In Vivo Models of Oncoproteins Encoded by Kaposi’s Sarcoma-Associated Herpesvirus

Kaposi’s sarcoma-associated herpesvirus (KSHV) is a human oncogenic virus. KSHV utilizes its proteins to modify the cellular environment to promote viral replication and persistence.

ABSTRACT

Kaposi’s sarcoma-associated herpesvirus (KSHV) is a human oncogenic virus. KSHV utilizes its proteins to modify the cellular environment to promote viral replication and persistence. Some of these proteins are oncogenic, modulating cell proliferation, apoptosis, angiogenesis, genome stability, and immune responses, among other cancer hallmarks. These changes can lead to the development of KSHV-associated malignancies. In this Gem, we focus on animal models of oncogenic KSHV proteins that were developed to enable better understanding of KSHV tumorigenesis.

INTRODUCTION

Viruses are linked to approximately 12% of all human cancers (1). Kaposi’s sarcoma-associated herpesvirus (KSHV), also known as human herpesvirus 8 (HHV-8), is a member of the Gammaherpesvirinae subfamily and is linked to three human cancers: the endothelial cancer Kaposi’s sarcoma (KS) and two B-cell lymphoproliferative diseases, primary effusion lymphoma (PEL), a type of non-Hodgkin lymphoma (NHL), and multicentric Castleman’s disease (MCD) (2–4). These malignancies develop primarily in immunocompromised individuals, although they can occur in HIV-naive individuals as well (5). KSHV is classified as a category 1 carcinogen by IARC/WHO (6).

Like other members of the Herpesviridae family, KSHV establishes a lifelong infection in the host, alternating between two distinct life cycles: latent and lytic. During latency, a limited number of genes are expressed, and the viral genome is maintained as an episome and is passed to daughter cells during cell division. During lytic infection, most of the viral genes are expressed, the viral genome is replicated, and new virions are produced.

The mechanisms by which KSHV infection leads to the development of cancer are not completely understood. However, it is believed that KSHV proteins play a critical role in tumorigenesis. KSHV expresses a myriad of lytic and latent proteins that alter the cellular environment to promote the survival of the infected cell and the establishment of latency; such alterations, including the induction of cell proliferation, evasion of apoptosis, and immune evasion, coincide with hallmarks of cancer (1).

Several transgenic mouse models have been developed to study the contributions of KSHV proteins to oncogenesis in vivo. These genetically modified mice express one or a few KSHV proteins shown to have oncogenic properties in cultured cells. The transgenic KSHV mouse models developed to date are discussed in this Gem (Table 1). Expression of KSHV proteins in mice often results in the development of a variety of diseases similar to those linked to KSHV, confirming the contributions of these proteins to KSHV-induced tumorigenesis. These models represent a suitable alternative for the study of KSHV pathogenesis, not only to provide insights into the mechanisms of KSHV-associated tumorigenesis but also to evaluate drugs to treat KSHV-associated infections by targeting these viral proteins.

TABLE 1.

