Oncogenic Transformation by the Jaagsiekte Sheep Retrovirus Envelope Protein

Home , HYAL2, Rous sarcoma virus

Oncogene (2007) 26, 789–801 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc REVIEW Oncogenic transformation by the jaagsiekte sheep retrovirus envelope protein

S-L Liu1 and AD Miller2

1Department of Microbiology and Immunology, McGill University, Montreal, Canada and 2Division of Human Biology, Fred Hutchinson Cancer Research Center, Seattle, WA, USA

Retroviruses have played profound roles in our under- sarcoma virus, both of which cause tumors in chickens standing of the genetic and molecular basis of cancer. (for a comprehensive review see Rosenberg and Jaagsiekte sheep retrovirus (JSRV) is a simple retrovirus Jolicoeur (1997)). Oncogenic retroviruses are historically that causes contagious lung tumors in sheep, known as classified into acute transforming retroviruses and ovine pulmonary adenocarcinoma (OPA). Intriguingly, nonacute retroviruses. Acute transforming retroviruses OPA resembles pulmonary adenocarcinoma in humans, induce tumors through acquisition and expression of and may provide a model for this frequent human cancer. cellular proto-oncogenes – a process referred to as Distinct from the classical mechanisms of retroviral oncogene capture. These retroviruses induce tumors oncogenesis by insertional activation of or virus capture rapidly, the tumors are of polyclonal origin, and the of host oncogenes, the native envelope (Env) structural viral oncogenes in these viruses are not essential for protein of JSRV is itself the active oncogene. A major virus replication. In fact, acute transforming retro- pathway for Env transformation involves interaction of viruses are often replication-defective (Rous sarcoma the Env cytoplasmic tail with as yet unidentified cellular virus is an exception) because of the loss or disruption of adaptor(s), leading to the activation of PI3K/Akt and essential viral genes during the oncogene capture MAPK signaling cascades. Another potential mechanism process. Nonacute retroviruses, on the other hand, do involves the cell-entry receptor for JSRV, Hyaluronidase not carry oncogenes and induce tumors slowly. This 2 (Hyal2), and the RON receptor tyrosine kinase, but the type of retrovirus causes tumors by activating cellular exact roles of these proteins in JSRV Env transformation proto-oncogenes close to the proviral DNA integration remain to be better understood. Recently, a mouse model sites – a process termed insertional activation. Nume- of lung cancer induced by JSRV Env has been developed, rous replication-competent retroviruses cause clonal and the tumors in mice resemble those seen in sheep tumors by proviral insertional activation. In addition infected with JSRV and in humans. In this review, we to oncogene capture and insertional activation, a few summarize recent progress in our understanding the replication-competent retroviruses induce tumors by molecular mechanisms of oncogenic transformation by expression of viral proteins. For example, human T-cells JSRV Env protein, and discuss the relevance to human leukemia virus types 1 and 2 (HTLV-1 and HTLV-2, lung cancer. respectively) induce adult T-cell immortalization and Oncogene (2007) 26, 789–801. doi:10.1038/sj.onc.1209850; leukemia in humans by expression of viral Tax proteins published online 14 August 2006 that transactivate several cytokines and cytokine receptors. It is worth noting that HTLV-1 and HTLV-2 are Keywords: Jaagsiekte sheep retrovirus; envelope pro- the only retroviruses identified so far, that directly cause tein; transformation; lung cancer; signaling pathways; human cancer. mouse model Jaagsiekte sheep retrovirus (JSRV) is simple retrovirus encoding Gag, Pro, Pol, and Env proteins (York et al., 1992; Palmarini et al., 1999; DeMartini et al., 2001), and is now classified as an ovine betaretrovirus. Sheep retroviral oncogenesis: a historical perspective This family also includes a closely related retrovirus of sheep and goats, enzootic nasal tumor virus (ENTV) For almost a hundred years, retroviruses have been (De las Heras et al., 2003). JSRV transforms peripheral known to induce tumors. The field started with the lung epithelial cells leading to ovine pulmonary adeno- discovery of avian erythroblastosis virus and Rous carcinoma (OPA) (Palmarini et al., 1999; DeMartini et al., 2001), while ENTV transforms nasal epithelial cells resulting in enzootic nasal tumor (ENT) (De las Correspondence: Professor S-L Liu, Department of Microbiology and Heras et al., 2003). OPA is a significant veterinary and Immunology, McGill University, 3775University Street, Lyman Duff economic problem in many countries, and in some Building Room 608, Montreal, QC Canada H3A 2B4. E-mail: [email protected] countries such as Britain and South Africa, the incidence Received 12 May 2006; revised 22 June 2006; accepted 24 June 2006; of OPA can be as high as 30% (Sharp and Angus, 1990). published online 14 August 2006 The clinical signs of OPA are associated with respiratory Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 790 dysfunction and include exaggerated breathing after JSRV and ENTV are members of a small family of exercise. Late stages of disease are often accompanied acutely transforming retroviruses in which an Env by the secretion of copious lung fluid that contains protein acts as an oncogene. infectious virus. Intriguingly, lung tumors induced by JSRV in sheep JSRV can induce lung tumors in newborn lambs in as morphologically resemble a subtype of human pulmonary little as 10 days following experimental inoculation adenocarcinomas previously referred to bronchiolo- (Verwoerd et al., 1980), showing that JSRV is an acutely alveolar carcinoma or BAC (Bonne, 1939). For exam- transforming retrovirus. However, JSRV is a simple ple, both OPA and BAC tumor cells express markers of retrovirus with no apparent oncogene present in its type II pneumocytes and Clara cells, and tumors are genome (Figure 1a). An alternate open reading frame, generally multifocal and localized to the periphery of the orf-X, which is present within the pol region of JSRV, lungs (Mornex et al., 2003). However, given the new was initially suspected to be an oncogenic factor but was definition of BAC by WHO (Travis et al., 1999), later excluded. Orf-X bears no resemblance to any important distinctions exist between OPA and BAC. known cellular oncogenes except sharing a weak BAC is now defined as a peripheral, well differentiated similarity with the adenosine A3 receptor (Bai et al., adenocarcinoma with pure bronchio-alveolar growth 1999). Disruption of orf-X has no effect on cell and no evidence of stromal, vascular and pleural transformation by JSRV (Maeda et al., 2001). Remark- invasion (Travis et al., 1999); while OPA is a mixed ably, expression of native JSRV Env protein alone is peripheral adenocarcinoma with acinar, papillary and sufficient to induce cell transformation in culture bronchio-alveolar growth (Palmarini and Fan, 2001). (Maeda et al., 2001; Rai et al., 2001; Allen et al., 2002; While the true frequency of pure BAC remains Danilkovitch-Miagkova et al., 2003; Liu and Miller, uncertain, tumors with histologically mixed BAC and 2005), indicating that Env is an active oncogene and is adenocarcinoma account for >20% all non-small cell likely to be the key factor in sheep oncogenesis. The only lung carcinomas (NSCLC) (Raz et al., 2006), and are other examples of retroviruses with oncogenic Env increasing in frequency (Barsky et al., 1994). proteins are ENTV (Alberti et al., 2002; Dirks et al., Similarities between OPA and human lung adenocar- 2002), a close relative of JSRV; avian hemangioma cinoma have led to the hypothesis that JSRV or a virus, which expresses an Env that induces proliferation related virus might be involved in human cancer. In one of monkey epithelial cells and NIH 3T3 cells (Alian study, antisera raised against the JSRV capsid protein et al., 2000); and spleen focus-forming virus, a replica- recognized B30% of human BAC and lung adenocar- tion-defective virus that expresses a recombinant non- cinoma samples but in general did not recognize other functional Env protein that induces cell proliferation in tumor types or normal lung tissue (De las Heras et al., cultured cells and in animals (Ruscetti, 1999). Therefore, 2000). In this case, PCR studies were not carried out to

a JSRV Gag Pro Orf-X Env

LTR Pol LTR

1 kb

b JSRV Env SU TM SP MCT

0.3 kb

c Env cytoplasmic tail (CT) 590

JSRV7 RGMVRDFLKMRVEMLHMKYRNMLQHQHLMELLKNKERGDAGD--D-P JSRV21 ...... JSRVSA ..L...... T....R...... A...--.P. ENTV ..LIK...Q..I.LI.....Y...Y.K..DFV.KR.GSCG.QPAEG- ESRV .SI.KE..H...-LI.---K...... A...--.-.

