcells compared to that in uninfected B cells. We observed asignificant increase in the percentage of aB cells expressingDC-SIGN (CD209) by 24 h after HHV-8 infection (Fig. 8)(P 0.05 compared to uninfected aB cells at 24 h or infectedaB cells at 0 h). In contrast to the enhancing effect of HHV-8infection on DC-SIGN expression in aB cells, there was down-regulation of the expression of CD20, as demonstrated both bythe percentage of positive cells and by the mean fluorescenceintensity (MFI) by 24 h postinfection (P � 0.05 compared touninfected aB cells at 24 h or infected aB cells at 0 h). We alsonoted the downregulation of the MFI for HLA-A, HLA-B, andHLA-C in HHV-8-infected aB cells by 24 h, i.e., from 1,617

MFI in the uninfected aB cells to 989 MFI in the virus-infectedaB cells by 24 h postinfection (P � 0.05). Finally, there wassome downregulation of the MFI for CD86 in the infected aBcells by 48 h postinfection (P � 0.05 compared to uninfectedaB cells at 48 h or infected aB cells at 0 h). No differences inthe expression of CD23, CD54, CD80, CD83, or CD86 be-tween uninfected and HHV-8-infected aB cells were observed.Taken together, these results indicate that HHV-8 infection ofaB cells results in the upregulation of DC-SIGN expressionand the downregulation of CD20, major histocompatibilitycomplex (MHC) class I, and CD86 expression. This result maybe related to the downmodulation of HLA and CD86 expres-sion by HHV-8 gene products MIR1 and MIR2 (12, 13).

HHV-8 enters B cells through endocytosis. The entry ofHHV-8 into human fibroblasts and endothelial cells occurs byan endocytotic pathway (1). Moreover, human immunodefi-ciency virus type 1 appears to enter DC by endocytosis afterattachment to DC-SIGN (29). To determine if the entry ofHHV-8 into B cells is mediated by endocytosis, we first utilizedRaji B cells, which do not express DC-SIGN and are resistantto infection with HHV-8 (33). The Raji cells were transfectedwith DC-SIGN (yielding Raji-DC-SIGN cells) or with a mu-tant form of DC-SIGN (yielding Raji-DC-SIGN-�20 cells), in

FIG. 6. Effect of anti-xCT and anti-DC-SIGN Ab treatment on HHV-8 infection. (A and B) K562-DC-SIGN cells (A) or aB cells (B) wereinfected with HHV-8 and stained for DC-SIGN (green), xCT (blue), or K8.1 (red). Each of the first three panels in each row shows the individualdual combinations, while the fourth panel shows the three-color overlay. (C and D) K562-DC-SIGN cells (C) or aB cells (D) were treated withanti-xCT rabbit antiserum (left panels) or anti-DC-SIGN MAb (right panels) or left untreated (A and B) and were infected with HHV-8 for 24 hand then stained for the expression of DC-SIGN, xCT, and K8.1. Each panel represents a three-color overlay as in panels A and B. (E) Relativelevels of HHV-8 DNA in the cell pellets and supernatants of cultures pretreated with either anti-DC-SIGN MAb or anti-xCT rabbit antiserum.The levels of inhibition were measured against those of cultures pretreated with mouse IgG or preimmune rabbit antiserum, respectively.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

FIG. 7. HHV-8 infects and replicates in fresh tonsillar B cells. (A) B cells were isolated from tonsils (as described in Materials and Methods)by magnetic bead purification. Cells were then stained for the expression of CD20 and DC-SIGN and analyzed by fluorescence-activated cellsorting. (B) Fresh tonsillar B cells were left untreated or treated with anti-DC-SIGN MAb or anti-xCT antisera prior to infection with HHV-8,cultured for 24 h, and stained with anti-DC-SIGN (green) or anti-K8.1 (red) MAb. Nuclei are stained blue with DAPI. (C) xCT-specific mRNAlevels in K562, K562-DC-SIGN, aB, and tonsillar B cells. The levels are expressed relative to mRNA levels for the �-GUS gene (an internalhousekeeping gene) in these cells. The results shown are representative of two independent experiments. (D) Quantitative DNA results fromreal-time PCR assays. HHV-8 DNA was extracted from tonsillar B cells and the corresponding DNase-resistant, concentrated cell culturesupernatants. Both uninfected cell pellets and supernatants were negative for HHV-8 DNA (data not shown).