Transgenic mouse models of KSHV oncoproteins

KSHV protein	Genetic backgrounda	Promoter	Disease(s) developedb	Age of mice (mo)c	Tumor incidence (%)	Tumor cell type	Reference
K1	C57BL/6J-CBA/J	SV40	Plasmablastic lymphoma	14	7.6	B	79
			Spindle-cell sarcomatoid tumor	14	7.6	Spindle
K1	C57BL/6J-CBA/J	SV40	Lymphoma	18	25	B	80
			Abdominal/hepatic tumors	18	50	Spindle-like
LANA	C57BL/6	LANA	MCD, PBL	NS	11	B	23
LANA	CD19–/– (C57BL/6)	LANA		NT			25
Latency locus	C57BL/6	Latency locus	Lymphoma	13	16	B	51
Latency locus	iMYCC/α	Latency locus	Frank lymphomas	4	27.5	B	54
			Lymphoid hyperplasia	4	50	B
Latency locus	FcγRIIBKO BALB/c		Plasmacytosis; lymphoid hyperplasia	NS	62	B	55
vCyclin	p53+/– C57BL/6	Eμ	NS	6–7	6.5–50	NS	36, 37
vCyclin	p53–/– C57BL/6	Eμ	NS	2.5–3	100	B or T	36, 37
vFLIP	ICR	H-2Kb	Lymphocytic, follicular, immunoblastic, and anablastic lymphoma	20	11.8	B	49
vFLIP	129/Sv-C57BL/6 or 129/Sv-C57BL/6-BALB/c	CD19.creKI or Cγ1.creKI	Histiocytic/DC sarcoma	7–19	High incidence (NS)	B	50
vGPCR	NS	CD2	KS-like lesions	1–1.5	100	Endothelial	112
vGPCR	TVAKI FVB/N	TIE2	KS-like lesions	>1	100	Spindle	113
vGPCR	C57BL/6	SV40	KS-like lesions	5–12	25	Spindle/no spindle	114
vGPCR	C57BL/6 × DBA/2	TRE-LacZ	KS-like lesions	1–2 (upon DOX induction)	100	Spindle	115
vIL-6	C57BL/6	MHC I	MCD-like	3	28	B	61
vIL-6	BALB/c	MHC I	MCD-like	8	70	B	62
vIL-6	C.iMycΔEμ	MHC I	PEL, PBL, PBM	5	100	B	62
vPK	C57BL/6	Ubiquitin	PEL-like lymphoma and hyperplasia	16	66	B	91

^aNS, not specified; TVA, avian leukosis virus receptor.

^bPBL, plasmablastic lymphoma; PBM, plasmablastic plasma cell myeloma; DC, dendritic cell.

^cNT, no tumors found; DOX, doxycycline.

MOUSE MODELS OF KSHV-ENCODED LATENT ONCOPROTEINS

The default phase of the KSHV life cycle is latency. Most KSHV-infected cells remain dormant throughout the lifetime of the host. During latent infection, few viral proteins are expressed; this ensures viral persistence by deregulating host oncogenic pathways, promoting cell survival, proliferation, and immune evasion. Such prolonged modulation of cellular homeostasis can lead to cell transformation and immortalization and to the subsequent development of KSHV-associated malignancies. In this section, we discuss the KSHV latent proteins that have been shown to have oncogenic properties in mouse models and the cellular pathways they modulate that potentially explain their roles in tumorigenesis (Fig. 1).

FIG 1.

KSHV-encoded oncoproteins and their modulation of cancer-inducing pathways leading to tumorigenesis (see reference 1) in transgenic mouse models. Created with BioRender.

LANA.

The latency-associated nuclear antigen (LANA) is known for its critical roles in latency maintenance, genomic viral DNA replication, and latent gene transcription (7–9). However, in this section, we focus on the oncogenic properties of LANA.

In cell culture models, LANA is implicated in KSHV-induced tumorigenesis by altering cellular pathways involved in cell cycle progression, proliferation, and apoptosis. LANA modulates retinoblastoma protein (Rb), p53 and its homolog p73, NOTCH, WNT, and mitogen-activated protein kinase (MAPK) signaling (10–14). In addition, LANA alters cell death, growth, and proliferation by modulating survivin and β-catenin as well as c-Myc and c-Jun signaling (15–19). Moreover, LANA modulates angiogenesis and induces chromosomal instability (20, 21). Given these oncogenic properties, it is not surprising that LANA transforms primary rat embryo fibroblasts (10), prolongs the life of primary human endothelial vein cells (22), and, as discussed below, induces lymphomagenesis in mice.

The LANA transgenic mouse model expresses LANA from its natural promoter (23), which has been shown to be B-cell specific (24). Eleven percent of LANA transgenic mice develop monoclonal tumors, which are of B-cell lineage and express LANA (23). In addition, mature B cells are activated and hyperproliferative. This expansion resembles KSHV-associated lymphomas, specifically MCD and plasmablastic lymphomas.

A follow-up study further elucidated that LANA lowers the activation threshold of B cells, leading to their continual activation without exhaustion (25). LANA transgenic mice have a significant increase in mature CD19⁺ IgM⁺ IgD⁺ B cells. In addition, early germinal center (GC) and activated follicular cells are increased. Crossing LANA mice with CD19^−/− mice shows that while LANA is not sufficient to compensate for the lack of CD19 (which is needed for GC development), it can partially restore marginal-zone B-cell development (25). Taken together, these data suggest that LANA lowers the B-cell receptor (BCR) signaling threshold in KSHV-infected cells.

vCyclin.

