Kaposi sarcoma (KS) is a tumor induced by infection with the oncogenic virus human herpesvirus-8/KS herpesvirus (HHV-8/KSHV). The endothelial cell culture model described here is uniquely suited for studying the mechanisms by which KSHV transforms host cells.
Kaposi sarcoma (KS) is an unusual tumor composed of proliferating spindle cells that is initiated by infection of endothelial cells (EC) with KSHV, and develops most often in the setting of immunosuppression. Despite decades of research, optimal treatment of KS remains poorly defined and clinical outcomes are especially unfavorable in resource-limited settings. KS lesions are driven by pathological angiogenesis, chronic inflammation, and oncogenesis, and various in vitro cell culture models have been developed to study these processes. KS arises from KSHV-infected cells of endothelial origin, so EC-lineage cells provide the most appropriate in vitro surrogates of the spindle cell precursor. However, because EC have a limited in vitro lifespan, and as the oncogenic mechanisms employed by KSHV are less efficient than those of other tumorigenic viruses, it has been difficult to assess the processes of transformation in primary or telomerase-immortalized EC. Therefore, a novel EC-based culture model was developed that readily supports transformation following infection with KSHV. Ectopic expression of the E6 and E7 genes of human papillomavirus type 16 allows for extended culture of age- and passage-matched mock- and KSHV-infected EC and supports the development of a truly transformed (i.e., tumorigenic) phenotype in infected cell cultures. This tractable and highly reproducible model of KS has facilitated the discovery of several essential signaling pathways with high potential for translation into clinical settings.
Kaposi sarcoma (KS) is a multi-focal angioproliferative tumor affecting dermal, mucosal, and visceral sites that develops most commonly in the setting of advanced immune suppression1. Four epidemiological forms have been described: classic, an indolent form that typically affects older people of Mediterranean and Middle Eastern heritage; iatrogenic, resulting from treatment with immunosuppressive drugs following organ transplantation; epidemic, an AIDS-defining cancer; and endemic, an HIV-independent form common in children in endemic regions in Africa. With the advent of effective combination anti-retroviral drug regimens for the treatment of HIV, epidemic KS is much less commonly diagnosed in developing countries. However, the clinically aggressive endemic and epidemic forms remain among the most commonly diagnosed cancers in many African countries2,3,4. Therefore, identification of effective pathogenesis-targeted drugs for treatment of KS is a research priority.
Histologically, KS lesions are characterized by extensive but abnormal neovascularization whereby spindle cells of EC origin form discontinuous vascular networks5. These abnormal vessels ("vascular slits") allow extravasation of erythrocytes, which give lesions their characteristic color. Additionally, lesions contain numerous leukocytes that characterize chronic inflammation (i.e., lymphocytes, macrophages, and plasma cells). Regression of KS lesions following immune reconstitution has been described, suggesting that KS has features of both a hyper-proliferative lesion and a true tumor6,7,8,9.
KS herpesvirus (KSHV), the causative agent of KS, was identified in 199410. Since then many in vitro cell culture models have been developed to enable pathogenesis studies, including cells explanted from tumor biopsy material and primary or telomerase-expressing EC infected with KSHV in vitro11,12,13,14,15,16,17,18. None of the currently available models fully recapitulates the KS tumor microenvironment, but all have contributed valuable knowledge to our understanding of the pathobiology of KSHV infection. Unlike the other known tumorigenic human herpesvirus Epstein-Barr virus (EBV), KSHV does not readily transform cells in culture following de novo infection19,20,21,22. However, this limitation has been overcome by transducing primary human EC of either mixed microvascular or lymphatic origin with the E6 and E7 genes from human papillomavirus type 16 prior to infection with KSHV23,24. Expression of these exogenous oncogenes dramatically increases the transforming potential of KSHV in vitro in part by providing further inhibition of the retinoblastoma protein and p5323,24. This EC transduction method has allowed multiple laboratories to identify key alterations in host cell gene expression that are induced by KSHV infection and that appear to facilitate KS cell survival and proliferation25,26,27,28,29,30,31,32. The protocols described herein are straightforward and highly reproducible, and will result in the generation of age- and passage-matched KSHV-infected EC and mock-infected controls that can be cultured for far longer than primary cells and will allow for the investigation of oncogenic mechanisms employed by KSHV. Although the protocol includes a method for production of wild type KSHV from the primary effusion lymphoma cell line BCBL-1, E6/E7-immortalized EC are also highly susceptible to infection with recombinant BACmid derived KSHV-BAC1630. Protocols for preparation of BAC16 are described elsewhere33,34.