4801

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

FIG. 8. Expression of cellular markers in activated, HHV-8-infected B cells. aB cells were left untreated (white bars) or infected with HHV-8(black bars) and cultured for 48 h. At the indicated time points, cells were stained for the expression of B-cell markers and analyzed by flowcytometry. rB and aB cells were also analyzed at the onset of infection (0 h). Bars and lines represent the means ( SE) of results from sixindependent experiments and refer to percentages of positive cells and MFIs, respectively.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

which 20 amino acids from the cytoplasmic, N-terminal domainof DC-SIGN were removed to eliminate the putativedileucine-based internalization motif (29). We confirmed ourprevious results that HHV-8 can enter and express viral pro-teins in Raji-DC-SIGN cells (33), as shown by the expressionof K8.1 by 24 h postinfection (Fig. 9A). In contrast, the Raji-DC-SIGN-�20 cells did not express K8.1 protein at 24 h (Fig.9A). These results support the conclusion that full-length DC-SIGN is required for cell entry by HHV-8 and suggest that viralentry involves endocytosis.

We next examined whether HHV-8 infects DC-SIGN-ex-pressing aB cells through endocytosis. Chlorpromazine HCl isa cationic amphiphilic drug that prevents clathrin-mediatedendocytosis, whereas NH4Cl and BFLA-1 inhibit the acidifica-

tion of endosomes (1). aB cells were infected with HHV-8 inthe presence of the different endocytosis inhibitors and com-pared to infected untreated aB cells. The endocytosis inhibi-tors blocked �80% of HHV-8 infection of aB cells comparedto that of untreated aB cells, as shown by the number of cellsexpressing viral proteins K8.1 and vIL-6 (Fig. 9B and C). Noeffect on cell viability after treatment with the inhibitors wasobserved (data not shown). These results support the conclu-sion that HHV-8 enters B cells through an endocytic pathway.

DISCUSSION

The infection of B cells by HHV-8 was first suggested by thepresence of viral DNA and particles within B cells from pa-

FIG. 9. HHV-8 endocytosis in B-cell lines and aB cells. (A) Raji-DC-SIGN or Raji-DC-SIGN-�20 cells were infected with HHV-8, culturedfor 24 h, and then stained for the expression of HHV-8 protein K8.1 (red) or DC-SIGN (green). Nuclei were stained with DAPI. (B and C) aBcells were left untreated or pretreated with BFLA, NH4Cl, or chlorpromazine HCl (as described in Materials and Methods) and infected withHHV-8. Cells were washed, cultured for 24 h, and stained for the expression of lytic proteins K8.1 and vIL-6.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

tients with body cavity PEL (6, 10, 39) and the isolation ofB-cell lines (termed BCBL or BC lines) harboring persistentHHV-8 infection from such patients (37). HHV-8 DNA hasalso been found in B cells of patients with KS (21, 23) andMCD (15, 43). To date, however, B cells have been relativelyresistant to in vitro infection with HHV-8. Early reports indi-cated that virus can be transferred to CD19� B cells (30) andcan coinfect B cells during the immortalization of B-cell linesby infection with EBV (27). However, attempts to infect im-mortalized B-cell lines have been successful only when viralDNA was introduced directly into the cells (11). In fact, thereis little previous evidence of productive HHV-8 infection ofcultures of freshly isolated B cells (7, 8).

