A Kaposi's Sarcoma-Associated Herpesvirus-Encoded Ortholog of MicroRNA miR-155 Induces Human Splenic B-Cell Expansion in NOD/LtSz-scid IL2Rγ^null Mice^,

Abstract

MicroRNAs (miRNAs) are small noncoding RNA molecules that function as posttranscriptional regulators of gene expression. Kaposi's sarcoma (KS)-associated herpesvirus (KSHV), a B-cell-tropic virus associated with KS and B-cell lymphomas, encodes 12 miRNA genes that are highly expressed in these tumor cells. One viral miRNA, miR-K12-11, shares 100% seed sequence homology with hsa-miR-155, an oncogenic human miRNA that functions as a key regulator of hematopoiesis and B-cell differentiation. So far, in vitro studies have shown that both miRNAs can regulate a common set of cellular target genes, suggesting that miR-K12-11 may mimic miR-155 function. To comparatively study miR-K12-11 and miR-155 function in vivo, we used a foamy virus vector to express the miRNAs in human hematopoietic progenitors and performed immune reconstitutions in NOD/LtSz-scid IL2Rγ^null mice. We found that ectopic expression of miR-K12-11 or miR-155 leads to a significant expansion of the CD19⁺ B-cell population in the spleen. Subsequent quantitative PCR analyses of these splenic B cells revealed that C/EBPβ, a transcriptional regulator of interleukin-6 that is linked to B-cell lymphoproliferative disorders, is downregulated when either miR-K12-11 or miR-155 is ectopically expressed. In addition, inhibition of miR-K12-11 function using antagomirs in KSHV-infected human primary effusion lymphoma B cells resulted in derepression of C/EBPβ transcript levels. This in vivo study validates miR-K12-11 as a functional ortholog of miR-155 in the context of hematopoiesis and suggests a novel mechanism by which KSHV miR-K12-11 induces splenic B-cell expansion and potentially KSHV-associated lymphomagenesis by targeting C/EBPβ.

INTRODUCTION

MicroRNAs (miRNAs) are small noncoding RNAs, 22 to 24 nucleotides in length, that mediate posttranscriptional gene repression by binding to the 3′ untranslated region (UTR) of target mRNAs (2). miRNAs are expressed by a diverse range of organisms that includes all metazoa and many plant species (21). Functionally, miRNAs are key regulators of many biological processes, including, but not limited to, embryonic development, hematopoiesis, immunity, and apoptosis. Their importance in regulating these processes is further underscored by their association with oncogenesis; for example, aberrant expression of miR-155 and members of the miR-17-92 family contributes to tumor formation in multiple types of leukemia and lymphomas (17).

Recently, DNA viruses were found to encode miRNAs, including all three families of herpesviruses (α, β, and γ) (for a review, see reference 4). Our group and others identified that the gammaherpesvirus Kaposi's sarcoma (KS)-associated herpesvirus (KSHV) encodes a total of 12 miRNA genes all located within the KSHV latency-associated region (5, 22, 40, 44). KSHV is lymphotropic, establishes latency in B cells (54), and is associated with the vascular tumor KS and two B-cell lymphoproliferative malignancies: primary effusion lymphoma (PEL) and multicentric Castleman's disease (MCD) (7, 8, 13, 50). The majority of the cells in these malignancies are latently infected, and during this stage, the viral genome expresses only a limited number of genes, including the viral miRNAs (11, 51). KSHV latent proteins regulate cellular pathways to inhibit apoptosis, induce cellular proliferation, and modulate cytokine responses, but the roles of KSHV miRNAs in pathogenesis are still being characterized (for a review, see reference 12). Insights into the pathogenic nature of these viral miRNAs have been provided by findings that they target host genes involved in tumorigenesis, cellular differentiation, immunity, and apoptosis (23, 32, 41, 45, 60).

The most essential parameter for miRNA regulation of mRNA expression is complementary base pairing between the miRNA “seed” sequence (5′ nucleotides 2 to 7) and the target transcript (2). Recently, we and others reported that KSHV miR-K12-11 shares 100% seed sequence homology with the human oncomir miR-155 and can regulate an overlapping set of genes in cell lines engineered to express miR-155 or miR-K12-11 (20, 49). This was an important finding because miR-155-dependent regulation is important during hematopoiesis of different lineages, including B cells (for a review, see reference 35), and deregulated miR-155 expression has been implicated in the formation of B-cell tumors (10). In addition to KSHV, the oncogenic avian alphaherpesvirus Marek's disease virus (MDV) also encodes a miRNA (mdv1-miR-M4) that shares seed sequence homology with miR-155 and, like miR-K12-11, is capable of regulating an overlapping set of miR-155 mRNA targets (30, 59). Moreover, in vivo functional analysis of mutant MDVs which contain a nonfunctional or deleted miR-M4 revealed that this miRNA plays an essential role in the induction of T-cell lymphomas in birds (58). Interestingly, two separate viruses that cause B-cell lymphomas, Epstein-Barr virus (EBV; a transforming human gammaherpesvirus closely related to KSHV) and oncogenic retrovirus reticuloendotheliosis virus strain T (REV-T), do not encode miR-155 orthologs but induce miR-155 expression during infection (3, 6, 18, 26, 31). Furthermore, a recent study found that inhibiting miR-155 function in two EBV-positive B-cell lines resulted in decreased proliferation and increased apoptosis, providing evidence that miR-155 plays an important role during B-cell immortalization (27). While these studies have confirmed the oncogenic potential of miR-155 and miR-M4 during viral infection, the miRNA targets responsible for these phenotypes have not been reported.

