Upregulation of xCT by KSHV-Encoded microRNAs Facilitates KSHV Dissemination and Persistence in an Environment of Oxidative Stress

Upregulation of xCT, the inducible subunit of a membrane-bound amino acid transporter, replenishes intracellular glutathione stores to maintain cell viability in an environment of oxidative stress. xCT also serves as a fusion-entry receptor for the Kaposi's sarcoma-associated herpesvirus (KSHV), the causative agent of Kaposi's sarcoma (KS). Ongoing KSHV replication and infection of new cell targets is important for KS progression, but whether xCT regulation within the tumor microenvironment plays a role in KS pathogenesis has not been determined. Using gene transfer and whole virus infection experiments, we found that KSHV-encoded microRNAs (KSHV miRNAs) upregulate xCT expression by macrophages and endothelial cells, largely through miR-K12-11 suppression of BACH-1—a negative regulator of transcription recognizing antioxidant response elements within gene promoters. Correlative functional studies reveal that upregulation of xCT by KSHV miRNAs increases cell permissiveness for KSHV infection and protects infected cells from death induced by reactive nitrogen species (RNS). Interestingly, KSHV miRNAs simultaneously upregulate macrophage secretion of RNS, and biochemical inhibition of RNS secretion by macrophages significantly reduces their permissiveness for KSHV infection. The clinical relevance of these findings is supported by our demonstration of increased xCT expression within more advanced human KS tumors containing a larger number of KSHV-infected cells. Collectively, these data support a role for KSHV itself in promoting de novo KSHV infection and the survival of KSHV-infected, RNS-secreting cells in the tumor microenvironment through the induction of xCT.


Introduction
Patients with immune deficiencies are at risk for lifethreatening illnesses caused by herpesviruses, including the Kaposi's sarcoma-associated herpesvirus (KSHV). Bone marrow failure [1], lymphoproliferative syndromes [2,3], and sarcoma [4] have all been etiologically linked to KSHV infection and occur with greater frequency in the setting of immune suppression related to HIV infection [5,6] or organ transplantation [7,8]. The most commonly encountered clinical manifestation of KSHV infection, Kaposi's sarcoma (KS), represents one of the most common tumors arising in the setting of HIV infection, one of the most common transplant-associated tumors, and a leading cause of morbidity and mortality [5][6][7]. Moreover, KS is the most common tumor arising in the general population in some geographic areas [9]. Despite the reduced incidence of KS in the modern era of highly active antiretroviral therapy (HAART) [10], KS is increasingly recognized in HIV-infected patients with suppressed HIV viral loads and elevated CD4 + T cell counts [11,12]. Clinical responses to cytotoxic agents for systemic KS vary widely in published trials, and these agents incur many side effects which may exacerbate or add to those already incurred by antiretroviral or immunosuppressive agents [10,13]. Given these shortcomings of existing therapies, novel targeted strategies are needed for the treatment or prevention of KS.
Published data support a role for KSHV-encoded genes in KS pathogenesis, including genes expressed primarily during lytic replication that facilitate angiogenesis and endothelial cell survival [14], and existing clinical data support this concept. An elevated KSHV viral load in the peripheral circulation predicts the onset and progression of both AIDS-and non-AIDS-related KS, and intralesional KSHV viral load correlates directly with tumor progression [15][16][17]. One retrospective clinical study demonstrated that ganciclovir, a nucleoside analog that inhibits viral DNA polymerase activity and reduces KSHV replication [18], reduced the incidence of KS in patients receiving organ transplants [19]. In addition, KS arising in the setting of well-controlled HIV infection may be explained in part by reduced KSHV-specific immunity despite general immune recovery with HAART [20,21]. Together, these data suggest a role for ongoing KSHV replication and infection of naïve target cells in the progression of KS. Interestingly, neither KS lesional spindle cells nor cultured endothelial cells infected by KSHV in vitro efficiently maintain viral episomes when passed in culture [22,23], suggesting the potential importance of additional microenvironmental factors within KS tumors for facilitating KSHV infection.
The amino acid membrane transport system x c 2 consists of a conserved heavy chain, 4F2hc, and an inducible subunit, xCT, that mediates amino acid exchange [24]. x c 2 exchanges intracellular glutamate for extracellular cystine at the cell membrane, and the latter is rapidly reduced in the intracellular space to cysteine and incorporated into glutathione (GSH) and other protein biosynthesis pathways [25]. This allows for restoration of intracellular GSH stores and protection of x c 2expressing cells from oxidative stress and cell death [25]. xCT expression is upregulated by physiological conditions that impact intracellular GSH levels, such as hypoxia, inflammation, and increased production of reactive species [25]. xCT was also recently identified as a fusion-entry receptor for KSHV and may mediate KSHV entry either in isolation or as part of a complex with other receptors for the virus [26,27]. KSHV establishes infection within multiple xCT-expressing cell types that have been implicated in KS pathogenesis, including intralesional or circulating monocytes, intralesional macrophages, and endothelial cells [26][27][28][29]. However, whether KSHV itself also regulates xCT expression to promote viral infection of new cell targets or increase the longevity of KSHV-infected cells in the local environment is unknown.
miRNAs are small (19)(20)(21)(22)(23) nucleotides in length), non-coding RNAs that bind target mRNAs, marking them for degradation or post-transcriptional modification, and KSHV encodes 17 mature miRNAs which are expressed within KSHV-infected cells and KS lesions [30][31][32][33]. xCT expression is regulated through competitive binding of positive and negative transcription factors to an ''Antioxidant Response Element'' (ARE) in the xCT promoter [34,35], and existing data suggest that negative transcription regulators of ARE may be targeted by KSHV miRNAs [36][37][38][39][40]. Therefore, using cell culture systems employing macrophages and endothelial cells, we sought to determine whether KSHV miRNAs regulate the expression of xCT, and if so, whether this infuences cell permissiveness for KSHV infection and protection of infected cells from oxidative stress.

