Long-term-infected telomerase-immortalized endothelial cells: a model for Kaposi's sarcoma-associated herpesvirus latency in vitro and in vivo

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV) is associated with Kaposi's sarcoma (KS), primary effusion lymphoma (PEL), and multicentric Castleman's disease. Most KS tumor cells are latently infected with KSHV and are of endothelial origin. While PEL-derived cell lines maintain KSHV indefinitely, all KS tumor-derived cells to date have lost viral genomes upon ex vivo cultivation. To study KSHV latency and tumorigenesis in endothelial cells, we generated telomerase-immortalized human umbilical vein endothelial (TIVE) cells. TIVE cells express all KSHV latent genes 48 h postinfection, and productive lytic replication could be induced by RTA/Orf50. Similar to prior models, infected cultures gradually lost viral episomes. However, we also obtained, for the first time, two endothelial cell lines in which KSHV episomes were maintained indefinitely in the absence of selection. Long-term KSHV maintenance correlated with loss of reactivation in response to RTA/Orf50 and complete oncogenic transformation. Long-term-infected TIVE cells (LTC) grew in soft agar and proliferated under reduced-serum conditions. LTC, but not parental TIVE cells, formed tumors in nude mice. These tumors expressed high levels of the latency-associated nuclear antigen (LANA) and expressed lymphatic endothelial specific antigens as found in KS (LYVE-1). Furthermore, host genes, like those encoding interleukin 6, vascular endothelial growth factor, and basic fibroblast growth factor, known to be highly expressed in KS lesions were also induced in LTC-derived tumors. KSHV-infected LTCs represent the first xenograft model for KS and should be of use to study KS pathogenesis and for the validation of anti-KS drug candidates.

FIG. 1.

Immunohistochemical analysis of hTert-immortalized HUVEC in comparison to primary HUVEC; staining for CD31, CD34, CD45, CD68, CD105, Flt, keratin, SMSA, UEA, S100A10, and factor VIII (vWF) expression. The left-hand panels show primary HUVEC at passage 2. The right-hand panels show TIVE cells. TIVE cells strongly express five endothelial-cell-specific markers (CD31, CD105, Flt, UEA, and factor VIII), but not CD45, keratin, or SMSA, which shows slight background staining compared to the primary antibody control (Negative mouse). Note the slightly more elongated morphology of TIVE in contrast to the characteristic cobblestone morphology of HUVEC.

FIG. 2.

TIVE cells are susceptible to KSHV and support lytic replication early after infection. (A) LANA IFA on TIVE cells 48 h postinfection with BCBL-1-derived KSHV. (B) Threefold serial dilutions of BCBL-1-derived cell-free virion preparations. The error bars indicate standard deviations. (C) KSHV-infected TIVE cells can be induced to lytic replication. TIVE cells were infected with KSHV or mock infected; 48 h later, the cells were infected with Ad-Orf50, and cell lysates were analyzed by Western blot analysis 72 h later. KSHV-infected SLK cells were used as positive controls. All KSHV-infected cells expressed LANA (lanes 1, 3, and 4). KSHV-infected and -induced TIVE cells expressed significantly more K8.1 than SLK cells. Tubulin was used as a loading control.

FIG. 3.

KSHV-infected TIVE cells lose viral genomes over time. TIVE cells were infected with BCBL-1-derived cell-free KSHV as described in Materials and Methods. At the indicated time points, the cells were analyzed for the expression of LANA and K8.1 and for the presence of viral DNA. (A) IFA for LANA and K8.1. The percentage of LANA-positive cells declined from greater than 95% at week 1 to less than 10% at week 4. A significant percentage of the cells expressed K8.1 at 1 week postinfection; this expression was undetectable at week 4. (B) Detection of viral DNA by PCR. Genomic DNA extracted at the indicated time points was amplified using 25 cycles. The DNA copy number decreased between weeks (W) 4 and 10.

FIG. 4.

KSHV-infected TIVE cells support long-term latency. (A) LANA IFA on KSHV-infected TIVE cells 5 and 8 weeks and 10 months postinfection. IFA at 10 months was analyzed by confocal microscopy and double stained with anti-LANA antibodies (green) and propidium iodide (red). (B) LANA detection by Western blot analysis in TIVE cells (lane 1) or KSHV-infected TIVE cells 3 months (lane 2) and 10 months (lane 3) postinfection. (C) Western blot analysis of KSHV-infected TIVE cells that were grown in the presence of 0.5 mM PFA for eight consecutive passages; as a loading control, membranes were stripped and incubated with an α-tubulin-specific antibody (P indicates passage numbers).

