If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Curriculum in Genetics and Molecular Biology, Lineberger Comprehensive Cancer Center, Center for AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
Address reprint requests to Dirk Dittmer, Ph.D., Department of Microbiology and Immunology, Lineberger Comprehensive Cancer Center, Center for AIDS Research, University of North Carolina at Chapel Hill, CB# 7290, 715 Mary Ellen Jones Bldg., Chapel Hill, NC 27599-7290
Curriculum in Genetics and Molecular Biology, Lineberger Comprehensive Cancer Center, Center for AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, North CarolinaDepartment of Microbiology and Immunology, Lineberger Comprehensive Cancer Center, Center for AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
Primary effusion lymphoma (PEL) is a non-Hodgkin's B-cell lymphoma driven by Kaposi's sarcoma-associated herpesvirus. It is uniquely sensitive to mTOR, PI3K, and Akt inhibitors; however, the basis of this requirement for the mTOR pathway remains to be elucidated. The phosphatase and tensin homolog gene (PTEN) on chromosome 10 controls the first step in the phosphatidylinositol 3 kinase (PI3K)-Akt-mammalian target of rapamycin (mTOR) pathway and is genetically inactivated in many solid tumors. We find an absence of PTEN deletions, mutations, or protein mislocalization in PEL. However, we find consistent hyperphosphorylation at serine 380 of PTEN, which is an inactivating modification, in PEL cell lines and in tumor xenografts. We also evaluated a large tissue microarray of Kaposi's sarcoma biopsies and observed concordant high levels of phospho-PTEN, phospho-Akt, and phospho-S6 ribosomal protein. Reintroduction of PTEN into PEL inhibited colony formation in soft agar, verifying the functional dependence of PEL on PI3K signaling. This was also true for PEL cell lines that carried mutant p53 and for KS-like cell lines. Activating PTEN in these cancers may yield a new treatment strategy for PEL, KS, and similar PTEN wild-type lymphomas.
Primary effusion lymphoma (PEL) was initially identified as an AIDS-associated non-Hodgkin's lymphoma with a poor prognosis.
KSHV, known formally as human herpesvirus 8 (HHV-8), shows primary tropism for B cells and endothelial cells. In addition to KSHV infection, PEL cells can be coinfected with Epstein-Barr virus [EBV; formally known as human herpesvirus 4 (HHV-4)].
KSHV is also the causative agent of Kaposi's sarcoma (KS), a tumor of endothelial cell lineage. KS is characteristically seen in HIV-infected patients, but also occurs with reduced frequency in other settings of immune suppression (eg, in solid organ transplant recipients).
In addition, KS is found in elderly patients in the absence of overt immune deficiency (classic KS) and in children in KSHV-endemic regions, such as Sub-Saharan Africa. Stallone et al
reported on a series of transplant patients who developed KS. On switching from cyclosporine to another immunosuppressant, rapamycin, the KS lesions regressed; however, the same patients maintained normal graft function, because rapamycin maintained T cell-targeted immunosuppression
Thus, there is a need to understand the molecular mechanism of both rapamycin and its analogs in PEL and KS, and to evaluate potential biomarkers that could help in predicting response.
Rapamycin binds to and inhibits the activity of mammalian target of rapamycin complex-1 (mTORC1),
a downstream effector of the phosphatidylinositol-3-kinase (PI3K) signaling cascade. Phosphorylation of receptor tyrosine kinases results in activation of the p110 catalytic subunit of phosphatidylinositol 3 kinase (PI3K), which then phosphorylates membrane-associated phospholipid phosphatidylinositol-4,5-biphosphate (PIP2) to become phosphatidylinositol-3,4,5-triphosphate (PIP3). PIP3 recruits the protein serine-threonine kinase Akt to the plasma membrane. Akt is phosphorylated on serine 473 by mTORC2 and by PI3K-dependent kinase 1 (PDK1) on threonine 308. Activated Akt then phosphorylates downstream targets, mTORC1 being one of its effectors.
The protein encoded by the phosphatase and tensin homolog gene (PTEN), located on band q23 of chromosome 10, constitutes a key negative regulator of the PI3K-mTOR signaling pathway. It can revert phospholipid PIP3 to PIP2, thereby counteracting the activation of Akt by PDK1.
Cell cycle-dependent nuclear export of phosphatase and tensin homologue tumor suppressor is regulated by the phosphoinositide-3-kinase signaling cascade.
PTEN expression and phosphorylation status thus serve as markers for PTEN activity. Various solid tumors that require PI3K-mTOR signaling have either a deletion or mutation in PTEN
; however, the status of PTEN in PEL and KS remains a subject of investigation. Our research group has previously reported that rapamycin inhibits PEL in culture, in a mouse model, and in a patient.
These findings suggest that PEL belongs to a class of cancers, such as Mantle Cell Lymphoma, that depend on PI3K-Akt-mTOR signaling. KSHV proteins may activate mTORC1 through cellular receptor engagement (by vIL6) or through PI3K activation (by viral proteins such as vGPCR and K1),
further corroborating the dependence of KSHV-infected cells on PI3K-mTOR signaling. In the present study, we examined whether, in addition to viral factors, host genetic aberrations in PTEN contribute to the required activity of the PI3K-Akt-mTOR pathway in PEL and KS.