The KSHV gene ORF72 (open reading frame 72) encodes a homolog of cellular cyclin D known as viral cyclin (vCyclin) (26, 27), which is expressed during both the latent and lytic replication stages (28). vCyclin interacts with several cyclin-dependent kinases (CDKs), though preferentially with CDK6. Such vCyclin-CDK complexes are resistant to known CDK inhibitors and have an increased number of substrates (29–32), making vCyclin a strong deregulator of cell cycle progression. In addition, vCyclin has been shown to modulate p53 and Bcl-2 (33, 34) and to induce the DNA damage response (DDR) (35).

To evaluate the oncogenic properties of vCyclin in vivo, a transgenic mouse model was generated in which vCyclin is expressed under the transcriptional control of the Eμ promoter, which allows for the expression of vCyclin in B and T cells (36). Approximately 17% of 6- to 9-month-old vCyclin transgenic mice develop B- or T-cell lymphomas (37).

Several lines of evidence suggest that p53 contributes to the suppression of vCyclin-induced lymphomagenesis: (i) expression of vCyclin in cell culture leads to p53-dependent cell growth arrest (36); (ii) vCyclin tumors exhibit elevated levels of p53 and p19^ARF, an inhibitor of p53 degradation (37); and (iii) crossing vCyclin transgenic mice with p53^−/− mice results in much more rapid and frequent development of T- and B-cell lymphomas (36, 37). The B-cell lymphomas are IgM^–, and the T-cell lymphomas are P/TCR^low, suggesting that vCyclin induces precursor B- and/or T-cell lymphoma in p53^−/− mice. vCyclin/p53^−/− mice show genome instability, demonstrated by a high frequency of aneuploidy.

To better study the contributions of vCyclin to KS development, a mouse model has been established which expresses vCyclin under the control of the vascular endothelial growth factor receptor 3 (VEGFR-3) promoter (38), resulting in vCyclin expression in lymphatic endothelial cells. In this model, 80% of vCyclin mice die prematurely by the age of 6 months and show pleural fluid accumulation, lymphatic dysfunction, and erythematous skin.

vFLIP.

KSHV expresses the latent viral FADD-like interleukin-1-converting enzyme (FLICE) inhibitory protein (vFLIP), also known as K13, in both KS and PEL (39, 40). Like other cellular and viral FLIPs, vFLIP inhibits Fas-induced apoptosis by preventing the recruitment and processing of caspase 8 (41–44). In addition, vFLIP regulates cell death by inhibiting autophagy (45). Uniquely among FLIPs, vFLIP activates NF-κB (46, 47). Given vFLIP’s inhibition of apoptosis, activation of NF-κB, and subsequent transformation of Rat-1 and BALB/c 3T3 cells (48), a role for vFLIP in KSHV-induced lymphomagenesis has been proposed. Below, several in vivo experiments supporting this hypothesis are discussed.

A transgenic mouse model was generated in which the vFLIP transgene is expressed under the control of the H-2K^b promoter and the Ig heavy chain enhancer (49). Surprisingly, vFLIP does not inhibit intrinsic or extrinsic apoptosis in transgenic mice. However, it induces constitutive activation of NF-κB. vFLIP transgene expression leads to an increase in cell proliferation and a higher incidence of lymphoma (11.8%) than that in age-matched wild-type mice (1.8%). vFLIP-induced lymphomas are B220⁺, express vFLIP, and show constitutive NF-κB activation. Given the low lymphoma incidence and the inability to recapitulate PEL in these mice, two transgenic mouse lines expressing vFLIP at different stages of B-cell development were generated (50). One transgenic mouse line expresses vFLIP under the control of the CD19 promoter, resulting in vFLIP expression in all B cells. A second line expresses vFLIP under the control of the Cγ1 promoter, restricting expression to IgG1 GC B cells. While neither line recapitulates PEL, vFLIP expression results in B-cell abnormalities similar to those observed in MCD, including lack of GC formation, increased frequency of marginal-zone and FAS⁺ B cells, and failure to produce class-switched IgG.