Figure 1 Schematic representations of JSRV genome, Env structure and sequence alignment of Env cytoplasmic tail. (a) JSRV is a simple retrovirus encoding Gag, Pol, Pro and Env that are ﬂanked by the long term repeats (LTRs). An additional open reading frame X (Orf-X) is present within the Pol region with unknown function. (b) JSRV Env is type I transmembrane protein, with B620 amino acids in length. It is composed of two subunits, surface (SU) and transmembrane (TM). SP is signal peptide. M is membrane-spanning domain. CT is cytoplasmic tail. (c) Sequence alignment of JSRV Env cytoplasmic tail and in comparison with that of ENTV (Y16627) and endogenous sheep retrovirus. Note that a YXXM motif, the putative binding site for PI3K/p85, is conserved among all transforming JSRV and ENTV Env, but is absent in that of endogenous sheep retrovirus. JSRV21 (AF105220) and JSRV7 (AF357971) are two JSRV strains isolated from UK. JSRVSA (M80216) is a JSRV strain isolated from South Africa. ESRV (AF153615) is endogenous sheep retrovirus (AF153615). Dots denote identical amino acids, dashes denote deletions.

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 791 conﬁrm the presence of a retrovirus related to JSRV. tissue morphology, cell–cell communication, and possi- Others have attempted to identify JSRV-related se- bly in tumor metastasis (Menzel and Farr, 1998). The quences in human lung cancer specimens by PCR. In normal biologic function of Hyal2 is not known, but one case, orf-X-like and gag-pro-like sequences were Hyal2 appears to play an important role in JSRV detected in 10–26% of lung cancer patients and healthy Env transformation of human airway epithelial cells individuals in Nigeria and Cameroon, but were not (Danilkovitch-Miagkova et al., 2003) (see below). Other present in samples from France, Germany and Russia GPI-anchored proteins have been shown to be involved (Morozov et al., 2004). In two other studies, JSRV-like in cell signaling, for example, a 150-KDa GPI-anchored sequences were not detected (Yousem et al., 2001; Hiatt TGF-b1-binding protein was recently found to regulate and Highsmith, 2002). Together these studies suggest TGF-b signaling in human keratinocytes (Tam et al., the possibility of JSRV involvement in human lung 2003). cancers, but more work is needed to resolve this issue.

JSRV Env is an oncogene: primary structure and JSRV and ENTV cell-entry receptor Hyal2: a putative oncogenic domains tumor suppressor The primary function of retroviral Env proteins is to Entry of retroviruses into target cells is dependent on mediate viral entry into host cells through binding to specific interactions between the Env proteins present on specific entry receptors on the cell surface. However, the virus surface and their corresponding cell-entry JSRV Env also transforms cultured cells. Given that receptors on the cell surface. The human receptor for diseased sheep often secrete a large amount of lung fluid JSRV and ENTV has been identified as hyaluronidase 2 containing infectious JSRV particles (Palmarini and (Hyal2) by phenotypic screening of human/hamster Fan, 2001) and the fact that JSRV expression is solely radiation hybrid cell lines, based on the fact that JSRV detected in tumor cells but not in surrounding normal vectors transduce human but not hamster cells (Rai cells (Palmarini et al., 1995), it is postulated that cell et al., 2000, 2001). Interestingly, unlike most retroviral transformation by JSRV Env may be advantageous for receptors or coreceptors which are single or multiple viral replication in vivo (Fan et al., 2003). transmembrane proteins (Overbaugh et al., 2001), the JSRV Env is a typical type-I transmembrane protein JSRV/ENTV entry receptor Hyal2 is attached to cell of approximately 620 amino acids in length before surface by a glycosylphosphatidylinositol (GPI)-anchor processing (York et al., 1992; Palmarini et al., 1999; (Rai et al., 2001). Hyal2 orthologs have been cloned DeMartini et al., 2001). The mature full-length Env is from several other species, including sheep, mouse, rat, composed of surface (SU) and transmembrane (TM) cow, pig and dog. Ovine Hyal2 functions best as a subunits that are linked by disulfide bonds (Figure 1b). receptor for JSRV and ENTV, as might be expected SU is responsible for receptor-binding, and TM is (Dirks et al., 2002). When overexpressed, rat Hyal2 can involved in virus–cell membrane fusion. As for other function as a receptor for JSRV but not for ENTV (Liu simple retroviruses, JSRV Env has a short cytoplasmic B et al., 2003a), whereas mouse Hyal2 is inactive as a domain of 44 amino acids in length (Figure 1c). receptor for either virus (Dirks et al., 2002; Liu et al., Interestingly, the cytoplasmic tail of JSRV Env harbors 2003a). The receptor activities of these Hyal2 orthologs a YXXM (Y is tyrosine, X stands for any amino acid, M are in accordance with the host range of JSRV infection; is methionine) peptide motif, a putative binding site for for instance, ovine and human cells are highly suscep- the PI3K/p85regulatory subunit (Songyang et al., tible to JSRV vector transduction compared to rodent 1993). This feature is distinct from most other retro- cells (Rai et al., 2000; Dirks et al., 2002). Domains or viruses that contain the YXXF motif (F stands for any amino-acid residues that are responsible for these amino acids with a bulky hydrophobic chain). The different receptor activities are complex, but appear to YXXF motif has been shown to play important roles in reside in the central third of the Hyal2 protein sequence endocytosis, trafficking, as well as retroviral pathoge- (Duh et al., 2005). nesis. The role of YXXM motif in JSRV Env transfor- Human Hyal2 is located in the p21.3 region of chro- mation has been investigated, and is summarized below. mosome 3 (Rai et al., 2000), a region that is deleted in a substantial proportion of human cancers (Zabarovsky The Env cytoplasmic tail: necessary and sufficient for cell et al., 2002). Hyal2 and other genes in this region are transformation? therefore regarded as putative tumor suppressors Studies from several groups demonstrate that the (Zabarovsky et al., 2002). Hyal2 belongs to the cytoplasmic tail of JSRV Env is essential for transfor- hyaluronidase gene family that includes Hyal1, PH20, mation, because replacement of this domain with that of Hyal3 and Hyal4 (Csoka et al., 2001). Hyal2 exhibits other retroviruses completely abolishes the transforming low yet detectable hyaluronidase activity compared to activity of JSRV Env in multiple cell lines, including that of Hyal1 and PH20 (Lepperdinger et al., 1998; Rai NIH 3T3 mouse fibroblasts (Palmarini et al., 2001), et al., 2001; Vigdorovich et al., 2005). Hyaluronidases 208F rat fibroblasts (Liu et al., 2003b), DF-1 chicken degrade hyaluronan, a major component of extracellular fibroblasts (Allen et al., 2002), and MDCK canine matrix that is important for the maintenance of normal epithelial cells (Liu and Miller, 2005). However, whether