Materials and Methods
Cell Culture
B-cell lines were cultured in RPMI Medium 1640 supplemented with 100 μg/mL streptomycin sulfate, 100 U/mL penicillin G (Invitrogen Life Technologies, Carlsbad, CA), 2 mmol/L l-glutamine, 0.05 mmol/L 2-mercaptoethanol, 0.075% sodium bicarbonate, 1 U/mL IL-6 (PeproTech, Rocky Hill, NJ), and 10% fetal bovine serum (Mediatech, Manassas, VA) at 37°C in 5% CO2. The cell lines are characterized in Table 1.
Table 1Summary of PEL Cell Lines, with Literature Citations
In vitro establishment and characterization of two acquired immunodeficiency syndrome-related lymphoma cell lines (BC-1 and BC-2) containing Kaposi's sarcoma-associated herpesvirus-like (KSHV) DNA sequences.
In vitro establishment and characterization of two acquired immunodeficiency syndrome-related lymphoma cell lines (BC-1 and BC-2) containing Kaposi's sarcoma-associated herpesvirus-like (KSHV) DNA sequences.
Establishment and characterization of a primary effusion (body cavity- based) lymphoma cell line (BC-3) harboring Kaposi's sarcoma-associated herpesvirus (KSHV/HHV-8) in the absence of Epstein-Barr virus.
Establishment and characterization of an Epstein-Barr virus (EBC)-negative lymphoblastoid B cell line (BJA-B) from an exceptional, EBV-genome-negative African Burkitt's lymphoma.
Genomic DNA was isolated from the PEL cell lines using a Wizard genomic DNA purification kit (Promega, Madison, WI) according to the manufacturer's protocol. Genomic profiles were obtained by hybridizing the DNA to GeneChip 250K Nsp and 250K Sty human mapping arrays (Affymetrix, Santa Clara, CA), which contain markers for copy number determination, in addition to single nucleotide polymorphisms. All data were analyzed using the Partek Genomics Suite version 6.0 (Partek, St. Louis, MO). Raw data were imported and adjusted for background using a robust multiarray algorithm. Copy numbers were assessed using the software suite's unpaired analysis and were compared with baseline created using normal tonsil control DNA. Copy number variation was represented in both the heat map and the dot-plot profile along the length of the individual chromosomes using a visualization scheme from the Partek software suite.
PCR and Sequencing
Genomic DNA was isolated using a Wizard genomic DNA purification kit (Promega). PCR amplification was performed using GoTaq PCR master mix (Promega). Specific primers, flanking individual exons, were as given in Table 2. The amplified fragments were subjected to Sanger sequencing. mRNA expression was determined using cDNA subject to RT-PCR, using primers as given in Table 2. cDNA was synthesized from total RNA isolated from PEL cells using TRIzol extraction and an ABI high capacity cDNA reverse transcription kit (Invitrogen Life Technologies) according to the manufacturer's protocol.
Table 2Primer Sequence and Annealing Temperature for PCR Analysis, with Literature Citations
The expression of PAX6, PTEN, vascular endothelial growth factor, and epidermal growth factor receptor in gliomas: relationship to tumor grade and survival.
The primary antibodies used were rabbit PTEN-specific and phospho-PTEN-specific (S380) (Cell Signaling Technology, Danvers, MA) diluted 1:1000 in blocking solution. Horseradish peroxidase-conjugated secondary antibody (Vector Laboratories, Burlingame, CA) was diluted 1:5000 and blots were visualized using Pierce ECL Western blotting substrate (Thermo Scientific, Rockford, IL). We used horseradish peroxidase-conjugated actin-specific monoclonal antibody (Abcam, Cambridge, MA) at 1:10,000 dilution as control.
Immunofluorescence Analysis
Cells were cultured overnight on glass coverslips in six-well plates (Falcon; BD Biosciences, San Jose, CA). They were washed in PBS and stained for immune fluorescence as described.
Primary antibodies used were rabbit PTEN-specific, rabbit phospho-PTEN-specific (S380) (Cell Signaling Technology) diluted 1:100 and mouse LANA-specific (Leica Microsystems-Novocastra Laboratories, Newcastle, UK), diluted 1:600. Images were captured with a Leica DM4000B fluorescence microscope (Leica Microsystems, Wetzlar, Germany) equipped with a 63/1.4 to 0.6 numerical aperture objective and a Q-Imaging Retiga 2000RV camera. Raw single microscopy images were deconvoluted using SimplePCI (Hamamatsu, Sewickley, PA) two-dimensional blind deconvolution, which iteratively applies an AutoQuant Imaging (Media Cybernetics, Bethesda, MD) proprietary algorithm to remove blur and generate high-clarity images. Images were stored as TIFF files.
Tissue Microarray
Unstained and LANA-stained tissue microarrays were obtained from the AIDS and Cancer Specimen Resource.