xCT is a principal determinant of macrophage permissiveness for KSHV infection
To first determine whether xCT expression correlates with macrophage susceptibility to KSHV infection, we utilized a BALB/c-derived murine macrophage cell line, 264.7 cells (''RAW'' cells). We chose RAW cells given the recent demonstration of KSHV infection of murine macrophages in vivo [41], the identity (89%) and similarity (93%) of the murine xCT protein to its human counterpart [25], and the utility of these cells for gene transfer studies. xCT expression is induced indirectly by substrates that compete for cystine uptake by xCT, like monosodium glutamate (Msg) [42] and sulfasalazine (Sul) [43]. We found that Msg or Sul significantly increased the number of RAW cells expressing the KSHV latency-associated nuclear antigen (LANA) following their incubation with purified KSHV (Fig. 1A-E). This increase in the number of infected cells was reflected in an increase in viral episome copies for cells from Msg-and Sul-treated cultures (Fig. 1F), although IFA suggested that the number of episomes (LANA dots) per cell was unchanged ( Fig. 1A-D). Msg and Sul also increased xCT transcript expression by RAW cells (Fig. 1G), and direct targeting of xCT with siRNA significantly reduced the number of LANA-positive cells following RAW cell incubation with KSHV ( Fig. 1H and I). These data confirm the role of xCT as a principal determinant of RAW cell susceptibility to KSHV infection.

KSHV miRNAs regulate xCT expression and macrophage permissiveness for KSHV infection
To explore whether KSHV miRNAs regulate xCT expression, we co-transfected RAW cells with a construct encoding 10 of the 17 mature KSHV miRNAs described elsewhere [44]. Using semiquantitative RT-PCR, we first confirmed upregulation of xCT with the collective expression of KSHV miRNAs encoded in the construct ( Fig. 2A). Using a KSHV miRNA target prediction algorithm validated previously [36], we identified KSHV miRNA binding sites within 3'UTR of several murine genes associated with the regulation of xCT. The majority of binding sites were identified for 3 KSHV miRNAs: miR-K12-1, miR-K12-9, and miR-K12-11 (data not shown). To first determine whether these miRNAs were expressed within KSHV miRNA transfectants, we co-transfected cells with KSHV miRNAs and pGL3 luciferase

