Kaposi's sarcoma (KS) is a vascular neoplasia etiologically associated with Kaposi's sarcoma-associated herpesvirus (KSHV) infection (1
). KSHV establishes a lifelong persistent latent infection following acute infection. Reactivation of the latent virus into productive lytic replication plays a pivotal role in the initiation and progression of KS, as viral load positively correlates with KS progression. Indeed, treatment of KS patients with antiherpesvirus drugs effectively leads to regression of KS tumors (2–6
Unlike iatrogenic or AIDS-associated KS, classic KS occurs predominantly in elderly men of Mediterranean or Jewish descent who have no apparent immune suppression (7
). The exact cause of the development of classic KS remains undefined. Asthma, allergies in males, topical corticosteroid use, and infrequent bathing have been suggested to be risk factors for classic KS (8
). Multiple studies have also documented a high prevalence of classic KS in patients with diabetes mellitus (10–13
), a metabolic syndrome that manifests with elevated levels of blood glucose and episodic ketoacidosis, due to either a lack of insulin (type 1 diabetes) or cellular resistance to insulin (type 2 diabetes). High levels of KSHV DNA and seropositivity have been seen in diabetic patients (14–16
). However, no study has ever determined if diabetes is the cause or an effect of KS and whether high glucose levels play a role in the development of KS.
In the present study, we found increased levels of viral lytic gene expression when KSHV-infected primary effusion lymphoma cells were cultured in medium containing high levels of glucose. To further examine the association between high blood glucose levels and KSHV replication, we generated hyperglycemic nude mice with streptozotocin (STZ), which damages pancreatic β cells, resulting in hypoinsulinemia and hyperglycemia (17
). We then xenografted telomerase-immortalized human umbilical vein endothelial cells (18
) that are reinfected with recombinant Kaposi's sarcoma-associated herpesvirus [TIVE-KSHV (BAC16) cells] (19
) into hyperglycemic and control healthy nude mice. The original TIVE-KSHV cells are malignantly transformed and grow “KS-like” tumors in nude mice (18
). Although hyperglycemia did not seem to enhance tumor growth, the injected TIVE-KSHV (BAC16) cells expressed significantly higher levels of KSHV lytic genes in hyperglycemic mice than in normal mice. Results from cells cultured in vitro
demonstrate that high glucose levels induce the production of H2
, which was previously shown to trigger the reactivation of latent KSHV through the activation of mitogen-activated protein kinase (MAPK) pathways (20
). Interestingly, H2
also mediates the downregulation of the class III histone deacetylase (HDAC) silent information regulator 1 (SIRT1) (22
) to induce histone hyperacetylation of viral chromatins, resulting in active transcription of KSHV lytic genes. Our results suggest that H2
mediates high-glucose induction of KSHV lytic gene expression and replication through multiple mechanisms.
To our knowledge, this study provides the first experimental evidence to support an association of diabetes with the development of KS, which was suggested by previous epidemiological studies.
MATERIALS AND METHODS
Cell culture, media, and reagents.
TIVE-KSHV cells, originally infected with native KSHV (18
), were cultured in Dulbecco's modified Eagle medium (DMEM) medium plus 10% fetal bovine serum (FBS). We reinfected these cells with recombinant KSHV BAC16 to obtain TIVE-KSHV (BAC16) cells that stably express green fluorescent protein (GFP). RPMI 1640 medium without glucose was purchased from Thermo Fisher Scientific (Waltham, MA, USA). BCBL1 cells were grown in RPMI 1640 medium with 1, 3, or 6 g/liter d
-glucose plus 10% FBS. Primary human umbilical vein endothelial cells (HUVECs) were grown in EBM-2 medium with growth factor supplements (Lonza, Allendale, NJ, USA).
