---

Research in the Ganem laboratory explores the molecular mechanisms of the replication of human viral pathogens and the biology and pathogenesis of their resulting diseases. Current projects focus on (i) Kaposi's sarcoma (KS) and its causative herpesvirus, KSHV; and (ii) the search for novel viral agents in acute and chronic human diseases suspected to have an infectious etiology.

Kaposi's sarcoma and KSHV

Background. KS is the leading neoplasm of patients with AIDS. Although HIV infection is an important risk factor for KS development, strong evidence implicates a novel human herpesvirus (termed KS-associated herpesvirus (KSHV) or human herpesvirus 8 (HHV8)) as the central determinant of KS etiology. KSHV was initially identified by the presence of its genomic DNA in KS tumors. It is a novel member of the lymphotropic herpesviruses. Like all herpesviruses, it can engage in 2 alternate genetic programs: lytic infection, in which most viral genes are expressed in a temporally regulated fashion and progeny virions are produced; and latency, in which the viral genome persists as a nuclear episome, but viral gene expression is strongly restricted. Fig 1 shows a schematic illustrating the large number of genes encoded by the viral genome, and the dramatic restriction of gene expression in latency.

Fig 1. The genome of KSHV. Latency genes are shown in green. From Cotter & Robertson, Frontiers Biosci 7:358 (2002)

In 1996, our laboratory developed the first in vitro culture system for KSHV replication, using B cells from a KSHV-related lymphoma. We exploited this discovery to generate serologic tests for KSHV infection and in 1996-98 we and others used these in epidemiologic studies to show that KSHV is indeed the agent predicted by the epidemiology of KS. It is always present in patients with KS, highly prevalent (30-60%) in groups at high risk for KS and uncommon (2-5%) in the general population. Infection antedates tumorigenesis and predicts a higher risk of KS development. Using in situ hybridization, we showed that in advanced KS lesions, all KS tumor cells display latent infection, but KS tumors also display low levels of lytic KSHV replication.
In the aggregate, our early studies and those of others indicated that KSHV is necessary for KS development: KS is never observed in the absence of KSHV infection. However, KSHV infection is not sufficient for tumorigenesis: 2-5% of the general population is latently infected by KSHV, but do not develop KS tumors. Clearly, cofactors are required, the best studied of which is HIV-related immune deficiency.

Current research: Our current work focuses on the molecular and cellular mechanisms by which KSHV infection engenders KS risk.
This work examines the contributions of both the latent and lytic viral gene expression programs to viral persistence and disease pathogenesis.

Latency The latency proteins of KSHV are thought to contribute to KS via promotion of cell proliferation and extension of cell survival. Our work on KSHV latency focuses on several areas. First, we are exploring the effects of KSHV infection on primary endothelial cell cultures (see Fig 2). In this system, originally described by Gary Hayward and colleagues at Johns Hopkins, latent infection engenders dramatic morphologic changes that resemble those seen in KS tumor cells in vivo.

Fig 2. Infection of primary human endothelial cells by KSHV triggers dramatic morphologic changes. Left panel, mock infected cells; Right panel, KSHV infection.

To determine which latency gene is responsible for this phenotype, we constructed retroviral vectors for each latent ORF and used them to express the corresponding protein in primary human endothelial cells. This revealed that only one viral gene (ORF71), encoding the v-FLIP protein, was required to induce this morphologic change. V-FLIP is known to activate NFkB by binding to the NEMO/IKK complex, and blockade of the NFkB activatin pathway prevents v-FLIP induced spindling. We are currently trying to determine the mechanism(s) by which NFkB provoke spindling, whichinvolves a large rearrangement of the actin cytoskeleton. Because of v-FLIP’s important role in both cell morphology and cell survival, we are also closely studying the regulation of v-FLIP gene expression.

We are also studying the mechanisms by which latent episomes are maintained in infected cells. In 2004 we reported that in KSHV-infected cells, latent viral episomes are unstable, and are frequently lost upon cell division. However, in a small subpopulation of cells, episomes can be stabilized by a process that involves cis-acting, epigenetic changes in the viral genome. Using genetically marked viruses encoding drug-resistance genes, we have been able to select for cells that stably maintain the episome and are now conducting detailed anlyses of these and other lines to try to understand the basis of episome stabilization.
Our recognition of the unstable nature of latency explains why most cultured KS cell lines appear to lack viral DNA, despite its ubiquity in primary KS tumors. More importantly, it helps to explain prior clinical observations that drugs that inhibit lytic replication block KS tumorigenesis at all stages of the natural history of infection. Our work suggests that one reason lytic infection is continuously required is to recruit new cells to latency to replace those lost to episome segregation.

We also study the biology of a group of latent proteins, the kaposins, that we initially discovered some years ago. The most abundant of these, kaposin B, is encoded by a string of GC-rich repeats that generates a polypeptide composed of 23aa proline-rich repeats. In 2005 we reported that this protein interacts with a cellular kinase known as MAPKAPK2 (or MK2). MK2 is a downstream effector of the p38 signaling pathway, and is responsible for upregulating the stability of AU-rich element (ARE-)containing mRNAs. Such mRNAs include a variety of cytokine and growth factor transcripts. Our work indicates that kaposin B activates MK2, resulting in a dramatic extension of cytokine mRNA stability and augmentation of cytokine production by kaposinB-expressing cells. This is an important result, since KS cells are known to both produce and respond to cytokine signals, and this has long been considered to reflect an important part of KS biology.