Solid tumors were removed from SCID mice injected subcutaneously with BC-1 cells in growth factor-reduced Matrigel and fixed in 5 mL 10% neutral buffered formalin (Fisher Diagnostics, Pittsburgh, PA) for 2 days, embedded in paraffin, and cut into sections (7 μm thick). Sections were deparaffinized and rehydrated using HistoChoice clearing agent (Sigma-Aldrich, St Louis, MO) and then incubated in 3% hydrogen peroxide diluted in 10% methanol to block the endogenous peroxidase activity. For antigen retrieval, samples were heated in a microwave oven for 20 minutes in Retrivagen A (pH 6.0) solution (BD Biosciences) and blocked with 10% horse serum (Vector Laboratories) diluted in PBS supplemented with 5% bovine serum albumin and 0.3% Triton X-100. Samples were then incubated with primary antibody diluted 1:100 (except 1:500 for Ki67) in blocking buffer at room temperature for 1 hour and then, after a series of PBS washes, were developed with VectaStain ABC kit (Vector Labs). Sections were developed with Vector NovaRed substrate (Vector Labs) for 5 to 10 minutes and counterstained with hematoxylin. Sections were dehydrated using an alcohol series and HistoChoice clearing agent washes and then mounted using CytoSeal mounting medium (Richard-Allan Scientific, Kalamazoo, MI). Primary antibodies were Ki67 (Novocastra Laboratories) phospho-PTEN serine 380 (S380) residue (Millipore, Billerica, MA), phospho-Akt threonine 308 (T308) residue, and phospho-S6-ribosomal protein serine 235/236 (S235/236) residues (Cell Signaling Technology). Sections were imaged using a Leica DM LA histology microscope equipped with a 10/0.25 numerical aperture or a 40/0.75 numerical aperture N Plan objective and a Leica DPC 480 camera. Images were stored as TIFF files under Macintosh OS X10.5. Image intensity was quantified using FRIDA software (Framework for Image Dataset Analysis; available at http://bui2.win.ad.jhu.edu/frida). Data were normalized to intensity per core, and the square root of the intensity count was used for linear regression analysis. Linear regression was conducted using GNU open-source R statistical software (version 2.12.1; available at http://www.r-project.org).
Growth Suppression Assay
The human PTEN cDNA pORF-hPTEN was obtained from InvivoGen (San Diego, CA) and cloned into pMONO-blasti-mcs, pDD1538 (InvivoGen), under the control of the constitutive ferritin promoter to yield pMONO-blasti-hPTEN. The construct was verified by restriction analysis and two independent clones, pDD1540 (clone 1) and pDD1541 (clone 7), were used for further analysis. pMONO-blasti-mcs carries the blasticidin selection marker. DNA was introduced into PEL cells (5 × 106 to 7 × 106) by nucleofection, according to the manufacturer's protocol (Lonza, Basel, Switzerland; Walkersville, MD). Nucleofected cells were recovered in complete medium overnight and suspended in 0.3% agar before plating on 0.5% soft agar in complete RPMI Medium 1640 supplemented with 6 μg/mL of blasticidin (selection marker) and growth was assessed 2 weeks later. In the case of KS-like endothelial cells, cells were transfected using SuperFect transfection reagent (Qiagen, Valencia, CA) according to the manufacturer's protocol and plated on 10-cm dishes. Colonies were allowed to form over a period of 2 weeks. Colonies were then stained with 200× Magic Stain (3 g Crystal Violet and 0.8 g ammonium oxalate in 20% ethanol) diluted to approximately 10× in water and counted. Data were plotted using the R statistics environment (v2.12.1) and significance was determined using the nonparametric Wilcoxon signed rank test.
Results
The PTEN Gene Is Wild Type in PEL
To determine whether PEL carries gross genomic abnormalities at the PTEN locus, we performed copy number analysis using the Affymetrix 500K array for a large set of PEL cell lines (GEO accession ID GSE28684). We did not detect any loss of heterozygosity or amplifications or deletions associated with the PTEN locus on chromosome 10, with the exception of the BC2 cell line (Figure 1A). On chromosome 3, which served as control; all PEL cell lines showed a loss specifically at a common fragile site locus encoding the fragile histidine triad gene (FHIT), in an otherwise normal region (Figure 1B). Amplification of the short p-arm of chromosome 3 was observed in 4/10 PELs. The normal distribution of markers along chromosome 10 indicates the absence of any gross genomic aberrations of the 10-kb PTEN locus in PEL.
Figure 1PTEN is wild type in PEL. A and B: Using an Affymetrix 500K single nucleotide polymorphism array, we observed that the majority of PELs do not show genomic gain or loss at the PTEN locus on chromosome 10 (A). Chromosome 3 served as control; distinct loss was noted at the FHIT locus, a known common fragile site gene (B). Heat maps indicate loss (blue) and gain (red). The dot profiles show the distribution of markers along the chromosome; an asterisk marks a small region that is specifically lost in an otherwise normal chromosomal region. On chromosome 10 (A), this does not correspond to any known transcripts; on chromosome 3 (B), it corresponds to FHIT. The labeled cytobands indicate location of PTEN (A) and FHIT (B). C and D: Genomic DNA was isolated from individual PEL cell lines and each of the exons was amplified using specific primers (listed in Table 2). C: Schematic of the organization of the primers used. D: A representative PCR gel showing detection of all nine exons in BCBL-1 cells.