We hypothesized that the paradox in which B cells are oneof the major cellular targets of HHV-8 infection in vivo whileB-cell lines and freshly isolated B cells are refractory to HHV-8infection in vitro was due to the differential expression ofDC-SIGN in these cells. This hypothesis was based on ourprevious finding that DC-SIGN expressed on DC and activatedmacrophages serves as a receptor for HHV-8 (33). The infec-tion of these cell types with HHV-8 results in abortive viralreplication, with the production of at least some lytic proteins,the establishment of latency, and little evidence of viral DNAreplication, similar to the limited replication of HHV-8 inhuman endothelial cells and fibroblasts (3, 31, 36, 45). More-over, we (34) and others (22) demonstrated previously thatwhereas DC-SIGN is constitutively expressed on only a smallnumber of rB cells and at a low level, the expression of thisC-type lectin can be enhanced by activating B cells in vitro witha combination of CD40L and IL-4. The present results indicatethat activated peripheral blood B cells support fully productiveinfection with HHV-8, as shown by the detection of the K8.1,ORF59, and vIL-6 lytic proteins in the infected B cells within24 to 72 h of infection. Furthermore, we were able to detectincreased levels of viral DNA in both the cell pellets andsupernatants of infected B cells by 24 h postinfection. In con-trast, rB cells did not support HHV-8 replication. Importantly,the increase in viral proteins and DNA reflected an increase inthe production of infectious virus, as shown by the transfer ofproductive viral infection to secondary B-cell cultures. Theproductive infection of aB cells and the lack of infection of rBcells were consistent findings for cells derived from multipledonors. Interestingly, Blackbourn et al. (8) reported that thetreatment of cultures of B cells with the lectin mitogen phyto-hemagglutinin results in limited HHV-8 replication. We havenoted that B cells treated with phytohemagglutinin have anincrease in the expression of DC-SIGN (our unpublished re-sults), suggesting that this increased expression may be respon-sible for the replication of HHV-8 in these aB cells.

We confirmed that DC-SIGN was essential for the infectionof aB cells with HHV-8 by blocking both viral DNA and pro-tein production by the pretreatment of the B cells with MAbspecific for DC-SIGN or with its natural ligand mannan. After3 days of infection, HHV-8 established latency, as shown by theexpression of LANA-1 protein. Our results on productiveHHV-8 infection of aB cells were supported by our finding thatHHV-8 productively infected the Raji B-cell and K562 celllines that had been transfected with DC-SIGN but not theparental strain of DC-SIGN-negative Raji B or K562 cells. Toour knowledge, this is the first in vitro model for consistent,

fully productive infection of cells naturally targeted by HHV-8in vivo. We are currently investigating if it is possible to reac-tivate HHV-8 from latency in the infected B cells and returnthe virus to the full lytic cycle.

Interestingly, tonsillar B cells were able to support a level ofproductive HHV-8 infection in vitro similar to that of aB cellswithout requiring in vitro activation with CD40L and IL-4. Agreater proportion of these tonsillar B cells than of peripheralblood B cells constitutively express DC-SIGN and B-cell acti-vation markers (22, 34). Moreover, HHV-8 infection of thetonsillar B cells could be blocked by anti-DC-SIGN MAb,similar to the blocking of HHV-8 infection of activated, blood-derived B cells. Thus, regardless of possible differences inactivation states between aB and tonsillar B cells, DC-SIGNwas required for productive HHV-8 infection of both celltypes. Interestingly, DC-SIGN is differentially expressed by Bcells in the parafollicular regions of lymphatic tissues (22).Several investigations have found HHV-8-infected B cells inthe parafollicular regions of lymph nodes from patients withKS, MCD, and PEL (15, 32, 43). This finding suggests thecolocalization of DC-SIGN and HHV-8 in B cells of the lym-phatics, which may be a source of virus persistence and devel-opment of HHV-8-related lymphomas.

We previously showed that HHV-8 utilizes DC-SIGN toinfect DC and activated macrophages (33). The outcome ofHHV-8 infection in these cell types was quite different thanthat of infection in aB cells. First, and most strikingly, the aBcells were permissive for complete virus replication, with theproduction of infectious virus. In contrast, DC and macro-phages, as well as endothelial cells, support the expression ofonly a limited number of HHV-8 lytic genes early after virusentry and little or no production of infectious virus (28, 33).There also was a limited amount of cell death over 72 h in theinfected B-cell cultures compared to that in uninfected cellcultures, suggesting that infectious virus production and sub-sequent destruction of the cell were restricted to a portion ofthe aB cells. Second, HHV-8 infection of DC and macrophagesresulted in a strong downmodulation of DC-SIGN, whereasDC-SIGN was significantly upregulated in B cells after 24 h ofinfection with HHV-8. A majority of the DC-SIGN-positive Bcells were infected with HHV-8 by 24 to 72 h. This resultsuggests that the increase in the number of B cells expressingDC-SIGN and the intensity of DC-SIGN expression per cellmay be due to direct effects of HHV-8 infection in B cells. Themechanisms for the regulation of DC-SIGN expression areunclear. IL-4 has been shown to upregulate DC-SIGN duringthe differentiation of monocyte-derived DC through the acti-vation of STAT6, and this effect is oppositely regulated byalpha interferon and gamma interferon (35). Regardless of thebasic mechanisms involved, increased expression of DC-SIGNmay serve to promote viral spread and replication in the Bcells.

