Role of Phosphatidylinositol 3-Kinase in Friend Spleen Focus-Forming Virus-Induced Erythroid Disease

Mouse and virus inoculation.p85α^+/− mice were obtained from RIKEN BRC. p85α^−/− mice (45, 47) were backcrossed to BALB/c mice for 12 generations before intercrossing them with heterozygous mice. The colony has been maintained in our animal facility under specific pathogen-free conditions by mating p85α^−/− male mice and p85α^+/− female mice. Two of the three isoforms of pik3r1, p55α and p50α, are intact in these mice (45, 47). p85α^+/− and p85α^−/− mice were inoculated intravenously with the following three different strains of SFFV: Fv-2-restricted Friend SFFV-P, obtained from SFFV_AP-L-transfected NIH 3T3 cells superinfected with Friend murine leukemia virus (F-MuLV) (49); Fv-2-restricted Friend SFFV-A, obtained from SFFV_A-L-transfected NIH 3T3 cells that had been superinfected with F-MuLV (49); and Fv-2-independent SFFV-BB6, obtained from infecting SFFV-BB6 clone 13 NIH 3T3 cells with F-MuLV (27). After 14 days, mice were sacrificed for further analysis. All experiments were performed in accordance with our universities' guidelines.

Phenylhydrazine (PHZ) treatment.Mice were treated intraperitoneally with phenylhydrazine (60 mg/kg at 0 and 24 h), and the mice were sacrificed and investigated on day 5.

Primary erythroleukemic cells and cell lines.Primary erythroleukemic cells were prepared by injection of F-MuLV/Friend SFFV-P into NIH Swiss mice, as previously described (41). 293T, NIH 3T3/SFFV-P, NIH 3T3/SFFV-A, and NIH 3T3/SFFV-BB6 cells were cultured in Dulbecco's minimal essential medium (DMEM) supplemented with 10% fetal calf serum (FCS). Mouse erythroleukemia cell lines NP1, NP4, and NP7 were established from mice infected with helper-free Friend SFFV-P (50). These cell lines were maintained in Dulbecco's minimal essential medium supplemented with 10% FCS.

Protein analysis.Cell lysates were prepared by resuspending cells in lysis buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA, 1 mM Na₃VO_4, 1 μg/ml each of aprotinin and leupeptin), followed by incubation on ice for 20 min. Insoluble components were removed by centrifugation, and protein concentrations were determined with a protein assay kit (Bio-Rad Laboratories). Proteins were separated by electrophoresis on Tris-glycine minigels and then transferred electrophoretically to nitrocellulose filters for Western blotting with rat anti-SFFV env monoclonal antibody (7C10) (48); rabbit anti-p85α, anti-phospho-Akt, and anti-Akt (Cell Signaling Technology, Beverly, MA); or goat anti-actin (Santa Cruz Biotechnology). Blots were incubated with anti-rat IgG antibody conjugated to horseradish peroxidase (HRP) (Cell Signaling Technology, Beverly, MA), anti-rabbit IgG antibody conjugated to HRP (GE Healthcare), or anti-goat IgG antibody conjugated to HRP (Santa Cruz Biotechnology), followed by visualization using enhanced chemiluminescence (ECL) (GE Healthcare). For immunoprecipitation assays, cell lysates were immunoprecipitated with mouse anti-hemagglutinin (HA) antibody (Roche Diagnostics, Indianapolis, IN) at 4°C overnight. Immune complexes were then collected with protein G-agarose (Santa Cruz Biotechnology) and transferred electrophoretically to nitrocellulose filters. The filters were then blotted with mouse anti-myc conjugated with HRP (Wako Pure Chemical Industries, Osaka, Japan) or mouse anti-HA followed by anti-mouse IgG conjugated to HRP (GE Healthcare) before using the ECL Western blotting system to detect bound proteins. Quantification of phospho-Akt was determined by scanning filters using ImageJ 1.37v software and determining the ratio of phospho-Akt to total Akt.

