3D culture conditions support Kaposi's sarcoma herpesvirus (KSHV) maintenance and viral spread in endothelial cells

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV) is a human tumorigenic virus and the etiological agent of an endothelial tumor (Kaposi's sarcoma) and two B cell proliferative diseases (primary effusion lymphoma and multicentric Castleman's disease). While in patients with late stage of Kaposi's sarcoma the majority of spindle cells are KSHV-infected, viral copies are rapidly lost in vitro, both upon culture of tumor-derived cells or from newly infected endothelial cells. We addressed this discrepancy by investigating a KSHV-infected endothelial cell line in various culture conditions and in tumors of xenografted mice. We show that, in contrast to two-dimensional endothelial cell cultures, KSHV genomes are maintained under 3D cell culture conditions and in vivo. Additionally, an increased rate of newly infected cells was detected in 3D cell culture. Furthermore, we show that the PI3K/Akt/mTOR and ATM/γH2AX pathways are modulated and support an improved KSHV persistence in 3D cell culture. These mechanisms may contribute to the persistence of KSHV in tumor tissue in vivo and provide a novel target for KS specific therapeutic interventions. KEY MESSAGES: In vivo maintenance of episomal KSHV can be mimicked in 3D spheroid cultures 3D maintenance of KSHV is associated with an increased de novo infection frequency PI3K/Akt/mTOR and ATM/ γH2AX pathways contribute to viral maintenance.

Keywords: 3D culture; Episomal viral genomes; KSHV infected endothelial cells; Viral maintenance; Xenograft model.

Fig. 1

KSHV loss in endothelial cell culture is not the result of cell proliferation a To assess the dependence of the proliferation capacity of HuARLT cells on doxycycline, the cells were cultured in the presence or absence of doxycycline for 3 days followed by fixation and the staining with an anti-Ki-67 antibody (red). Nuclei were visualized by DAPI (blue), scale bar 100 μm. Representative images upon immunofluorescence staining are shown from one out of two independent experiments. For quantification, the Ki-67/DAPI ratio was assessed in 3 biological replicates with 3 fields of view per replicate b The percentage of GFP positive cells was measured by flow cytometry of rKSHV-HuARLT cells cultured with and without puromycin and doxycycline for up to 33 days as indicated. GFP is encoded by episomal KSHV. The figure depicts one representative experiment out of more than 5 experiments performed c The viral copy number of rKSHV-HuARLT cells was assessed by qPCR after 14 days of 2D culture in absence of puromycin and with or without doxycycline as indicated. Data from rKSHV-HuARLT cells cultured in presence of puromycin are shown as a control. Depicted are the results of one out of 4 experiments with 3 biological and 3–4 technical replicates. d The viral load was assessed by counting LANA dots per cell upon immunofluorescence staining of rKSHV-HuARLT cells cultured for 14 days in the absence of puromycin and with or without doxycycline as indicated. As control, viral copy numbers of rKSHV-HuARLT cells cultured for 14 days in presence of doxycycline and puromycin are shown. Three replicates were analyzed per condition and 3 independent fields of view were analyzed per replicate. The experiment was performed twice

Fig. 2

Viral maintenance in vivo and in 3D cell culture. a rKSHV-HuARLT cell spheroids were transplanted to RAG^−/−gC^−/− mice. Twenty-eight days after transplantation, plugs were reisolated and stained for hematoxyline and eosine (H&E), for ALU-positive nuclei (ALU), and for GFP and LANA expression. Representative immunohistochemistry sections are shown, magnification × 250. The experiment was performed more than three times. b To assess viral maintenance upon transplantation, cells were reisolated from plugs day 28 post transplantation and stained for human vimentin. The frequency of GFP⁺ cells within the vimentin⁺ population was assessed by flow cytometry (d28); it reflects the frequency of KSHV positive cells. An aliquot of the cells before transplantation (d0) is presented as a control. The experiment was performed twice, using 3 mice with two plugs containing 10–20 lesions each. c The number of viral copies per cell was determined by qPCR upon isolation from xenograft-transplanted mice on day 28 (d28). The viral copy number was assessed from KSHV-HuARLT cells cultured in 2D and 3D conditions in the presence of doxycycline and in the absence of puromycin for 14 and 28 days as indicated. The viral copy number of KSHV-HuARLT cells in 2D cultures in presence of puromycin is shown as a control. The figure compiles the results of 2 in vivo experiments (3 mice and 1–2 plugs per mouse) and 3 in vitro experiments with 3 replicates. Note that a prolonged cultivation of spheroids in 3D beyond 14 days was not possible because of matrix degradation which compromised stability of the spheroids