We observed a moderate downregulation of MHC class Imolecules in HHV-8-infected B-cell cultures. This finding is inagreement with the observations of Ishido et al. (24) andCoscoy and Ganem (12), which implicated the MIR1 andMIR2 products of the K3 and K5 genes in the downregulationof MHC class I expression on target cells, a common immuneevasion strategy employed by herpesviruses. This immuno-modulatory effect of the virus in B cells may contribute to the


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

escape of these cells from targeting by cytotoxic T cells, en-hancing the development of HHV-8-related cancers. The im-munomodulatory effect of HHV-8 infection on antigen presen-tation may be further enhanced by HHV-8 infection of DC andmacrophages. We (46) and others (20) have shown previouslythat T-cell responses to HHV-8 antigens in infected individualsare not very robust and that DC infected with HHV-8 in vitroare poor antigen-presenting cells (33). We are currently inves-tigating the effect of infection with fully replicating HHV-8 onantigen presentation by B cells.

The 12-transmembrane light chain of the human cystine/glutamate exchange transporter system x�

c, xCT, has beenreported previously to be involved in cell fusion and the entryof HHV-8 (26). To address the potential role of xCT in ourB-cell system, we determined the pattern of expression of thismolecule in B cells and cell lines utilized in our study. For thisinvestigation, we used rabbit antiserum raised against the xCT-derived peptide P218-232, which is the same type of antiserumused to block virus-cell fusion in the previous study of xCT(26). We could not directly distinguish a role for DC-SIGNcompared to xCT in blood B cells, as aB cells that could beinfected with HHV-8 expressed both xCT and DC-SIGN.However, Raji B-cell and K562 erythroleukemia cell lines ex-pressed xCT yet were nonpermissive for HHV-8 infection. Incontrast, these same cell lines transfected to express DC-SIGNwere susceptible to HHV-8 infection, as measured by the ex-pression of viral proteins and the detection of increased levelsof viral DNA in the infected cells over time. Thus, xCT waspresent on DC-SIGN-expressing aB cells and cells of a B-cellline and a non-B-cell line that could be infected with HHV-8but also on the parental, non-DC-SIGN-expressing cell linesthat could not be infected with the virus. Furthermore, block-ing experiments showed that the pretreatment of B cells withanti-DC-SIGN MAb resulted in a 99% decrease in the detect-able amount of viral DNA in the supernatants of the infectedcell cultures by 24 h, as opposed to an 8% reduction observedwhen cells were pretreated with xCT antisera. It is possible thatthe lack of inhibitory effect of xCT antisera may be due tolower affinity of the anti-xCT polyclonal antibodies in the rab-bit sera than of the MAb specific for DC-SIGN. However,further support for the conclusion that DC-SIGN but not xCTwas essential for infection with HHV-8 is that we were not ableto detect the expression of xCT mRNA in most of the tonsillarB-cell samples that supported productive HHV-8 replication.Further work is necessary to delineate a possible role for xCTin HHV-8 infection of B cells.

Based on the known endocytic internalization of HHV-8 inhuman endothelial cells and fibroblasts (1) and the finding thatother pathogens that bind DC-SIGN enter cells via endocytosis(16), we reasoned that HHV-8 may be internalized into endo-cytic vacuoles in B cells after attachment to DC-SIGN. To testthis hypothesis, we utilized chemical blockers of endocytosisand also cells transduced to express a form of DC-SIGN lack-ing the transmembrane domain, essential for its endocyticfunction (29). Our data indicate that HHV-8 enters B cellsthrough an endocytic pathway, as demonstrated by the lack ofinfection of cells expressing the truncated form of DC-SIGNversus the full-length DC-SIGN. Previous studies showed thatin human foreskin fibroblasts, HHV-8 enters through endocy-tosis via clathrin-coated pits and that infection can be blocked

by interfering with the acidification of the endosomes or theformation of clathrin (1). To confirm that this also was themechanism used by HHV-8 in aB cells, we employed the sameantiendocytotic chemicals used in the study by Akula et al. (1)in our B-cell system. Our data showed that the inhibition ofclathrin and endosome pH modifications affected the outcomeof infection; i.e., we observed strong inhibition of the expres-sion of viral lytic proteins in the infected B cells. Taken to-gether, these data support the conclusion that HHV-8 enters Bcells through endocytosis that involves the DC-SIGN cytoplas-mic region.