FIG. 5.

KSHV-infected TIVE cells display phenotypic changes indicative of transformation. (A) Colony formation assay of TIVE cells or LTC at 3 months and 10 months postinfection; 10,000 cells each were seeded as single-cell suspensions into semisolid media containing EMEM and 5% FCS, and colony growth was scored 2 weeks postinfection. (B) LTC divide faster than TIVE cells. Comparative cell cycle analysis after propidium iodide staining of TIVE cells and LTC. A significantly higher proportion of LTC (39%) than uninfected TIVE cells (10%) were in S phase. (C) Gardella gel analysis of LTC. Five LTC clones established from colonies shown in panel A were analyzed for the presence of episomal KSHV genomes. TPA-induced BCBL-1 cells were loaded as controls to indicate circular and linear genome migration, as previously described (24).

FIG. 6.

Comparative genomewide gene expression profiling of LTC and SLK and BCBL-1 cells during latency and after induction of lytic replication; genomewide real-time RT-PCR analysis of LTC-derived tumors. (A) Cluster analysis of multiple experiments, comparing the induction profiles of orf50 and TPA for SLK (lane 1), BCBL-1 (lane 2), and two different LTC (lanes 3 and 4). Each cell line is represented by three columns: uninduced, TPA treated, and Ad-Orf50 infected. The arrows denote the most induced mRNAs, and the bars indicate mRNAs that are not induced. The grayscale indicates the relative level of transcription, normalized to GAPDH. Black indicates the most abundant mRNAs on a log₂ scale. (B, C, and D) Comparative gene expression analysis of BCBL-1 and SLK cells and LTC in response to TPA or Ad-Orf50. The vertical axis indicates dC_T values, normalized to GAPDH, for actin (open triangles), orf50 (open circles), and orf57 (closed squares). A decrease in deltaC_T represents increased levels on a log₂ scale. Panel C plots dC_T values on the vertical axis for mock-treated (gray squares), TPA-treated (open circles), and Ad-Orf50-treated (black triangles) cells relative to mock-treated cells on the horizontal axis for BCBL-1 cells (B), LTC (C), or SLK cells (D). (F) LTC do not efficiently reactivate from latency. Western blot analysis for K8.1 on BCBL-1 cells and long-term-infected LTC clones. LTC clones or BCBL-1 cells as controls were infected with recombinant Ad-Orf50 for 5 days or treated with TPA for 48 h. The cell lysates were tested for K8.1 expression. K8.1 was highly induced in either TPA- or Ad-Orf50-infected cells (lanes 2 and 3). LTC did not express any detectable level of K8.1 (shown is one representative result from a total of eight similar experiments). (E) LTC are susceptible to adenovirus infection. TIVE cells express GFP 48 h postinfection with Ad-GFP.

FIG. 7.

LTC are highly tumorigenic in NUDE mice. Uninfected TIVE cells or LTC (10⁵ cells in growth factor-depleted Matrigel) established at 3 or 10 months postinfection were injected subcutaneously into nude mice. (A, B, and C) Mice injected with LTC at 3 (5/5) and 10 (5/5) months developed tumors, while none of the control mice injected with TIVE cells (0/3) did. (D) Graph indicating the size and distribution of resulting tumors. Pairwise comparisons using sum-rank statistics demonstrate statistical significance. Boxes represent interquartile ranges; lines within boxes represent the medians; T bars indicate highest and lowest observed values.

FIG. 8.

LTC-derived tumor cells express LANA and LYVE-1 and show a more permissive viral expression pattern than LTC grown in vitro. LTC-derived tumors were dissected and analyzed for protein and viral-mRNA expression. (A) Hematoxylin and eosin staining of L1 tumor displaying a mixture of elongated-spindle-cell and undifferentiated morphologies with prominent mitotic figures. The inset shows tumor cells closely surrounding a blood vessel. The arrowheads indicate erythrocytes extravasated into the tumor. (B and C) Immunohistochemical detection of LANA, LYVE-1, and PCNA. Shown is one representative panel out of five tumor samples. Panel C shows an enlarged tissue section from panel B to emphasize typical LANA speckled staining. (D) Genomewide real-time RT-PCR analysis of LTC-derived tumors. The graph shows mean expression levels from pooled mRNA samples taken from a total of five LTC-derived tumors (two at 3 months and 3 at 10 months postinfection). The data are plotted as dC_T in comparison to GAPDH and LTC. A much broader gene expression pattern in tumors than in LTC grown in vitro is shown.