Plasmids.pEFsf-Stk (wild-type) and pEFsf-Stk K190M (kinase-inactive mutant) expression vectors tagged with N-terminal myc were described previously (27). pEFsf-StkY429F and pEFsf-StkY436F mutant vectors, with each tyrosine (Y)-to-phenylalanine (F) substitution within sf-Stk, were generated by PCR amplification with the following primers (mutated codons are underlined): for pEF sf-StkY429F construction, Y429F-S (5′-TCGGATCCAACATGACCGTGGGTGGTGAGGTCTGCC-3′) and Y429F-R (5′-TAAGCTGCTGTCAGCTGCACAAAGTGGTCCCCAAGCAGTG-3′); and for pEFsf-StkY436F construction, Y436F-S (5′-GTGCAGCTGACAGCAGCTTTTGTGAACGTAGGCCCCAG-3′) and Y436F-R (5′-AGAATTCCAAGTGGGCAGGGGTGGCTCTGAGAGAGGC-3′). Each PCR-mutated fragment was replaced with each normal fragment of pEFsf-Stk. HA-tagged p85α was kindly provided by T. Kadowaki (University of Tokyo).

Transfection.293T cells plated in 60-mm dishes were transfected with SFFV_AP-L (49) and p85α plasmids along with either the pEFsf-Stk, pEFsf-StkK190M, pEFsf-StkY429F, or pEFsf-StkY436F plasmids using Lipofectamine 2000 reagent (Invitrogen). Approximately 48 h after transfection, cells were collected and further analyzed.

Inhibitors.The protein kinase inhibitor LY294002 was purchased from Calbiochem (Darmstadt, Germany). The inhibitor was reconstituted in dimethyl sulfoxide (DMSO) (Sigma, St. Louis, MO).

Cell proliferation assay.Cell proliferation was assayed using the WST-1 reagent (Roche Diagnostics, Indianapolis, IN), according to the manufacturer's instructions and as previously described (30). Cells were plated in 96-well microtiter plates with various concentrations of the inhibitor in vehicle or with vehicle alone and incubated for 24 h before the addition of the cell proliferation reagent WST-1. After incubation of the samples in a humidified atmosphere, the absorbance of samples was measured at 450 nm against that of background controls using an enzyme-linked immunosorbent assay reader.

Flow cytometric analysis.Phycoerythrin (PE)-conjugated anti-Ter119 antibody (eBioscience), fluorescein isothiocyanate (FITC)-conjugated anti-Ter119 antibody (Miltenyi Biotec), and their PE rat IgG2b isotype controls (eBioscience) and FITC rat IgG2b isotype controls (Medical & Biological Laboratories) were used for cell surface phenotyping. Spleen mononuclear cells prepared using Histopaque (Sigma-Aldrich) were incubated with antibody for 30 min at 4°C before being washed in phosphate-buffer saline (PBS) containing 1% fetal bovine serum (FBS) and fixed with 1% formaldehyde in PBS. The cells were analyzed using the FACScan flow cytometer.

Statistical analysis.Statistical analysis was performed by using the nonparametric Mann-Whitney U test.

Role of p85α in steady-state and stress erythropoieses in adult mice.It is known that the p85α regulatory subunit of class IA PI3-kinase is required for normal murine fetal erythropoiesis (10), but it is not clear if it is required for normal adult erythropoiesis in the mouse. The availability of a nonlethal knockout mouse model for pik3r1 in which only the p85α regulatory subunit of PI3-kinase is deleted allowed us to investigate the role of p85α in steady-state and stress erythropoieses in adult mice. When untreated, age-matched p85α^+/− and p85α^−/− mice were examined for spleen weights and the percentages of Ter119-positive erythroid cells in the spleens (Fig. 1A), as well as for hematocrit (Ht) levels (data not shown). No significant differences could be detected, indicating that steady-state erythropoiesis was not affected by p85α status. In contrast, p85α^+/− and p85α^−/− mice treated with phenylhydrazine (PHZ), which causes hemolytic anemia, showed a significant difference in their response to erythropoietic stress (Fig. 1B) While p85α^+/− mice treated with PHZ showed a large increase in spleen weights (range, 0.47 to 0.69 g; average, 0.56 g) and frequencies of Ter119⁺ cells (range, 46.4 to 73%; average, 57.7%) in the spleen to compensate for the hemolytic anemia induced, the spleens of p85α^−/− mice treated with PHZ failed to support efficient erythroid hyperplasia, as indicated by significantly lower spleen weights (range, 0.2 to 0.32 g; average, 0.26 g) and frequencies of Ter119⁺ cells (range, 5.0 to 11.6%; average, 9.6%). Thus, p85α-deficient mice are unable to support stress-induced erythropoiesis, consistent with previous data indicating that a functional p85α gene is required for normal murine fetal erythropoiesis (10) and that BMP4/Smad5-dependent stress erythropoiesis is required for the expansion of erythroid progenitors during fetal development (39).