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 792 or not this domain is sufficient for JSRV Env how it affects transformation in these cells is poorly transformation still remains elusive. We and others understood. Several groups have investigated whether have attempted to address this issue by building the SU is important for rodent fibroblast transformation by cytoplasmic domain of JSRV Env into other retroviral JSRV Env, but have obtained inconsistent results. In envelope proteins, but in general these proteins fail to be one study, Chow et al. (2003) replaced the receptor- expressed well and fail to transform. One study reported binding domain and proline-rich region of JSRV SU that the C-terminal 141-amino-acid, but not the 44 with that of Moloney murine leukemia virus, or deleted amino-acid cytoplasmic domain, of JSRV Env protein this receptor-binding domain from SU, but found no was able to efficiently transform 208F rat fibroblasts dramatic effect on Env transformation of 208F cells (Chow et al., 2003). This was accomplished by attaching (Chow et al. 2003). This finding is consistent with our the C-terminus of JSRV Env sequence to a myristoyla- data that JSRV SU protein does not bind the mouse tion signal derived from the Rasheed sarcoma virus gag ortholog of Hyal2 receptor in NIH 3T3 cells, and that protein. However, we could only detect minimal if any overexpression of mouse Hyal2 has no effect on JSRV transforming activity of this construct in 208F cells or Env transformation in NIH 3T3 and 208F cells (Liu other cell lines (unpublished results). Similar results were et al., 2003a). By contrast, Hofacre and Fan (2004) also obtained by other investigators, including this recently showed that large deletions or small insertions group (Varela et al., 2006). Thus, it is possible that the in the JSRV SU region could abolish JSRV Env cytoplasmic tail of JSRV Env is necessary but not transformation of NIH 3T3 and 208F cells. Interest- sufficient for cell transformation, but more work is ingly, these investigators observed that co-transfection required to resolve this issue. of transformation-defective SU and TM mutants was able to rescue the Env transforming activity in 208F cells (Hofacre and Fan, 2004), suggesting that efficient Role of the YXXM motif in cell transformation transformation of rodent fibroblasts by JSRV Env It has been hypothesized that the YXXM peptide motif may require both SU and TM. present in the cytoplasmic tail of JSRV Env, a putative While it is easy to understand the direct role of the PI3K/p85binding motif, might be essential for cell cytoplasmic domain of JSRV Env in cell transforma- transformation, especially given its absolute conservation tion, how SU subunit might be involved in this process among all transforming JSRV and ENTV strains yet remains unclear. One possibility is that SU transmits absence in the nononcogenic endogenous sheep retro- independent signals that synergize with those of the Env viruses (Palmarini et al., 2000b). Indeed, an early study cytoplasmic domain, as proposed by Hofacre and Fan showed that mutation of this tyrosine residue to (2004). Alternatively, SU may contribute to cell phenylalanine or to aspartic acid completely abolished transformation indirectly by modulating the overall the transforming activity of JSRV Env in NIH 3T3 cells structure of Env protein. In both cases, cellular proteins (Palmarini et al., 2001). However, we and others later might be involved. Thus, it will be important to identify observed that all of these tyrosine mutants were still able cellular adaptor molecules that directly interact with to transform NIH 3T3 (Liu et al., 2003b), 208F (Liu JSRV SU and/or TM subunits, and to investigate how et al., 2003b), DF-1 (Allen et al., 2002), and MDCK (Liu these molecules mediate intracellular signaling, leading and Miller, 2005) cells, but with reduced efficiency in to cell transformation. most cases. The difference in results is likely due to Unlike the situation in rodent fibroblasts, JSRV SU different transformation procedures used by each group, and its interaction with receptor Hyal2 appear to be including differences in the duration of transformation critical for the Env transformation of BEAS-2B human assay and use of G418 selection after transfection. In bronchial epithelial cells. BEAS-2B is an immortalized addition, despite repeated attempts, no tyrosine phos- cell line established by infection with a hybrid adeno- phorylation has been detected for the JSRV Env protein virus-12 and SV40 virus (Reddel et al., 1988). Transfec- in the transformed cells (Liu et al., 2003b), which argues tion of BEAS-2B cells with a plasmid encoding JSRV against a direct role of the YXXM motif in activating the Env was shown to result in piled-up colonies with fork- PI3K/Akt signaling pathway (see below). Furthermore, like structures, indicative of transformation (Danilkovitch- mutation of the methionine residue of the YXXM motif Miagkova et al., 2003). Interestingly, Hyal2 and its also failed to have significant effect on cell transformation interaction with JSRV Env seem to play a critical role in (Allen et al., 2002; Liu et al., 2003b). Moreover, the the Env transformation of this cell line. In normal YXXM Env mutant-transformed cells still show robust BEAS-2B cells, Hyal2 was found in a complex with PI3K-dependent Akt activation (Liu et al., 2003b). Thus, RON receptor tyrosine kinase and therefore inhibit it is now generally believed that the tyrosine residue of RON activation, expression of JSRV Env resulted in YXXM motif is important for Env transformation, Hyal2 proteasomal degradation, thereby liberating possibly by modulating the Env protein expression or its RON and leading to its activation (Danilkovitch- overall configuration, but the YXXM motif apparently Miagkova et al., 2003). As the receptor-binding domain does not directly activate PI3K in transformed cells. that interacts with Hyal2 resides in the SU subunit of Env protein (Liu et al., 2003a), the SU subunit is Is SUinvolved? presumably responsible for the Env transformation of The role of SU subunit in JSRV Env cell transformation BEAS-2B cells. However, whether or not SU or the appears to depend on the target cell type, but exactly receptor-binding domain of SU is necessary and

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 793 sufficient for BEAS-2B cell transformation, or if other and RK3E rat epithelial cells (Maeda et al., 2005). The regions of JSRV Env are also required in the transfor- third is referred to as RON-Hyal2 pathway, which is mation process, need to be addressed. In addition, these primarily activated in the transformed BEAS-2B human experiments have been difficult to reproduce because of epithelial cells (Danilkovitch-Miagkova et al., 2003). a high background of transformed foci and the poor Studies from several labs have shown that multiple transfectability of BEAS-2B cells. signaling pathways can be activated in the same cell types transformed by JSRV Env, and the same pathway can be operative in different transformed cell lines. Below we highlight evidence that support for the roles of Signaling pathways involved in JSRV Env-mediated cell each pathway in JSRV Env transformation, and discuss transformation some unsolved problems. Three signaling pathways have been reported to be involved in cell transformation by JSRV Env protein PI3K-dependent and independent Akt pathways (Figure 2). The first is Akt pathway that is either The PI3K-dependent Akt signaling pathway is well dependent or independent of PI3K. This pathway is known for its role in cell survival, oncogenic transfor- activated in Env-transformed fibroblasts derived from mation and cancer development (Datta et al., 1999). rodent (NIH 3T3 and 208F) (Palmarini et al., 2001; This pathway can be activated by many growth factors, Alberti et al., 2002; Chow et al., 2003; Maeda et al., hormones, cytokines, as well as by mutations or 2003, 2005; Liu et al., 2003b) and chicken (CEF and overexpression of receptor or nonreceptor tyrosine DF-1) (Allen et al., 2002; Zavala et al., 2003), as well as kinases (Datta et al., 1999). Many viral proteins, in MDCK dog epithelial cells (Liu and Miller, 2005). including retroviral oncogenes, can also activate this The second is Raf-MEK-MAPK pathway, which is pathway. Importantly, PI3K and Akt themselves were activated in Env-transformed NIH 3T3 mouse fibroblasts originally discovered from the acutely transforming retroviruses, avian sarcoma virus 16 (ASV16) (Chang et al., 1997) and AKT8 (Bellacosa et al., 1991), JSRV Env respectively. The vial counterparts of PI3K and Akt, v-p3K and v-Akt, are also potent oncogenes that transform cells in culture (Chang et al., 1997; Aoki et al., 1998). The PI3K-independent Akt activation, on the other hand, is less common with only limited study.

Hyal2 In the latter case, Akt is activated by cAMP, an agonist of PKA pathway, and is directly phosphorylated by PI3K Ras calcium/calmodulin dependent kinase (CaMKK), in a PDK1/2 PI3K-independent of manner (Sable et al., 1997; Filippa et al., 1999). Akt RON Several lines of evidence indicate that PI3K-dependent Akt activation is critical for cell transformation by p38 the JSRV Env protein. First, Akt is activated in several Raf Env-transformed cell lines, as shown by Akt phosphor- TSC1-TSC2 ylation and elevated in vitro Akt kinase activity (Palmarini et al., 2001; Zavala et al., 2003; Liu et al., MEKs 2003b; Liu and Miller, 2005). Second, PI3K speciﬁc RheB inhibitor LY294002 or wortmannin inhibits transformation by JSRV Env in these cells, and inhibits their Akt phosphorylation and in vitro kinase activity in a dose-

mTOR ERKs dependent manner (Palmarini et al., 2001; Zavala et al., 2003; Liu et al., 2003b). Third, PI3K-speciﬁc inhibitors are able to reverse the transformed phenotypes of Env- transformed 208F cells (Liu et al., 2003b). In addition, there appears to be a correlation between the trans- Transformation formed phenotype and the extent of Akt phosphorylation in transformed cells, with highly transformed cells Figure 2 Signaling pathways involved in cell transformation by by JSRV Env exhibiting higher Akt phosphorylation JSRV Env. Three major pathways have been uncovered in different (Liu et al., 2003b). Despite these important correlations, cell lines transformed by JSRV Env, including the Akt pathway no interaction between JSRV Env protein and PI3K/p85 that is dependent or independent of PI3K, the Hyal2-RON has been observed, indicating JSRV Env may activate pathway that activates both PI3K/Akt and Ras-MEK-ERK PI3K/Akt indirectly, possibly through as-yet unidenti- pathways, and the MAPK pathway that includes p38. Except for the RON-Hyal2 pathway, how JSRV Env engages other signaling ﬁed cellular adaptor molecules or by other signaling networks is not known. Note that MAPK/p38 activation plays an pathways that are directly triggered by JSRV Env. inhibitory role in cell transformation. See text for details. Interestingly, however, Akt phosphorylation was not