To further investigate the status of the PTEN locus, we amplified each of the nine exons of the PTEN gene (Figure 1C) individually from 10 PEL and 2 non-PEL lymphoma cell lines. This represents a majority of all PEL cells lines that can successfully be grown in culture and form tumors in immunodeficient mice (Refs.
and D.R., unpublished data). All nine exons were present in the BCBL-1 cell line (Figure 1, C and D). We found no gross deletions in any of the PTEN exons in any of the 10 independent PEL cell lines. We sequenced each of the PTEN exons for the individual PEL cell lines and were unable to detect any aberrations in the genomic sequence. These data have been submitted to GenBank (submission ID 1367943).
The PTEN Protein Is Expressed and Hyperphosphorylated in PEL
Because PTEN can be epigenetically silenced in some cancers,
we wanted to determine whether PTEN mRNA and PTEN protein are expressed in PEL. Using RT-PCR, we detected PTEN mRNA in all of our PEL cell lines (Figure 2). These data led us to exclude the possibility of promoter silencing of the PTEN locus in PEL.
Figure 2PTEN is phosphorylated at the serine 380 site in PEL. RT-PCR was performed to determine PTEN mRNA expression in PEL; representative gels are shown for PTEN (A) and GAPDH (B). NT, nontemplate control. C: Immunoblots were performed on protein isolated from whole-cell lysate; representative immunoblots are shown for total PTEN and phospho-PTEN (S380). β-actin was used as loading control. Equivalent amounts of protein were loaded in each case, except for HEK293, which served to indicate the molecular weight of PTEN.
Next, we used Western blot analysis to detect total PTEN and PTEN phosphorylated at serine 380. This modification renders the protein inactive and unable to dephosphorylate PIP3 to PIP2.
PTEN protein was detectable at the appropriate molecular weight (∼54kDa) in every PEL cell line tested at similar levels to those observed in HEK293-positive control cells. PTEN was present in a majority of the cell lines (Figure 2C). In addition to total PTEN, phospho-PTEN (S380) was also detected in all PEL cell lines (Figure 2C). As negative control, we used PTEN-null BJAB cells,
As expected, protein extracts from these cells did not show a band for either PTEN or phospho-PTEN (S380). We were able to verify actin expression in BJAB and all PEL cell lines.
Additionally, phosphorylation of PTEN at serine 380 results in nuclear localization of PTEN, where it is involved in mediating cell cycle arrest.
Nuclear-cytoplasmic partitioning of phosphatase and tensin homologue deleted on chromosome 10 (PTEN) differentially regulates the cell cycle and apoptosis.
Using immunofluorescence analysis, we determined that, independent of its phosphorylation status, PTEN was found in both the nucleus and the cytoplasm (see Supplemental Figure S1 at http://ajp.amjpathol.org). We used KSHV-LANA staining as a positive nuclear control.
LANA was distributed in the characteristic nuclear speckles. There was no colocalization between PTEN and LANA, irrespective of PTEN phosphorylation status. After examining the gene locus, gene sequence, mRNA levels, and protein levels, we concluded that PTEN protein expression is intact in PEL; however, PTEN was heavily phosphorylated at serine 380, which represents an inactivating modification.
Phospho-PTEN Is Readily Detectable in PEL Xenograft Tumors
To test the hypothesis that the phosphorylation status of PTEN is maintained in a more demanding tumor microenvironment, compared with the nutrient-rich culture conditions, we analyzed expression of phospho-PTEN in previously established PEL tumor grafts.
We stained sections of PEL xenograft tumors for total PTEN (Figure 3, A and E) and phospho-PTEN (S380) (Figure 3, C and G). Both total PTEN and phospho-PTEN (S380) were expressed in a majority of tumor cells. The staining pattern of total PTEN matched that of phospho-PTEN (S380): both were present in the nucleus and cytoplasm. We found no evidence of heterogeneity. The sections were also stained for LANA, to determine presence of KSHV (Figure 3, D and H). Because LANA was exclusively localized to the nucleus, it served as an additional control, showing that the cytoplasmic signal we observed with total PTEN and phospho-PTEN (S380) using two independent PTEN antibodies was specific. Analogous to our observation in culture, and irrespective of phosphorylation status, there was no colocalization between PTEN and LANA and the staining patterns were distinct. The no-primary-antibody controls confirmed the absence of staining in the absence of primary antibody (Figure 3, B and F). These findings demonstrate that the high phosphorylation status of PTEN is not a result of optimal (10% serum) growth conditions in culture, but is maintained even in a more stringent tumor xenograft microenvironment.