Based on the roles of miR-155 and its ortholog miR-M4 in virally induced immortalization and lymphomagenesis, we hypothesize that KSHV miR-K12-11 plays a similar role in promoting KSHV pathogenesis. To directly address this, we examined the effects of ectopic miR-K12-11 and miR-155 expression in human hematopoietic stem cells (HSCs) during immune reconstitution using the NOD/LtSz-scid IL2Rγ^null mouse model. This is the first in vivo study using a humanized mouse model to examine the function of miR-K12-11 during hematopoiesis. In brief, human cord blood (CB)-derived CD34⁺ progenitors were retrovirally transduced with miRNA-green fluorescent protein (GFP) expression vectors and transplanted into sublethally irradiated mice. Fluorescence-activated cell sorter (FACS) and histology results show that ectopic expression of either miR-K12-11 or miR-155 leads to a significant expansion of the hCD19⁺ B-cell population in the spleen. To gain further insight into the mechanisms contributing to this expansion, we analyzed RNA from harvested splenocytes for the expression of validated miR-155 targets involved in lymphomagenesis and B-cell development and found that CCAAT enhancer-binding protein β (C/EBPβ), a negative regulator of interleukin-6 (IL-6), is repressed (9). Moreover, inhibition of miR-K12-11 function with specific antagomirs in two separate PEL cell lines (BCBL1 and BC3) resulted in derepression of C/EBPβ. These data suggest that miR-K12-11 contributes to human B-cell expansion in part by regulating the miR-155 target C/EBPβ and provides further evidence that this miRNA plays an important role in promoting KSHV B-cell pathogenesis.

MATERIALS AND METHODS

Ethics statement.

All animals were handled in strict accordance with good animal practice as defined by the relevant national and/or local animal welfare bodies, and all animal work was approved by IACUC no. 299801489.

Cell culture.

The 293T cell line (human embryonic kidney fibroblasts) was obtained from the American Type Culture Collection (Manassas, VA). Cryogenically preserved primary CD34⁺ human CB cells were purchased from StemCell Technologies, Vancouver, British Columbia, Canada. The 293T cells were grown in Dulbecco's modified Eagle's medium supplemented with streptomycin (100 μg/ml), penicillin (100 μg/ml), and 10% fetal calf serum. CD34⁺ human CB cells were cultured for transduction in Iscove's modified Dulbecco's medium supplemented with 10% fetal bovine serum containing 1 μg/ml each human stem cell factor, human thrombopoietin, human Flt-3 ligand, and IL-11 (Peprotech, Rocky Hill, NJ).

Foamy virus vector construction.

To produce miR-155 or miR-K12-11 vector constructs, we first amplified the miR-K12-11 or miR-155 miRNA sequences containing a region of approximately 200 nucleotides surrounding each pre-miRNA hairpin from pCDNA3.1/V5/HisA expression vectors that were previously described (45). We inserted individual miRNA cassettes downstream of enhanced GFP (EGFP) into a simian foamy virus type 1 (SFV-1) vector backbone (pCCEGFPL) that was previously described (38, 39).

Luciferase assays and reporter construction.

miRNA sensor vectors were created using the pGL3 promoter vector from Promega (Madison, WI). Synthetic oligonucleotides containing two complete complementary copies of a miRNA sequence separated by a 9-bp-long spacer were inserted into the 3′ UTR of the luciferase gene upstream of the polyadenylation signal as previously described (49). To construct a luciferase reporter plasmid containing the full-length 3′ UTR sequence of C/EBPβ, the following primers were first designed with Vector NTI (Invitrogen) (the forward primer contained an NdeI site [bold], and the reverse primer contained an Fse1 site [bold]): fwd, 5′-CATATGGAACTTGTTCAAGCAGCTGC-3′; rev, 5′-GGCCGGCCGGCTTTGTAACCATTCTC-3′. PCR products were cloned into pCRII-TOPO (Invitrogen), excised, and inserted into the 3′ UTR of the pGL3 promoter at the NdeI and Fse1 sites. All constructs were confirmed by sequencing. 293T cells were cotransfected with luciferase reporter constructs, foamy virus vectors, and/or miRNA expression vectors in 24-well plates for 72 h using Mirus TransIT-293 reagent (Madison, WI) according to the manufacturer's instructions. Luciferase activity was quantified using the Luciferase assay system (Promega) according to the manufacturer's protocols. Briefly, transfected 293T cells were lysed in cell culture lysis reagent (Promega) and 20% of each cell lysate was assayed for firefly luciferase activity. Light units were normalized to Renilla luciferase using a dual-luciferase reporter kit (Promega).

Foamy virus production and CD34⁺ human CB cell transduction.