Author Summary
Herpesviruses are the most common etiologic agents of cancer in patients with suppressed immune function, and the Kaposi's sarcoma-associated herpesvirus (KSHV) is one of the most common causes of cancer in this setting. KSHV infection of new cell targets is critical for tumor progression, and a better understanding of how viral receptors on the surface of cells are regulated in the tumor microenvironment may lead to new therapies. KSHV encodes unique RNAs called microRNAs (KSHV miRNAs) that regulate a variety of cell functions. In this study, we show that KSHV miRNAs increase the susceptibility of cells to KSHV infection and protect infected cells from death induced by cancer-promoting reactive nitrogen species (RNS). They accomplish this in large part by increasing cell surface expression of a transport protein subunit called xCT. We also show that KSHV miRNAs increase secretion of RNS by infected cells, and that blocking RNS secretion reduces the ability of KSHV to infect cells. Therefore, by regulating xCT and RNS, we find KSHV is able to ''finetune'' cell function in order to maintain a stable population of infected cells which secrete cancer-promoting factors in the local environment. This work has important implications for developing new therapies to target xCT and reduce survival of KSHV-infected tumor cells.
KSHV miR-K12-11 suppresses BACH-1 to induce xCT expression and cell permissiveness for KSHV infection in macrophages and endothelial cells Our bioinformatics analyses revealed putative binding sites for miR-K12-11 within the 3'UTR of the murine BACH-1 gene (Fig. 4A). We subsequently confirmed KSHV miRNA suppression of BACH-1 within RAW cells, an effect largely reversed through direct targeting of KSHV miR-K12-11 (Fig. 4B). In addition, direct siRNA targeting of BACH-1 significantly increased basal levels of xCT expression in these cells (Fig. 4C-E) as well as macrophage permissiveness to KSHV infection (Fig. 4F), although to a lesser degree than the collective expression of miR-K12-1, miR-K12-9, and miR-K12-11 (Fig. 3).
Published data have confirmed direct targeting of BACH-1 by miR-K12-11 in human cells [36]. To validate our observations and to determine their broader significance for human cells with known relevance to KS pathogenesis, we repeated our experiments using primary human umbilical vein endothelial cells (HUVEC). We found that Msg, Sul or KSHV miRNA transfection significantly increased xCT transcript expression and, based on IFA, KSHV episome copy number per cell following subsequent de novo infection ( Fig. 5A-J). In contrast to what was observed for RAW cells, IFA indicated that the total number of infected HUVEC remained unchanged with these interventions (Fig. S1). In addition, either direct suppression of xCT by siRNA or concurrent inhibition of miR-K12-11 reduced xCT expression and intracellular viral load in KSHV miRNA transfectants ( Fig. 5H-J). Collective expression of KSHV miRNAs also reduced BACH-1 expression in HUVEC, an effect suppressed with concurrent inhibition of miR-K12-11 (Fig. 5K).
KSHV miRNAs upregulate macrophage secretion of reactive nitrogen species (RNS) and protect macrophages from RNS-induced cell death xCT restoration of intracellular glutathione and increased scavenging of free radicals reduces cell death resulting from nitration of proteins, lipids, and nucleic acids by RNS [25,45,46].
Because we observed increased xCT expression induced by KSHV miRNAs, we hypothesized that KSHV may also increase RNS secretion by macrophages and that xCT upregulation would serve as an auto-protective mechanism in this environment. Using a standard Greiss reaction assay for quantifying nitrite in culture supernatants as a surrogate measure of RNS secretion, we found that KSHV infection of RAW cells induced a ,20-fold increase in RNS secretion and that the majority of this effect was mediated through the collective expression of miR-K12-1, miR-K12-9, and miR-K12-11 (Fig. 6A). A similar pattern was observed following overexpression of KSHV miRNAs ( Fig. 6B and C). Non-specific TLR activation, as might be initiated by mature miRNAs or their precursors, is capable of inducing RNS production [47,48]. However, specific inhibitors of MyD88-independent and -dependent toll-like receptor (TLR) pathways failed to reduce induction of RNS secretion by KSHV miRNAs, suggesting that this effect is not mediated through TLR activation (Fig. S2).
To determine whether upregulation of xCT by KSHV miRNAs offers a protective mechanism for macrophages in an either control or miRNA-expressing vectors. 48 h later, luciferase expression was determined for miRNA transfectants relative to controls (RLU). (C) Cells were transfected with control or miRNA-expressing vectors with or without 2'OMe RNA antagomirs targeting miR-K12-1, miR-K12-9, and miR-K12-11 (mi1/9/11). 48 h following subsequent incubation with KSHV (K), qRT-PCR was used to determine relative xCT transcript expression. For all assays, error bars represent the S.E.M. for three independent experiments. * * = p,0.01. doi:10.1371/journal.ppat.1000742.g002 environment rich in RNS, we first established that provision of the nitric oxide (NO) donor S-nitroso-N-acetylpenicillamine (SNAP) [49] increased RNS concentrations within RAW cell culture supernatants and induced cell death in a dose-dependent manner ( Fig. 7A and B). Subsequently, we found that either KSHV infection or overexpression of KSHV miRNAs significantly increased macrophage resistance to SNAP-induced cell death. Moreover, siRNA experiments confirmed that this effect was mediated primarily through the upregulation of xCT (Figs. 7C and D).