Figure 3Immunohistochemical analysis of PEL xenograft tumors shows similar localization of total PTEN and phospho-PTEN. BC-1 xenograft tumor sections were stained for total PTEN (A and E), phospho-PTEN (S380) (C and G), and LANA (D and H) (red; NovaRed substrate). Sections were counterstained using hematoxylin (blue), which identifies the nucleus. Sections shown were incubated with dilution buffer either with primary antibody at 1:100 dilution (A, C, D, E and H) or without primary antibody (B and F). No red staining was observed in the no-primary-antibody control (B and D). Original magnification: ×100 (A–D); ×400 (E–H).
To test the hypothesis that inactive PTEN contributes to hyperactivation of Akt and PEL growth, we ectopically expressed human PTEN in PEL and evaluated its effect on cell growth using the classic soft agar assay.
If PEL evolved downstream mutations (eg, activating mutations in Akt) to be resistant to the activity of PTEN, we would expect continued growth. Instead, PTEN reduced colony formation, as determined independently in three distinct PEL cell lines (BC1, BCBL1, BCP1) that differ in p53 status.
Fewer to no colonies were formed by PTEN-overexpressing BCP1 cells (Figure 4B), compared with vector-only control (Figure 4A). Given that a 1/10th amount of EGFP-Max plasmid (Lonza) was nucleofected in tandem with vector and hPTEN plasmids, we used GFP expression to demonstrate equivalent efficiency of nucleofection (Figure 4, C and D). PTEN protein expression from our PTEN-expressing plasmid on nucleofection, was verified by Western blot using BJAB cells that lack endogenous PTEN (Figure 4E). PTEN protein was expressed at the appropriate molecular weight after nucleofection of PTEN, but not vector control. HEK293 cells served as positive control, showing a band of the expected molecular weight. Inhibition of colony formation in three PEL cell lines was significant, with P ≤ 0.005 for BC1 and BCP1 and P ≤ 0.05 for BCBL1 (Figure 4F). This functional assay suggests that signaling components downstream of PTEN still respond to PTEN.
Figure 4Growth suppression of BCP-1 and SLK cells upon overexpression of PTEN. BCP1 cells colonies formed in soft agar after vector-only nucleofection (A), but no colonies formed on nucleofection of human PTEN clone 1 (pDD1540) (B). GFP positive control for nucleofection is shown in bright light (C) and fluorescence (D) fields. E: Expression of protein in PTEN-null BJAB cells on expression of the hPTEN only and not the vector control. F: Quantification of colonies observed in the PEL cell lines BC1, BCBL1, and BCP1 (P = 0.003, 0.02, and 0.005, respectively) and in the KS-like cell lines L1T2 and SLK (P = 0.003 and 0.03, respectively). *P ≤ 0.05; **P ≤ 0.005. G–I: Representative images of colonies upon overexpression of hPTEN, clones 1 (H) and 7 (I) compared with vector only (G) in SLK cells.
PTEN Is Expressed in Primary KS Tumor Biopsies and Reduces Colony Formation in KS Tumor Models
In addition to PEL, KSHV is also the causative agent of KS, a tumor of endothelial cell lineage. We hypothesized that this cancer too, would inactivate PTEN by phosphorylation. Unlike PEL, for which multiple cell lines are available, only a few KS-derived tumor cell lines are available: SLK,
The beta-core fragment of human chorionic gonadotrophin inhibits growth of Kaposi's sarcoma-derived cells and a new immortalized Kaposi's sarcoma cell line.
We used the SLK cell line and the newly established L1T2 cell line, explanted from L1-TIVE xenograft tumor (unpublished data), to test the effect of introducing PTEN on tumor cell growth. We tested colony formation of SLK cells on transfection with vector alone expressing the blasticidin selection marker or two independent clones expressing either human PTEN or the blasticidin selection marker. After 2 weeks in selective medium, the number of colonies was reduced in cells transfected with PTEN, but not with vector alone (Figure 4, G–I). We observed a statistically significant reduction in the number of colonies formed (Figure 4F), with P ≤ 0.05 for SLK and P ≤ 0.005 for L1T2.
To verify these experimental results in biopsies from patients with KS and to investigate PTEN phosphorylation, we used KS-TMAs. The main advantages of a tissue microarray are that all cores are exposed to the same exact conditions of antibody staining and that the cores represent adjacent sections. This facilitates a direct comparison of protein expression. In our first array, we detected total PTEN in 141/176 (80%) cases and intense phospho-PTEN (S380) in 75/130 (58%) cases (Figure 5, A and B). Often the same core also stained positive for phospho-Akt (T308) (Figure 4C); that is, many KS tumors exhibited activated Akt despite robust PTEN expression. Phospho-Akt (T308) requires PI3K activation and PTEN inactivation. No staining was observed in the no-primary-antibody control of the same core or control tissue (Figure 5D). The control tissue was from an unaffected area of the lung of an HIV+ KS patient. It appears nontypical, in that the tissue may have partially collapsed and shows alterations in response to episodes of non-neoplastic disease in AIDS patients, such as past or ongoing pneumocystis pneumonia. The correspondence between phospho-PTEN (S380) and phospho-Akt (T308) was not observed in normal tissue. We detected total PTEN in a normal control lung sample (Figure 5A). Nonetheless, the same core did not stain for phospho-PTEN (S380) (Figure 5B) or for phospho-Akt (T308) (Figure 5C). This could be expected for normal cells, in which high levels of hypophosphorylated PTEN counteract the phosphorylation of Akt. In a representative skin KS lesion, total PTEN, phospho-PTEN (S380), and phospho-Akt (T308) were detected in the same biopsy (Figure 5, A–C). These data support our hypothesis that, in the majority of KS tumors, the tumor suppressor PTEN is not deleted but instead is post-translationally inactivated to allow phosphorylation of Akt.