• Open in new tab
• Download powerpoint

FIG. 1.

Erythropoiesis in untreated and phenylhydrazine-treated p85α-deficient mice. Age-matched p85α^+/− mice or p85α^−/− mice were treated with phenylhydrazine (PHZ), and after 5 days, the increase in splenic erythropoiesis due to hemolytic anemia was investigated by determining the spleen weights (g) and frequencies of Ter119-positive cells (%). (A) Untreated p85α^+/− mice (n = 9) or p85α^−/− mice (n = 6); (B) PHZ-treated p85α^+/− mice (n = 5) or p85α^−/− mice (n = 4). Statistical analysis was performed by using the nonparametric Mann-Whitney U test.

Role of p85α in SFFV-induced erythroleukemia.Many signal transduction pathways are constitutively activated in hematopoietic cells infected with Friend SFFV, including the PI3-kinase pathway. Since Friend SFFV has been shown to utilize the BMP4-dependent stress erythropoiesis pathway to induce erythroleukemia (44), and our data shown above indicate that p85α is required for the response of mice to erythropoietic stress, we carried out studies to determine if the p85α regulatory subunit of PI3-kinase plays a role in SFFV-induced erythroid hyperplasia. For these studies, adult p85α^+/− and p85α^−/− mice (both strains carry the Fv-2^ss genotype that expresses sf-Stk) were inoculated intravenously with the following three different strains of SFFV: sf-Stk-dependent Friend SFFV-P and SFFV-A and the sf-Stk-independent mutant SFFV-BB6. After 14 days, mice were analyzed for erythroid hyperplasia, as indicated by spleen weights and the percentages of Ter119-positive erythroid cells in the spleens. As an indicator of erythroid cell differentiation, hematocrit (Ht) values were also determined. Uninfected adult p85α^+/− mice and p85α^−/− mice exhibited no difference in spleen weights and the percentages of erythroid cells in the spleens (Fig. 1A).

p85α^+/− mice infected with SFFV-P showed both splenomegaly (spleen size, 1.6 to 2.8 g; average, 2.04 g) and polycythemia (Ht, 37 to 55%) (Fig. 2A). Although the spleens from SFFV-P-infected p85α^−/− mice also showed an increase in size (range, 1.1 to 1.6 g; average, 1.4 g), this increase was significantly (P = 0.003) smaller than that in virus-infected p85α^+/− mice. Like SFFV-P-infected p85α^+/− mice, p85α^−/− mice infected with SFFV-P also showed polycythemia (Ht, 44 to 61.5%), and interestingly, these values were even higher than those in SFFV-P-infected p85α^+/− mice (Fig. 2A). When the spleen cells were analyzed for the presence of the erythroid cell surface marker Ter119, those from SFFV-P-infected p85α^+/− mice expressed higher percentages of Ter119-positive cells (range, 63.9 to 88.7%; average, 78.8%) than virus-infected p85α^−/− mice (range, 40.1 to 70.4%; average, 60.7%) (Fig. 2A). The reduction in the frequencies of Ter119-positive cells correlates with the significantly reduced spleen sizes in the p85α^−/− mice, suggesting that p85α may regulate the SFFV-P-induced proliferation of erythroleukemic cells.

• Open in new tab
• Download powerpoint

FIG. 2.

SFFV-induced erythroleukemia in p85α-deficient mice. Adult p85α^+/− or p85α^−/− mice were injected with 3 different strains of SFFV and analyzed 14 days later for spleen weights in grams (g), frequencies of Ter119-positive cells (%), and hematocrit (Ht) levels (%). (A) p85α^+/− mice (n = 6) or p85α^−/− mice (n = 9) were inoculated intravenously at 5 to 8 weeks of age with 0.3 ml of SFFV-P-containing virus preparations; (B) p85α^+/− mice (n = 11) or p85α^−/− mice (n = 6) (all 7 to 11 weeks old) were inoculated intravenously with 0.3 ml of SFFV-A-containing virus preparations; (C) p85α^+/− mice (n = 10) or p85α^−/− mice (n = 5) (all 7 weeks old) were inoculated intravenously with 0.3 ml SFFV-BB6-containing virus preparations. Statistical analysis was performed by using the nonparametric Mann-Whitney U test. Horizontal bars indicate mean values.