NOTE: All procedures described in this protocol should be performed under BSL-2 conditions.
1. KSHV Stock Preparation
2. Transduction of EC with E6 and E7 Papillomavirus Genes
NOTE: We routinely use standard tissue culture flasks for growing primary EC. However, if unsatisfactory growth of EC are obtained then the use of commercial culture flasks should be considered.
3. Infection of Transduced EC with KSHV
4. Confirming Infection with KSHV by Immunofluorescence
NOTE: Detection of the KSHV latency-associated nuclear antigen (LANA-1/ORF73) provides a reliable quantitative measure of infection. Anti-LANA antibodies are commercially available.
The morphology of primary EC is classically described as "cobble stone", and this morphology is not altered by expression of the papillomavirus E6 and E7 genes (Figure 1A). Expression of the E6 and E7 genes alone does not induce a transformed phenotype; thus, cells are susceptible to contact inhibition and will cease dividing upon reaching confluence in culture. The cells will however proliferate and regrow to confluence upon trypsinization and replating at a lower density, allowing for the maintenance of age- and passage-matched cultures for use as mock-infected controls.
Infection of E6/E7-transduced EC with KSHV causes dramatic morphological changes in culture that are reminiscent of the "spindle cell" phenotype observed in KS lesions (Figure 1B). A transformed phenotype is also induced upon infection with KSHV, which is manifest in part by loss of contact inhibition when cells are cultured post-confluence (Figure 1C) and anchorage-independent growth in soft agar (Figure 2). Loss of both contact inhibition and the dependence on extra-cellular matrix interactions (anchorage) during oncogenesis are two of the hallmarks of cellular transformation36. It is important to note that primary EC are also susceptible to infection with KSHV and primary EC systems have been extremely valuable for elucidating diverse aspects of the virus-host interaction, and its consequences. Such studies are well-represented in the literature and several examples are cited herein37,38,39,40,41,42,43,44.
Immunofluorescent staining demonstrates that the spindle morphology of KSHV-infected cells is associated with expression of the viral protein LANA-1/ORF73 (Figure 3A and 3B, red). As the concentration of virus in stocks will vary between preparations, the volume of virus added per well should be adjusted so that KSHV-infected cultures reach >90% infection within one or two passages. With serial passage the percentage of infection will approach 100% and will be maintained for the duration of the culture (approximately 20 passages). Lower amounts of virus can be used if a study of paracrine influences is intended, or if adjacent uninfected cells are desirable as controls (e.g., for immunofluorescence studies involving other proteins of interest). Infected cultures also support expression of lytic viral proteins but, as is observed in KS lesions, only a minority of cells in culture will spontaneously undergo lytic reactivation (Figure 3B, green). Cells infected with KSHV-BAC16, which is tagged with GFP, also develop a spindle morphology (Figure 3C and 3D). When KSHV-BAC16 is used, viral titers can be obtained by infecting cells with a serially-diluted concentrated virus stock and evaluating cells for GFP expression at 48 h post infection using flow cytometry30.
For both recombinant and WT virus, viral genomes in stock preparations can also be measured by purifying DNA from infected cells followed by PCR quantification of KSHV DNA levels; however, one should not assume that all genomes are infectious45. If the study of the KSHV lytic cycle is intended, virus reactivation can be achieved by stimulating EC with inducing agents as described above for BCBL-1 cells, although the efficiency of reactivation is lower23. The lytic cycle can also be induced by transduction of EC with the KSHV transactivator protein ORF5046. Consistent with findings in other cell culture models of KS, both spontaneous expression and chemical induction of KSHV lytic genes in E6/E7-expressing EC decreases over time despite maintenance of a stable latent infection11,16.
Figure 1: Morphology of mock- and KSHV infected EC. (A) Mock-infected EC monolayers exhibit the classic cobble stone appearance by phase contrast vital microscopy. Without the actions of KSHV genes these cells are not transformed and will therefore cease dividing and exhibit contact inhibition upon reaching confluence in culture. (B) KSHV-infected cells, in contrast, develop an elongated "spindle cell" morphology reminiscent of KS spindle cells. This panel shows the morphology of KSHV-infected cells at day 5 PI upon reaching confluence. (C) In KSHV-infected cells, extensive changes in host cell gene expression lead to cellular transformation, which becomes evident when KSHV-infected cells are cultured without passage. Under such conditions, KSHV-infected cells continue proliferating even after reaching confluence leading to the development of multi-layered foci of cells. This panel shows the morphology of KSHV-infected cells at day 14 PI, cultured for 9 days post-confluence. Please click here to view a larger version of this figure.