A mouse monoclonal antibody to the KSHV lytic protein replication and transcription activator (RTA) was a gift from the Pasteur Research Institute in Shanghai, China. A mouse monoclonal antibody to the KSHV lytic protein K8α was purchased from MyBiosource, Inc. (San Diego, CA, USA). A rat antibody to KSHV latent nuclear antigen (LANA) was purchased from Advanced Biotechnologies, Inc. (Columbia, MD, USA). A mouse monoclonal antibody to SIRT1 was purchased from EMD Millipore (Temecula, CA, USA). d-Glucose and l-glucose were purchased from Sigma-Aldrich.
Generation of hyperglycemic mice and xenografting of TIVE-KSHV (BAC16) cells.
A total of 32 athymic nude mice (4 weeks old and female) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). The blood glucose level of each mouse was measured before any treatment by using a glucose meter. The mice were then randomly separated into two groups, with one being treated with an intraperitoneal (i.p.) injection of STZ (Sigma-Aldrich) at a dose of 200 mg/kg body mass, twice a week for 2 weeks. The other group of untreated mice was used as a control. Two weeks after the last STZ treatment, the blood glucose level of each mouse from both groups was measured again to confirm the development of hyperglycemia in the treated mice. Equal numbers of TIVE-KSHV (BAC16) cells at 5 × 106
cells per injection site, with 2 sites per mouse, were then subcutaneously injected into each mouse in the abdominal region. Tumor volumes (length by width by height) were measured once a week with a caliper. At the end of the experiments, all tumors were surgically removed from the mice. All procedures were carried out in strict accordance with recommendations in the Guide for the Care and Use of Laboratory Animals
of the National Institutes of Health (23
) and according to a protocol (2011-0802) that was approved by the Institutional Animal Care and Use Committee (IACUC) at Case Western Reserve University.
Immunochemical staining and imaging.
Fresh frozen sections were prepared from the surgically removed tumors. A standard procedure for the preparation and staining of acetone-fixed frozen tissue sections was performed, using primary antibodies to KSHV small capsid protein (open reading frame 65 [ORF65]), the latent protein LANA, and control IgG. After multiple washes with phosphate-buffered saline (PBS), primary antibody-antigen signals were revealed with a biotinylated secondary antibody, streptavidin-horseradish peroxidase, and a DAB (3,3′-diaminobenzidine) detection system (BioLegend, San Diego, CA, USA). DAPI (4′,6-diamidino-2-phenylindole) was used for nuclear staining. Images were captured under a microscope (Carl Zeiss, Inc., Thornwood, NY).
Isolation of total RNAs and quantification of mRNA by qRT-PCR.
Total RNAs were isolated by using an RNA purification kit from Qiagen, which includes a step to remove residual genomic DNA prior to RNA purification. Reverse transcription (RT) of total RNA was performed by using Superscript transcriptase II (Invitrogen, Carlsbad, CA). RT-quantitative PCR (qPCR) (qRT-PCR) was conducted to quantify the levels of different viral transcripts by using primers described previously (24
). The mRNA level of the housekeeping gene β-actin was used as a reference for normalization, using the primers 5′-ATTGCCGACAGGATGCAGA-3′ (forward) and 5′-GAGTACTTGCGCTCAGGAGGA-3′ (reverse). All qRT-PCRs were carried out in triplicates.
KSHV virion production and titration.
The culture supernatants of BCBL1-BAC36 cells were collected 5 days after culturing in RPMI 1640 medium plus 10% FBS and various concentrations of d-glucose, followed by low-speed centrifugation (4,000 × g for 15 min) to remove cellular debris. To determine the relative viral titers in the supernatants, 1 ml of the supernatant was used to infect HUVECs in 6-well plates. At 72 h postinfection, cells were harvested and counted with a hemocytometer under a fluorescence microscope. The numbers of KSHV-infected GFP-positive cells and the numbers of total cells from 8 independent readings were used to calculate the average percentage of GFP-positive cells, which was used as the relative viral titer of the supernatant in question.
Chromatin immunoprecipitation assay.