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 794 detected in the lung sections of sheep OPA, nor in the understood. Among the diverse array of phospho-Akt JSRV Env-transformed DF-1 cells in vitro, yet was (pAkt) substrates that have been identified so far, present in the ENTV-induced nasal tumors of sheep mTOR, the mammalian target of rapamycin, was (Zavala et al., 2003). These data indicate that the Akt recently found to be involved in JSRV Env transforma- phosphorylation in the Env-transformed cells may be tion of NIH 3T3 cells (Maeda et al., 2005). This was cell-type dependent. shown by the fact that mTOR inhibitor, rapamycin, PI3K-independent Akt activation has also been partially inhibits the Env transformation efficiency reported for JSRV Env transformation. Fan and co- (Maeda et al., 2005). However, to what extent the workers co-transfected NIH 3T3 cells with plasmids mTOR activation contributes to transformation of NIH encoding JSRV Env and a dominant-negative mutant of 3T3 cells by JSRV Env protein has not been firmly p85, or transfected the Env-coding plasmid to mouse determined. This issue is crucial in light of a recent study embryo fibroblasts derived from p85a/p85b-double- showing that mTOR can also function in the upstream knockout mice, and observed that Env transformation of Akt and directly phosphorylates Akt at the Ser473 efficiency in these cells were not significantly affected site in Drosophila and mammalian cells (Sarbassov et al., (Maeda et al., 2003). Interestingly, however, Akt 2005). In this sense, it might be important to further phosphorylation was still readily detected from the examine the relationship between PI3K/Akt and mTOR Env-transformed cells, suggesting a PI3K-independent pathways in the JSRV Env-mediated cell transforma- Akt activation by JSRV Env. Further evidence support- tion, in particular to investigate how Akt is activated in ing this scenario came from the fact that PI3K inhibitor transformed cells. On the other hand, the Akt down- LY294002 or wortmannin failed to specifically inhibit stream effectors have not been characterized. Although the Env-mediated transformation in NIH 3T3 cells and these molecules are unlikely to be directly or specifically to reverse the transformed phenotypes (Maeda et al., associated with JSRV Env protein, elucidation of the 2003). These results are quite provocative, because they downstream pathways should promote our understand- stand in contrast with other studies and with the notion ing of the molecular basis of JSRV Env transformation that Akt activation involved in cell survival and and may provide new insights into the role of Akt transformation is generally PI3K-dependent. pathway in cell survival and transformation. While it is not currently clear what factors may have contributed to these discrepancies, several possibilities exist. First, the mechanism of JSRV Env transformation Raf-MEK-MAPK pathway might be cell type-dependent, and this has been proven The importance of MAPK signaling pathways for cell to be the case for several viral oncogenes. For example, proliferation, differentiation, survival, and oncogenic transformation of NIH 3T3 cell by viral Src (v-Src) transformation, as well as for programmed cell death or oncogene is dependent on cellular Ras (c-Ras) activa- apoptosis, is becoming increasingly clear (see a compre- tion, whereas v-Src transformation of Rat-2 fibroblasts hensive review by Pearson et al. (2001)). To date, five does not involve c-Ras (Aftab et al., 1997). Second, members of MAPK family have been identified in different experimental procedures used by each group, mammals, each playing distinct yet critical roles in a for example, the duration of drug treatment or the drug variety of cellular processes. The first is extracellular concentration, may account for the distinct results. signal-regulated kinases (ERKs), the classical members Third, although p85a and p85b, two major regulatory of MAPK family that are activated by the upstream units of PI3K, were knocked out from the NIH 3T3 cells kinases, MAPKKKs (A-Raf, B-Raf, and Raf-1) and in the study by Maeda et al. (2003), other PI3K MAPKKs (MEK1 and MEK2). This pathway is usually regulatory subunits, such as p55g, may still be expressed activated by extracellular mitogens, and is thus impor- in these cells that could mediate a PI3K-dependent Akt tant for cell proliferation and differentiation (reviewed activation. Future work should address if Akt activation by Ballif and Blenis (2001)). The second and third is mediated by other PI3K regulatory subunits in the families are c-Jun amino-terminal kinases (JNKs) p85a and p85b double-knockout cells transformed by (JNKs 1, 2 and 3) and p38 (isoforms a, b, g and d), JSRV Env, and if the PI3K-independent Akt activation respectively. These two pathways are often activated by is operative in other Env-transformed cell lines. environmental stress, UV radiation and cytokines, and While Akt is activated in rodent fibroblasts trans- therefore are critical for normal immune and inflamma- formed by JSRV Env protein, either in a PI3K- tory responses (reviewed by Wada and Penninger dependent or independent fashion, the signaling mole- (2004)). The other two members are ERKs 3/4 and cules that directly interact with JSRV Env protein and ERK5, respectively, with biological function having not initiate the signaling network are still not known. been well characterized. Notably, ERK5, also called big Attempts have been made by several groups to identify MAPK or BMK1 because of its large size, has been the Env-interacting cellular molecules without success. shown to be important for neuronal survival (reviewed Recently, toll-like receptor 4 (TLR4) was found to by Cavanaugh (2004)). Among these MAPK family interact with JSRV Env protein, but the relevance of this members, the Raf-MEK-REK1/2 pathway has been interaction to cell transformation remains uncertain shown to play the most critical roles in cell transforma- (Hung Fan, personal communication). tion by oncogenes, including oncogenic retroviruses. Likewise, the downstream Akt effector targets in- The roles of the MAPK pathway in JSRV Env volved in JSRV Env transformation are also poorly transformation has just begun to be appreciated. We

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 795 initially examined ERK1/2 phosphorylation in 208F by the JSRV Env protein (Danilkovitch-Miagkova cells transformed by JSRV Env protein, but failed to et al., 2003). In normal BEAS-2B cells, Hyal2 is detect any ERK phosphorylation in these cells (Liu constitutively associated with RON and inhibits RON et al., 2003b). A similar observation was also recently activation. Expression of JSRV Env leads to Env made by Fan and colleagues in the Env-transformed binding to Hyal2, Hyal2 degradation, liberation of NIH 3T3 cells (Maeda et al., 2005). Interestingly, the RON from Hyal2 inhibition, and RON activation and latter group found that the MEK1-specific inhibitor, cell transformation (Danilkovitch-Miagkova et al., PD98059, as well as an H/N-Ras inhibitor, FTI-277, can 2003). In this case, Hyal2 functions as a tumor inhibit Env transformation of NIH 3T3 and RK3E rat suppressor that negatively regulates RON activity and epithelial cells in a dose-dependent manner (Maeda Env-mediated cell transformation. Importantly, a domi- et al., 2005), indicating that the Ras-Raf-MEK-ERK nant-negative kinase-dead RON mutant was able to pathway might be involved in JSRV Env transformation block transformation by Env, indicating that the RON- of these cells. However, it is still not known how this Hyal2 pathway was critically important for Env pathway is activated by the JSRV Env protein and why transformation of BEAS-2B cells. ERK phosphorylation is not detected in Env-trans- RON belongs to the Met proto-oncogene family, and formed cells. One possibility is that ERK phosphoryla- is widely expressed in human tissues, in particular those tion is transient or unstable in the transformed cells, of epithelial origin and in immune cells (reviewed by either due to the presence of negative feedback loops or Comoglio and Boccaccio (1996)). RON is also over- cross-talks with other signaling pathways (see details expressed in a variety of human tumors, in particular below). breast and colon cancers (reviewed by Wang et al. MAPK/p38 pathway was also recently shown to be (2003)). Overexpression of RON has been shown to involved in JSRV Env-mediated cell transformation. induce distal lung tumors in transgenic mice and This is directly demonstrated by the effect of p38 synergize in mammary tumor formation and metastasis inhibitor SB203580 that can increase the Env transfor- (Wang et al., 2003). The role of RON in BEAS-2B mation efficiency in NIH 3T3 and RK3E cells (Maeda transformation by JSRV Env is supported by RON et al., 2005). Interestingly, p38 phosphorylation is only tyrosine phosphorylation in transformed cells and by the slightly increased in the Env-transformed cells compared fact that the dominant-negative kinase-dead RON to that of negative controls (Maeda et al., 2005), which mutant could block cell transformation by JSRV Env suggests additional pathways that may regulate p38 (Danilkovitch-Miagkova et al., 2003). phosphorylation. Intriguingly, treatment of transformed Several critical questions need to be addressed NIH 3T3 cells with SB203580 was shown to significantly concerning the RON-Hyal2 pathway. The first is potentiate MEK1/2 and ERK1/2 phosphorylation in whether a constitutively activated RON can directly these cells (Maeda et al., 2005), implying a cross-talk transform BEAS-2B cells in the absence of JSRV Env. between ERK and p38 pathways. While this may We have attempted to address this issue by over- explain why ERK phosphorylation is not detectable in expressing RON in BEAS-2B cells, but failed to observe the Env-transformed NIH 3T3 and 208F cells (Liu et al., morphological transformation. We also expressed RON 2003b; Maeda et al., 2005), exactly how p38 pathway in several cell lines, including NIH 3T3, 208F and negatively modulates ERK phosphorylation remains to MDCK, again, we only detected low transforming be better defined. One possibility is that p38 activates activity (Miller et al., 2004; unpublished results). These PP1/2A, a serine/threonine phosphatase that depho- results indicate that human RON has low if any sphorylates MEK1/2, leading to reduced ERK1/2 transforming activity, and are consistent with a pre- phosphorylation (Westermarck et al., 2001). Alterna- viously published study (Santoro et al., 1996). However, tively, ERK1/2 may be dephosphorylated by other constitutively activated RON mutants have been shown unknown phosphatases. Additional question that needs to induce transformation in vitro, cause tumors in nude to be seriously addressed is how JSRV Env protein mice, and promote tumor metastasis (Wang et al., 2003). activates the p38 pathway in NIH 3T3 and RK3E cells, The involvement of Hyal2 in JSRV Env transforma- and whether or not this pathway is important for Env tion of BEAS-2B cells is supported by the disassociation transformation in other cell lines. It is noteworthy that of the Hyal2-RON complex that is normally detected in although ERK/MAPK phosphorylation is readily pre- the untransformed cells (Danilkovitch-Miagkova et al., sent in the natural and experimentally-induced OPA, 2003). This finding is intriguing, because it indicates that however, p38 phosphorylation is not detected in these Hyal2 can function as a tumor suppressor through tumors (Maeda et al., 2005). RON. We have further investigated this issue by overexpressing human Hyal2 in NIH 3T3, 208F or MDCK cells that express human RON (note that these RON-Hyal2 pathway cells do not normally express an ortholog of human The mechanism of JSRV Env transformation in the RON), but unfortunately we found no obvious effect of immortalized human bronchial epithelial cell line BEAS- Hyal2 on the RON-induced scattering and tyrosine 2B is quite different from that in fibroblasts and other phosphorylation (Miller et al., 2004; unpublished epithelial cells. Several lines of evidence indicate that the results). We also overexpressed human Hyal2 in several receptor tyrosine kinase RON and the cell-entry receptor human lung cancer cell lines by retroviral vector Hyal2 are important for BEAS-2B cell transformation transduction, and again, no significant change in cell