Latency locus.

The KSHV latency locus (LL) includes LANA, vCyclin, vFLIP, K12, and all viral microRNAs (miRNAs). The genes in the LL are constitutively expressed from a single B-cell-specific promoter in all KSHV-infected cells. B cells from LL mice, which express the LL from its natural promoter, exhibit chronically activated mature B cells, leading to hyperglobulinemia due to increases in the frequencies of CD138⁺ plasma cells and marginal-zone B cells (51). In addition, the augmentation of marginal-zone and GC B-cell responses suggests that the LL drives both T-independent and T-dependent B-cell activation, respectively. The B-cell activation abilities of the LL were further demonstrated by showing that the LL can drive B-cell development even in the absence of miRNA-155, which is normally required for B- and T-cell function and is expressed in many cancers (52). In addition, the LL can compensate for the lack of interleukin 6 (IL-6) (53). Not only is IL-6 critical for B-cell function, but its signaling is deregulated in KSHV-associated malignancies. Collectively, KSHV latent genes might compensate for a lack of IL-6 in early B-cell development.

Given that LL mice do not display lymphomas until they age (51), it was suspected that additional cellular effects were needed to accelerate lymphomagenesis. One candidate was MYC deregulation, which occurs frequently in lymphomas. To test this hypothesis, LL transgenic mice were crossed to iMYCCα transgenic mice (54), in which the MYC coding region was inserted into the IgG Cα locus under the control of its natural promoter and the Eα enhancer. The iMYCCα/LL mice develop lymphomas that resemble PEL at a low rate and, like LL mice, after a long latency. This allowed study of the synergism between MYC and the KSHV LL. iMYCCα/LL mice display higher frequencies of plasmablasts and plasma cells, and more proliferation and GC formation, than iMYCCα mice or LL mice. Moreover, double transgenic mice develop lymphomas at a higher rate. Together, these findings suggest that MYC and the KSHV LL cooperate to promote KSHV-induced lymphomagenesis.

An additional LL transgenic mouse model is available in which the LL is expressed (i) in a BALB/c background and (ii) in the absence of the FCγRIIB inhibitory receptor (55). BALB/LL mice develop hyperglobulinemia, plasmacytosis, and B-cell hyperplasia. In addition, in accord with the lower B-cell activation threshold induced by the LL, these mice exhibit an augmented antibody response against a secondary viral infection.

MOUSE MODELS OF KSHV-ENCODED LYTIC ONCOPROTEINS

Under certain still-elusive physiological conditions, KSHV undergoes lytic reactivation. Initially, it was thought that only KSHV latent proteins contributed to tumorigenesis, while lytic proteins contributed to the production of infectious virions. However, it was found that latently KSHV infected cells sometimes express low levels of lytic genes in the absence of full-blown lytic replication. Additionally, abortive replication can lead to the expression of lytic genes. Furthermore, lytic genes may be expressed alone in the context of certain environmental stimuli, e.g., hypoxia. Interestingly, many KSHV lytic proteins display oncogenic properties, such as induction of cell proliferation, inhibition of apoptosis, and evasion of immune responses, suggesting a role in tumorigenesis. This gave rise to the concept that lytic proteins might contribute to tumor progression in a paracrine fashion. Several transgenic mouse models, each expressing one lytic gene, confirm the oncogenic potential of such proteins and are discussed below (Fig. 1).

vIL-6.

KSHV encodes a structural homolog of mouse IL-6 (mIL-6) and human IL-6 (hIL-6). It is expressed from ORF K2 during lytic replication and at low levels during latency. Viral IL-6 (vIL-6) supports the proliferation of a mouse B-cell hybridoma and a human myeloma cell line, and it is expressed in all three KSHV-associated cancers (2, 56–58). vIL-6 activates the JAK/STAT signaling pathway similarly to hIL-6. However, it requires only one receptor subunit, gp130, rather than both the gp130 and IL-6Rα subunits (59). Stable expression of vIL-6 in NIH 3T3 cells induces hematopoiesis, plasmacytosis, and angiogenesis when the cells are injected into athymic mice (60).