RNS facilitate KSHV infection of macrophages
RNS are expressed within KS lesions [50], but whether RNS themselves influence de novo KSHV infection is unknown. To reduce macrophage secretion of RNS, we incubated RAW cells with L-N 6 -monomethyl-arginine (L-NMMA), an inhibitor of all forms of nitric oxide synthase (NOS) [51] that induces no discernable toxicity for RAW cells over a wide range of concentrations (Fig. S3). Interestingly, L-NMMA significantly reduced secretion of RNS initiated by KSHV miRNAs and reduced de novo KSHV infection of macrophages in a dosedependent manner (Fig. 8), suggesting a role for NOS and RNS in facilitating de novo KSHV infection.
xCT expression is increased within more advanced KS lesions KSHV miRNAs are expressed within KS lesions [30][31][32] but to our knowledge, expression of xCT within KS tissue has never been demonstrated. To address this, we used immunohistochemistry to quantify xCT expression within KS skin lesions representing the full spectrum of histopathologic progression of KS. We found that stage I tumors (patches) and stage II tumors (plaques) exhibited either no or minimally discernable xCT expression, respectively (Fig. 9). In contrast, stage III tumors (nodules) exhibited easily discernable membrane expression of xCT by the majority of cells in these lesions, including nearly all spindle-shaped cells (Fig. 9). Moreover, we confirmed that stage III tumors exhibited significantly more LANA + cells than stage I lesions (Fig. S4) in agreement with published data [52,53] as well as our