Fig. 3

KSHV maintenance in 3D cell culture depends on de novo infection. a To assess de novo infection, KSHV-free BFP-HuARLT cells (BFP in the nucleus) were co-cultured with rKSHV-HuARLT cells carrying KSHV-encoded GFP in a ratio of 10:1 for 14 days in 2D in presence of doxycycline and in absence of puromycin. KSHV-infected cells were assessed upon staining for LANA and by monitoring GFP and BFP fluorescence. Draq5 was used to lable nuclear DNA. Fluorescence microscopic pictures using specific filteres are shown for a representative section; scale bar 20 μm. Representative pictures are depicted. b KSHV-free BFP-HuARLT were cocultured with rKSHV-HuARLT cells (carrying KSHV-encoded GFP) in indicated ratios in 3D or 2D cell culture for 3 days. The percentage of newly infected BFP⁺GFP⁺ cells out of the BFP⁺ cells was determined by flow cytometry. The figure compiles the results of 4 and 5 independent experiments with technical triplicates for the 2D and 3D conditions, respectively. See Fig. S5 for flow cytometry plots of a representative experiment. c The percentage of newly infected BFP + GFP+ cells out of BFP+ cells was determined on day 3 and day 14 of cocultivation in 3D condition based on flow cytometry analysis. See Fig. S5 for flow cytometry plots of a representative experiment. d The viral copy number was determined by qPCR upon cultivation of rKSHV-HuARLT cells in 2D and 3D conditions for 14 days in presence of 100 μM phosphonoformic acid (PhA) and normalized to the copy number of cells in the same culture condition in absence of PhA (Control). Non-infected cells (Non-inf) were used as a negative control. n.d., not detected. The graph compiles the data of 2 independent experiments. A representative experiment is shown in Fig. S6

Fig. 4

Role of the PI3K/mTOR and ATM/γH2AX pathways in KSHV maintenance. a Volcano plot showing differentially expressed genes in rKSHV-HuARLT cells upon cultivation in 2D conditions or in 3D spheroids for 3 days. The genes upregulated in 3D are shown in green; downregulated genes are shown in red. The data are based on one RNA sequencing experiment with two replicates per condition. b Top 10 canonical pathways enriched in KEGG pathways by differentially expressed genes from the RNA-Seq analysis of rKSHV-HuARLT cell cultured either in 2D or 3D cell culture for 3 days. c rKSHV-HuARLT cells were cultivated in 2D and 3D conditions for 14 days in absence of puromycin and doxycycline and in presence or absence of the following compounds: 2.5 μM LY294002 (PI3Ki), 2.5 μg/ml rapamycin (mTORi), 2.5 μg/ml FK506, 10 μM U0126, 2.5 μM Bay7085 (NF-kBi), 50 μM DAPT, and 5 μM manumycin. Subsequently, the relative viral copy number was measured by qPCR. The viral load in the respective non-treated control samples was set to 1 to account for the regular loss of viral copies during cultivation. Depicted are the compiled results of four independent experiments with 3 biological and 3 technical replicates. See Fig. S8B for a representative experiment. c rKSHV-HuARLT cells were cultivated in 2D and 3D conditions for 14 days in absence of puromycin and doxycycline and in presence or absence of the following compounds: 10 μM Ku55933 (ATMi), 1 μM Nu7441 (DNA-PKi), 10 μM mirin (MRNi), or 10 μM VE-821 (ATRi). Subsequently, the relative viral copy number was measured by qPCR. Viral load in the respective non-treated control samples was set to 1 to account for the regular loss of viral copies during cultivation. Depicted are the compiled results of four independent experiments with 3–4 biological and 3 technical replicates. See Fig. S8C for a representative experiment. d The abundance of the MRN complex proteins MRE11A, RAD50, and NBN was analyzed by Western blot analysis of rKSHV-HuARLT cells cultured either in 2D conditions or in 3D spheroids for 3 days. A representative blot out of 5 independent experiments is shown. For statistical analysis of RAD50 abundance, see Fig. S9. e The presence of γH2AX was analyzed by Western blot of HuARLT and rKSHV-HuARLT cells cultured either in 2D conditions or in 3D spheroids for 3 days (Ctrl). The abundance of γH2AX in rKSHV-HuARLT cells was evaluated in standard conditions (Ctrl) or upon 3d treatment with 2.5 μM LY294002 (PI3Ki), 10 μM Ku55933 (ATMi), and 2.5 μg/ml rapamycin (mTORi). A representative blot out of 2 and 3 independent experiments for 2D and 3D conditions, respectively, is shown