vIL-6 transgenic mice were generated by expressing a codon-optimized vIL-6 gene under the control of the major histocompatibility complex (MHC) class I H-2 promoter in the B6D2 strain (61). vIL-6 is detected in mouse serum, with intolerably high concentrations (>130 ng/ml) resulting in death/euthanasia. B6D2/vIL-6 mice were crossed to C57BL/6 mice, giving rise to vIL-6 mice, which express vIL-6 in serum and in various tissues. At the age of 15 weeks, vIL-6 transgenic mice develop enlarged spleens and lymph nodes. While no malignant tumors are observed, vIL-6 mice show a massive accumulation of plasma cells, as well as enlarged B-cell follicles with hyperplastic GCs and vascular proliferation, which are hallmarks of plasma cell-type MCD. The spleens of vIL-6 mice show increased extramedullary hematopoiesis and plasma cell hyperplasia, which are also characteristic of MCD. GCs in lymph nodes show high expression of the proliferation marker Ki67, and phosphorylation of STAT3 is observed in the lymph nodes and spleens of vIL-6 mice. Taking these findings together, expression of vIL-6 in vivo leads to the induction of symptoms that resemble human MCD. Moreover, vIL-6 induces high expression levels of mIL-6, which has been shown to be essential for MCD-like symptom development in mice.

Since vIL-6 on the C57BL/6 genetic background acts as a weak oncogene, this transgenic model was refined by crossing vIL-6 mice with BALB/c mice (vIL-6/BALBc mice) (62), which are highly susceptible to neoplasms of plasmablasts and plasma cells (63). These mice were further intercrossed with C.iMYCΔEμ mice, which have deregulated expression of the oncoprotein MYC. vIL-6/BALBc mice show much faster proliferation of MCD-like plasma cells, with a higher tumor incidence (70%), than vIL-6 mice, while vIL-6/BALBc mice with deregulated MYC expression show much faster tumor development and a 100% tumor incidence. Tumors consist mostly of plasmablastic neoplasms similar to PEL, plasmablastic lymphoma (PBL), and immunoglobulin-producing, extramedullary plasmablastic plasma cell myeloma (PBM).

K1.

The K1 protein is a viral immunoreceptor encoded by ORF-K1 (64, 65). It is expressed at high levels during the lytic stage of viral replication and at low levels during latency (66, 67).

The K1 protein has structural and functional similarity to the BCR and can activate B cells (64, 68). When K1 is stimulated, it recruits and activates signaling molecules such as the tyrosine kinase Lyn, the spleen tyrosine kinase (SYK), and the p85α subunit of phosphatidylinositol 3-kinase (PI3K) (69). K1 also induces the downregulation of BCR surface expression (70), a phenomenon observed in several cancers (71). Thus, K1, and not the BCR, controls the signaling environment of infected B cells, providing a survival advantage to virus-infected cells. K1 also promotes cell survival by activating the PI3K/Akt/mammalian target of rapamycin (mTOR) pathway (72, 73) and by protecting cells from Fas-mediated apoptosis (74, 75) in a heat shock protein 90 (HSP90)- and HSP40-dependent fashion (76). In addition, K1 induces cell migration and angiogenesis by inducing the upregulation of cytokines and angiogenic factors (68, 69, 77). Given the ability of K1 to induce multiple signaling pathways, it is not surprising that K1 plays a role in KSHV tumorigenesis. In cell culture, K1 immortalizes primary endothelial cells and transforms rodent fibroblasts (65, 66). In addition, marmosets develop lymphomas when inoculated with a recombinant herpesvirus saimiri (HVS) expressing the K1 protein instead of the saimiri transforming protein (STP) (65), which is essential for HVS oncogenicity (78). While K1 shares no amino acid sequence similarity with STP, the latter is located at an equivalent position in the HVS genome.