Figure 2: Anchorage independent growth. Untransformed cells typically undergo apoptotic cell death following loss of contact with a substrate, a process called anoikis. (A) When E6/E7-transduced mock-infected EC, which are not transformed, are suspended in soft agar and cultured for two weeks they do not form colonies. (B) KSHV-infected EC, which are virus-transformed, are resistant to anoikis and will form colonies in soft agar. Please click here to view a larger version of this figure.
Figure 3: Demonstration of viral protein expression. (A) The spindle morphology is evident following infection with KSHV (40X magnification). (B) Immunofluorescent staining of the same field shown in (A) demonstrates that the spindle morphology is associated with expression of the viral latency protein ORF73 (red). Infected cells support lytic viral replication as well, as demonstrated by a small percentage of cells expressing the viral processivity factor ORF59 (green). (A and B modified from reference26.) The spindle morphology of cells infected with KSHV-BAC16 virus (C) is associated with expression of GFP (D) (20X magnification). Please click here to view a larger version of this figure.
Oncogenesis is a multistep process that circumvents important safeguards within an organism36. As KS lesions exist along a spectrum of chronic inflammation to true sarcomas, elucidation of certain pathophysiological processes mediated by KSHV requires that some studies be conducted in cell culture models that support transformation9. It should be noted that loss of contact inhibition and anchorage-dependent growth, phenotypes indicative of cellular transformation, do not readily develop following infection of primary EC or EC immortalized by exogenous expression of telomerase. Therefore, while the in vitro model of KS described here does not recapitulate all aspects of the KS tumor microenvironment or behavior of explanted KS cells, it is uniquely suited for studying changes in host cell gene expression induced by infection with KSHV that contribute to KS tumorigenesis.
The first reported gene expression profiling study using the culture system described herein identified the receptor tyrosine kinase c-Kit as a contributor to the transformed phenotype of KSHV-infected cells25. Knock down of c-Kit expression using siRNA or inhibition of c-Kit signaling using a dominant negative construct or tyrosine kinase inhibitor STI 571 (Imatinib) interfered with the transforming ability of KSHV manifest by loss of foci formation after prolonged culture of infected cells. Subsequent clinical evaluation of Imatinib has demonstrated efficacy of this drug in patients with epidemic KS47,48,49. Other potential treatment targets have also been identified, including heme oxygenase-126,50, CXCR727, and the beta adrenergic receptors28, all of which play a role in the growth or transformation of EC in vitro. Confirmation of these gene expression patterns in primary EC as well as KS biopsy tissue confirms the physiologic relevance of this system and underscores its value as a pre-clinical model for KS therapeutics25,26,51,52,53,54,55.
Another strength of the E6/E7 transduction system is that the extended life span of immortalized as compared to primary EC enables the long-term maintenance of age and passage-matched mock-infected control cells for comparative evaluation of the outcomes of KSHV infection. It is critical to note that in order to maximize the number of passages of EC transduced with E6 and E7 it is critical to split cultures before reaching confluence (~85 to 90%). Furthermore, these cells tolerate serum-free conditions for extended periods of time. Primary EC rapidly undergo programmed cell death upon growth factor withdrawal and therefore cannot be cultured for long periods without serum or recombinant growth factors56,57. This dependence makes autocrine signaling pathways induced by infection with KSHV difficult to identify and study, as prior to full transformation KSHV-infected cells will also undergo programmed cell death upon growth factor withdrawal (unpublished observation). The presence of the E6 and E7 proteins in this model, however, prevent programmed cell death following growth factor withdrawal and allow for prolonged (e.g., 48 h) culture in low serum or serum free medium25,26,28.
KS is a complex and unusual disease with features of both chronic inflammation and cellular transformation. The clinical need for novel therapeutic strategies is great, and the cell culture model described herein has unique properties that allow assessment of candidate drug targets.
The authors have nothing to disclose.
This work was supported by K12 HD068322 (SCM); R01 CA179921 and P51 OD011092 (AVM); and award 14PRE20320014 from the America Heart Association (SB).
BCBL-1 cells | NIH AIDS Reagent Program | 3233 |
PA317 cells | ATCC | CRL-2203 |
Neonatal dermal microvascular endothelial cells | Lonza | CC-2505 |
EBM-2 Basal Medium | Lonza | CC-3156 |
EGM-2 BulletKit | Lonza | CC-3162 |
anti-KSHV LANA/ORF 73 | Advanced Biotechnologies | 13-210-100 |
TrypLETMExpress, no phenol red | ThermoFisher | 12604013 |
RPMI | ||
DMEM | ||
PBS with calcium and magnesium |