Equal numbers of BCBL1-BAC36 cells (8 × 106 cells) were cultured in RPMI 1640 medium with various concentrations of d-glucose with and without catalase (400 U/ml) for 24 h, followed by fixation with 0.5% formaldehyde for 15 min. Chromatin suspensions were prepared, and chromatin immunoprecipitation (ChIP) assays were performed by using a ChIP assay kit (Invitrogen) with antibodies to RNA polymerase II (Pol II), acetylated histone H4 (H4K12-Ac), acetylated histone H3 (H3K9-Ac), histone H3 (H3), LANA, and rabbit IgG, all from Millipore, as well as a rat monoclonal antibody to LANA and rat IgG (as a control), as described above. DNAs from the input and the end ChIP products were isolated by using a DNA purification kit (Qiagen). The purified DNA was resuspended in 200 μl sterile water and used for qPCR quantification of specific viral chromatin with the primers 5′-CTCATCGTCGGAGCTGTCACACG-3′ (RTA promoter forward) and 5′-TCTCCCGATGGCGACGTGCACTAC-3′ (RTA promoter reverse) from the RTA (ORF50) promoter region.
Measurement of intracellular H2O2.
BCBL1-BAC36 cells were cultured in RPMI 1640 medium plus 10% FBS with 1, 3, and 6 g/liter d-glucose for 24 h. The cells were collected, washed twice with ice-cold PBS, and resuspended in 1× assay buffer from the OxiSelect hydrogen peroxide-peroxidase assay kit from Cell Biolabs, Inc. (San Diego, CA, USA), at a concentration of 2 × 106 cells/ml. The cells were homogenized by sonication, followed by high-speed centrifugation (10,000 × g for 15 min at 4°C). The supernatants and 1× assay buffer (as the background) were loaded into a 96-well plate for measurement of the relative levels of H2O2, with 6 repeats per sample. In parallel, a series of different concentrations (0 to 10 μM) of H2O2 were loaded into the same plate. The florescent H2O2 detection reaction was read under a fluorescence microplate reader (Bio-Tek, Winooski, VT, USA) at 560 nm (excitation) and 600 nm (emission). Upon the subtraction of each reading from that of the background, a standard curve was established with relative fluorescence units (RFU) from the different concentrations of H2O2, and the concentrations of H2O2 in the samples were determined by comparing the RFU of the samples with the standard curve.
Multiple studies have reported a high prevalence of classic KS in patients with diabetes mellitus (10–13
), and KSHV DNA was detected in >50% of type 2 diabetes patients (14–16
). These clinical studies seem to suggest that diabetes patients are more prone to KSHV infection and that diabetes is a risk factor for the development of classic KS. However, whether this metabolic syndrome truly contributes to KS tumor development has never been experimentally tested.
Type 1 diabetes results from insulin deficiency due to the lack of insulin-producing β cells in the pancreas. In contrast, type 2 diabetes occurs in adults as a consequence of the development of cellular resistance to insulin. Despite the different mechanisms, a common outcome of both types of diabetes is high glucose levels in plasma. In the present study, we found increased levels of KSHV lytic gene expression when KSHV-infected BCBL1 and TIVE-KSHV cells were cultured in medium containing diabetic levels of glucose. In full support of data from the in vitro study, TIVE-KSHV cells also displayed substantially higher expression levels of KSHV lytic genes in hyperglycemic mice than in mice with normal levels of blood glucose. These results strongly suggest that high levels of blood glucose promote the development of KS by inducing productive KSHV lytic replication.