To generate infectious viral particles, we cotransfected 293T cells with the individual miRNA expression vectors and the previously described packaging plasmid pCIenv (38). Transfections were carried out in T75 cell culture flasks (5 × 10⁶ 293T cells per flask) by the calcium phosphate method. Viral supernatants were harvested 4 days posttransfection and clarified by centrifugation at 5,000 rpm for 20 min and then by passage through a 0.45-μm filter. The vector particles were further concentrated 100-fold by using 70-kDa Apollo Centrifugal Spin Concentrators (Orbital Biosciences, Topsfield, MA). The SFV-1 vector titers produced were determined on fresh 293T cells plated at a density of 2.5 × 10⁴ per well in 24-well plates. Seventy-two hours after infection, cells were monitored and scored for GFP fluorescence under a microscope with a UV light source. Transduction of CD34⁺ cells was carried out by spin inoculation as previously described (61). Briefly, 3 × 10⁶ CD34⁺ cells (a heterogeneous mixture of cells from two separate donors) were seeded into 15 wells of a 24-well human fibronectin plate (BD Biosciences, San Jose, CA) at a density of 1.5 × 10⁵ cells per well. At 24 h after the initial seeding, viral supernatant was added to cells at a multiplicity of infection of 50. The plates were then spun at 1,200 rpm for 1.5 h, and the infection procedure was repeated 24 h later. Following the last transduction, 2 × 10⁵ transduced CD34⁺ cells were transplanted by injection into the lateral tail vein of each of 4 to 8 sublethally irradiated (250 rads from a cesium 137 source at 65.7 rads/min) NOD/LtSz-scid IL2Rγ^null mice. For colony-forming assays, transduced cells were plated in serum-free methylcellulose culture (Methocult 04236; StemCell Technologies, Vancouver, British Columbia, Canada) in the presence of 1 μg/ml each human Flt-3 ligand, human stem cell factor, human granulocyte-macrophage colony-stimulating factor, human IL-3, and human erythropoietin for 14 days.

Mice.

NOD/LtSz-scid IL2Rγ^null mice were obtained from The Jackson Laboratory. All experiments involved male mice and were performed according to IACUC-approved protocols.

miRNA detection and absolute quantification of miR-K12-11.

RNA was extracted from samples using the RNA-Bee kit in accordance with the manufacturer's instructions (AMS Biotechnology, Milton, United Kingdom). cDNA was synthesized from 10 ng total RNA using the TaqMan MicroRNA reverse transcription (RT) kit (Applied Biosystems, Foster City, CA). To detect miR-155 and miR-K12-11, the TaqMan miRNA detection assay was run in triplicate using human miR-155 and KSHV miR-K12-11 TaqMan probes according to the manufacturer's instructions (Applied Biosystems, Foster City, CA). Relative miRNA expression was determined using the Applied Biosystems Relative Quantification (RQ) Manager software v2.1 with human miR-16 set as the endogenous control. The absolute copy number of miR-K12-11 in both splenocytes and BCBL1 cells was calculated by using a standard curve of known quantities of a miR-K12-11 synthetic miRNA mimic (Thermo Fisher Scientific, Lafayette, CO). To determine the miR-K12-11 copy number in GFP-positive CD19⁺ splenocytes, we assumed that 10 ng of total RNA equals 10,000 cells. Furthermore, we expressed the absolute copy number per GFP-positive CD19⁺ splenocyte by taking into account the percentage of GFP-positive CD19⁺ cells as determined by FACS.

Flow cytometry cell lineage analysis.

GFP expression and phenotypic markers were analyzed by flow cytometry using an LSR-II cytometer and FACSDiva software (BD Biosciences, San Jose, CA).

Fluorophore-conjugated monoclonal antibodies specific for human CD45, CD19, CD33, and CD3 (BD Biosciences, San Jose, CA BD) were used to stain red blood cell-depleted splenocytes and bone marrow (BM) cells. Background staining was determined using a murine monoclonal IgG1 isotype control (BD Biosciences, San Jose, CA).

Necropsy, histology, and IHC analysis.

Mice were necropsied, and all tissues were evaluated for gross lesions. Portions of the spleen, liver, and femur were fixed in 10% buffered formalin for 18 to 24 h, dehydrated, and embedded in paraffin. Sections were cut at 5 μm for routine hematoxylin and eosin (H&E) staining and at 3 μm onto positively charged slides (Probe On Plus; Fisher Scientific, Springfield, NJ) for immunohistochemistry (IHC) analysis against CD19, a marker for human B lymphocytes. Deparaffinized tissue sections were subjected to heat-induced antigen retrieval by microwaving in citrate buffer solution (Antigen Unmasking Solution; Vector Laboratories, Burlingame, CA). The primary antibody for IHC analysis was mouse monoclonal anti-human antibody CD19 (BIOCARE Medical, LLC, Concord, CA) used at a dilution of 1:150. Sites of primary antibody binding were identified by high-affinity immunocytochemistry analysis with STAT-Q (Innovex Biosciences; Richmond, CA) using a secondary antibody and streptavidin-horseradish peroxidase. The chromogen was diaminobenzidine with Mayer's hematoxylin counterstain.

Antagomir derepression assays and real-time quantitative RT-PCR (qRT-PCR) analysis.

For inhibition of miR-K12-11, previously described 2′OMe RNA antagomirs (49) were used. PEL cells (1 × 10⁶) were transfected with 25 nM antagomir in 24-well plates using Mirus TransIT-TKO (9 μl/250 ml total medium). After 6 h of incubation, cells were pelleted, transfection medium was removed, and cells were plated in fresh RPMI 1640 medium supplemented with 10% fetal bovine serum and 5% penicillin-streptomycin (Gibco) for 48 h before RNA was harvested.

RNA from splenocytes, BC-3, and BCBL1 cells was reverse transcribed using SuperScript III reverse transcriptase (Invitrogen) in the presence of random hexamers according to the manufacturer's protocols. Quantitative PCR (qPCR) was carried out using an ABI StepOne Plus system along with ABI Fast SYBR (Applied Biosystems, Carlsbad, CA). Primers for C/EBPβ were designed across exon boundaries and were previously described (29). Primer pair efficiencies for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), β-actin, and C/EBPβ were validated before analysis for C/EBPβ expression. PCR signals were normalized to both GAPDH and β-actin to check for accuracy and reported as RQ values using StepOne software.