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 796 morphology and proliferation was observed (unpub- A mouse model of lung cancer for JSRV Env oncogenesis lished results). Together, these experiments failed to support for the role of Hyal2 in inhibiting RON activity, The inability of JSRV Env to mediate infection of in agreement with a previous study showing that Hyal2 mouse cells (Rai et al., 2000, 2001) or to bind mouse was unable to suppress the growth of normal human Hyal2 (Liu et al., 2003a), and the apparently essential bronchial epithelial cells and lung cancer cell lines (Ji role of Hyal2 in Env transformation of BEAS-2B et al., 2002). Interestingly, the latter group did observe a human lung epithelial cells (Danilkovitch-Miagkova 70% reduction of lung tumor metastasis following et al., 2003) initially discouraged an effort to establish expression of Hyal2 in an experimental A549 metastatic a mouse model for JSRV oncogenesis. However, lung cancer model (Ji et al., 2002). Thus, whether or not growing evidence that JSRV Env protein could trans- Hyal2 is a general tumor suppressor cannot be easily form a variety of cell types, including epithelial cells, answered at present. independently of Hyal2 (Liu and Miller, 2005; Maeda While apparently essential for BEAS-2B cell et al., 2005; Varela et al., 2006), suggested that transformation by JSRV Env, the RON-Hyal2 pathway development of such a mouse model might be possible. is unlikely to operate in NIH 3T3 cells. First, the mouse Indeed, a mouse model of lung cancer induction by ortholog of RON, Stk, is not expressed in NIH JSRV Env has been recently developed by using a 3T3 cells, as determined by showing that radioactively- replication-defective adeno-associated virus type 6 labeled MSP, the RON/Stk ligand, does not bind (AAV6) vector that expresses the JSRV Env protein to NIH 3T3 cells, and by showing that Stk protein (Wootton et al., 2005). AAV6 vectors promote long- cannot be detected in lysates from NIH 3T3 cells by term transgene expression in all epithelial cell types of Western blotting analysis (Wang et al., 1995; Miller mouse lungs (Halbert et al., 2001), including type II et al., 2004). Second, mouse Hyal2, although expressed pneumocytes, a presumed target of JSRV oncogenesis in in NIH 3T3 cells, does not appear to bind JSRV sheep. Mouse lung tumors induced by JSRV Env Env protein, as shown by FACS assay using a hybrid appeared as early as 9 weeks after vector exposure, with protein consisting of JSRV SU fused to a human IgG size increasing over time, and showed morphology and Fc fragment (Liu et al., 2003a). Third, overexpression localization similar to that in sheep (Wootton et al., of mouse Hyal2 has no effect on JSRV Env transforma- 2005) (Figure 3). tion efficiency of NIH 3T3 cells, nor has any effect on AAV6 vectors are known to transduce all types of the transformed phenotype (Liu et al., 2003a). Likewise, epithelial cells in the mouse lung and airway. Strikingly, the RON-Hyal2 pathway is unlikely to play a role in tumors induced by JSRV Env-expressing AAV6 vector Env transformation of MDCK cells because RON all expressed the type II pneumocyte marker, surfactant expression is not detected in MDCK cells (Wang protein C (Wootton et al., 2005). It has been previously et al., 1994), and overexpression of human RON in reported that the JSRV long-terminal repeat (LTR) MDCK cells has no effect on transformation by JSRV promoter and enhancer is preferentially active in the Env (Liu and Miller, 2005). In fact, we have recently type II pneumocytes (Palmarini et al., 2000a; McGee- shown that JSRV Env protein transforms MDCK cells Estrada et al., 2005), due to a requirement for the by the same mechanism as that in rodent fibroblasts (Liu presence of hepatocyte nuclear factor-3 beta (HNF-3b) and Miller, 2005). Thus the RON-Hyal2 pathway and other factors expressed in type II cells (McGee- appears to be specific for JSRV Env transformation of Estrada et al., 2002) that can transactivate the JSRV BEAS-2B cells. LTR and enhance the downstream gene expression Interestingly, the RON-Hyal2 pathway might be (McGee-Estrada and Fan, 2006). While this proposed functionally operative in some human lung cancer cell mechanism may explain the tumor type observed in lines. This was suggested by the finding that RON is sheep, it cannot explain the tumor type observed in the overexpressed and constitutively activated in some mouse model because in this case, expression of JSRV human lung cancer cell lines, in particular BAC cells Env protein is not driven by the native JSRV LTR derived from BAC patients (Danilkovitch-Miagkova promoter but by the broadly-active Rous sarcoma virus et al., 2003). In addition, the receptor activity of human promoter (Wootton et al., 2005). One possibility, yet to Hyal2 is inhibited or lost in these cells, as shown by be proved, is that the restricted JSRV Env expression in reduced or undetectable titers of JSRV retroviral vectors cells expressing type II pneumocyte markers in mouse as compared to that of normal human lung epithelial lungs reflects a general feature of mouse lung biology. It cells (unpublished results). One intriguing possibility is has been noticed for years that the majority of murine therefore that Hyal2 protein is degraded in these cells, lung tumors induced by oncogenes or chemical carcino- due to the presence of JSRV or a JSRV-like virus or gens contain cells having the phenotypes of type II sequence. Alternatively, defects may exist in the Hyal2 pneumocytes or Clara cells, but the exact mechanism is genomic DNA or mRNA of these BAC cells that not well understood (Tuveson and Jacks, 1999). preclude the Hyal2 protein synthesis. Future effort Another important finding from the mouse model for should focus on addressing these possibilities, and JSRV Env oncogenesis is that host immunity can results from these investigations should lead to a better provide a significant level of protection against tumori- understanding the mechanism of RON activation in genesis. This is manifested by the fact that AAV6 vector these BAC cells and a possible relationship between expressing JSRV Env induced multiple large lung JSRV and human lung cancer. tumors in the immune-deficient C57BL/6 Rag2 mice,

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 797

Figure 3 JSRV Env induces lung tumors in mice with morphology similar to that of ovine pulmonary adenocarcinoma and human peripheral adenocarcinoma. (a) Multifocal lung tumors in mice 6 months after exposure to an AAV6 vector encoding JSRV Env. (b) Adenoma induced in mice by the AAV6 vector. (c) Adenocarcinoma induced in mice by the AAV6 vector. (d) Human lung peripheral adenocarcinoma. (e) Ovine pulmonary adenocarcinoma induced by JSRV. (d and e) courtesy of JC DeMartini. Adapted and modiﬁed from Wootton et al., 2005.