Figure 5Expression of PI3K activation markers and phospho-PTEN (S380) in KS primary biopsies. Immunohistochemistry showed the presence of total PTEN (A), phospho-PTEN (S380) (B), phospho-Akt (T308), a PI3K-dependent site (C), and no-primary-antibody negative control (D). Original magnification: ×100 (upper row); ×400 (lower row). In each panel (A–D), images at the left are of control lung tissue, which stains positive only for PTEN but not for phospho-PTEN (S380) or phospho-Akt (T308); images at the right are of a representative skin biopsy, in which the same tissue section is positive for total PTEN, phospho-PTEN (S380), and phospho-Akt (T308). E: Staining of a second TMA incorporating AIDS-KS from tissues of three different origins for phospho-PTEN (S380), PI3K activation markers phospho-Akt (T308) and phospho-S6r (S235/236), KSHV marker LANA, and proliferation marker Ki67. Sections were selected such that a region of the biopsy was negative (asterisks), in an otherwise positive section, serving as an internal control. Original magnification, ×100.
We used a newer, second independent AIDS-KS TMA to statistically verify the correlation of phospho-PTEN (S380) with activated Akt (Figure 5E). For each core, the KS diagnosis and tissue of origin was reverified (images of H&E stains for each core are available on request). Altogether, we analyzed 130 cores of AIDS-KS and normal controls. Of these, 29 were from lymph node KS, 40 from oral KS (including tonsillar and adenoid), and 71 from skin KS lesions. This TMA was stained with antibodies to phospho-PTEN (S380), phospho-Akt (T308), phospho-S6 ribosomal (S6r) protein (S235/236), LANA, and the proliferation marker Ki67. Sections from three different sites (skin, lymph node, and head and neck) were stained for phospho-PTEN (S380), phospho-Akt (T308) phospho-S6r (S235/236), KSHV marker LANA and Ki67; all sections were counterstained with hematoxylin (Figure 5E). The sections were selected to include regions of no tumor and no staining within an otherwise positive biopsy, thus providing an internal negative control.
To support our visual impressions, we conducted a quantitative analysis (Figure 6). Images were recorded from each core, and staining intensity was quantified as described under Materials and Methods. Linear regression analysis was performed to determine the correlation between phospho-PTEN and markers of PI3K-mTOR activation (phospho-Akt (T308) and phospho-S6r (S235/236)), KSHV-infection (LANA), and active proliferation (Ki67). Each of the plots in Figure 6 represents the relative staining intensity for different markers on the y axis (phospho-Akt, phospho-S6r, LANA and Ki67) against phospho-PTEN (S380) on the x axis. We found a strong linear correlation between phospho-PTEN (S380) and both phospho-Akt (T308) and phospho-S6r (S235/236). This indicates that the PI3K-mTOR signaling pathway is activated in KS and that PTEN is phosphorylated at S380 and thus is inactive in KS (P ≤ 10−14), across all three sites (Figure 6, A and B). By contrast, we found no significant correlation between phospho-PTEN (S380) and LANA or Ki67 (Figure 6, C and D). There were no significant differences in staining patterns across the different tumor sites. Regardless of tumor site, KS presented with characteristic spindle cell morphology and a similar pattern of PI3K-Akt-mTOR pathway member status. This suggests that in KS, as in PEL, post-translational inactivation of PTEN contributes to the activation of the PI3K-Akt-mTOR pathway.
Figure 6Significant correlation of phospho-Akt (T308) and phospho-S6 ribosomal protein (S235/236) with phospho-PTEN (S380) in KS. The square root of the intensity count for phospho-Akt (A), phospho-S6r (B), LANA (C), and Ki67 (D), is plotted on the vertical axis and that for phospho-PTEN is plotted on the horizontal axis for KS sections from three body sites: lymph node, oral, and skin. The solid line indicates the linear regression; the shaded region indicates the 95% confidence interval of the fit. There was a significant correlation of both phospho-Akt (T308) and phospho-S6r (S235/236) with phospho-PTEN (S380), with P values of 3.09 × 10−13 and 2.6 × 10−13 respectively. There was no significant correlation between LANA positivity (denoting viral infection) and Ki67 (denoting active proliferation) (P > 0.05).