Statistics.

All statistical analyses used a Student two-tailed t test performed on Microsoft Excel software.

RESULTS

Transduction of CD34⁺ human CB cells with miR-K12-11 and miR-155 expressing foamy virus vectors and their engraftment into NOD/LtSz-scid IL2Rγ^null mice.

To ectopically express the miRNAs in CD34⁺ human CB progenitors, we constructed foamy virus vectors that contained miR-K12-11 or miR-155 pri-miRNA sequences downstream of EGFP (Fig. 1A). miRNA expression from these vectors in 293T cells was analyzed by performing luciferase reporter assays as previously described (49). Transfection of the miR-K12-11 or miR-155 expression vector resulted in dose-dependent inhibition of luciferase activity, while transfection of a control vector did not, confirming that miR-K12-11 and miR-155 pre-miRNAs are efficiently processed into mature miRNAs (Fig. 1C).

Fig. 1.

The foamy virus vectors expressing miR-K12-11 or miR-155 and the empty-vector control used in this study. (A) Foamy virus vectors were constructed by inserting the pri-miRNA sequence downstream of a GFP cassette and a cytomegalovirus (CMV) promoter. LTR, long terminal repeat. (B) Schematic of miRNA sensor vectors containing two perfectly complementary binding sites. (C) miRNA expression and sensor vectors were cotransfected into 293T cells, and luciferase activity was measured at 72 h posttransfection. Results show that both miR-K12-11 and miR-155 expression vectors repressed luciferase activity >2-fold compared to no repression by the control.

CD34⁺ human CB progenitors were retrovirally transduced and monitored for GFP expression in colony-forming assays. GFP-expressing colonies were detected 14 days later, indicating successful transduction in vitro. For immune reconstitution, 2 × 10⁵ transduced CB progenitors were transplanted by injection via the tail vein into groups of sublethally irradiated NOD/LtSz-scid IL2Rγ^null mice (8 mice for each miRNA and 4 for the EGFP vector control). At 14 weeks postreconstitution, BM and spleens were harvested from mice for FACS and histological analysis. GFP expression was detected in hCD45⁺ leukocytes harvested from both the BM and spleen, indicating successful human hematopoietic engraftment of transduced cells in all of the mice (Fig. 2).

Fig. 2.

Engraftment of transduced CB CD34⁺ cells. Cells harvested from BM were analyzed by FACS using a human CD45-specific antibody. Human CD45⁺ cells were detected in all mice reconstituted with CD34⁺ human CB progenitors expressing either miR-K12-11, miR-155, or an empty control vector. A large percentage of the CD45⁺ cells also expressed GFP, as shown in the upper right quadrant of each histogram. Shown are representative dot plots for one animal from each group.

Ectopic expression of miR-K12-11 and miR-155 in cells harvested from BM and spleen.

To validate miR-K12-11 and miR-155 expression in the BM and spleens of engrafted mice, total RNA was analyzed by stem-loop qRT-PCR assays. As expected, miR-K12-11 was detected only in the BM and spleens of miR-K12-11-engrafted mice (Fig. 3A), while ectopic miR-155 expression was highest in the BM and spleens of miR-155-engrafted mice (Fig. 3B). Interestingly, the relative increase in ectopic miR-155 expression in the spleen was higher (1- to 1.5-fold) than the increase detected in BM (0.4- to 1.2-fold), possibly indicating that the majority of the cells ectopically expressing miR-155 had already migrated to the spleen at that time point. We next compared the absolute levels of miR-K12-11 expression in splenocytes (hCD19⁺ GFP-positive) versus PEL cells. miR-K12-11 miRNA expression was at levels similar to or lower than those observed in BCBL1 cells (Fig. 3C). These data confirm the ectopic expression of miR-K12-11 and miR-155 in engrafted mice and furthermore demonstrate that we did not overexpress these miRNAs in our model system.

Fig. 3.

Ectopic miR-K12-11 and miR-155 expression in engrafted mice. (A) Ectopic miR-K12-11 was detected only in the miR-K12-11-engrafted animals in both BM and the spleen. (B) Ectopic miR-155 was detected above endogenous levels in miR-155-engrafted animals in both BM and the spleen. (C) The absolute copy numbers of miR-K12-11 in GFP-positive CD19⁺ splenocyte populations from engrafted mice are comparable to or lower than endogenous miR-K12-11 expression in the PEL cell line BCBL1, and therefore it is not overexpressed.

Expression of either miR-155 or miR-K12-11 does not affect cell lineage populations in BM at 14 weeks posttransplantation.

To ask whether ectopic expression of miR-155 or miR-K12-11 affects hematopoiesis in BM, we performed cell lineage analysis by FACS. Results indicated that the majority of the cells in all of the mice were human CD45⁺ leukocytes, indicating successful engraftment. Although we observed a modest increase in hCD45⁺ leukocytes in miR-K12-11 (80.2% ± 10.6%)- and miR-155 (83.7% ± 4.5%)-expressing mice compared to those in vector controls (70.8% ± 18.3%), these differences were not statistically significant across all of the animals (Fig. 4A).

Fig. 4.