but yielded a few small tumors in immune-competent occurring or experimentally induced sheep (Sharp and C57BL/6 mice (Wootton et al., 2005). In addition, Herring, 1983; Palmarini et al., 1996; Ortin et al., 1998; serum harvested from the C57BL/6 mice exposed to the Summers et al., 2002). Thus, it is possible that host AAV6 vector encoding JSRV Env is able to block immune response can protect other species, including transduction by a JSRV retroviral vector (Wootton humans, from JSRV tumorigenesis. et al., 2005), indicating that a host humoral immune Several important questions remain to be answered response was indeed mounted against JSRV Env. In with respect to the mouse model of JSRV Env contrast, normal sheep are immunotolerant of JSRV oncogenesis. First, and as mentioned above, it is still and are susceptible to JSRV oncogenesis (DeMartini unknown why tumor formation is restricted to peri- et al., 2003) because of the presence and expression of pheral cells that express type II pneumocyte markers, several closely related endogenous retroviruses in their and it will be important to determine whether this is genomes (Hecht et al., 1996; Palmarini et al., 2000b; JSRV-Env-speciﬁc or a reﬂection of the default mode of Sanna et al., 2002; Klymiuk et al., 2003; Spencer et al., mouse lung cancer. Second, the molecular mechanisms 2003). Indeed, no humoral or cellular immune response of oncogenic transformation by JSRV Env in the mouse has been detected against JSRV, either in the naturally model are not known. It will be essential to determine

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 798 whether the oncogenic domains and signaling pathways significance of these integration sites for JSRV oncoge- identified in vitro are also active in the mouse model of nesis remains to be further defined. JSRV Env oncogenesis. While it is possible that PI3K/ As discussed in this review, significant progress has Akt and Ras-Raf-MEK-MAPK pathways are still been made over the last few years in understanding important for JSRV Env tumorigenesis in vivo, other JSRV oncogenesis. However, our understanding in the signaling pathways may also play a role and therefore mechanisms of cell transformation by JSRV Env protein should be explored. On the other hand, although mouse is still far from complete. For instance, we still do not Hyal2 does not interact with JSRV Env, making a role know how the JSRV Env protein engages the cell for the RON-Hyal2 pathway unlikely, this assumption signaling network, leading to transformation, and should be experimentally tested. Third, it is well identification of cellular molecules that directly interact accepted that tumorigenesis in vivo is a complicated with JSRV Env protein will be of great interest. On the process involving series of signaling events, and it would signaling side, three major pathways have been un- be important to examine if other factors, such as cellular covered, but how they are activated independently yet oncogenes or tumor suppressor genes, might be also cross-talk to each others is not known. In addition, we involved in JSRV oncogenesis in mice. Lastly, it will be know little about JSRV Env transformation in primary interesting to determine if tumor metastasis occurs in the lung epithelial cells and in vivo, and future research mouse model of lung cancer induced by JSRV Env. should be directed towards this end. Further exploration of the mouse model for JSRV Env JSRV presents an attractive model for the study of oncogenesis will provide critical insights into the human lung cancer, given the strong similarities between mechanism of sheep retroviral oncogenesis and may the sheep tumors induced by JSRV and human illuminate the biology of human lung cancer. adenocarcinoma. The convergence and interplay among JSRV Env, human Hyal2, and RON, as revealed in the BEAS-2B cell transformation by JSRV Env Concluding remarks and perspectives (Danilkovitch-Miagkova et al., 2003), are very interesting and may have important implications in this respect. Cancer is a multistep process involving complex inter- While traditional methods such as PCR or immunohisto- actions between activation of oncogenes, inactivation of chemistry may be valuable, future efforts should take tumor suppressor genes, as well as genetic or epigenetic advantage of other recently developed technologies, alterations associated with tumor progression and such as microarray or proteomics, to explore whether metastasis (Hanahan and Weinberg, 2000). While JSRV JSRV, or a related virus, is involved in human lung is an acute transforming retrovirus, with its Env protein cancer. Discovery of such a virus, although challenging, acting as a potent oncogene, it must be born in mind would have a profound impact on lung cancer preven- that tumor development in sheep by JSRV usually takes tion, diagnosis and treatment. Even if no such virus is months to years in natural settings (Sharp and found, these investigations should lead to a deeper DeMartini, 2003). This paradox highlights, once again, understanding of retroviral oncogenesis as well as the multistep nature of tumorigenesis by JSRV in sheep; human lung tumorigenesis. more importantly, it suggests other mechanisms, such as insertional mutagenesis, might also play a role in JSRV oncogenesis. Several groups have investigated the Acknowledgements possibility that JSRV provirus may have common integration sites in the sheep genomes, and obtained We apologize to those colleagues whose original work is not interesting but inconsistent results (DeMartini et al., cited due to space limitations. We thank Michael Lerman, Alla 2001; Cousens et al., 2004; Philbey et al., 2006). Danilkovitch-Miagkova and James DeMartini for collabora- tions and discussions. This work was supported by funds from Notably, one common integration site was recently the Canadian Institutes of Health Research and McGill identified on the sheep chromosome 16 (Cousens et al., University to S-L Liu, and grants from the United States 2004), and another clonal integration was observed in National Institutes of Health and the Fred Hutchinson Cancer the receptor protein tyrosine phosphatase g genes of Research Center to AD Miller, S-L Liu is a Canada Research chromosome 19 (Philbey et al., 2006). However, the Chair in Virology and Gene Therapy.

References

Aftab DT, Kwan J, Martin GS. (1997). Ras-independent Allen TE, Sherrill KJ, Crispell SM, Perrott MR, Carlson JO, transformation by v-Src. Proc Natl Acad Sci USA 94: DeMartini JC. (2002). The jaagsiekte sheep retro- 3028–3033. virus envelope gene induces transformation of the avian Alberti A, Murgia C, Liu S-L, Mura M, Cousens C, Sharp M ﬁbroblast cell line DF-1 but does not require a et al. (2002). Envelope-induced cell transformation by ovine conserved SH2 binding domain. J Gen Virol 83: betaretroviruses. J Virol 76: 5387–5394. 2733–2742. Alian A, Sela-Donenfeld D, Panet A, Eldor A. (2000). Avian Aoki M, Batista O, Bellacosa A, Tsichlis P, Vogt PK. (1998). hemangioma retrovirus induces cell proliferation via the The akt kinase: molecular determinants of oncogenicity. envelope (env) gene. Virology 276: 161–168. Proc Natl Acad Sci USA 95: 14950–14955.