Discussion
In this study, we found that KSHV-encoded miR-K12-11 upregulates the expression of xCT in macrophages and endothelial cells, in part through suppression of a negative regulator of gene transcription, BACH-1. We also found that KSHV miRNAs induce macrophage secretion of RNS and protect these cells from RNS-induced cell death through the upregulation of xCT. Moreover, reducing NOS activity and RNS secretion by macrophages reduces their permissiveness for KSHV. Finally, we found that cells within more advanced human KS tumors express more xCT than cells from early-stage lesions. We hypothesize, therefore, that KSHV miRNAs facilitate KS pathogenesis through cooperative mechanisms that regulate xCT and RNS secretion to promote ongoing de novo infection and survival of infected cells in the tumor microenvironment.
Implications of the regulation of BACH-1, xCT expression, and cell susceptibility to KSHV infection by KSHV miRNAs xCT expression is differentially regulated during oxidative stress through transcription factor binding to the cis-acting ARE in its promoter [34,35]. Transcription factors that bind to the ARE include a positive regulator known as Nuclear factor erythroid 2related factor-2 (Nrf-2) [35] and negative regulators, including BACH-1 and c-Maf, which competitively reduce Nrf-2 binding to the ARE thereby repressing ARE-mediated gene expression [37,38]. KSHV miRNAs are expressed within KSHV-infected cells and KS lesions [30][31][32], and existing data suggest that both BACH-1 and c-Maf are targeted by KSHV miRNAs [36,39,40]. More specifically, KSHV miR-K12-11, an ortholog of cellular miR-155, targets and reduces expression of BACH-1 [36]. miR-155 downregulates c-Maf expression by T cells [40], and KSHV miRNAs downregulate c-Maf expression in endothelial cells [39]. Therefore, we hypothesized that miR-K12-11, in cooperation with other KSHV miRNAs, regulates xCT expression. We found that miR-K12-11 downregulated BACH-1 and induced xCT expression in both macrophages and endothelial cells, although additional experiments using site-directed mutagenesis are needed to confirm direct interactions between BACH-1 and the xCT ARE in murine cells. Further validating our findings, we found that BACH-1 targeting by siRNA increased macrophage permissiveness for infection by approximately 70% (Fig. 4F), although overexpression of multiple miRNAs increased permissiveness by approximately 250% (Fig. 3E). Differences in transfection efficiency for siRNA and the KSHV miRNA constructs could be partially responsible for this discrepancy, but we hypothesize that it is due in part to the effect of multiple KSHV miRNAs, including miR-K12-1 and miR-K12-9, and the cooperative targeting of multiple genes. Our initial screen for KSHV miRNA binding sites within murine and human genes known to regulate xCT expression and RNS secretion revealed multiple binding sites for miR-K12-1, miR-K12-9 and miR-K12-11 (data not shown). These analyses also revealed binding sites for miR-K12-4, although not miR-K12-1 or miR-K12-9, within both murine and human BACH-1 3'UTR sequences (not shown), although we have not yet confirmed the functional impact of miR-K12-4 expression on BACH-1 or xCT expression. Additional studies are needed to confirm direct targeting of BACH-1 or other genes by these KSHV miRNAs and to characterize the functional impact of this targeting for expression of xCT and other AREcontaining genes regulated by BACH-1, including those involved in the generation of RNS (see below).
To our knowledge, these data are the first to suggest a role for a herpesvirus in the autocrine upregulation of its own receptor, although whether increased cell permissiveness for KSHV entry following initial infection and miRNA expression is ''accidental'' or ''purposeful'' in the context of KSHV-host evolution remains debatable. In addition to promoting cell survival (see below), autocrine upregulation of xCT may provide evolutionary advantages for the virus achieved through an increase in intracellular viral load that were not addressed by our studies. This concept is supported by several reports revealing that a significant proportion of KSHV-infected tumor cells, including those within KS lesions, contain multiple viral clones [54][55][56]. Another study showed that the downregulation of MHC Class I (MHC-I) in KSHV-infected cells is directly proportional to the level of expression of the KSHV modulator of immune recognition 2 (MIR2) and intracellular KSHV episome copy number [57], implying that increasing intracellular viral copies may promote reduced KSHV epitope presentation to CD8 + T cells as a mechanism for immune evasion. Our IFA indicated that for RAW cells, miRNA upregulation of xCT increased the permissiveness of uninfected cells for KSHV, although not viral episome copies within individual cells. In contrast, miRNA upregulation of xCT increased HUVEC viral episome copies per cell following subsequent infection, although not the total number of infected cells. Our studies did not directly address whether the observed increase in episome copies per cell for HUVEC is the result of intracellular episome replication or ''superinfection'' with exogenous virions. Experiments utilizing limiting dilution PCR [58] or single cell imaging techniques [41,59] could be used to confirm whether intracellular KSHV viral load and miRNA expression correlate with xCT expression on a single cell level and to address the possibility that KSHV miRNA upregulation of xCT increases cell permissiveness for subsequent virion entry. It is interesting to speculate whether autocrine regulation of surface receptors by KSHV miRNAs differs depending on the cell type, and whether soluble factors released by infected cells differentially influence xCT expression for different cell types.