To further investigate the oncogenic properties of K1 in vivo, a transgenic mouse model was generated in which the K1 ORF is ubiquitously expressed under the transcriptional control of the simian virus 40 (SV40) promoter (79). B and T lymphocytes isolated from 6- to 8-month-old K1 transgenic mice show higher mRNA expression levels of NF-κB and basic fibroblast growth factor (bFGF) than wild-type mice. In contrast, K1 mice have decreased IL-12 transcripts in their B cells and lower levels of IL-12 in serum. In addition, K1 induces activation not only of NF-κB, but also of the B-cell transcription factor Oct-2. A majority of K1 mice develop hyperplasia and splenomegaly at the ages of 8 and 10 months, respectively (80). In addition, splenocytes from K1 mice display higher levels of IL-6 and CXCR5 mRNAs. At the age of 18 months, 87.5% of K1 mice show signs of lymphoid hyperplasia, with 25% having confirmed lymphoma and 50% showing abdominal and/or hepatic tumors. In accordance with the inhibition of Fas-mediated apoptosis by K1 in cell culture models, splenocytes from 6-month-old transgenic mice are significantly less responsive to Fas activation than those from wild-type mice. K1-induced tumor growth is also dependent on VEGF production and NF-κB activation, since both are necessary for the expansion of K1 lymphoma cells injected into nude mice (81).

ORF36/vPK.

Viral protein kinase (vPK) is a serine-threonine kinase encoded by KSHV ORF36. (82). The subcellular localization of vPK is predominantly nuclear, and it is expressed as an early-late gene during the lytic stage of KSHV infection. However, its mRNA has been found in 59% of KS biopsy specimens (83). vPK expression can be induced by hypoxia, since its promoter contains hypoxia-inducible elements (84).

vPK alters many cellular signaling pathways by phosphorylating a myriad of cellular substrates. Mitogen-activated kinases 4 and 7 (MKK4 and MKK7) are phosphorylated by vPK, resulting in the phosphorylation of c-Jun N-terminal kinase (JNK) and the subsequent phosphorylation of c-Jun transcription factor (85). One study suggests that vPK does not induce the DDR, given the absence of phosphorylation of the histone protein H2AX in vPK-expressing cells (86). However, vPK coimmunoprecipitates with and phosphorylates the histone acetyltransferase TIP60 (87), which is thought to be a tumor suppressor that participates in the oncogene-induced DDR (88).

vPK has been shown to have functional similarities to CDKs. For example, vPK phosphorylates the CDK substrates retinoblastoma and lamin A/C (89). In addition to functional similarity to CDKs, vPK mimics ribosomal protein S6 kinase (S6KB1) structurally and functionally, sharing several phosphorylation substrates, including ribosomal S6 protein (90). In the same study, vPK was shown to phosphorylate eukaryotic initiation factor 4E (eIF4E). vPK-induced S6 phosphorylation enhances several hallmarks of cancer, including cellular proliferation, protein synthesis, anchorage independence, and angiogenesis.

Two transgenic mouse lines were generated in which the vPK transgene is expressed under the control of a ubiquitin promoter (91). vPK expression alone increases the incidence of B-cell hyperproliferative disorder and/or a high-grade B-cell non-Hodgkin lymphoma (NHL) 8-fold over that for age-matched wild-type mice. Although vPK is expressed ubiquitously, the mice develop only B-cell tumors. vPK mice show increased B-cell activation in the absence of antigen stimulation and increased levels of IL-1β and IL-12 p40. Lymphomas from vPK mice consist predominantly of B220⁺ GL-7⁺ IgM^– IgD^– B cells, suggesting a GC origin. In addition, lymphomas express vPK, display monoclonal or polyclonal IgH rearrangements, and show robust expression of phosphorylated S6 and eIF4E, suggesting that vPK phosphorylation of these cellular substrates is important for tumorigenesis.

vGPCR.

KSHV encodes a constitutively active G protein-coupled receptor (vGPCR) (26) functionally homologous to the human interleukin 8 receptor (IL-8R). It binds to the CXC chemokine IL-8 and signals through phosphoinositide phospholipase C (92). In contrast to IL-8R, vGPCR binds both CXC and CC chemokines (92). vGPCR can transform endothelial cells (93, 94) due to its deregulation of cellular signaling via a variety of ligand-independent paracrine and autocrine mechanisms (95). vGPCR directly modulates several intracellular signaling pathways, such as the JNK, extracellular signal-regulated kinase (ERK), p38, PI3K/AKT/mTOR, and WNT/β-catenin pathways (93, 94, 96–101). Subsequently, vGPCR activates key transcription factors, such as NF-κB, nuclear factor of activated T cells (NFAT), AP-1, and CREB (96, 102–106). Such modulation results in increased cell survival and proliferation and in the expression of angiogenic and inflammatory molecules such as CCL2 and VEGF. These cytokines and growth factors can activate receptors in neighboring infected cells, amplifying vGPCR signaling (92, 93, 107–111).