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 799 Bai J, Bishop JV, Carlson JO, DeMartini JC. (1999). Sequence Dirks C, Duh FM, Rai SK, Lerman MI, Miller AD. (2002). comparison of JSRV with endogenous proviruses: envelope Mechanism of cell entry and transformation by enzootic genotypes and a novel ORF with similarity to a G-protein- nasal tumor virus. J Virol 76: 2141–2149. coupled receptor. Virology 258: 333–343. Duh FM, Dirks C, Lerman MI, Miller AD. (2005). Ballif BA, Blenis J. (2001). Molecular mechanisms mediating Amino acid residues that are important for Hyal2 function mammalian mitogen-activated protein kinase (MAPK) as a receptor for jaagsiekte sheep retrovirus. Retrovirology kinase (MEK)-MAPK cell survival signals. Cell Growth 2:59. Differentiation 12: 397–408. Fan H, Palmarini M, DeMartini JC. (2003). Transformation Barsky SH, Cameron R, Osann KE, Tomita D, Holmes EC. and oncogenesis by jaagsiekte sheep retrovirus. In: Fan H (1994). Rising incidence of bronchioloalveolar lung carci- (ed). Jaagsiekte sheep retrovirus and lung cancer. Curr Top noma and its unique clinicopathologic features. Cancer 73: Microbiol Immunol, Vol. 275, pp. 139–177. 1163–1170. Filippa N, Sable CL, Filloux C, Hemmings B, Van Obberghen Bellacosa A, Testa JR, Staal SP, Tsichlis PN. (1991). A E. (1999). Mechanism of protein kinase B activation by retroviral oncogene, akt, encoding a serine-threonine kinase cyclic AMP-dependent protein kinase. Mol Cell Biol 19: containing an SH2-like region. Science 254: 274–277. 4989–5000. Bonne C. (1939). Morphological resemblance of pulmonary Halbert CL, Allen JM, Miller AD. (2001). Adeno-associated adenomatosis (jaaksiekte) in sheep and certain cases of the virus type 6 (AAV6) vectors mediate efficient transduction lung in man. Am J Cancer 35: 491–501. of airway epithelial cells in mouse lungs compared to that of Cavanaugh JE. (2004). Role of extracellular signal regulated AAV2 vectors. J Virol 75: 6615–6624. kinase 5in neuronal survival. Eur J Biochem 271: Hanahan D, Weinberg RA. (2000). The hallmarks of cancer. 2056–2059. Cell 100: 57–70. Chang HW, Aoki M, Fruman D, Auger KR, Bellacosa A, Hecht SJ, Stedman KE, Carlson JO, DeMartini JC. (1996). Tsichlis PN et al. (1997). Transformation of chicken cells by Distribution of endogenous type B and type D sheep the gene encoding the catalytic subunit of PI 3-kinase. retrovirus sequences in ungulates and other mammals. Proc Science 276: 1848–1850. Natl Acad Sci USA 93: 3297–3302. Chow YH, Alberti A, Mura M, Pretto C, Murcia P, Albritton Hiatt KM, Highsmith WE. (2002). Lack of DNA evidence for LM et al. (2003). Transformation of rodent fibroblasts by jaagsiekte sheep retrovirus in human bronchioloalveolar the jaagsiekte sheep retrovirus envelope is receptor indepen- carcinoma. Hum Pathol 33: 680. dent and does not require the surface domain. J Virol 77: Hofacre A, Fan H. (2004). Multiple domains of the Jaagsiekte 6341–6350. sheep retrovirus envelope protein are required for Comoglio PM, Boccaccio C. (1996). The HGF receptor family: transformation of rodent fibroblasts. J Virol 78: 10479– unconventional signal transducers for invasive cell growth. 10489. Genes Cells 1: 347–354. Ji L, Nishizaki M, Gao B, Burbee D, Kondo M, Kamibayashi Cousens C, Bishop JV, Philbey AW, Gill CA, Palmarini M, C et al. (2002). Expression of several genes in the human Carlson JO et al. (2004). Analysis of integration sites of chromosome 3p21.3 homozygous deletion region by an Jaagsiekte sheep retrovirus in ovine pulmonary adenocarci- adenovirus vector results in tumor suppressor activities in noma. J Virol 78: 8506–8512. vitro and in vivo. Cancer Res 62: 2715–2720. Csoka AB, Frost GI, Stern R. (2001). The six hyaluronidase- Klymiuk N, Muller M, Brem G, Aigner B. (2003). Charac- like genes in the human and mouse genomes. Matrix Biol 20: terization of endogenous retroviruses in sheep. J Virol 77: 499–508. 11268–11273. Danilkovitch-Miagkova A, Duh FM, Kuzmin I, Angeloni D, Lepperdinger G, Strobl B, Kreil G. (1998). HYAL2, a human Liu S-L, Miller AD et al. (2003). Hyaluronidase 2 negatively gene expressed in many cells, encodes a lysosomal hyalu- regulates RON receptor tyrosine kinase and mediates ronidase with a novel type of specificity. J Biol Chem 273: transformation of epithelial cells by jaagsiekte sheep retro- 22466–22470. virus. Proc Natl Acad Sci USA 100: 4580–4585. Liu S-L, Duh FM, Lerman MI, Miller AD. (2003a). Role of Datta SR, Brunet A, Greenberg ME. (1999). Cellular survival: virus receptor Hyal2 in oncogenic transformation of rodent a play in three Akts. Genes Dev 13: 2905–2927. fibroblasts by sheep betaretrovirus env proteins. J Virol 77: De las Heras M, Barsky SH, Hasleton P, Wagner M, Larson E, 2850–2858. Egan J et al. (2000). Evidence for a protein related Liu S-L, Lerman MI, Miller AD. (2003b). Putative phospha- immunologically to the jaagsiekte sheep retrovirus in some tidylinositol 3-Kianse (PI3K) binding motifs in ovine human lung tumours. Eur Respir J 16: 330–332. betaretrovirus Env proteins are not essential for rodent De las Heras M, Ortin A, Cousens C, Minguijon E, Sharp JM. fibroblast transformation and PI3K/Akt activation. J Virol (2003). Enzootic nasal adenocarcinoma of sheep and goats. 77: 7924–7935. In: Fan H (ed). Jaagsiekte sheep retrovirus and lung cancer. Liu S-L, Miller AD. (2005). Transformation of madin-darby Curr Top Microbiol Immunol, Vol. 275, pp. 201–223. canine kidney epithelial cells by sheep retrovirus envelope DeMartini JC, Bishop JV, Allen TE, Jassim FA, Sharp JM, proteins. J Virol 79: 927–933. de las Heras M et al. (2001). Jaagsiekte sheep retrovirus Maeda N, Fu W, Ortin A, de las Heras M, Fan H. (2005). proviral clone JSRV(JS7), derived from the JS7 lung tumor Roles of the Ras-MEK-mitogen-activated protein kinase cell line, induces ovine pulmonary carcinoma and is and phosphatidylinositol 3-kinase-Akt-mTOR pathways integrated into the surfactant protein A gene. J Virol 75: in Jaagsiekte sheep retrovirus-induced transformation of 4239–4246. rodent fibroblast and epithelial cell lines. J Virol 79: DeMartini JC, Carlson JO, Leroux C, Spencer T, Palmarini M. 4440–4450. (2003). Endogenous retroviruses related to jaagsiekte sheep Maeda N, Inoshima Y, Fruman DA, Brachmann SM, Fan H. retrovirus. In: Fan H (ed). Jaagsiekte sheep retrovirus and (2003). Transformation of mouse fibroblasts by Jaagsiekte lung cancer. Curr Top Microbiol Immunol, Vol. 275, sheep retrovirus envelope does not require phosphatidyli- pp. 117–137. nositol 3-kinase. J Virol 77: 9951–9959.