Cell lineage differentiation of human progenitors was not significantly altered by miRNA expression in BM. Cells harvested from the BM of mice expressing the empty vector (n = 4), miR-K12-11 (n = 7), or miR-155 (n = 8) were stained with antibodies specific for human (A) CD45⁺ leukocytes, (B) CD19⁺ B cells, (C) CD33⁺ monocytes, and (D) CD3⁺ T cells and analyzed by FACS. Each dot represents FACS analysis of one animal from each group, and the mean score of each group is shown as a solid horizontal line.

We further characterized the various subpopulations of human leukocytes based on the cell surface expression of hCD19 (B cells), hCD33 (myeloid cells), and hCD3 (T cells) (Fig. 4B to D). The hCD19⁺ B-cell population represented the predominant lineage, with higher levels found in mice expressing miR-K12-11 (61% ± 12.6%) and miR-155 (62.1% ± 4.9%) than in those expressing the control vector (52.2% ± 17.4%), but again this trend was not statistically significant across all of the animals. In contrast to the large number of hCD19⁺ B cells in BM, the fraction of hCD33⁺ myeloid cells in the miR-K12-11 (14% ± 5.7%), miR-155 (13.4% ± 3%), and empty-vector control (12.7% ± 3.1%) mice was much lower, regardless of miRNA expression. Across all of the animals, we detected less than 1% hCD3⁺ T cells. Except for the modest but nonsignificant increase in hCD45⁺ leukocytes and hCD19⁺ B-cell populations in the miR-K12-11- and miR-155-expressing mice, these values represent a normal distribution of hematopoietic cell lineages as previously reported after engraftment of CD34⁺ human CB progenitors into NOD/LtSz-scid IL2Rγ^null mice (19, 47).

Because miR-155 has been implicated in B-cell development (9, 10), we also analyzed B cells for expression of CD10 (B-cell precursors) and surface IgM (mature B cells) (see Fig. S1 in the supplemental material). In all of the animals, the majority of the hCD19⁺ cells expressed CD10⁺ (miR-K12-11, 98.5% ± 0.9%; miR-155, 97.5% ± 1.0%; vector control, 97.7% ± 1.2%) compared to lower levels of IgM expression (miR-K12-11, 66.2% ± 11.2%; miR-155, 55.7% ± 11.6%; vector control, 63.5% ± 11.6%). These data indicate that the majority of the hCD19⁺ B cells in BM represent an immature phenotype whose differentiation was not affected by ectopic miRNA expression.

Ectopic expression of miR-K12-11 and miR-155 induces B-cell proliferation in the spleen.

To further evaluate human hematopoietic development in the engrafted mice, we removed the spleens for histology and harvested splenocytes for cell lineage analysis by FACS. Results indicated a significant increase in the percentage of hCD45⁺ leukocytes in the miR-K12-11 (49.6% ± 8.7%)- and miR-155 (46.3% ± 9.5%)-expressing mice compared with that in the empty-vector control mice (33.6% ± 5.7%) (Fig. 5A). Furthermore, splenocytes were significantly enriched for hCD19 (B cells) in the miR-K12-11 (45.7% ± 12.6%)- and miR-155 (42.6% ± 10.1%)-expressing mice compared to those of the vector control mice (29.3% ± 6.1%) (Fig. 5A). In contrast, the hCD33⁺ monocyte and hCD3⁺ T-cell populations were not significantly altered in the presence of miRNA expression (Fig. 5A). The increased percentages observed in the hCD45⁺ and hCD19⁺ populations were due to an increase in the absolute cell numbers of these populations and not a reduction in the absolute cell numbers of the hCD33⁺ and hCD3⁺ populations (data not shown). Based on these observations, the increased hCD45⁺ leukocyte counts in the spleen are caused by an expansion of the hCD19⁺ B-cell population. This was further supported by the observation that the percentage of GFP-positive miRNA-expressing cells in the hCD45⁺ and hCD19⁺ populations represented a significantly larger fraction of the total cell population in mice expressing miR-K12-11 (14.4% ± 5.2% CD45⁺ and 14.3% ± 4.0% CD19⁺) and miR-155 (17.0% ± 5.4% CD45⁺ and 17.1% ± 5.4% CD19⁺) than in those expressing the empty control vector (6.5% ± 0.9% CD45⁺ and 7.9% ± 1.2% CD19⁺) (Fig. 6). Interestingly, there was also an increase in the GFP-negative hCD45⁺ and hCD19⁺ populations in mice ectopically expressing miR-K12-11 or miR-155, but this increase was not statistically significant (data not shown). Together, these data show that ectopic miR-K12-11 and miR-155 expression during hematopoiesis in NOD/LtSz-scid IL2Rγ^null mice led to a marked increase in B-cell proliferation within the spleen.

Fig. 5.

Ectopic expression of miR-K12-11 or miR-155 in human leukocytes during hematopoiesis leads to increased CD19⁺ B-cell expansion in the spleen. Splenocytes harvested from mice expressing the empty vector (n = 4), miR-K12-11 (n = 5), or miR-155 (n = 7) were stained with antibodies specific for human CD45⁺ leukocytes, CD19⁺ B cells, CD3⁺ T cells, and CD33⁺ monocytes and analyzed by FACS. (A) The fraction of human CD45⁺ leukocytes and CD19⁺ B cells was significantly higher (*, P < 0.05) in mice expressing either miR-K12-11 or miR-155 than in those expressing the empty control vector. No change was detected in the CD33⁺ monocyte or CD3⁺ T-cell population when either miRNA was expressed. Each dot represents FACS analysis of one animal from each group, and the mean score of each group is shown as a solid horizontal line. A P value (*) of 0.05 or less after a Student two-tailed t test was considered statistically significant. (B) Representative dot plots for flow cytometry analysis of splenocytes using hCD45⁺ and hCD19⁺ antibodies.