During the past few years, several lines of evidence havepointed to HHV-8’s using different receptors for attachment,binding, and entry depending on the cell type it is targeting,similar to all other herpesviruses (41). Heparan sulfate servesas an attachment molecule, probably concentrating virus onthe cell surface to position it for subsequent binding and entry(2). Integrins are involved in as-yet-undetermined postbindingevents when HHV-8 infects adherent fibroblasts and endothe-lial cells (3, 17), but their role in the infection of other celltypes has not been investigated. Our data demonstrate that inthe three major, professional antigen-presenting cell types, i.e.,myeloid DC, macrophages, and B cells, HHV-8 requires DC-SIGN to result in either abortive or fully productive infection.DC-SIGN is known to internalize its ligands through endocy-tosis, and our data suggest that this is the mechanism of entryof this virus into B cells. We believe that the use of DC-SIGN-expressing aB lymphocytes and cell lines provides a new andunique model to study in vitro viral entry and replication for amajor in vivo cell target of HHV-8. This model may providevaluable insights into the biology of this virus and help in thedesigning of therapeutic strategies to interfere with virus rep-lication in infected individuals.

ACKNOWLEDGMENTS

We thank R. Jaffe and L. Schmitt for help in the collection of tonsilspecimens and Amgen for the gift of CD40L (material transfer agree-ment 200312724).

This work was supported by NIH grants R01 CA 82053 and U01 AI35041.

REFERENCES

1. Akula, S. M., P. P. Naranatt, N. S. Walia, F. Z. Wang, B. Fegley, and B.Chandran. 2003. Kaposi’s sarcoma-associated herpesvirus (human herpesvi-rus 8) infection of human fibroblast cells occurs through endocytosis. J. Vi-rol. 77:7978–7990.

2. Akula, S. M., N. P. Pramod, F. Z. Wang, and B. Chandran. 2001. Humanherpesvirus 8 envelope-associated glycoprotein B interacts with heparansulfate-like moieties. Virology 284:235–249.

3. Akula, S. M., N. P. Pramod, F. Z. Wang, and B. Chandran. 2002. Integrin�3�1 (CD 49c/29) is a cellular receptor for Kaposi’s sarcoma-associatedherpesvirus (KSHV/HHV-8) entry into the target cells. Cell 108:407–419.

4. Akula, S. M., F. Z. Wang, J. Vieira, and B. Chandran. 2001. Human her-pesvirus 8 interaction with target cells involves heparan sulfate. Virology282:245–255.

5. Ambroziak, J. A., D. J. Blackbourn, B. G. Herndier, R. G. Glogau, J. H.Gullett, A. R. McDonald, E. T. Lennette, and J. A. Levy. 1995. Herpes-likesequences in HIV-infected and uninfected Kaposi’s sarcoma patients. Sci-ence 268:582–583.

6. Arvanitakis, L., E. A. Mesri, R. G. Nador, J. W. Said, A. S. Asch, D. M.Knowles, and E. Cesarman. 1996. Establishment and characterization of aprimary effusion (body cavity-based) lymphoma cell line (BC-3) harboringKaposi’s sarcoma-associated herpesvirus (KSHV/HHV-8) in the absence ofEpstein-Barr virus. Blood 88:2648–2654.

7. Blackbourn, D. J., J. Ambroziak, E. Lennette, M. Adams, B. Ramachandran,and J. A. Levy. 1997. Infectious human herpesvirus 8 in a healthy NorthAmerican blood donor. Lancet 349:609–611.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

8. Blackbourn, D. J., E. Lennette, B. Klencke, A. Moses, B. Chandran, M.Weinstein, R. G. Glogau, M. H. Witte, D. L. Way, T. Kutzkey, B. Herndier,and J. A. Levy. 2000. The restricted cellular host range of human herpesvirus8. AIDS 14:1123–1133.

