Oncogenic herpesviruses sending mixed signals.

Human herpesvirus infections are generally asymptomatic in immune-competent hosts; however, immune dysfunction can unveil inherent oncogenic properties associated with Epstein-Barr virus (EBV) and Kaposi’s sarcoma herpesvirus (KSHV) infection. Although these viruses were first discovered in human tumor specimens in 1964 and 1994, respectively, it was discovered later on that large proportions of the human population are persistently infected with EBV and to a lesser extent KSHV. To colonize immunecompetent hosts, EBV and KSHV drive proliferation of newly infected cells by latent transcription programs. EBV infection is principally limited to B cells, and few infected epithelial cells in the oral mucosa that produce infectious virus for host–host transmission via the saliva. In contrast, KSHV has a broader spectrum of target cells while transmission and life cycle are less well understood. However, these viruses share a hostcolonization and persistence strategy that involves activation of NF-κB, a critical survival factor for developing human B cells. Deregulated NF-κB activation can cause uncontrolled B-cell proliferation and secretion of proinflammatory factors that support cancer development. Apart from protumorigenic soluble factors, many tumor cells release endosome-derived vesicles, called exosomes, which mediate intercellular communication. In PNAS, Meckes et al. describe that EBVand KSHV-infected lymphoma cells secrete exosomes with distinct repertoires of cellular proteins (1). In light of recent studies demonstrating that tumor cellderived exosomes exert protumorigenic signaling properties (2), the data presented by Meckes et al. suggest that virus and tumor cells share common strategies in shaping a permissive microenvironment by modulating the cargo of secreted exosomes (Fig. 1). Because KSHV and EBV are lymphotropic gamma herpesviruses associated with B-cell lymphomas, similar mechanisms may underlie their oncogenicity. EBV infects naïve resting B cells in vitro, leading to immortalized proliferating lymphoblastiod cell lines (LCLs), but KSHV in isolation does not immortalize lymphocytes. Only B cells from EBV-infected hosts and subsequent KSHV infection can promote LCLs that harbor both viral genomes (3). Thus, the KSHV latency program is not immortalizing in vitro and presumably requires one or more cofactors to transform B cells. Studies in transgenic mouse models showed that the EBV-encoded latent-membrane protein 1 (LMP1) is critical for B-cell lymphoma development (4), but FLICE-inhibitory protein (vFLIP) acts as an important driver of KSHV-induced lymphomagenesis (5). Both proteins constitutively activate NF-κB (6), the function of which is frequently deregulated in human cancer-mediating prosurvival functions and tumor-associated inflammation (7). Fig. 1. Possible effects on the human B-cell lymphoid microenvironment by EBVand KSHV-modified exosomes. In this schematic representation of “normal” B-cell activation, the left image represents a human B-cell that is activated by CD40 ligand (CD40L) carrying T cells. This leads to NF-κB transcription (black arrow) in the nucleus of the B cells, which is also a signal for the production and release of IL-4 and exosomes from specialized endosomes known as MVBs (16). The released B cell exosomes may be internalized by other cell types, such as macrophages, or could “decorate” T cells by binding to the surface. In the B-/T-cell contact area (immunological synapse), responding T cells polarize and may proliferate, secreting exosomes that seem to be preferentially internalized by the activated B cells (17). The right schematic represents a coinfected B cell carrying latent EBV and KSHV (represented by the circular blue and green DNA episomes in the nucleus, respectively). EBV-encoded LMP1 and KSHV-encoded vFLIP activate NF-κB without the external involvement of T cells (6). LMP1 localizes to the limiting membrane of MVBs (represented in red), and presumably activates NF-κB from this site. As a consequence, LMP1 can be sorted into ILVs inside these compartments. Besides LMP1-associated vesicles, distinct types of vesicles carrying different proteins, lipids, or RNAs may also be produced (indicated by light and dark blue color variations). It is currently unknown whether the trafficking, localization, and intracellular sorting of vFLIP follows a similar path as LMP1.The infected cells with activated NF-κB produce IL-6, which serves as an autocrine growth factor promoting B-cell proliferation and could contribute to a sustained positive inflammatory feedback loop (18). The infected B cells release virus-modified exosomes, of which some carry LMP1 (indicated with a red membrane) when derived from LMP1 MVBs. These exosomes can activate signaling PI3K, ERK, and STAT signaling pathways in target cells, either upon internalization or alternatively by activating surface receptors on the target cell (12). Both EBV and KSHV impact the proteome of the produced exosomes that are distinct from exosomes produced by noninfected counterparts (represented as brown vesicles in the left) that may primarily function in immune cell–cell interactions. However, the EBV and KSHV vesicles (indicated by blue and green vesicles, respectively) are suspected to render the behavior of distinct target cells in the vicinity, including endothelial cells, macrophages, and T cells that, as a whole, provide a permissive environment for virus-infected tumor cells. Author contributions: D.M.P. wrote the paper.