One of the manifestations of metabolic syndromes such as obesity and diabetes is the production of excessive levels of ROS (26–31
). By using a previously established BCBL1 cell line that stably expresses the H2
sensor protein pHyper-cyto (21
), we demonstrated that cells cultured with high concentrations of glucose produce increased levels of intracellular H2
. This result was further confirmed by another intracellular H2
measurement assay. Notably, the addition of catalase, which converts H2
O and O2
, and the antioxidants NAC and glutathione effectively blocked high-glucose induction of KSHV lytic gene expression in a dose-dependent manner. Therefore, H2
mediates high-glucose induction of KSHV lytic gene expression, which further supports data from previous reports that H2
is an important physiological factor involved in the reactivation of latent KSHV (20
). Interestingly, H2
has also been shown to enhance viral entry (44–46
). It is therefore highly possible that the hyperglycemic environment in diabetic KS patients contributes to the development of KS by promoting both productive KSHV replication and recurrent de novo
Similarly to stimulation with H2
), culture of cells in medium containing high concentrations of d
-glucose also activates ERK1/2, JNK, and p38, and inhibitors of these MAPK pathways inhibit high-glucose induction of KSHV gene expression. Interestingly, we found that high glucose levels and H2
also cause the downregulation of the class III HDAC SIRT1, leading to increased levels of histone acetylation in the promoter region of the KSHV key lytic gene RTA. Thus, high glucose concentrations also engage this epigenetic mechanism to promote KSHV lytic gene expression. SIRT1 is well known for its antiaging, anti-oxidative-stress, and anti-inflammation properties (22
), and downregulation of SIRT1 has been linked to the development of diabetes (49
). Suppression of SIRT1 has been shown to trigger the reactivation of latent KSHV (42
). SIRT1 is a member of the sirtuin protein family that couples histone lysine deacetylation to NAD hydrolysis (50–52
). The dependence of SIRT1 on NAD links its enzymatic activity directly to the energy status of cells via the cellular NAD-to-NADH ratio; the absolute levels of NAD, NADH, or nicotinamide; or a combination of these variables.
The development of KS is a complex process. KSHV infection resulting from productive lytic replication plays an essential role in the initiation and progression of KS. However, in already formed KS tumors, KSHV-infected tumor cells are predominantly latent (53
). Inflammatory cytokines, stress, and ROS are known to stimulate KSHV reactivation from latency (54
). Under highly inflammatory and stressful conditions, such as diabetes, it is expected that the latent virus undergoes reactivation. In this study, we xenografted the KS tumor model cell line TIVE-KSHV (BAC16) into normal and hyperglycemic nude mice. While no obvious difference in tumor growth was seen between the two groups of mice, we found significantly higher numbers of TIVE-KSHV (BAC16) cells undergoing lytic replication in hyperglycemic mice than in normal mice. Nevertheless, the majority of TIVE-KSHV cells in tumors from hyperglycemic mice remained latently infected, suggesting that the virus might have evolved unique mechanisms to overcome highly inflammatory and stressful conditions to maintain latency. One possible mechanism might be through modulation of the cellular metabolic status. In support of this hypothesis, our recent study demonstrated that KSHV inhibits cellular aerobic glycolysis and oxidative phosphorylation by inhibiting the expression of GLUT1 and GLUT3, thus preventing overflow of the metabolic pathways and maintaining the homeostasis of latently infected and malignantly transformed cells (55
In summary, our study provides the first evidence for a link between diabetes and higher levels of KSHV replication, which may lead to the development of classic KS. Our results highlight H2O2 as the mediator of high-glucose induction of KSHV lytic replication through multiple mechanisms, which may shed light on the development of new strategies to prevent KSHV infection and KS development in diabetic patients.
The present work was supported by a grant from the Immunology Alliance Fund from Case Western Reserve University to Fengchun Ye. This work was also supported by grants from the NIH to Shou-Jiang Gao (CA096512, CA124332, CA132637, DE025465, and CA197153) and to Charles Wood (CA65903, P30 GM103509, and Fogarty D43 TW01492).
We thank Ke Lan from the Pasteur Research Institute in Shanghai, China, for providing antibodies to the KSHV lytic protein RTA. We are grateful to Jae U. Jung from the University of Southern California for providing recombinant KSHV BAC16.
We declare no conflict of interest.