Activation of the PI3K-Akt-mTOR signaling cascade is a defining phenotype of the KSHV-associated cancers PEL and KS. KS has been reported to respond to rapamycin,
Other virus-associated lymphoproliferative diseases such as EBV-associated post-transplant lymphoproliferative disease also rely on activation of the PI3K-Akt-mTORC1 pathway
Activation of mammalian target of rapamycin in transformed B lymphocytes is nutrient dependent but independent of Akt, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase, insulin growth factor-I, and serum.
The immunosuppressive macrolide RAD inhibits growth of human Epstein-Barr virus-transformed B lymphocytes in vitro and in vivo: a potential approach to prevention and treatment of posttransplant lymphoproliferative disorders.
mTOR inhibitor monotherapy is insufficient to suppress viremia and disease progression in Epstein-Barr virus-driven lymphoproliferative disorders (EBV-LPD).
Thus, there is urgent need to understand the mechanism of rapamycin, and discover potential biomarkers of response. To activate this pathway, either a positive regulator has to be hyperactivated, or a negative regulator has to be inactivated, or both. To date, most studies in PEL and KS have concentrated on positive regulators of the PI3K-Akt-mTOR pathway. In the present study, we investigated the principal negative regulator of PI3K-Akt-mTOR signaling, the tumor suppressor PTEN.
PTEN antagonizes PI3K in the first step in the PI3K-Akt-mTOR cascade by dephosphorylating PIP3 to PIP2 and thus constitutes a major step of regulation. Diminished PTEN activity in human cancer results in hyperactivation of the PI3K signaling pathway, a condition that has also been termed oncogene addiction to PI3K-mTOR. In nonviral tumors, PTEN is typically inactivated by deletion or mutation. The PTEN gene is deleted in many solid tumors
and in ∼40% of cell lines derived from prostate, endometrial, and nervous system cancers (Cancer Genome Project, Wellcome Trust Sanger Institute; http://www.sanger.ac.uk/perl/genetics/CGP/core_line_viewer?action=bygene&ln=PTEN&start=1&end=404&coords=AA:AA. Accessed August 11, 2011). In the context of hematological malignancies, it was reported that the majority of primary acute leukemias and non-Hodgkin's lymphomas, both primary and cell lines, show abnormalities in PTEN gene and protein expression.
The situation is different for PEL, which is a viral lymphoma. We conclude that genetic aberrations of PTEN are infrequent in PEL and our data show that KSHV-associated PELs invariably express high levels of PTEN protein (Figure 1, Figure 2) and at the same time exhibit high levels of phosphorylated Akt.
We observed the same pattern in KS biopsies (Figure 5, Figure 6). We found no evidence of point mutations in PEL cell lines, and no evidence of mislocalization. Rather, PTEN appeared to be hyperphosphorylated in PEL and KS. Furthermore, we showed that ectopic expression of PTEN in PEL or KS-tumor derived cell lines inhibited colony formation (Figure 4). This suggests that downstream effectors of the PI3K-Akt-mTOR pathway cannot override the growth-inhibitory signal of PTEN in PEL and KS. This phenotype is in contrast to the behavior of cells carrying activating mutations in downstream players such as Akt.
reported similarly on the absence of PTEN mutations in archived biopsies from PEL patients, although they found mutations in some of their PEL cell lines. For instance, they reported a homozygous deletion of PTEN exons 6 to 9 in the BCP-1 cell line, whereas we detect PTEN protein in this cell line. Uddin et al
had changed as a result of prolonged passage in culture. Such discrepancies are not unexpected, given that PEL cell lines change on passage in culture (Roy and Dittmer, unpublished data). Of note, BCP-1 is unique among PELs, because it carries homozygous mutant p53 and would thus be expected to be less genetically stable than other PEL cell lines, all of which carry wild-type p53.
At present, we do not know which kinase phosphorylates PTEN in PEL. KSHV proteins may mediate the phosphorylation of PTEN. Alternatively, casein kinase 2 (CK2), glycogen synthase kinase 3β (GSK3β), LKB1, and rhoA-associated kinase (ROCK) have all been implicated in PTEN phosphorylation.
which would eliminate GSK3β from mediating PTEN phosphorylation in PEL. Further studies are needed to elucidate the players in phosphorylation of PTEN in KSHV-associated malignancies.
What is the significance of our findings? Viral cancers express viral oncogenes. Viral lymphomas in particular express viral oncoproteins that involve the PI3K-Akt-mTOR pathway. Several KSHV proteins, such as K1 and vGPCR, can activate PI3K.
This would lessen the selective pressure to mutate and genetically inactivate PTEN. EBV, a human herpesvirus closely related to KSHV, is capable of silencing PTEN protein expression by CpG island and promoter methylation in gastric cancers,
but no such observations have been reported for EBV-associated lymphomas. It appears that KSHV silences PTEN protein function by inducing constitutive phosphorylation, which, like deletion, relieves the negative regulation of PTEN on the PI3K-mTOR signaling pathway. Phosphorylation of PTEN contributes to PI3K-Akt-mTOR activation in KSHV-infected cells and can thus be speculated to contribute to the proliferative advantage of PEL and KS tumor cells. In addition, we found two reports of nonviral lymphomas in which PTEN was also inactivated post-translationally: cases of mantle cell lymphoma (MCL) and some cases of T-cell acute lymphoblastic leukemia (T-ALL) show hyperactivation of PI3K in the context of wild-type PTEN.