Implications of the regulation of RNS secretion, xCT expression, and survival of KSHV-infected cells by KSHV miRNAs
Our data indicate a role for KSHV miRNAs in the induction of RNS secretion and the protection of cells from RNS-induced cell death through the upregulation of xCT. Of additional relevance, we found that L-NMMA, an inhibitor of NOS, reduced KSHV miRNA-induced secretion of RNS and de novo KSHV infection. Additional studies are currently underway to elucidate the OMe RNA antagomirs targeting miR-K12-1, miR-K12-9, and miR-K12-11 or all three together (mi1/9/11). Cells were subsequently incubated with purified KSHV (K) and nitrite quantified within culture supernatants as described in Methods. (B) Cells were transfected with control or miRNA-expressing vectors and nitrite quantified within culture supernatants at the times indicated. (C) Cells were co-transfected with either control or miRNAexpression constructs without (mock) or with specific 2'OMe RNA antagomirs. As an additional control, some cells were transfected with an antagomir targeting miR-K12-12 which is not expressed by the miRNA-expressing construct. For all assays, error bars represent the S.E.M. for three independent experiments. * = p,0.05, * * = p,0.01. doi:10.1371/journal.ppat.1000742.g006 mechanism for these observations. Through the nitration of either extracellular or intracellular proteins, RNS activate Nrf-2 [34,35,60] and, therefore, may upregulate xCT expression through both autocrine and paracrine mechanisms. It is also conceivable that miR-K12-11 downregulation of BACH-1 increases expression of other ARE-containing genes involved in the induction of RNS or the protection of cells from oxidative stress, including heme oxygenase-1 (HO-1) [34,35]. Interestingly, HO-1 is expressed within KS lesions, and KSHV infection of endothelial cells induces activation of HO-1 [61]. It is probable that RNS secretion and downstream consequences are mediated through the collective targeting of multiple genes by KSHV miRNAs, and that this may occur within a variety of KSHV-infected cells with the capacity to secrete RNS, including endothelial cells and dendritic cells [25]. Characterization of miRNA regulation of RNS for a broader array of cell types relevant to KS pathogenesis is underway. Furthermore, our studies do not address whether KSHV miRNAs regulate secretion of reactive oxygen species (ROS) by infected cells, and it is conceivable that ROS play a role in the paracrine regulation of xCT or other events pertaining to KSHV infection. Studies are ongoing in our laboratory to define the relative importance of specific reactive species in the regulation of xCT expression and KSHV dissemination in the microenvironment.
Multiple studies implicate RNS in KS pathogenesis [46,50,[62][63][64][65][66]. RNS and NOS are expressed within KS lesions [50,62], and RNS induce endothelial cell migration, proliferation and angiogenesis [46] as well as T cell apoptosis [63]. Moreover, existing data support a role for KSHV in the regulation of superoxide dismutase (SOD) in the KS microenvironment [50,67], and cytokines associated with KS pathogenesis have been implicated in the activation of RNS secretion by macrophages [64,65]. Interestingly, a recent publication demonstrated that Rac1 transgenic mice overexpressing NADPH-oxidase-dependent reactive species developed KS-like lesions and that systemic administration of the antioxidant N-acetylcysteine reduced KS formation in this model [66]. Preliminary experiments performed in our laboratory have revealed that inhibition of NADPH-oxidase using diphenylene iodonium (DPI) also reduces KSHV miRNA-induced RNS secretion and infection of naïve cells (data not shown). In addition, at least one study has implicated cellular miRNAs in the regulation of Rac1 [68]. Therefore, our findings have important implications for paracrine regulation of cellular events pertaining to KS pathogenesis, and systemic inhibition of RNS may interfere

Clinical implications of xCT expression within KS lesions
We have demonstrated that xCT is expressed to a greater extent within more advanced KS lesions containing a greater number of KSHV-infected cells. To our knowledge, these are the first data to demonstrate xCT expression in clinical samples from KSHVinfected patients and are consistent with published data documenting higher KSHV intratumoral viral loads within more advanced KS lesions [52,53]. Importantly, they also support our hypothesis that KSHV upregulation of xCT facilitates expansion of the KSHV reservoir in the microenvironment and KS progression. Moreover, our observation that spindle-shaped cells within stage III tumors express xCT is consistent with our data revealing upregulation of xCT and KSHV permissiveness for endothelial cells in vitro by KSHV miRNAs. Additional studies are needed to confirm whether KSHV miRNAs, BACH-1 and other putative xCT regulatory factors are differentially expressed during different stages of KS progression. Future translational studies may also shed light on whether quantifying xCT in clinical samples provides additional prognostic information for patients at risk for KS, and whether targeting xCT or its regulatory pathways will offer a useful approach for the treatment or prevention of this disease.