Fig. 6.

GFP-positive (miRNA-expressing) cells accounted for the overall increase in human CD45⁺ leukocytes and CD19⁺ B cells. Splenocytes harvested from mice expressing the empty vector (no-miRNA control, n = 4), miR-K12-11 (n = 5), or miR-155 (n = 7) were stained with antibodies specific for human CD45⁺ leukocytes or CD19⁺ B cells and analyzed for GFP expression by FACS. (A) The fraction of GFP-positive CD45⁺ human leukocytes was significantly higher (*, P < 0.05) in mice expressing either miR-K12-11 or miR-155 than in those expressing the empty control vector. (B) The fraction of GFP-positive CD19⁺ human B cells was significantly higher (*, P < 0.05) in mice expressing either miR-K12-11 or miR-155 than in those expressing the empty control vector. Each dot represents FACS analysis of one animal from each group, and the mean score of each group is shown as a solid horizontal line. A P value (*) of 0.05 or less after a Student two-tailed t test was considered statistically significant.

Next, we asked whether the observed expansion of hCD19⁺ B cells in the spleen was due to increased frequencies of B-cell subsets expressing CD10 or surface IgM. In all of the animals, regardless of ectopic miRNA expression, the hCD19⁺ B-cell population was significantly enriched for IgM expression (miR-K12-11, 84.8% ± 4.4%; miR-155, 87.9% ± 4.1%; vector control, 88.1% ± 3.4%), indicating that the majority of the cells had differentiated into a more mature phenotype after migrating from the BM to the spleen (see Fig. S2A in the supplemental material). Furthermore, when hCD19⁺ cells were gated for GFP (miRNA expressing) and analyzed for IgM expression, there was no significant difference between the groups (miR-K12-11, 86.9% ± 3.6%; miR-155, 86.7% ± 3.5%; empty-vector control, 89.8% ± 3.1%) (see Fig. S2B). Compared to IgM expression, CD10 expression was lower in hCD19⁺ B cells, but again there was no difference between the groups (miR-K12-11, 67.9% ± 6.6%; miR-155, 69.7% ± 8.1; vector control mice, 68.2% ± 8.8%) (see Fig. S2C). Gating for GFP also revealed no significant difference in CD10 expression between the miR-K12-11 (67.3% ± 9.6%), miR-155 (73.6% ± 8.2%), and empty-vector control mice (76.0% ± 7.1%) (see Fig. S2D). Together, these data suggest that while ectopic expression of both miR-K12-11 and miR-155 had a significant effect on B-cell proliferation, B-cell differentiation, as assessed by the distribution of CD10- and IgM-expressing cells, was not affected in this model.

miR-155 and miR-K12-11 expression leads to hCD19⁺ B-cell infiltrates in splenic red pulp.

Histopathological examination of BM from femurs and tibias after H&E staining revealed no major differences in cellularity, with the majority of the animals displaying large numbers of nucleated cells. We also found no significant differences in the hCD19⁺ B-cell population in the BM of mice when it was examined by IHC analysis using an hCD19 antibody, which supports the FACS data (Fig. 4B). Initial gross analysis of the spleen did not indicate any abnormalities in weight or size in any of the mice examined. However, H&E staining and IHC analysis of the spleen for hCD19⁺ B cells confirmed the significant expansion of human B cells in the miRNA-expressing mice (Fig. 7), as observed by FACS analysis (Fig. 5A). Furthermore, we observed peculiar differences in the splenic localization of hCD19⁺ B cells in the miRNA-expressing mice. While the majority of the B cells from empty-vector control mice were localized interior to the periarteriolar lymphoid sheaths (PALS), reflecting normal spleen architecture, we observed large numbers of hCD19⁺ cells from the miR-K12-11- and miR-155-expressing mice infiltrating and expanding into the splenic red pulp regions outside the PALS (Fig. 7). These B-cell infiltrates appear to disrupt the normal architecture of the PALS and may either indicate a homing defect or be a direct result of aberrant B-cell proliferation. Interestingly, a similar immunophenotype of splenic red pulp B-cell infiltrates was previously reported for studies where miR-155 was overexpressed in the Eμ-miR-155 transgenic mouse model (9, 10).

Fig. 7.

Immunohistochemical analysis of spleens revealed an increase in human CD19⁺ B-cell infiltrates in the splenic red pulp of mice expressing miR-K12-11 or miR-155. For IHC analysis, spleens were fixed, sectioned, and stained with a monoclonal antibody against human CD19. Photomicrographs of splenic sections at a ×40 magnification are shown at the top. The splenic red pulp regions are further magnified (×200) in the bottom panels to show the increased hCD19⁺ B-cell infiltrates (red staining) in the miRNA-expressing animals versus those in the no-miRNA control. Shown are representative sections from one animal in each group.

C/EBPβ is targeted by miR-K12-11 in splenocytes and PEL cells.