What seems to be the default route for viral lymphomas (ie, post-translational or epigenetic inactivation of PTEN) is also seen in the evolution of nonviral cancers. It is intriguing that of all non-Hodgkin's lymphomas, only mantle cell lymphoma responds clinically to mTOR inhibitors.
We speculate that phospho-PTEN (S380) expression may serve (assuming confirmation in a clinical trial) as a novel biomarker for rapamycin sensitivity of human cancers.
A key finding of the present study is that expression of wild-type PTEN in PEL inhibited growth in semisolid medium.
This result fuels the speculation that PTEN activation by small molecules may present a novel treatment modality for wild-type PTEN tumors, similar to the rationale for p53 activation by Nutlin-3
as a possible treatment modality for post-translationally inactivated but genetically wild-type p53 cancers.
Acknowledgments
We thank Sang-Hoon Sin for providing p53 sequence data for BC2, BC5, and TY1 cell lines and Jessica L. Poisson for evaluating the TMA cores for tissue of origin. The initial KS array was constructed by Barbara Hackman with the help of Leona Ayers (pathologist) at the Ohio State University AIDS Cancer Specimen Resource (ACSR) site. The second KS-TMA was constructed by Alanna Morris with the help of Nancy Ciau (pathologist) at the University of California at San Francisco ACSR site under the direction of Michael S. McGrath. We thank Blossom Damania and members of her laboratory for helpful discussion and critical reading. We apologize to our colleagues whose work we could not cite because of space limitations.
Immunofluorescence analysis showing the distribution of total PTEN and phospho-PTEN (S380) in PEL cells. Green punctate staining marks either total PTEN (A) or phospho-PTEN (S380) (B). Independent of its phosphorylation status, PTEN localizes to both the cytoplasm and nucleus of the cells. LANA staining (red) was exclusively localized to the nucleus, as expected, and therefore was used both as a control for nuclear staining and to denote the presence of viral episome in the PEL cells. Merge of total PTEN or phospho-PTEN (S380) illustrates the lack of colocalization of PTEN with LANA, independent of PTEN phosphorylation. DAPI (blue) nuclear counterstain added to the merge of LANA and PTEN confirms LANA localization to the nucleus. Immunofluorescence analysis was performed in three PEL cell lines, BC1, BCBL1, and BCP1, that were generated independently and show different KSHV and EBV infection, as well as distinct p53 status (as shown in Table 1). All images were processed through deconvolution software, as described under Materials and Methods. Original magnification, ×630. Scale bars, 10 µm.
References
Cesarman E.
Chang Y.
Moore P.S.
Said J.W.
Knowles D.M.
Kaposi's sarcoma-associated herpesvirus-like DNA sequences in AIDS-related body-cavity-based lymphomas.
Cell cycle-dependent nuclear export of phosphatase and tensin homologue tumor suppressor is regulated by the phosphoinositide-3-kinase signaling cascade.
In vitro establishment and characterization of two acquired immunodeficiency syndrome-related lymphoma cell lines (BC-1 and BC-2) containing Kaposi's sarcoma-associated herpesvirus-like (KSHV) DNA sequences.
Establishment and characterization of a primary effusion (body cavity- based) lymphoma cell line (BC-3) harboring Kaposi's sarcoma-associated herpesvirus (KSHV/HHV-8) in the absence of Epstein-Barr virus.
Establishment and characterization of an Epstein-Barr virus (EBC)-negative lymphoblastoid B cell line (BJA-B) from an exceptional, EBV-genome-negative African Burkitt's lymphoma.
The expression of PAX6, PTEN, vascular endothelial growth factor, and epidermal growth factor receptor in gliomas: relationship to tumor grade and survival.
Nuclear-cytoplasmic partitioning of phosphatase and tensin homologue deleted on chromosome 10 (PTEN) differentially regulates the cell cycle and apoptosis.
The beta-core fragment of human chorionic gonadotrophin inhibits growth of Kaposi's sarcoma-derived cells and a new immortalized Kaposi's sarcoma cell line.
Activation of mammalian target of rapamycin in transformed B lymphocytes is nutrient dependent but independent of Akt, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase, insulin growth factor-I, and serum.
The immunosuppressive macrolide RAD inhibits growth of human Epstein-Barr virus-transformed B lymphocytes in vitro and in vivo: a potential approach to prevention and treatment of posttransplant lymphoproliferative disorders.
mTOR inhibitor monotherapy is insufficient to suppress viremia and disease progression in Epstein-Barr virus-driven lymphoproliferative disorders (EBV-LPD).
Present study.
00634-1&id=gr1.jpg){kind=link}
00634-1&id=gr2.jpg){kind=link}
00634-1&id=gr3.jpg){kind=link}
00634-1&id=gr4.jpg){kind=link}
00634-1&id=gr5.jpg){kind=link}
00634-1&id=gr6.jpg){kind=link}