Antibodies and reagents
Antibodies recognizing BACH-1 (H-130) and b-Actin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and Transfection assays A 2.8 Kbp construct encoding 10 individual KSHV micro-RNAs (pcDNA-miRNA, containing miR-K12-1/2/3/4/5/6/7/ 8/9/11), and luciferase reporter constructs encoding complimentary sequences for individual miRNA (pGL3-miRNA sensors), have been validated previously in transfection assays for expression of KSHV miRNAs [44]. These constructs were used to transiently transfect RAW cells and HUVEC. For inhibition of mature miRNAs, 2'OMe RNA antagomirs were designed and purchased from Dharmacon (Chicago, IL) as previously described [44]. BACH-1, xCT, and non-target (control) siRNAs (ON-TARGET plus SMART pool) were also purchased from Dharmacon. Cells were transfected with 1 mg pcDNA-miRNA, 0.5 mg pGL3-miRNA sensors, 300 pmol 2'OMe RNA antagomirs, siRNAs, and/or 1 mg pcDNA for negative controls in 12-well plates using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) and/or Dharma-FECT Transfection Reagent (Dharmacon, Chicago, IL) for 48 h prior to their incubation with KSHV. For miRNA inhibitor assays, control cells were transfected with a 2'OMe RNA antagomir targeting miR-K12-12, a KSHV miRNA not encoded by the pcDNA-miRNA construct. For luciferase expression assays, cells were incubated with 100 mL lysis buffer (Promega, Madison, WI), and luciferase activity determined within lysates using a Berthold FB12 luminometer (Titertek, Huntsville, AL). Light units were normalized to total protein levels for each sample using the BCA protein assay kit according to the manufacturer's instructions (Pierce, Rockford, IL). Transfection efficiency was assessed through co-transfection of a lacZ reporter construct kindly provided by Dr. Yusuf Hannun (Medical University of South Carolina, Charleston, SC), and b-galactosidase activity determined using a commercially available b-galactosidase enzyme assay system according to the manufacturer's instructions (Promega, Madison, WI). 3 independent transfections were performed for each experiment, and all samples were analyzed in triplicate for each transfection.

Nitrite quantification
Nitrite concentrations within culture supernatants were determined using the Griess Reagent System (Promega, Madison, WI) according to the manufacturer's instructions.

Immunoblotting
Cells were lysed in buffer containing 20 mM Tris (pH 7.5), 150 mM NaCl, 1% NP40, 1 mM EDTA, 5 mM NaF and 5 mM Na 3 VO 4 . 30 mg of total cell lysate was resolved by SDS-10% PAGE and transferred to nitrocellulose membranes prior to incubation with antibodies for proteins of interest as well as b-Actin for loading controls. Immunoreactive bands were developed by enhanced chemiluminescence reaction (Perkin-Elmer), visualized by autoradiography, and quantified using Image-J software.

Bioinformatics analysis
The 3'UTR sequences of BACH-1 and other RNS-associated genes were obtained from Ensembl (http://www.ensembl.org). 3'UTRs were analyzed to extract all potential KSHV miRNA binding sites using an ad-hoc scanning program specifically developed to assess 3'UTR KSHV miRNA seed sequence matching, as validated previously [36].

KSHV purification and infection
BCBL-1 cells were incubated with 0.6 mM valproic acid for 4-6 days, and KSHV was purified from culture supernatants by ultracentrifugation at 20,0006g for 3 h, 4uC. The viral pellet was resuspended in 1/100 the original volume in the appropriate culture media, and aliquots were frozen at 280uC. Target cells were incubated with concentrated virus in the presence of 8 mg/ mL polybrene (Sigma-Aldrich) for 2 h at 37uC. Inactivated KSHV used for negative controls was prepared by incubating viral stocks with ultraviolet (UV) light (1200 J/cm 2 ) for 109 in a CL-1000 Ultraviolet Crosslinker (UVP). The concentration of infectious viral particles used in each experiment (multiplicity of infection [MOI]) was calculated as previously described [57,69].
Immunofluorescence assays and determination of relative infection rates Biochemical assays 1610 4 RAW cells or HUVEC were seeded per well in eightwell chamber slides and incubated with 10 mM Msg, 0.3 mM Sul or 0.01-1.0 mM L-NMMA for 12 h at 37uC, then incubated with cell-free KSHV for 2 h at 37uC. After 12 h, LANA expression within RAW cells was determined by IFA as outlined above.

Toll-like receptor inhibition
RAW cells were transfected with 1 mg of pcDNA-miRNA or empty vector control, and after 24 h, incubated with 10 mM of either drug vehicle or a double-stranded RNA-activated protein kinase (PKR) inhibitor 2-Aminopurine (InvivoGen, San Diego, CA) for an additional 3 h. In parallel experiments, transfectants were incubated with 100 mM of MyD88 inhibitory peptide or control peptide (Imgenex, San Diego, CA) for an additional 24 h. Nitrite concentration was quantified within culture supernatants as detailed previously.