9. Cerimele, F., F. Curreli, S. Ely, A. E. Friedman-Kien, E. Cesarman, and O.Flore. 2001. Kaposi’s sarcoma-associated herpesvirus can productively infectprimary human keratinocytes and alter their growth properties. J. Virol.75:2435–2443.

10. Cesarman, E., Y. Chang, P. S. Moore, J. W. Said, and D. M. Knowles. 1995.Kaposi’s sarcoma-associated herpesvirus-like DNA sequences in AIDS-re-lated body-cavity-based lymphomas. N. Engl. J. Med. 332:1186–1191.

11. Chen, L., and M. Lagunoff. 2005. Establishment and maintenance of Kapo-si’s sarcoma-associated herpesvirus latency in B cells. J. Virol. 79:14383–14391.

12. Coscoy, L., and D. Ganem. 2000. Kaposi’s sarcoma-associated herpesvirusencodes two proteins that block cell surface display of MHC class I chains byenhancing their endocytosis. Proc. Natl. Acad. Sci. USA 97:8051–8056.

13. Coscoy, L., and D. Ganem. 2001. A viral protein that selectively downregu-lates ICAM-1 and B7-2 and modulates T cell costimulation. J. Clin. Investig.107:1599–1606.

14. Curtis, B. M., S. Scharnowske, and A. J. Watson. 1992. Sequence andexpression of a membrane-associated C-type lectin that exhibits CD4-inde-pendent binding of human immunodeficiency virus envelope glycoproteingp120. Proc. Natl. Acad. Sci. USA 89:8356–8360.

15. Dupin, N., C. Fisher, P. Kellam, S. Ariad, M. Tulliez, N. Franck, E. vanMarck, D. Salmon, I. Gorin, J. P. Escande, R. A. Weiss, K. Alitalo, and C.Boshoff. 1999. Distribution of human herpesvirus-8 latently infected cells inKaposi’s sarcoma, multicentric Castleman’s disease, and primary effusionlymphoma. Proc. Natl. Acad. Sci. USA 96:4546–4551.

16. Engering, A., T. B. Geijtenbeek, S. J. van Vliet, M. Wijers, E. van Liempt, N.Demaurex, A. Lanzavecchia, J. Fransen, C. G. Figdor, V. Piguet, and Y. vanKooyk. 2002. The dendritic cell-specific adhesion receptor DC-SIGN inter-nalizes antigen for presentation to T cells. J. Immunol. 168:2118–2126.

17. Garrigues, H. J., Y. E. Rubinchikova, C. M. DiPersio, and T. M. Rose. 2008.Integrin �V�3 binds to the RGD motif of glycoprotein B of Kaposi’s sarcoma-associated herpesvirus and functions as an RGD-dependent entry receptor.J. Virol. 82:1570–1580.

18. Geijtenbeek, T. B., R. Torensma, S. J. van Vliet, G. C. van Duijnhoven, G. J.Adema, Y. van Kooyk, and C. G. Figdor. 2000. Identification of DC-SIGN, anovel dendritic cell-specific ICAM-3 receptor that supports primary immuneresponses. Cell 100:575–585.

19. Godfrey, T. E., S.-H. Kim, M. Chavira, D. W. Ruff, R. S. Warren, J. W. Gray,and R. H. Jensen. 2000. Quantitative mRNA expression analysis from for-malin-fixed, paraffin-embedded tissues using 5� nuclease quantitative reversetranscription-polymerase chain reaction. J. Mol. Diagn. 2:84–91.

20. Guihot, A., N. Dupin, A.-G. Marcelin, I. Gorin, A.-S. Bedin, P. Bossi, L.Galicier, E. Oksenhendler, B. Autran, and G. Carcelain. 2006. Low T cellresponses to human herpesvirus 8 in patients with AIDS-related and classicKaposi sarcoma. J. Infect. Dis. 194:1078–1088.

21. Harrington, W. J., Jr., O. Bagasra, C. E. Sosa, L. E. Bobroski, M. Baum,X. L. Wen, L. Cabral, G. E. Byrne, R. J. Pomerantz, and C. Wood. 1996.Human herpesvirus type 8 DNA sequences in cell-free plasma and mono-nuclear cells of Kaposi’s sarcoma patients. J. Infect. Dis. 174:1101–1105.