In addition, kaposin expression upregulates p38 activity itself, most likely by an amplification mechanism summarized in Fig 3.

Fig. 3. Potential amplification loops in Kaposin B action. Activation of MK2 by kaposin B binding triggers upregulation of target ARE-mRNAs. These include (i) MKK6 mRNA; and (ii) cyokine mRNAs (e.g. IL6). Upregulation of MKK6 can directly induce p38 activation; released IL6 can also upregulate the p38 pathway in an autocrine or paracrine fashion.

However, we have good reason to believe that p38/MK2 activation is not the sole function of kaposin B. We have recently shown that the protein interacts with several other signaling molecules known to be downstream of B and T cell antigen receptors, whose signaling can be modulated by kaposin B expression. In addition, we continue to hunt for new interaction partners of Kaopsin B, using genetic and biochemical strategies.

Finally, the viral latency program includes the production of 17 microRNAs (miRNAs), processed from 12 pre-miRNA hairpins located in the kaposin transcript (10 in an intron, 2 in the body of the mRNA; Fig 4).

Fig 4. KSHV latent miRNAs. The 12 pre-miRNAs of KSHV are depicted as arrowheads. The region labeled ORFK12-DR3-DR4 represents an alternative nomenclature for the Kaposin locus discussed above.

A major effort in the lab focuses on the function of these miRNAs. MicroRNAs funtion by imperfect hybridization of their so-called seed regions to complementary sequences in target mRNAs (of host or virus); this annealing results in impaired translation of the targeted mRNA, by still-uncertain mechanisms. One correlate of miRNA targeting is a modest reduction in the accumulation of many targeted mRNAs. We have used these small abundance changes as a screen for putative targets of the viral miRNAs, using comprehensive transcript profiling coupled with bioinformatic analysis of resulting candidates. This approach has recently borne fruit: one host gene that regulates apoptosis has been shown to be targeted by 4 different KSHV miRNAs! We are now systematically working through the potential targets of all 17 viral miRNAs.Lytic infection. Our work on the lytic cycle centers on three areas:

Lytic infection. Our work on the lytic cycle centers on three areas:
(i) The mechanisms by which viral replication is reactivated from latency. We and others have shown that a single viral gene product, called RTA, controls the switch from latency to lytic reactivation. RTA is a viral immediate-early transcription factor. Current work centers on analysis of the DNA binding properties of RTA and the host proteins with which it interacts to activate transcription. We discovered that in addition to direct, sequence-specific DNA binding, RTA can also interact with host transcription factors, notably RBP-Jk (CSL). In uninfected cells, RBP-Jk is a repressor that is the target of Notch signaling. Association with RTA converts RBP-Jk into an activator with numerous targets in the viral genome. By examining infection in cells in which RBP-Jk has been inactivated, we found that recognition of these targets is essential for the progression of lytic infection. Ongoing work is designed to define the specificity of direct RTA-DNA binding and to further characterize RTA-associated complexes.

We are also studying the roles of transcripts antisense to RTA that are abundantly transcribed during lytic growth. These RNAs have been annotated as noncoding, and were presumed to regulate RTA synthesis. However, we find that they do not affect RTA proiduction, and appear to be localized to polysomes, where they direct the production of tiny peptides. The function of the latter is now under study.

(ii) The effects of KSHV lytic replication on host gene expression. These studies involve microarray analysis of host gene expression and biochemical studies of the viral genes that influence the program of host gene expression. Recently, these studies have led to the recognition that lytic KSHV replication engenders a profound block to host gene expression. This block is at the level of mRNA accumulation, and is mediated by a single viral protein, called SOX (shut-off exonuclease), encoded by open reading frame 37. Interestingly, homologs of SOX are found in all other herpesviruses, but play no role in host shutoff in those viruses. Evidently, viral evolution has endowed KSHV SOX with a novel actvity in mRNA metabolism, and we are now intensively investigating how this activity functions.

(iii) The mechanisms by which KSHV escapes from immune surveillance. Genetic and biochemical studies in our lab have defined two viral genes that are involved in the evasion of host T cell-mediated immunity, via downregulation of surface MHC-I, B7. 2 and ICAM-1 molecules (Fig 5).



Fig 5. Downregulation of MHC-I in cells expressing MIR1 (blue) and MIR2 (red). In black are the levels of MHC-I in cells expressing KSHV genes that do not regulate this protein.

These proteins act to enhance endocytosis of their targets from the plasma membrane, in a ubiquitin-dependent fashion. We recently showed that they define a novel family of membrane-bound E3 ubiquitin ligases, and are now pursuing their biochemical function in detail by characterizing complexes in which they are found, as well as by careful mutagenesis of their functional domains. More recently, we discovered that these proteins can also modulate a protein, CD1d, that is key to recognition by NKT cells. These cells are specialized to recognize host-derived lipids, and are thought to play roles in autoimmunity and perhaps in immune regulation. Little is known of their roles in host defense against viral infection, but our finding that KSHV infection modulates this system suggests that CD1d-restricted T cells do indeed play hitherto unsuspected roles in antiviral immunity.
Currently, we are focusing on identifying viral strategies that modulate the activity of the innate immune system. This has led us to discover at least 2 proteins that impair signal transduction by type I interferons (IFNs) – key effectors of antiviral immunity. One of these, RIF (the product of ORF 10) appears to act very early in the signaling cascade, by forming inhibitory complexes on the cytosolic tail of the IFN receptor, blocking Jak and STAT activation.