A number of miR-155 targets have previously been identified, including C/EBPβ, a transcription factor involved in B-cell lymphomagenesis (9). C/EBPβ is a negative regulator of IL-6, a cytokine associated with the proliferation of KSHV-infected B-cell malignancies (1, 16, 24, 37, 48). Hence, we investigated whether miR-K12-11 also targets C/EBPβ, thereby providing a possible mechanism for the observed splenic B-cell expansion.

The 3′ UTR of C/EBPβ contains one putative binding site for both miR-K12-11 and miR-155 (Fig. 8A). Previous studies have shown that miR-155 can directly target and repress reporter constructs containing portions of the C/EBPβ 3′ UTR with the miR-155 binding site (9, 36, 57). To test the ability of miR-K12-11 to target and repress C/EBPβ, we inserted the full-length C/EBPβ 3′ UTR into a reporter vector downstream of the firefly luciferase gene. Cotransfection of the C/EBPβ reporter construct with either a miR-K12-11 or a miR-155 expression vector in 293T cells resulted in a 50% repression of luciferase activity compared to that in the no-miRNA control, indicating that both miRNAs can target C/EBPβ (Fig. 8B).

Fig. 8.

C/EBPβ is targeted by both miR-K12-11 and miR-155 in splenocytes and is regulated by miR-K12-11 in PEL. (A) The C/EBPβ 3′ UTR contains 1 seed match site for miR-K12-11 and miR-155. CDS, coding sequences. (B) The full-length C/EBPβ 3′ UTR was cloned (nucleotides 1233 to 1836) downstream of luciferase (pGL3-C/EBPβ) and cotransfected into 293T cells with increasing amounts (400 ng and 800 ng) of miR-155 or miR-K12-11 expression plasmids and a Renilla luciferase control vector. Transfection was normalized to Renilla luciferase values, and firefly values were graphed as relative light units. (C) RNA harvested from splenocytes from two separate animals from each group (empty-vector control, miR-K12-11, or miR-155) was analyzed by qRT-PCR for expression of C/EBPβ mRNA and normalized to GAPDH. (D) miR-K12-11 function in PEL cell lines BCBL1 and BC1 was inhibited using 25 nM 2′OMe antagomir specific for miR-K12-11. RNA was harvested from these cells, and derepression of C/EBPβ mRNA was analyzed by qPCR and normalized to GAPDH. Mock-transfected cells were used as a control. All experiments represent the average of three independent replicates and were repeated at least two times.

Next, we wanted to determine if ectopic miR-155 and miR-K12-11 expression correlates with reduced endogenous C/EBPβ mRNA levels in harvested splenocytes. Using qRT-PCR, we found that C/EBPβ transcript levels were reduced (miR-K12-11, 0.4-fold; miR-155, 0.5-fold) compared to those in empty-vector control mice, indicating that these miRNAs regulate C/EBPβ expression in our mice (Fig. 8C). To investigate the ability of endogenous miR-K12-11 to regulate C/EBPβ in PEL cells, we inhibited miR-K12-11 function with specific antagomirs. Inhibition of miR-K12-11 in two PEL cell lines (BCBL1 and BC3) resulted in moderate derepression of C/EBPβ mRNA levels (BCBL1, 0.25-fold; BC3, 0.26-fold) measured by qRT-PCR (Fig. 8D). These analyses validate C/EBPβ as a miR-K12-11 target and suggest one possible mechanism to explain the observed splenic B-cell expansion.

DISCUSSION

miR-155 was one of the first described “oncomirs” (a miRNA with tumorigenic activity) based on its aberrant expression in B-cell lymphomas (14). Within this context, the finding that miR-K12-11 and miR-155 have identical seed sequences immediately led to the hypothesis that miR-K12-11 could mimic miR-155, thereby contributing to KSHV tumorigenesis (20, 49).

To determine whether miR-K12-11 can phenocopy miR-155 activity in vivo, we utilized the humanized NOD/LtSz-scid IL2Rγ^null mouse model. In summary, we demonstrate that ectopic expression of miR-K12-11 or miR-155 led to an increased expansion of human B cells in the spleen. Furthermore, this increase was accompanied by B-cell infiltrates within the splenic red pulp, a phenotype which was previously described in miR-155-overexpressing mice using the Eμ-miR-155 transgenic mouse model (9, 10).

This study describes the first phenotype for a KSHV-encoded miRNA in the context of human hematopoiesis and more specifically B-cell development. The ability of miR-155 to induce lymphoproliferative diseases when overexpressed in hematopoietic cells during differentiation has been previously documented in studies using nonhumanized mouse models (10, 36). Interestingly, the phenotypes observed in these studies differed, depending on the type of progenitor cell and mouse model used. miR-155 overexpression in a B-cell-restricted manner induced B-cell proliferation, while ubiquitous expression in adult murine HSCs induced deregulated myeloproliferation (10, 36), suggesting that miR-155 plays a role in regulating several differentiation pathways during hematopoiesis (for a review, see reference 35).

In our NOD/LtSz-scid IL2Rγ^null mouse model, ectopic miR-155 or miR-K12-11 expression, but not overexpression, in CD34⁺ human CB progenitors induced a splenic expansion of mature B cells without a marked inhibition of myeloid lineages. Our observations resemble the splenic B-cell proliferation reported in the Eμ-miR-155 transgenic mouse but do not correlate with the reduction of mature IgM⁺ B cells seen in that model (9, 10). We also observed no increase in myelopoiesis, which was previously reported during inflammatory responses and during ectopic expression of miR-155 in murine BM-derived HSCs (34, 36). In our model, the absence of an increased B-cell population in BM may suggest that the cells ectopically expressing either miR-K12-11 or miR-155 had already migrated from the BM at this point of differentiation and/or that the miRNAs in our system might only be affecting later time points of differentiation in the spleen.