Cell viability assays
Cell viability was assessed using a standard MTT assay as previously described [70]. A total of 5610 3 RAW cells were incubated in individual wells in a 96-well plate for 24 h. Serial dilutions of L-NMMA were then added and after 24-48 h, cells were incubated in 1 mg/ml of MTT solution (Sigma-Aldrich) at 37uC for 3 h then 50% DMSO overnight and optical density at 570 nm determined by spectrophotometer (Thermo Labsystems). For assessing cell viability in an environment of oxidative stress, we transfected or infected RAW cells with pcDNA-miRNA (pcDNA control) or KSHV (or UV-KSHV control) in the presence of siRNA targeting xCT or control non-target siRNA. Thereafter, cells were incubated for 12 h with SNAP and cell viability determined using 0.4% trypan blue (MP Biomedicals, Solon, Ohio) to identify dead cells under light microscopy. Relative differences for cell viability between groups was determined as follows: relative cell viability = dead cells per 200 cells in experimental group/dead cells per 200 cells in vehicle control group.

PCR
Total DNA was isolated using the QIAamp DNA Mini kit (QIAGEN). Briefly, cells were trypsinized for 59 at 37uC and collected with 1 mL of ice-cold DMEM. Cells were pelleted at 2,000 rpm for 59, washed, and resuspended in 200 mL of 1phosphate-buffered saline (PBS), and total DNA was prepared according to the manufacturer's instructions. Total RNA was isolated using the RNeasy Mini kit (QIAGEN) as previously demonstrated [41]. cDNA was synthesized from total RNA using SuperScript III First-Strand Synthesis SuperMix Kit (Invitrogen) according to the manufacturer's instructions. Coding sequences for genes of interest and b-actin (loading control) were amplified from 200 ng input cDNA and using iQ SYBR Green Supermix

Immunohistochemistry
Archived, paraffin-embedded KS skin lesions were collected from the Medical University of South Carolina (MUSC) Hollings Cancer Center Tumor Bank and the Maize Center for Dermatopathology (Charleston, S.C.). The diagnosis of KS and the histopathologic stage of each lesion were verified by an independent dermatopathologist. Histopathologic staging was determined using published criteria [71] to characterize lesions as patches, plaques or nodules (with nodules representing the most advanced lesional stage). Tissue sections were deparaffinized and hydrated through xylene and graded alcohol series, rinsed for 59 in distilled water, incubated for 109 in 3% hydrogen peroxide, and following PBS wash, incubated for 309 in a commercial antigen retrieval solution (Vector Laboratories, Burlingame, CA) at 100uC. Thereafter, all sections were incubated in 100% rabbit serum for 309 to reduce non-specific staining then for 1 hour with either control preimmune rabbit sera or xCT antisera diluted 1:400. In parallel, representative sections were incubated with 1:1200 dilution of the anti-LANA rat monoclonal antibody (ABI). All sections were subsequently incubated for 309 with a commercially available biotinylated secondary antibody and reagents according to the manufacturer's instructions (Vector Laboratories, Burlingame, CA). Bound antibodies were recognized using a 3,39-Diaminobenzidine (DAB) substrate and nuclei identified using either hematoxylin to contrast xCT expression, or Methyl Green to contrast LANA expression. xCT expression was determined for at least 8 independent tumors representing each of three histopathologic stages of KS (patches, plaques, and nodules). LANA expression was determined for patch and nodular lesions in this cohort.

Statistical analysis
Significance for differences between experimental and control groups was determined using the two-tailed Student's t-test (Excel 8.0), and p values ,0.05 or ,0.01 were considered significant or highly significant, respectively.  Figure S4 Advanced KS lesions contain more KSHV-infected cells relative to early-stage lesions. Representative early (I) and late (III) stage lesions were processed for immunohistochemistry as described in Methods using secondary antibodies alone (control) or anti-LANA antibodies followed by secondary antibodies (LANA). LANA expression is indicated by dark brown, punctate intranuclear staining seen best at higher power (representative LANA + cells are identified with red arrows). Images are shown at original magnification 620 or 50.