Similar results were obtained from mice infected with SFFV-A (Fig. 2B). Spleens from SFFV-A-infected p85α^+/− mice showed a significant increase in size (range, 1.1 to 2.5 g; average, 1.9 g) compared with those from uninfected p85α^+/− mice. Although p85α^−/− mice also showed an increase in spleen size (range, 0.9 to 1.5 g; average, 1.2 g) after infection with SFFV-A, the increase was significantly (P = 0.011) less than that in SFFV-A-infected p85α^+/− mice. The frequencies of Ter119-positive cells in the spleens of SFFV-A-infected p85α^+/− mice ranged from 53.0 to 75.6% but in p85α^−/− mice, ranged only from 31.5 to 53.4% (Fig. 2B). Thus, the reduction in Ter119-positive cells correlates with the decrease in spleen size in the SFFV-A-infected p85α^−/− mice. All of the mice infected with SFFV-A were anemic (Fig. 1B), with p85α^+/− mice showing Ht values of 20 to 39% and p85α^−/− mice showing values of 27 to 41%. Thus, p85α expression appears to play a role in regulating the proliferation, but not the differentiation, of erythroleukemia cells induced by both SFFV-P and SFFV-A.

The induction of erythroleukemia by both SFFV-P and SFFV-A is dependent upon the Fv-2 genotype of the mouse and occurs only in Fv-2^ss mice, which express the receptor tyrosine kinase sf-Stk. To determine if p85α also plays a role in the erythroleukemia induced by an sf-Stk-independent mutant of SFFV, we infected p85α^+/− and p85α^−/− mice with SFFV-BB6. As shown in Fig. 2C, both groups of mice developed splenomegaly, with similar spleen weights (0.9 to 1.6 g for p85α^+/− mice and 0.8 to 1.3 g for p85α^−/− mice). The frequencies of Ter119-positive cells between the SFFV-BB6-infected p85α^+/− and p85α^−/− groups (25.8 to 58.5% and 30 to 46%, respectively) were not significantly different. Both groups showed polycythemia, with no significant differences in Ht levels in the p85α^+/− and p85α^−/− mice (36 to 61% and 45 to 70%, respectively) (Fig. 3C), indicating that p85α status did not affect erythroid cell differentiation. Thus, unlike erythroleukemia induced by sf-Stk-dependent SFFV-P and SFFV-A, erythroleukemia induced by sf-Stk-independent SFFV-BB6 does not appear to be regulated by p85α.

• Open in new tab
• Download powerpoint

FIG. 3.

Expression of viral and cellular proteins in SFFV-infected p85α-deficient mice. Spleen cell lysates were prepared from p85α^+/− or p85α^−/− mice infected with SFFV-P, SFFV-A, or SFFV-BB6 and examined by Western blot analysis for expression of p85α (A), expression of phosphorylated Akt (pAkt), as represented by the mean expression level of phosphorylated Akt/expression level of total Akt ratio in individual spleens ± SD (B), and expression of SFFV Env (C).

Examination of splenic lysates using Western blot analysis for p85α confirmed that p85α was expressed in the virus-infected p85α^+/− mice but not in the p85α^−/− mice (Fig. 3A). Furthermore, the levels of expression of phosphorylated Akt, a downstream component of the PI3-kinase signaling pathway, were reduced in the spleens of the virus-infected p85α^−/− mice compared with those levels in the spleens of p85α^+/− mice (Fig. 3B). Importantly, spleens from the virus-infected mice showed high levels of expression of their respective SFFV envelope proteins, with no significant differences in the levels of expression in p85α^+/− and p85α^−/− mice (Fig. 3C).