22. He, B., X. Qiao, P. J. Klasse, A. Chiu, A. Chadburn, D. M. Knowles, J. P.Moore, and A. Cerutti. 2006. HIV-1 envelope triggers polyclonal Ig classswitch recombination through a CD40-independent mechanism involvingBAFF and C-type lectin receptors. J. Immunol. 176:3931–3941.

23. Huang, Y. Q., J. J. Li, B. J. Poiesz, M. H. Kaplan, and A. E. Friedman-Kien.1997. Detection of the herpesvirus-like DNA sequences in matched speci-mens of semen and blood from patients with AIDS-related Kaposi’s sarcomaby polymerase chain reaction in situ hybridization. Am. J. Pathol. 150:147–153.

24. Ishido, S., C. Wang, B. S. Lee, G. B. Cohen, and J. U. Jung. 2000. Down-regulation of major histocompatibility complex class I molecules by Kaposi’ssarcoma-associated herpesvirus K3 and K5 proteins. J. Virol. 74:5300–5309.

25. Kaleeba, J. A., and E. A. Berger. 2006. Broad target cell selectivity ofKaposi’s sarcoma-associated herpesvirus glycoprotein-mediated cell fusionand virion entry. Virology 354:7–14.

26. Kaleeba, J. A., and E. A. Berger. 2006. Kaposi’s sarcoma-associated herpes-virus fusion-entry receptor: cystine transporter xCT. Science 311:1921–1924.

27. Kliche, S., E. Kremmer, W. Hammerschmidt, U. Koszinowski, and J. Haas.1998. Persistent infection of Epstein-Barr virus-positive B lymphocytes byhuman herpesvirus 8. J. Virol. 72:8143–8149.

28. Krishnan, H. H., P. P. Naranatt, M. S. Smith, L. Zeng, C. Bloomer, and B.Chandran. 2004. Concurrent expression of latent and a limited number of

lytic genes with immune modulation and antiapoptotic function by Kaposi’ssarcoma-associated herpesvirus early during infection of primary endothelialand fibroblast cells and subsequent decline of lytic gene expression. J. Virol.78:3601–3620.

29. Kwon, D. S., G. Gregorio, N. Bitton, W. A. Hendrickson, and D. R. Littman.2002. DC-SIGN-mediated internalization of HIV is required for trans-en-hancement of T cell infection. Immunity 16:135–144.

30. Mesri, E. A., E. Cesarman, L. Arvanitakis, S. Rafii, M. A. Moore, D. N.Posnett, D. M. Knowles, and A. S. Asch. 1996. Human herpesvirus-8/Kapo-si’s sarcoma-associated herpesvirus is a new transmissible virus that infects Bcells. J. Exp. Med. 183:2385–2390.

31. Naranatt, P. P., S. M. Akula, C. A. Zien, H. H. Krishnan, and B. Chandran.2003. Kaposi’s sarcoma-associated herpesvirus induces the phosphatidylino-sitol 3-kinase-PKC- -MEK-ERK signaling pathway in target cells early dur-ing infection: implications for infectivity. J. Virol. 77:1524–1539.

32. Parravicini, C., B. Chandran, M. Corbellino, E. Berti, M. Paulli, P. S.Moore, and Y. Chang. 2000. Differential viral protein expression in Kaposi’ssarcoma-associated herpesvirus-infected diseases: Kaposi’s sarcoma, primaryeffusion lymphoma, and multicentric Castleman’s disease. Am. J. Pathol.156:743–749.

33. Rappocciolo, G., F. J. Jenkins, H. R. Hensler, P. Piazza, M. Jais, L.Borowski, S. C. Watkins, and C. R. Rinaldo, Jr. 2006. DC-SIGN is a receptorfor human herpesvirus 8 on dendritic cells and macrophages. J. Immunol.176:1741–1749.

34. Rappocciolo, G., P. Piazza, C. L. Fuller, T. A. Reinhart, S. C. Watkins, D. T.Rowe, M. Jais, P. Gupta, and C. R. Rinaldo. 2006. DC-SIGN on B lympho-cytes is required for transmission of HIV-1 to T lymphocytes. PLoS Pathog.2:e70.