Since KSHV, is a human pathogen, we chose to transduce CD34⁺ human CB progenitors and not murine BM-derived adult HSCs. Furthermore, our experiment was carried out under steady-state conditions without the use of either inflammatory inducers or IL-6, which has been shown to increase myelopoiesis and suppress lymphopoiesis at early stages of differentiation in the BM (33). Importantly, in our system, miR-K12-11 and miR-155 were not overexpressed but were expressed at levels similar to those in the PEL cell line BCBL1, eliminating potential off-target consequences due to miRNA oversaturation.

Although the consequences of miR-155 expression for the hematopoietic system vary, depending on the model system used, our study clearly demonstrates that miR-K12-11 can phenocopy the lymphoproliferative activity of miR-155 during hematopoiesis in vivo. The ability to induce B-cell proliferation strongly indicates a role for miR-K12-11 in promoting KSHV lymphomagenesis and provides evidence supporting previous studies with MDV, EBV, and REV-T that targeting of the miR-155 regulatory pathway is conserved among transforming herpesviruses (3, 27, 28, 30, 55, 58, 59).

To delineate the underlying molecular mechanisms contributing to the observed B-cell expansion/proliferation, we searched for miR-155 targets that could also be regulated by miR-K12-11 in B-cell malignancies. Our search identified C/EBPβ as a potential candidate based on its regulation by miR-155 in B-cell lymphoproliferation (9). We confirmed direct targeting of the C/EBPβ 3′ UTR by miR-K12-11 using luciferase reporter constructs and correlated repression of C/EBPβ mRNA in splenocytes ectopically expressing miR-K12-11 or miR-155. Lastly, regulation of C/EBPβ in PEL cell lines was validated by inhibiting miR-K12-11 with antagomirs, leading to derepression of C/EBPβ mRNA.

C/EBPβ is a negative regulator of IL-6, and its deficiency in mice has been shown to induce a B-cell lymphoproliferative disorder that closely resembles human MCD, a malignancy closely associated with KSHV infection (46, 50). The development of MCD in C/EBPβ-deficient mice has been linked to dysregulated IL-6 production (46), while the clinical presentation of KSHV-associated MCD is correlated with high plasma IL-6 and IL-10 levels (37, 57). IL-6 and IL-10 are cytokines that function in an autocrine and paracrine fashion to promote the proliferation and survival of B cells, including PEL (1, 16, 24, 25, 37, 43, 48). To our knowledge, there has been no reported correlation between KSHV B-cell lymphomagenesis and C/EBPβ repression, but a recent study bioinformatically predicted that C/EBPβ could be targeted by multiple KSHV miRNAs, including miR-K12-11 (41). Qin et al. also showed that these KSHV miRNAs induce IL-6 and IL-10 production in monocytes and macrophages but did not confirm that this was due to direct miRNA regulation of C/EBPβ. Because lack of C/EBPβ has been shown to lead to deregulated IL-6 in MCD, we believe that miR-K12-11 induces IL-6 expression in KSHV-infected B cells by repressing C/EBPβ, thereby promoting B-cell proliferation. The ability of IL-6 to stimulate B-cell proliferation may also explain the increase in GFP-negative CD19⁺ cells that we observed in our mice. Further studies are ongoing to determine the potential role of miR-K12-11 induction of IL-6 in B-cell proliferation. In this study, we have shown that C/EBPβ is indeed a direct target of miR-K12-11 and further establish a direct correlation between KSHV miRNA regulation of C/EBPβ and KSHV B-cell lymphomagenesis in vivo.

In addition to C/EBPβ, a number of other miR-155 targets that play roles in hematopoietic malignancies and B-cell function have been identified (3, 20, 28, 34, 42, 49, 52, 56). While we have identified one gene regulated by miR-K12-11 both in our mouse model and in PEL cells, it is highly probable that this is not the only miR-155 gene deregulated by miR-K12-11 that contributes to KSHV B-cell lymphomas. Additional work is still needed to identify those targets which have functional relevance in KSVH-associated malignancies.

During latency, KSHV expresses a small set of viral genes, including that for V-cyclin, a cyclin D homolog; that for V-Flip, a potent inducer of NF-κB; that for LANA, a modulator of host gene expression; and that for kaposin, which stabilizes cytokine mRNAs (for a review, see reference 12). While ectopic expression has unmasked limited transforming potential for each of these genes, in vitro KSHV infection of either lymphoid or endothelial cells rarely leads to outgrowth of transformed cells (15, 53). Since all KSHV miRNAs and the above proteins are coexpressed during latency, it is plausible that they work synergistically to deregulate host transcriptional networks promoting cell proliferation and transformation (4, 24). Here, we show that miR-K12-11 expression alone induces human B-cell proliferation in the context of hematopoiesis. Other KSHV miRNAs have been found to repress proapoptotic, antiangiogenic, and immunostimulatory factors, thereby potentially contributing to lymphomagenesis, a notion that is testable using our NOD/LtSz-scid IL2Rγ^null mouse model (20, 23, 32, 41, 45, 49, 60).

In summary, this in vivo study further validates miR-K12-11 as a functional mimic of miR-155. The discovery that miR-K12-11 can promote B-cell proliferation suggests a novel mechanism by which a KSHV miRNA contributes to lymphomagenesis.