Interaction of sf-Stk with p85α.Since our data indicate that p85α status is more important for the induction of erythroleukemia by SFFV isolates that are sf-Stk-dependent than by those that are sf-Stk independent, we carried out studies to determine whether sf-Stk was interacting with p85α. 293T cells transfected with plasmids expressing myc-tagged sf-Stk, SFFV_AP-L, and HA-tagged p85α were immunoprecipitated with anti-HA antibody to precipitate p85α and then immunoblotted with anti-myc antibody conjugated with HRP for detection of sf-Stk. As shown in Fig. 4, when anti-HA p85α immunoprecipitates were immunoblotted with anti-myc, a band corresponding to the sf-Stk wild type was detected (lane 1), indicating that sf-Stk and p85α are interacting. p85α failed to interact strongly with a kinase-inactive mutant of sf-Stk (K190M) (lane 4) or with an sf-Stk construct in which tyrosine 436 in the multifunctional docking site of sf-Stk had been mutated to phenylalanine (lane 3), a change which was previously shown to be important for the induction of Epo independence by sf-Stk in conjunction with SFFV (5, 41). Mutation of tyrosine 429 in sf-Stk, which was previously shown to have little effect on the induction of Epo independence by sf-Stk in conjunction with SFFV (5, 41), did not affect its ability to interact with p85α (lane 2). These results indicate that sf-Stk interacts with p85α and that kinase activity and tyrosine 436 of sf-Stk contribute to the interaction.

• Open in new tab
• Download powerpoint

FIG. 4.

Interaction sf-Stk with p85α. 293T cells were transfected with SFFV_AP-L, HA-tagged p85α, and the indicated myc-tagged sf-Stk-expressing plasmids. Cell lysates were immunoprecipitated (IP) with anti-HA to precipitate p85α, followed by immunoblotting (Western blotting [WB]) with HRP-conjugated anti-myc antibody to detect sf-Stk or with anti-HA- and anti-HRP-conjugated anti-rabbit IgG to detect p85.

Proliferation of both primary and immortal erythroleukemia cells from Friend SFFV-P-infected mice is inhibited by treatment with a PI3-kinase inhibitor.Friend SFFV-induced erythroleukemia occurs in two stages. In the first stage, Friend SFFV causes Epo-independent erythroid proliferation and differentiation. Primary erythroleukemic splenocytes from these animals can grow in liquid culture for only a short time but can both proliferate and differentiate in semisolid medium, forming hemoglobin-positive erythroid colonies. In the second stage, Friend SFFV-infected erythroid cells become blocked in differentiation, resulting in the outgrowth of transformed erythroleukemia cells that can be grown as cell lines. While we could establish erythroleukemia cell lines from SFFV-P-infected p85α heterozygous mice, we failed to do so with SFFV-P-infected p85α-deficient mice (data not shown), suggesting that PI3-kinase activity is important for establishing and maintaining immortal erythroleukemia cell lines from SFFV-infected mice. In support of this, we showed that the PI3-kinase inhibitor LY294002 not only inhibited the growth of primary erythroleukemic splenocytes from a Friend SFFV-P-infected mouse but also inhibited the growth of immortal erythroleukemia cell lines derived from these mice (Fig. 5). Growth inhibition occurred as early as 24 h after exposure in a dose-dependent fashion, with kinetics similar to that seen using Friend SFFV-P-infected HCD-57 cells, as previously reported (29). These results indicate that blocking PI3-kinase activity in primary and immortal erythroleukemia cells results in inhibition of cell growth and suggests that PI3-kinase plays an important role in leukemic cell survival. Interestingly, PI3-kinase also appears to play an important role in the survival of leukemic cells from SFFV-BB6-infected mice, whose proliferation was inhibited in vitro by the PI3-kinase inhibitor LY294002 (data not shown). Since p85α status had no effect on erythroid hyperplasia induced in mice by SFFV-BB6, the PI3-kinase inhibitor results suggest that other subunits of p85 may be contributing to PI3-kinase activity in SFFV-BB6-infected p85α^−/− mice.

• Open in new tab
• Download powerpoint

FIG. 5.

Effects of a PI3-kinase inhibitor on the proliferation of primary erythroleukemia cells and erythroleukemia cell lines from Friend SFFV-infected mice. Erythroleukemia cells were cultured with different concentrations of the PI3-kinase inhibitor LY294002 for 24 h. Proliferation was then measured using the WST-1 reagent. The cells analyzed were primary leukemic spleen cells (○) and the erythroleukemia cell lines NP1 (⧫), NP4 (□), and NP7 (▪). The graph represents the mean results obtained from using triplicate samples. The standard error was less than 0.05. OD₄₅₀, optical density at 450 nm.