To date, four vGPCR transgenic mouse models have been developed. The first model expresses the vGPCR gene under the control of the T-cell-specific CD2 promoter, resulting in the expression of vGPCR predominantly in T and NK cells (112). By 30 to 90 days after birth, all vGPCR mice develop KS-like lesions with strong expression of the endothelial cell marker CD34, suggesting that vGPCR acts in a paracrine fashion. The second mouse model employs mice expressing the avian retroviral receptor under the control of the endothelial-cell-specific TIE2 promoter (113). This allows for the infection of mouse endothelial cells with avian leukosis virus (ALV) expressing a KSHV gene. vGPCR-expressing mice show a decreased median survival from that of uninfected mice. Mice infected with 10⁷ copies of vGPCR-expressing ALV die at an incidence of 100% within 6 weeks after infection, and numerous microscopic tumors develop. Mice infected with 10⁵ copies of vGPCR-expressing ALV develop angiogenic tumors by the age of 4 months. vGPCR-induced pathogenesis is abolished when an inactive vGPCR mutant is transduced. The third mouse model expresses vGPCR ubiquitously under the control of the SV40 promoter (114). These mice develop internal and/or external tumors at an incidence of 30% to 40%. The tumors are highly angiogenic, resembling KS given the presence of spindle cells, CD31⁺ cells, and expression of VEGF-C. In the fourth vGPCR model, vGPCR is conditionally expressed to bypass the potential adverse effects of vGPCR expression prior to birth (108, 115). Induction of vGPCR with doxycycline results in highly angiogenic lesions with spindle cells of endothelial origin, similar to KS. Angiogenic factors are significantly upregulated in vGPCR⁺ cells. In addition, proangiogenic chemokine receptors and chemokines are upregulated in doxycycline-induced mice only. The development of this angioproliferative disease is independent of T and B cells, and the continued expression of vGPCR is required.

CONCLUSION

KSHV encodes multiple proteins that potentially contribute to KSHV-related human cancers. The modulation of cellular signaling pathways by KSHV oncogenic proteins allows infected cells to bypass growth suppressors and apoptotic signals, often leading to cell proliferation and immortality. In addition, KSHV oncogenic proteins lead to other hallmarks of cancer, such as angiogenesis, immune evasion, and genome instability. One possible reason why KSHV encodes several oncoproteins is so that the virus can keep the infected cell alive and thereby itself survive and be transmitted to other hosts. However, the pathways that are required to keep the infected cell alive are the very same pathways that go awry in cancer. Hence, although the goal of the virus is to survive in its host cell, the unintended consequence is the development of cancer in the host.

While many of the oncogenic properties of these viral proteins have been discovered and further studied in a variety of cell culture models, the oncogenic potential of just a few have been confirmed in vivo by means of the transgenic mouse models discussed here. In such models, we can see how the expression of a multifunctional KSHV protein initiates the multistep process of cancer development, with mice showing symptoms that mimic KSHV-associated malignancies. Although these transgenic mouse models are an excellent tool for studying the mechanisms by which KSHV induces oncogenesis and the functional role of each protein, no model yet entirely recapitulates KSHV-associated cancers in humans. One reason is that each of these proteins is expressed individually or in a small group, as in the LL mice, and as outlined by this review, each protein contributes to KSHV tumorigenesis in a unique fashion. In addition, given that KSHV is a strictly human pathogen, some human factors present or absent in mice can affect disease development. Despite such caveats, these transgenic mouse models can be used not only to study how KSHV leads to tumorigenesis but also to evaluate drugs that target KSHV proteins with the aim of treating KSHV-related cancers.