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 800 Maeda N, Palmarini M, Murgia C, Fan H. (2001). Direct Palmarini M, Maeda N, Murgia C, De-Fraja C, Hofacre A, transformation of rodent fibroblasts by jaagsiekte sheep Fan H. (2001). A phosphatidylinositol 3-kinase docking site retrovirus DNA. Proc Natl Acad Sci USA 98: 4449–4454. in the cytoplasmic tail of the Jaagsiekte sheep retrovirus McGee-Estrada K, Fan H. (2006). In vivo and in vitro analysis transmembrane protein is essential for envelope-induced of factor binding sites in Jaagsiekte sheep retrovirus long transformation of NIH 3T3 cells. J Virol 75: 11002–11009. terminal repeat enhancer sequences: roles of HNF-3, NF-I Palmarini M, Sharp JM, de las Heras M, Fan H. (1999). & C/EBP for activity in lung epithelial cells. J Virol 80: Jaagsiekte sheep retrovirus is necessary and sufficient 332–341. to induce a contagious lung cancer in sheep. J Virol 73: McGee-Estrada K, Palmarini M, Fan H. (2002). HNF-3beta is 6964–6972. a critical factor for the expression of the Jaagsiekte sheep Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar retrovirus long terminal repeat in type II pneumocytes but M, Berman K et al. (2001). Mitogen-activated protein not in Clara cells. Virology 292: 87–97. (MAP) kinase pathways: regulation and physiological McGee-Estrada K, Palmarini M, Hallwirth C, Fan H. (2005). functions. Endocrine Rev 22: 153–183. A Moloney murine leukemia virus driven by the Jaagsiekte Philbey AW, Cousens C, Bishop JV, Gill CA, Demartini JC, sheep retrovirus enhancers shows enhanced specificity for Sharp JM. (2006). Multiclonal pattern of Jaagsiekte sheep infectivity in lung epithelial cells. Virus Genes 31: 257–263. retrovirus integration sites in ovine pulmonary adenocarci- Menzel EJ, Farr C. (1998). Hyaluronidase and its substrate noma. Virus Res 117: 254–263. hyaluronan: biochemistry, biological activities and thera- Rai SK, DeMartini JC, Miller AD. (2000). Retrovirus vectors peutic uses. Cancer Lett 131: 3–11. bearing jaagsiekte sheep retrovirus Env transduce human Miller AD, Van Hoeven NS, Liu S-L. (2004). Transformation cells by using a new receptor localized to chromosome and scattering activities of the receptor tyrosine kinase 3p21.3. J Virol 74: 4698–4704. RON/Stk in rodent fibroblasts and lack of regulation by Rai SK, Duh FM, Vigdorovich V, Danilkovitch-Miagkova A, the jaagsiekte sheep retrovirus receptor, Hyal2. BMC Cancer Lerman MI, Miller AD. (2001). Candidate tumor suppres- 4: 64. sor HYAL2 is a glycosylphosphatidylinositol (GPI)- Mornex JF, Thivolet F, De las Heras M, Leroux C. (2003). anchored cell-surface receptor for jaagsiekte sheep retro- Pathology of human bronchioloalveolar carcinoma and its virus, the envelope protein of which mediates oncogenic relationship to the ovine disease. In: Fan H (ed). Jaagsiekte transformation. Proc Natl Acad Sci USA 98: 4443–4448. sheep retrovirus and lung cancer. Curr Top Microbiol Raz DJ, He B, Rosell R, Jablons DM. (2006). Current Immunol, Vol. 275, pp. 225–248. concepts in bronchioloalveolar carcinoma biology. Clin Morozov VA, Lagaye S, Lower J, Lower R. (2004). Detection Cancer Res 12: 3698–3704. and characterization of betaretroviral sequences, related to Reddel RR, Ke Y, Gerwin BI, McMenamin MG, Lechner JF, sheep Jaagsiekte virus, in Africans from Nigeria and Su RT et al. (1988). Transformation of human bronchial Cameroon. Virology 327: 162–168. epithelial cells by infection with SV40 or adenovirus-12 Ortin A, Minguijon E, Dewar P, Garcia M, Ferrer LM, SV40 hybrid virus, or transfection via strontium phosphate Palmarini M et al. (1998). Lack of a specific immune coprecipitation with a plasmid containing SV40 early region response against a recombinant capsid protein of Jaagsiekte genes. Cancer Res 48: 1904–1909. sheep retrovirus in sheep and goats naturally affected by Rosenberg N, Jolicoeur P. (1997). Retroviral pathogenesis. enzootic nasal tumour or sheep pulmonary adenomatosis. In: Coffin JM, Hughes SH and Varmus HE (eds) Retro- Vet Immunol Immunopathol 61: 229–237. viruses. Cold Spring Harbor Laboratory Press: New York, Overbaugh J, Miller AD, Eiden MV. (2001). Receptors and pp. 475–585. entry cofactors for retroviruses include single and multiple Ruscetti SK. (1999). Deregulation of erythropoiesis by the transmembrane-spanning proteins as well as newly de- Friend spleen focus-forming virus. Int J Biochem Cell Biol scribed glycophosphatidylinositol-anchored and secreted 31: 1089–1109. proteins. Microbiol Mol Biol Reviews 65: 371–389. Sable CL, Filippa N, Hemmings B, Van Obberghen E. (1997). Palmarini M, Datta S, Omid R, Murgia C, Fan H. (2000a). cAMP stimulates protein kinase B in a Wortmannin- The long terminal repeat of Jaagsiekte sheep retrovirus is insensitive manner. FEBS Lett 409: 253–257. preferentially active in differentiated epithelial cells of the Sanna E, Sanna MP, Loddo C, Sanna L, Mura M, Cadelano T lungs. J Virol 74: 5776–5787. et al. (2002). Endogenous jaagsiekte sheep retrovirus RNA Palmarini M, Dewar P, De las Heras M, Inglis NF, Dalziel is expressed by different cell types in ovine foetus and RG, Sharp JM. (1995). Epithelial tumour cells in the lungs placenta. Eur J Histochem 46: 273–280. of sheep with pulmonary adenomatosis are major sites Santoro MM, Collesi C, Grisendi S, Gaudino G, Comoglio PM. of replication for Jaagsiekte retrovirus. J Gen Virol 76: (1996). Constitutive activation of the RON gene pro- 2731–2737. motes invasive growth but not transformation. Mol Cell Palmarini M, Fan H. (2001). Retrovirus-induced ovine Biol 16: 7072–7083. pulmonary adenocarcinoma, an animal model for lung Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. (2005). cancer. J Natl Cancer Inst 93: 1603–1614. Phosphorylation and regulation of Akt/PKB by the rictor- Palmarini M, Hallwirth C, York D, Murgia C, de Oliveira T, mTOR complex. Science 307: 1098–1101. Spencer T et al. (2000b). Molecular cloning and functional Sharp JM, Angus KW. (1990). Sheep pulmonary adenoma- analysis of three type D endogenous retroviruses of tosis: studies on its etiology. In: Hoff-Jogensen PGaR (ed). sheep reveal a different cell tropism from that of the highly Maedi-visna and related diseases. Kluwer Academic Publi- related exogenous jaagsiekte sheep retrovirus. J Virol 74: shers: Boston (MA), pp. 177–185. 8065–8076. Sharp JM, DeMartini JC. (2003). Natural history of JSRV in Palmarini M, Holland MJ, Cousens C, Dalziel RG, Sharp JM. sheep. In: Fan H (ed). Jaagsiekte sheep retrovirus and lung (1996). Jaagsiekte retrovirus establishes a disseminated cancer. Curr Top Microbiol Immunol, Vol. 275, pp. 55–79. infection of the lymphoid tissues of sheep affected by Sharp JM, Herring AJ. (1983). Sheep pulmonary adenoma- pulmonary adenomatosis. J Gen Virol 77: 2991–2998. tosis: demonstration of a protein which cross-reacts with the

Oncogene Oncogenesis by the JSRV envelope protein S-L Liu and AD Miller 801 major core proteins of Mason-Pfizer monkey virus and Wang MH, Iwama A, Skeel A, Suda T, Leonard EJ. (1995). mouse mammary tumour virus. J Gen Virol 64: 2323–2327. The murine stk gene product, a transmembrane protein Songyang Z, Shoelson SE, Chaudhuri M, Gish G, Pawson T, tyrosine kinase, is a receptor for macrophage-stimulating Haser WG et al. (1993). SH2 domains recognize specific protein. Proc Natl Acad Sci USA 92: 3933–3937. phosphopeptide sequences. Cell 72: 767–778. Wang MH, Ronsin C, Gesnel MC, Coupey L, Skeel A, Spencer TE, Mura M, Gray CA, Griebel PJ, Palmarini M. Leonard EJ et al. (1994). Identification of the ron gene (2003). Receptor usage and fetal expression of ovine endoge- product as the receptor for the human macrophage nous betaretroviruses: implications for coevolution of endoge- stimulating protein. Science 266: 117–119. nous and exogenous retroviruses. JVirol77: 749–753. Wang MH, Wang D, Chen YQ. (2003). Oncogenic and Summers C, Neill W, Dewar P, Gonzalez L, van der Molen R, invasive potentials of human macrophage-stimulating pro- Norval M et al. (2002). Systemic immune responses tein receptor, the RON receptor tyrosine kinase. Carcino- following infection with Jaagsiekte sheep retrovirus and genesis 24: 1291–1300. in the terminal stages of ovine pulmonary adenocarcinoma. Westermarck J, Li SP, Kallunki T, Han J, Kahari VM. (2001). J Gen Virol 83: 1753–1757. p38 mitogen-activated protein kinase-dependent activation Tam BY, Finnson KW, Philip A. (2003). Glycosylphos- of protein phosphatases 1 and 2A inhibits MEK1 and phatidylinositol-anchored proteins regulate transforming MEK2 activity and collagenase 1 (MMP-1) gene expression. growth factor-beta signaling in human keratinocytes. J Biol Mol Cell Biol 21: 2373–2383. Chem 278: 49610–49617. Wootton SK, Halbert CL, Miller AD. (2005). Sheep retrovirus Travis WD, Colby TV, Borrin B, Shimosato Y, Brambilla E. structural protein induces lung tumours. Nature 434: (1999). International histological classfication of tumors. 904–907. World Health Organization: Geneva (Switzerland). York DF, Vigne R, Verwoerd DW, Querat G. (1992). Tuveson DA, Jacks T. (1999). Modeling human lung cancer Nucleotide sequence of the jaagsiekte retrovirus, an in mice: similarities and shortcomings. Oncogene 18: exogenous and endogenous type D and B retrovirus of 5318–5324. sheep and goats. J Virol 66: 4930–4939. Varela M, Chow YH, Sturkie C, Murcia P, Palmarini M. Yousem SA, Finkelstein SD, Swalsky PA, Bakker A, Ohori (2006). Association of RON tyrosine kinase with the NP. (2001). Absence of jaagsiekte sheep retrovirus DNA Jaagsiekte sheep retrovirus envelope glycoprotein. Virology and RNA in bronchioloalveolar and conventional human 350: 347–357. pulmonary adenocarcinoma by PCR and RT-PCR analysis. Verwoerd DW, De Villiers EM, Tustin RC. (1980). Aetiology Hum Pathol 32: 1039–1042. of jaagsiekte: experimental transmission to lambs by means Zabarovsky ER, Lerman MI, Minna JD. (2002). Tumor of cultured cells and cell homogenates. Onderstepoort J Vet suppressor genes on chromosome 3p involved in the Res 47: 13–18. pathogenesis of lung and other cancers. Oncogene 21: Vigdorovich V, Strong RK, Miller AD. (2005). Expression and 6915–6935. characterization of a soluble, active form of the jaagsiekte Zavala G, Pretto C, Chow YH, Jones L, Alberti A, Grego E sheep retrovirus receptor, Hyal2. J Virol 79: 79–86. et al. (2003). Relevance of Akt phosphorylation in cell Wada T, Penninger JM. (2004). Mitogen-activated protein transformation induced by Jaagsiekte sheep retrovirus. kinases in apoptosis regulation. Oncogene 23: 2838–2849. Virology 312: 95–105.

Oncogene