35. Relloso, M., A. Puig-Kroger, O. M. Pello, J. L. Rodriguez-Fernandez, G. dela Rosa, N. Longo, J. Navarro, M. A. Munoz-Fernandez, P. Sanchez-Mateos,and A. L. Corbi. 2002. DC-SIGN (CD209) expression is IL-4 dependent andis negatively regulated by IFN, TGF-beta, and anti-inflammatory agents.J. Immunol. 168:2634–2643.

36. Renne, R., D. Blackbourn, D. Whitby, J. Levy, and D. Ganem. 1998. Limitedtransmission of Kaposi’s sarcoma-associated herpesvirus in cultured cells.J. Virol. 72:5182–5188.

37. Renne, R., W. Zhong, B. Herndier, M. McGrath, N. Abbey, D. Kedes, and D.Ganem. 1996. Lytic growth of Kaposi’s sarcoma-associated herpesvirus (hu-man herpesvirus 8) in culture. Nat. Med. 2:342–346.

38. Rinaldo, C. R., Jr., and P. Piazza. 2004. Virus infection of dendritic cells:portal for host invasion and host defense. Trends Microbiol. 12:337–345.(Erratum, Trends Microbiol. 12:483.)

39. Said, W., K. Chien, S. Takeuchi, T. Tasaka, H. Asou, S. K. Cho, S. de Vos,E. Cesarman, D. M. Knowles, and H. P. Koeffler. 1996. Kaposi’s sarcoma-associated herpesvirus (KSHV or HHV8) in primary effusion lymphoma:ultrastructural demonstration of herpesvirus in lymphoma cells. Blood 87:4937–4943.

40. Soilleux, E. J., L. S. Morris, J. Trowsdale, N. Coleman, and J. J. Boyle. 2002.Human atherosclerotic plaques express DC-SIGN, a novel protein found ondendritic cells and macrophages. J. Pathol. 198:511–516.

41. Spear, P. G., and R. Longnecker. 2003. Herpesvirus entry: an update. J. Vi-rol. 77:10179–10185.

42. Stamey, F. R., M. M. Patel, B. P. Holloway, and P. E. Pellett. 2001. Quan-titative, fluorogenic probe PCR assay for detection of human herpesvirus 8DNA in clinical specimens. J. Clin. Microbiol. 39:3537–3540.

43. Staskus, K. A., R. Sun, G. Miller, P. Racz, A. Jaslowski, C. Metroka, H.Brett-Smith, and A. T. Haase. 1999. Cellular tropism and viral interleukin-6expression distinguish human herpesvirus 8 involvement in Kaposi’s sar-coma, primary effusion lymphoma, and multicentric Castleman’s disease.J. Virol. 73:4181–4187.

44. van Kooyk, Y., and T. B. Geijtenbeek. 2003. DC-SIGN: escape mechanismfor pathogens. Nat. Rev. Immunol. 3:697–709.

45. Vieira, J., P. O’Hearn, L. Kimball, B. Chandran, and L. Corey. 2001. Acti-vation of Kaposi’s sarcoma-associated herpesvirus (human herpesvirus 8)lytic replication by human cytomegalovirus. J. Virol. 75:1378–1386.

46. Wang, Q. J., X. L. Huang, G. Rappocciolo, F. J. Jenkins, W. H. Hildebrand,Z. Fan, E. K. Thomas, and C. R. Rinaldo, Jr. 2002. Identification of an HLAA*0201-restricted CD8� T-cell epitope for the glycoprotein B homolog ofhuman herpesvirus 8. Blood 99:3360–3366.

47. Watanabe, T., K. Yoshizaki, T. Yagura, and Y. Yamamura. 1974. In vitroantibody formation by human tonsil lymphocytes. J. Immunol. 113:608–616.

48. Wu, L., T. D. Martin, M. Carrington, and V. N. KewalRamani. 2004. Raji Bcells, misidentified as THP-1 cells, stimulate DC-SIGN-mediated HIV trans-mission. Virology 318:17–23.


Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 13

Oct

ober

202

1 by

183

.179

.115

.79.

Documents

Human Herpesvirus 8 Infects and Replicates in Primary Cultures of Activated B Lymphocytes